Peter G. Bryant/Marlene A.

Smith

V 0 L U Md E T W 0

AMX
NED'S BOOKSTORE
No Return w/o label

Practical

Data Analysis
Case Studies in Business Statistics
 PRACTICAL
DATA ANALYSIS
CASE STUDIES IN
BUSINESS STATISTICS
VOLUME II
 PRACTICAL
DATA ANALYSIS
CASE STUDIES IN
BUSINESS STATISTICS
VOLUME II
SECOND EDITION

Peter G. Bryant
Marlene A. Smith
College of Business and Administration
and
Graduate School of Business Administration
University of Colorado at Denver

ta Irwin
find McGraw-Hill
Boston Burr Ridge, IL Dubuque,IA Madison,WI New York San Francisco St. Louis
Bangkok Bogoté Caracas Lisbon London Madrid
Mexico City Milan NewDelhi Seoul Singapore Sydney Taipei Toronto
Irwin/McGraw-Hill
A Division of TheMcGraw-Hill Companies
PRACTICAL DATA ANALYSIS: CASE STUDIES IN BUSINESS STATISTICS VOLUME II
Copyright ©1999 by The McGraw-Hill Companies, Inc. All rights reserved. Previous edition ©1995 by Richard D. Irwin, Inc. Printed in
the United States of America. Except as permitted under the United States Copyright Act of 1976, no part of this publication may be
reproduced or distributed in any form or by any means, or stored in a data base or retrieval system, without the prior written permission
of the publisher.
This book is printed on acid-free paper.
1234567890 DOC/DOC 9321098
ISBN 0-256-23872-3
Vice president/Editor-in-chief: Michael W. Junior
Publisher: Jeffrey J. Shelstad
Senior sponsoring editor: Scott Isenberg
Developmental editor: Wanda J. Zeman
Marketing manager: Zina Craft
Senior project manager: Susan Trentacosti
Senior production supervisor: Heather D. Burbridge
Cover Designer: Steven Vena/SrV Unlimited Design
Compositor: Carlisle Communications, Ltd.
Typeface: 10/12 Times Roman
Printer: R. R. Donnelley & Sons Company
Library of Congress Cataloging-in-Publication Data
Bryant, Peter G.
Practical data analysis : case studies in business statistics /
Peter G. Bryant and Marlene A. Smith.—2nd ed.
p. cm.
Includes bibliographical references (p. ).
ISBN 0-256-23871-5 (vol. 1). —ISBN 0-256-23872-3 (vol. 2). —
ISBN 0-07-365488-4 (vol. 3)
1, Commercial statistics—Case studies. 2. Commercial statistics-
—Data processing. 3. Statistics—Data processing. I. Smith,
Marlene A. IL. Title.
HF1017.B79 1999
650’.01’5195—dc21 98-43159
http://www.mhhe.com
 To L.S.B. and L.G.B.
In memory of Reuben and Charlotte Aschenbach

And he who is versed in the science of numbers can tell of the regions of weight
and measure, but he cannot conduct you thither.
Kahlil Gibran
The Prophet

Was man nicht nitzt, ist eine schwere Last.

Johann Wolfgang Goethe
Faust I, 684

Some of the blocks to dealing comfortably with numbers . . . are due to quite
natural psychological responses to uncertainty, to coincidence, or to how a
problem is framed. Others can be attributed to . . . romantic misconceptions about
the nature and importance of mathematics.
John Allen Paulos
Innumeracy
 2 Onis 6.8.3 oT
ong :
" Teednerioed etighiart) bas advert to yromem nt
eft -*. 3)

‘ae 1

= _a P ) “ 7 ;
- 7 ~~ .

Hir flora
"
TH ry Ty :

en aes ERIN S er 32Het ngs acirmunt hy BOTHIOS ert! Ni DEN aiitededb

: ‘sf f ’ VON wey tube fone erif vasern b
wu
PAL Bis felon amserupts Kk WGA STATICS VOLIAES
' v4 oe 7 7 5 ) alike BD) oye w dieecd i.

ar ercghu Doral 1074,tap jut 'l heal

ri, we ; ' re eyes » i Me old wether
YY : —_

roar mpd evawrne onietas Hsin Stole tae SAE

cai 4 putes M
weit
e Tee a
Sr "ertiend: QeapticW ongiibk = +
b80 } ceae $.
+s, iASpoor: 8.via
. >
re “ehhup of eub sic. eck) Ure gaan ctmdo gellaab of eavoid ert to smoz ‘. ee
a _
y_ chee B aet.OF aeerabink Ko OF NX? vin wan oO} Chega tay olor ag tua
rte # Mba ancien 2TN | .0) het inte ed nea war 7 ‘bamed 3 realy
otic af =o a | reomiesertien to sihatioan’" = ante ort
ie a wee
inrmes ire a
ok aus a Seki wer v
hes ogee uw? Ee wale 2
baatSere Wisi’ Untertet oS
Carta) Ccncnanis uci
PASThorees K euamhgst

B.
tert ?.& hone Cemeur

Comers Ovtuleptny

na) rule ©wary “4 ;

y— Marluce
 INTRODUCTION FOR
THE STUDENT

INTRODUCTION
This book contains a collection of cases or situations. Most cases contain some data.
In others, we ask you to gather data or figure out how you would gather the data. Pre-
sumably the data can shed some light on the situation. Your job is to decide what light
the data shed. To do this, you will have to
¢ Understand the situation and context.
¢ Choose an appropriate technique to summarize or analyze the data.
¢ Decide what you think the data have to say.
Then you will have to communicate your findings, in a way that’s effective in the con-
text of the situation described, whether that be a presentation to a board of directors, a
memo to your boss, or something else.

CASES
In some fields of education—law or medicine, for example—you often “solve” a case
by applying your knowledge to the situation given in a systematic way, and if you ap-
ply your knowledge correctly, you arrive at a single correct, well-defined answer, like
a legal opinion or a diagnosis. Different people applying the course material correctly
will arrive at the same answer.
In business education, cases usually play a different role: they are vehicles to raise
issues, point out inherent conflicts, and stimulate debate. At the end of the debate, you
may have settled some issues, but others remain open. Your answers or conclusions will
not necessarily agree with those of your colleagues, for they will have brought differ-
ent assumptions, interpretations, and solution methods to the debate.

ti—vii
 ll-viii INTRODUCTION FOR THE STUDENT

We expect that you will debate these cases. We designed them that way. You may
debate various aspects of the situation:
e The analysis method to use.
¢ The quality of the data.
* The proper interpretation of the analysis results.
¢ The recommendations or conclusions to draw.
For example, you may find you agree with your colleagues about the appropriate analy-
sis of the data, but disagree about whether the data are of a quality that justifies a con-
clusion. Sometimes the analyses will be easy, sometimes hard, Sometimes a single
method of attack seems obvious; at other times the situation yields to various different
approaches. Some cases seem to require relatively sophisticated methods for full solu-
tion, but simple techniques will lead to at least part of the message. All the cases can
profit from discussion and debate, though; they are not purely statistics or computa-
tional homework exercises.

ee,
REMARKS
Your instructor will specify the formats he or she wants you to use for your presenta-
tion and written summary, and will guide any debate. We often ask students to prepare
a one-page summary (perhaps augmented by a chart or graph) and to come to class pre-
pared to present and/or defend their findings in a general class discussion. This works
particularly well when you prepare your report in a group. You will find it takes a lot
of work to get a group of three to five people to agree on what the message in the data
is! When you have to boil it down to one page, it puts even more pressure on you to
know exactly why you did what you did and what you think it means. Often, other stu-
dents will use different techniques and/or come to different conclusions, and you will
have to defend your approach and conclusions to them. That’s part of the fun and the
frustration: statistics is not meant to be a sterile exercise at the end of a textbook chap-
ter. It’s supposed to be techniques for real people to better inform themselves about real
situations by using real data. That sometimes means ambiguity, controversy, assump-
tions, disagreement and hard work. We think you’ll find it helpful, worthwhile, and rel-
evant, though. That’s the message from our students who’ ve tried it this way.
There’s no one “right” approach to any of these cases, although often we had a par-
ticular approach in mind when we wrote it. Almost every time we assign one of them,
though, some student points out a valid alternative we hadn’t thought about. And just
because the case is in the book doesn’t mean we think the results will be overwhelm-
ing. Sometimes, the data don’t really have much to say. That happens in real life, and
it may happen here, too.
The cases are given in random order, so you can’t tell what technique we had in mind
by where the case is in the book or by how you approached the case next to it.
Most of the cases include computer files with data, and most of the time you will
want to use a computer but not always. The presence of a computer file doesn’t neces-
sarily mean we think that you should use a computer. For those instances in which you
would like to use a computer, we have prepared data files in a variety of formats. Your
instructor will give you more information about acquiring and using the data files.
In our classes, we place a lot of emphasis on how well students write up or present
their findings. In practice, a good analysis may influence nobody if it isn’t presented
well, and we think you should make it a habit, right from the start, to put as much en-
ergy into your presentation as you do into the analysis. We don’t particularly advocate
 INTRODUCTION FOR THE STUDENT _ II—ix

fancy, color graphics or other “fashionable” approaches. Good hand-drawn graphs can
be wonderfully effective, too. The key is to communicate clearly. We discuss report
writing more in the next section.
We hope you have fun with the cases. If you come up with an example from your
own life or career that you think others would enjoy or profit from, please let us know.
Maybe the next edition will have room for it!

Peter G. Bryant
Marlene A. Smith
 ar / es ae Z r

2 sislone it wintebas J GO iene rsaat: anni <>

-,gris
eat ib ok’ vbebts dient ae a rc ant oy sv
m™) ors
. 4 The wel Conethy ufo otTao it
~ aan 2ine Fr) cuss Cetinrae arty keene Th,eis HO eliw Lala enaf We
ae hte ml ome? ener ew,je BAYA LM BNO a ahs ues rece sient
; + The enc prrarl MANO 0 Ey wi ™ [ae

fail Oa: For nda, cna Bet x nave th mane stngng iy i

/ ai Sel

a ie sna: e 7 Oe dota, hale eagice ‘ Ms ot 6 guallty Sa ;

oe ) PSasien, Soares! i. ; L fe nam es
rrentant | Te : ; 7 Rian Lh iu a tag
ay vite ay wre G é' 6 wet 7 ' ois a ; elie
4 om, has Serpe d » ih det rf its fy wage All
Feet ire 7) ib i , g. e 4 { Te AY,etuiMiss OF

neural | Ont eer, * as

»
S a YoU WOUsS Ne peer

es Dhaot i

Enel “ + disern “ The La Las

i) ‘ou. wi int’ Wikies @ W -
we rei » howalae ae
Ra (> SURre& rye

brit litwir Trt

; iy ‘ve ried & es
We a
(one ee worNan 8 ee
yeS) Lome Wwe j Ng Ps Bi
7
pac’ Cywnwit ew Ageveers.
pil
LoMi li 4 suite “— we 7 wher las mi ?

Fite Wa SW. 4 Tae — i.

a Shae Sita, ot .
7 as
pcan i fel, wilt hates
= a healban
(rome~
thee ce a
“eal & i.
a i
a
served it oth me €
ie
ol . ae Wes ke ite

J Hy Alte ite altace <aen inh oF:

a ‘mse ‘ si 5 0h wu
D

i a ro Us mas e hi a il
2 welll ial nf eet
ji
+ t>
° ig a ‘wo ee
‘« 0%

7 serene :
5 P 7
 ACKNOWLEDGMENTS

Many colleagues have reviewed individual cases and the manuscript of Practical Data
Analysis. We thank specifically:
Sung K. Ahn, Washington State University
Randy Anderson, California State University at Fresno
Bridget G. Heidemann, Seattle University
Raj Jagannathan, University of Iowa
Jerzy J. Letkowski, Western New England College
Walter J. Mayer, University of Mississippi
Dale McFarlane, Oregon State University
J. B. Orris, Butler University
Paul Paschke, Oregon State University
John R. Pickett, Georgia Southern University
Kipling M. Pirkle, Washington and Lee University
Al Schainblatt, San Francisco State University
Milo Schield, Augsburg College
William L. Seaver, University of Tennessee
Herbert Spirer, University of Connecticut
Jeffrey W. Steagall, University of North Florida
Stanley A. Taylor, California State University at Sacramento
Betty M. Thorne, Stetson University
Thomas K. Tiemann, Elon College
Wayne Winston, Indiana University at Bloomington

II-xi
II—xii ©ACKNOWLEDGMENTS

We acknowledge also the founders, organizers and participants in the annual con-
ference entitled Making Statistics More Effective in Schools of Business for their hard
work and interest in changing the way in which we teach statistics in our business
classrooms.
For specific permissions we thank the American Statistical Association for permis-
sion to reproduce items from The American Statistician; the Rocky Mountain News for
copyright permission; and Stanley D. Elofson for assisting us in obtaining copyright
permission for the 1992 Arapahoe County property assessment study. Further ac-
knowledgments are listed with individual cases.
Several friends, colleagues, and relatives helped in various ways. Edward J. Conry
advised us on the preparation of our prospectus. Lucinda Bryant, Louise Bryant, and
Lynwood Bryant read the book and commented on content, style, and presentation.
Kathryn E. Chmelir and Adriano Pimenta helped us with research, data input and
analysis, and other related tasks. The staff at Irwin/McGraw-Hill, including Scott Isen-
berg and Wanda Zeman, have also been helpful.
Finally, we thank our students and other contributors to the cases in the book. Where
appropriate, we acknowledge them individually in the source notes of their respective
cases. All of our students contributed indirectly to this project: they told us what did
and didn’t work, complained about bad cases (and rightly so!), and had fun with the
good ones.

Peter G. Bryant
Marlene A. Smith
 CONTENTS

PREPARING WRITTEN CASES AND BUSINESS REPORTS Il—xv

THE DATA DISK Il—xxi

CASE 26 Dave’s Truck Washing Service I-14

CASE 27 Housing Prices—| lI-5

CASE 28 Housing Prices—ll ll-9

CASE 29 Books Il-13

CASE 30 ELISA Plate Uniformity I-17

CASE 31 Violence on TV—I I-25

CASE 32 Overdue Bills I-27

CASE 33 The Brazing Operation I-29

CASE 34 Leading Indicators II-33

CASE 35 Interrupting Tina I-37

CASE 36 Violence on TV—Il il-43

CASE 37 Commercial Property I-47

CASE 38 AIDS Rates 1-51

CASE 39 Software Service Calls I-53

CASE 40 Titanic Shipping Company lI-55

I-xiii
 ll-xiv CONTENTS

CASE 41 Campaign Statistics lI-59

CASE 42 Hospital Charges Il-61

CASE 43 Coal Testing \|-67

CASE 44 Performance Evaluations il-71

CASE 45 Amchem’s Tax Credit I-83

CASE 46 Natural Gas Prices \|-87

CASE 47 Westar Beverage Sales I-89

CASE 48 Price-Fixing Cases lI-91

CASE 49 Property Crimes I-93

CASE 50 Rockville Physicians’ Group I-97
 PREPARING WRITTEN
CASES AND
BUSINESS REPORTS

The cases in this book require you to communicate your statistical results somehow.
You may discuss your results in class with your instructor and your other classmates.
You may give formal classroom presentations. Often, you will prepare a written report.
Writing and speaking effectively are important to your professional career. Here are
some thoughts on writing.
There are many ways to organize a business memo, but formal structure is probably
less important than clear, concise communication. Writing the summary of your results
will probably be just as difficult as doing the computations. Our students say it takes
about as long to write the results as it does to finish the statistical work. You might want
to plan accordingly.

WRITING STYLE
Once you are done with the computer work, you must interpret these findings for some-
one, perhaps a manager or your boss.

Avoid Jargon
Write so that anyone can understand it. Don’t use technical language to intimidate or
overwhelm the recipient. Unless you are writing to someone well-versed in statistical
techniques, don’t use statistical and mathematical jargon in your report. If you were the
vice president of marketing, and had never studied statistics, which of the following
two summaries would you find most valuable?
° The regression model is given by: SALES = 13,000 + 5*ADVERTISING, which
means that the y-intercept is $13,000 and the slope is $5.

Il—-xv
 ll-xvi PREPARING WRITTEN CASES AND BUSINESS REPORTS

¢ According to the model: (a) if the firm does no advertising, sales will be $13,000,
and (b) each additional dollar spent on advertising will bring in five additional dollars
in sales.
Your instructor and the textbook will use jargon. You will see specific statistical terms
like “least squares” and “p-values” in class, but don’t presume that the reader of your
memo understands them. You must translate statistical concepts, methods, and out-
comes for the uninitiated. By virtue of spending time in your statistics class, you may
forget that certain statistical concepts don’t exist in most people’s vocabularies. Here
are some guidelines.
¢ Don’t use words or phrases in your report unless you used them before signing up
for statistics.
¢ Would your next door neighbor understand the essence of your report?
¢ What would the editor of your local newspaper think about publishing your report
in tomorrow’s newspaper?

Use the Active Voice

Use the active voice. “I ate my dinner” says more than “The dinner was eaten by me,”
and says it better.

Be Clear
Say what you mean, unambiguously. “Nothing is capable of being well set to music
that is not nonsense,” is a quotable epigram, but does it mean “Nothing is capable of
being well set to music-that-is-not-nonsense”? Or does it mean “Nothing-that-is-not-
nonsense is capable of being well set to music”? If you mean “only nonsense can be set
to music,” say so.
Similarly,
“My face looks like a wedding cake left out in the rain.”
is great poetry, but leaves room for various interpretations.

ORGANIZATION
Introductory Material
To help your reader, provide some background information, such as a statement of the
problem or situation and a description of the data available to answer this question.
Since your report involves statistical analyses, you might also provide initial descriptive
statistics (e.g., means and/or standard deviations) of the more important variables, un-
less that information would simply distract from your point. Such background informa-
tion puts the problem in perspective and eases the reader into the upcoming material.

"Joseph Addison (1711), quoted in The Columbia Dictionary of Quotations. Copyright © 1993 by
Columbia University Press. All rights reserved.
2w. H. Auden (1907-73), quoted in Humphrey Carpenter, W. H. Auden, pt. 2, ch. 6 (1981). The
Columbia Dictionary of Quotations. Copyright ~ 1993 by Columbia University Press. All rights
reserved.
 PREPARING WRITTEN CASES AND BUSINESS REPORTS __ II—xvii

Good Grammar

An effective report uses good grammar, correct spelling and appropriate punctuation.
You can’t convince a reader that your statistical results are valid if your writing is poor.
Sloppy, misspelled, or disorganized reports send a message: you don’t think the report
* is very important. Brush up on your writing skills. It’s fun and valuable to write effec-
tively. Word processors may help. A spelling checker is useful, too.

Junked-Up Appendixes
Consider putting supporting statistical documentation, such as graphs, tables, and other
statistical output, at the end of your written report. Within the report, refer to these ap-
pendixes (e.g., “See Exhibit IT’) for guidance. On the other hand, if one particular table
or graph contains the essence of the point you’re making, put it with the text, where
your reader can see it quickly. Ask yourself: If I put it in with the text, will it distract
the reader more than if the reader has to turn to an appendix? A critical graph belongs
in with the text. A table giving relatively minor support to your argument belongs in an
appendix.
Don’t append every statistical printout you produced. Include the important ones. If
you can’t decide which is important and which is not, how can your readers? In fact, if
you haven’t decided which are important, you’re not done with your analysis yet.
“Don’t appendastatistical exhibit if you don’t refer to it in the report, and, of course,
don’t refer to an exhibit that’s not there” might be a good guideline.

Length
We generally limit reports to one or two pages. Your instructor may impose different
limits, but write as succinctly as you can. Present the essence of your statistical find-
ings, not a comprehensive validation of every step of your work. Managers won’t wade
through pages and pages of information. It’s easy to overwhelm a reader who wants
nothing more than a summary of the major findings, but why do so?
You must identify the fine line between too much detail and not enough detail. Don’t
make your report so short that it’s deceptive, but don’t bore or intimidate your intended
reader with unnecessary details, either. You, the reporter of the data, must decide what
to include.
Even if your report is a lengthy project, we suggest that you make the first page an
“executive summary,” written for someone who has never hadastatistics class, and
doesn’t care to have statistics explained. Include only the important results and impli-
cations derived from the data, and any necessary caveats or limitations of your find-
ings. For instance, you might not trust your results because of a small sample size, or
some confidence interval might suggest inaccuracy of a point estimate. Try to limit this
executive summary to one page, single-spaced. It’s hard to do!
To start you off, we’ve included sample reports on the next three pages. Preparing
reports will become increasingly straightforward with practice. Good luck!
 \I—-xviii ©PREPARING WRITTEN CASES AND BUSINESS REPORTS

A PRETTY LOUSY BUSINESS REPORT

TO: Tom Jones, Director of Personnel

FROM: Jana Smith, Statistician

DATE: 5/19
SUBJECT: Analysis of Personnel Data

Per your request, I have analyzed the data on sick days and the company’s new
wellness program. The results are summarized below.

The regression model suggests that there is a statistically significant relationship

between the two variables. The correlation between dollars contributed to the
wellness program and absenteeism is a positive 0.63. The standard error is 0.073.
The R-squared statistic on the simple regression model is 56%, which is pretty
good for cross-sectional data. Moreover, the F-statistic measuring the statistical
significance of the model as a whole is 54.90, indicating a good model.

When analyzing the relationship by gender, there is no statistically significant

impact here. The p-value on the categorical variable called GENDER (see
EXHIBIT ID) is .46. This means that absenteeism does not seem to be correlated
with the gender of the employee. On the other hand, it just might be a
multicollinearity problem with the two independent variables.

I'd be happy to assist in any further analysis of this data at your request.
 PREPARING WRITTEN CASES AND BUSINESS REPORTS __ II-xix

A PRETTY GOOD BUSINESS REPORT

TO: Tom Jones, Director of Personnel

FROM: Jana Smith, Statistician

DATE: 5/19
SUBJECT: Analysis of Personnel Data

As you asked, I have analyzed the data on sick days and the company’s new
wellness program. Here are my results.

Data

The personnel department provided a sample of 125 randomly chosen employee

files. From those files, we obtained:

e the company’s payment for that employee’s participation in our wellness

program,

¢ that employee’s absentee record (measured in number of absent days)

over the past two years, and

e the gender of the employee.

72% of the sample group was female.

Results

* On average, our employees missed 15 days of work in the first year. The
typical fluctuation around the average of 15 days was seven days.

¢ In the second year, after starting the wellness program, the average number
of absent days declined to 10 days, and the typical fluctuation also decreased to
2 days. We committed about $50 per employee to the wellness program last year.

¢ A statistical model (see EXHIBIT below) of the relationship between

dollars committed to wellness, absenteeism, and gender indicates that the wellness —
program hasastatistically significant relationship to absenteeism. The model
suggests that:

e each additional dollar committed to our wellness program was

associated with a two hour decline in absenteeism, and

¢ there is no statistically significant relationship between gender

and absenteeism.
Il-xx | PREPARING WRITTEN CASES AND BUSINESS REPORTS

The model would generally be considered a statistically strong one, since 56%
of the variation in absenteeism is explained by the model. Although this leaves
44% of the variation unexplained, it is difficult to do much better with the type of
data available for this analysis.

Recommendation

Because of the decline in absenteeism after the institution of the wellness

program, I recommend that the program be continued. ‘

Limitations

Consider redoing this study in another year with a larger sample size. It is not
clear whether one year is enough to observe the full benefits of the wellness
program. Also, there are some troubling aspects of the statistical results that might
be alleviated with a larger sample size. For instance, many of the standard errors in
the model (that is, measures of the accuracy of the estimates) are quite large in my
opinion.

I'd be happy to answer any further questions that you might have about my
analysis or report.

EXHIBIT

The regression equation is

ABSENT = 12.2 — 0.26 DOLLARS + 0.07 GENDER

Predictor Coef Stdev t-ratio p

Constant bere Byes 3)I S02 0.00
DOLLARS -0.26 0.04 6.50 0.00
GENDER OOF 0242)5 0.28 0.80

R-sq = 56.2% R-sq(adj) = 55.5%

Analysis of Variance

SOURCE DF So
Regression 2 9770
Error 12 7614
Total 124 17384
 THE DATA DISK

Most of the cases in Practical Data Analysis require you to analyze a data set. A printed
version (hard copy) of the data set may be found at the end of each case. The printed
version lets you get a quick, informal sense of the data. At some point, you will prob-
ably want to use astatistical or spreadsheet software package for more formal analy-
ses of these numbers. Some data sets are small enough to enter the data yourself. For
others, that task would be very time consuming.
Your instructor has access to the data files for the cases in Practical Data Analysis.
She or he will distribute those files to you in a form that is appropriate for the comput-
ing environment at your school. You may also receive instructions on how to copy the
files for your own use and storage, and how to feed the data into one or more software
programs.
Once you receive the data in machine-readable form from your instructor, you
should be able to get up and running quickly.

I—xxi
 a |
7 7 es%
ieee wena ¥ _ ar

Pines”.: *
<a eh
ai 3
ae
a on
Je X ttn we <3 “ oy Cun, Sik
- ~
a ite
Pe wh a shee
resfe 2 |Aibatgt|inkg Hae.
oe 1s 4a Ae Corinthm ieonhaatad AR
——
ow
2
oT
:
ai quienav eitetitefoithis eppe i. cons h- Aa iene Gee Sik _
fas: , be = a ’ = 7 - ’

n yeaATAC| oy , = — sake
0 me ame
mers
ae 1 = : —s

er
i ms pee singe
PSgees t ebecarniee? fim Napu sas aene
= 2 a + a .
ti 7 - ?hanthe thas + : :

ee Be =
a Ses Cimmaddhyd aang tins em, atecbear peas viestoil mas |
Chen? obi vw ar yy ewm=g @ gimagt( cee full henotite oftheeeltahetigr
‘Foran.Alan,Gare Ws Gas co ing ppett Mf the ele ea
Se Bi dated aes | apet~iwigee Bo. Ay ipa yee, [eaty ou iw r
'* Geb retet | iad A, tices > toon of the cota om antsLae

oP
;
ae
s :
f. iat, a : r ed coed
= in re > Pl te Yipee &* ane wey ay Foren questions “has you eigie ketene
eeeee ae . oe > .

al be eae naa nv ylants od vow PeLcaya Ni (pnd oreO, lech pt porno cdo oli ;
BS5 La reyat 980 dn) Veybere sat!tn bait ae ynfit te -aeab od tetyqo> Raed) apigie ——
, ae ihe Woy iniog snes tA swab oft to seems tno lor taiape
aayuoy “ial Onley —
Me daaJafar! spon 96) est204 ow thos teidtabaonqs W leoliniere a Seu oF Ioew Yide- 731
e SPT Hariyoy atnh
| ol) Pa Viigvons Hops Mh, Rhs4 ej anode apdiuin sao} ig aoa aa 4
=|i : anurans amp veyed bhugw Aail4 eTagle .
a oe Biavinlak octIona what ennaSe ya 9 abll wah ait) GI HROQuE€a a > _
oe emit!
GUb) 207 saniweyowgin ai fait) (readl A oi7
Dry ag eal seodt siudiielb ive a (Te
tons BitYeon wast ao obi rien: nie alu vain BOY Joodoeawoy ty Jasmine rok er
.i eens 2 300ort}adil petiFool ot
ol wortbas RENO? Hate, au TO Wey ae zal? m ere =o
ie: | HOBO y: o yg +, ek
a _ Hr Agta wor trodferwil "aidebest anidcinen ai atnty oath ison woe sent)
bintagy uy '
; : ; Vitsio grt HEU Le a1 2k i] aie pa
ee%
os im oe - 1 4 -
> Z : : : ; - '
ta ot EY
' ch 7 | ie . rvy ‘ ;

a 7 a g cc Gf 7
y ad ; j 4 ae ®
: | z ; \
i o , a
ie
oleae a - a

wei : :
ie 7
; :
aoe me =
-
. , ee ne
So a
 PRACTICAL
DATA ANALYSIS
CASE STUDIES IN
BUSINESS STATISTICS
VOLUME Il
 26
DAVE’S TRUCK
WASHING SERVICE

Dave is the owner of the Truck Washing Service (TWS). TWS provides cleaning ser-
vices to truckers and trucking operations within a moderately sized metropolitan area.
Dave is interested in understanding the factors that influence monthly fluctuations in
total sales of TWS.
TWS offers two types of services. Type 1 sales are from Dave’s mobile truck wash-
ing service. Dave also provides a mobile chemical treatment service for removing de-
cals and logos from trucks, called type 2 sales. Most of TWS revenues come from the
type 1 operations. Occasionally, there will be an upcoming auction of used trucks in
town. Trucks are generally cleaned up prior to auctions, and Dave can pick up extra
business at auction time.
Dave does all of the marketing for TWS and also handles the administrative and fi-
nancial side of the business. When necessary, he occasionally works in the field. Dave
started out with only one other full-time employee. Later, he was able to add an addi-
tional part-time employee to the payroll.
Dave has recently become sensitive to the old adage, “While the cat’s away, the mice
will play.” Sales seem to drop pretty dramatically when Dave is on vacation. Dave has
also recently enrolled in an MBA program, which reduces the amount of time that he
can commit to his business. During the semester, Dave suspects that business drops off.
The data file is named TWS. This file contains 46 monthly time-series observations
and eight variables. The time-series begins in May. The definitions of the variables on
that file are SALES, the total monthly sales of TWS in dollars; SALES2, the monthly
sales of type 2 in dollars; TEMP, the average monthly temperature (degrees Fahrenheit);

These data are real, but the name of the company has been changed. The data were made avail-
able to us by David S. Larson and Carmin Hunter-Siegert. Carmin is with the Colorado Student Oblig-
ation Bond Authority. We would also like to thank Dave and Carmin for coming to several of our
classes to discuss the data set and their study with our students.

i-1
 II-2 CASE26 DAVE’S TRUCK WASHING SERVICE

and PREC, the total monthly precipitation in the metro area in inches. CAT and SCHOOL
are binary variables indicating whether Dave was unavailable to work at TWS that
month. CAT takes on a value of 0 if Dave was not vacationing that month, and 1 if he
was; SCHOOL is 0 if Dave was not in school that month, and 1 if he was. EMPLOY rep-
resents the number of employees on the payroll that month. If there was an auction in
town, the variable AUCTION is 1 (0 for no auction).
Dave has hired you to look at the data for him. He thinks the data can shed some
light on two general areas:
1 Factors that influence fluctuations in revenues of TWS. Provide an overview of
the data. Does it appear that Dave is right about his absenteeism Hurting the business?
What other factors influence fluctuations in revenues of TWS?
2 Future values of TWS revenues. Dave would like to try his hand at forecasting fu-
ture values of TWS revenues. If he is successful at doing this, budgeting for employee
hours, equipment, and chemicals will be much easier.
For the second task, construct a forecasting model of total sales for TWS. Use this
model to predict total sales for the upcoming months of March and April.
Write a business memo to Dave that addresses the issues raised above. Without re-
lying on any jargon, explain to Dave how he might use your forecasting model to pre-
dict revenues later in the year.

DATA SET
SALES ~SALES2 TEMP PREC CAT SCHOOL EMPLOY AUCTION
4214 262 59°..0 4.26 0 0 Bw) 0
DoS 560 TI L250 0 0 2.0 0
3274 iS 74.2 aoe ik 0 Za) 0
6075 420 3 u0 estes 0 0 Boh *
4354 Ze C2 29 0 0 20) 0
3994 200 135 0.06 1 0 250 0
6256 Soin) 42.8 0.47 0 0 29 al
4150 540 Bf 1.04 0 0 2.0 0
4277 23 So nD 1.14 0 0 220 0
SD 812 22.4 0.66 0 0 AO 0
6986 1486 43.3 DO 0 0 2.60 0
5940 939 utd: Led 0 0 460 0
2973 261 Day soo 0 0 2.0 0
5687 332 65.4 2.04 0 0 ae 0
4779 95 ees 1.64 1 0 Leu 0
SS 2k 25), nea eo 0) 0 130 0
4849 1085 Oe ees 0 0 Lo 0
6484 525 Slee Said 0 0 i 0 0
Bog 89 42.8 Oe i 0 152.0 0
3834 1245 Bre 0.81 0 0 2.0 L
5469 188 36.4 0.74 il 0 2.0 0
5863 580 oo Ds Oo 0 0 2.0 0
6148 LID 397.5 Ja0 0 0 2,0 0
DoS 335 49.1 Oe 0 0 260 0

(concludes
on the next page)
 CASE 26 DAVE’S TRUCK WASHING SERVICE lI-3

SALES SALES2 TEMP PREC CAT SCHOOL EMPLOY AUCTION

8318 i337 56. OH 0 0
1142 663 Wa ee
6888 25 TOR i,
8301 200 ale =
6838 432 66. -46
6145 230 Sais 403
6470 225 44. 20
4580 0) 25% wad
6555 100 ils no
6540 205 40. .08
5874 230 42. 76
5692 0 47. 94
5922 475 60. ~43
5189 1100 69% oa)
5458 347 Sis od
10165 2657 PAs «69
5829 63 bar aS)
4210 460 50: 10
5066 784 34. 5
6960 885 335. 219
5484 378 sAy 249
5892 1020 40. FW
O}
OV
WO
WW
Oo
1
WO
O@OW ee)
OGCFONGCOWRNNPFPODCOORPKRRPEWOR
FPOrRSTWOHUrP
oom
tat
oto.
On
S88
eee
a or
Cor
Oe
CO
i
Eh
Oe
oe
Oe DONS
NO)
DS)
NS
DST
NTN
DEN
NT
DSU
NDS
DO
DNasCE
IG
OT
Cs
OO
Rt
a
Ce
eC CO
Oo
oS
OH
OO
Oe
ao
OG
Ce
oo
 \y

364 iti

r Ty
bbe hull
roe, ul om Thi wie hs . Pp : ‘ *,
tj [=the 5 eS : oa
: ‘ Livy » Sqrgn ny Tiere To

"be vom wren" a hatin eq?

nee ninegse® eayee

dion magPheu ppp td ee he
Sry Cee oop ‘ni gti otc ad re
Rae eh a sien e) a in a Tt & 2
‘ Cara | ak) 3
* bbtiet it OA Soue few PWS Cnet
ine we ij fait lcdAgee
a
2

Po
ws
*)
uwvow
rm
o>
™ he
nw
P
 udl
HOUSING PRICES—I

You currently own a house in Eastville, Oregon, and are planning on moving sometime
in the next year. A friend of yours has suggested that you try to sell the house on your
own. The decision is a difficult one. Real estate commissions can be quite hefty, but
brokers bring a great deal of expertise to the process of selling a house, particularly the
ability to set a reasonable asking price. Real estate professionals can quantify a variety
of unobservable or unmeasurable characteristics of a house, such as that shiny purple
wallpaper in the upstairs bathroom and the overall desirability of the neighborhood.
To get an initial feel for the types of houses in your area, you have collected the data
in the file named HOUSES. This file contains information such as the selling price of
the house, square feet, numbers of bedrooms and bathrooms, and the presence ofa fire-
place for 108 homes sold in your neighborhood last winter. In the data set:
SQ_FT is the variable measuring the total square feet in the house. BEDS and
BATHS are number of bedrooms and bathrooms, respectively. HEAT and STYLE are
categorical variables. HEAT takes on the value of 0 for gas forced air heating and 1 for
electric heat. STYLE is the architectural style of the home: 0 indicates a trilevel, 1 in-
dicates a two-story house, and 2 indicates that the house is a ranch-styled home.
GARAGE is the number of cars that can fit into the garage. We don’t know whether it
is an attached garage. AGE is the age of the home in years. FIRE and BASEMENT in-
dicate the presence (1) or absence (0) of at least one fireplace or basement. PRICE is
the selling price of the house in thousands of dollars and SCHOOL is the school dis-
trict (0 = Eastville school district; 1 = Apple Valley school district). All else the same,
Apple Valley is viewed as the preferred school district.

These data were used by Ellen L. Chilikas, David S. Abelson, and Brian R. Landry to price a house
owned by one of the group members as part of a student project. The name of the suburb from which
the data were taken has been changed.

lI-5
 II-6 CASE 27 HOUSING PRICES—I
‘

Your task is to use these data to answer a couple of questions.

e What is the nature of the selling prices of these houses?
¢ What is the breakout of selling prices in terms of architectural style?
¢ How does the existence of a basement contribute to selling price?
¢ How does the existence of a fireplace add to the selling price?
¢ What other interesting or useful information do you see in this data set?

DATA SET
BEDS HEAT GARAGE AGE . PRICE
SOhE BATHS STYLE BASEMENT FIRE SCHOOL
1238 2 0 ee 59 2200
LOY 64. 000 i
1296 66 .500 ey
120 66 500 dea
2A 66. 900 6
E250 68. 000 dey
L7OS 68. 500 20
eZ} 697 000 ne.
1794 70 950 18
1294 Ae 000 ae
T3872 Aes 692 9
I 62 hes 801 8
2996 U5 -207 13
1764 TCX 000 13
1416 LG 000 8
dWis0 elise 500 5
MSO 79 200 8
1664 a3 - 900 dau
IBZ SL. 000 14
SZ B2. 800 18
2167 84. 900 13
1664 85 -000 9
LOTS 86 000 153
1384 89 280 5
1431 89 -900 a
950 90 000 ‘LS
1452 a2 -000 4
1829 92 439 10
.
16 52 94. 646
1516 10 Sy ses
1998 LT 93. 100
1984 9 98% 149
1840 3 LOS 000
1823 3 Le Ole 000
2150 ne Dradton 900
2096 9 PPPLis 000
PPPPPPRPPORRPRRPOR
PRP
eee FwWOPWRER
PRWW
WwW
PB
BW
PP
WWW
KP
PBPWWWWwWwWwnPWdP
PW
WW GS
Go"
Gor
CO)
GO"
GI
GO)
IN
(GO)
(SS
DOG
PO)
(OO)
WD
I)
Go
NS
NON
Go)
NN)
NW)
DS
RS)
DS)
NM
SS)
Sw
NS
GS S411
70
O76
SS)
One
Ole
Or
Ore
SO)
Ono
OO
>
OOO
2
tS
OOS
SS OS
SoS
1S.
SR
Se)
C1)
Gyo
O'S)
1.
(OO
O11
SS
OO
oOo
Oe
@
2
SO
© We
NMI
MWh
NM
DNM
FP
NM
NNW
NM
HM
NF
wh
DMN
S
NN
NM
wD
DMM le 124. 000
PRPRPRPPRPPPPRPOFPOFPGAGCDOFROF
rF
CO
COO
oOoOrFOCO
FP
FFE
Co
co
lH
Oo
oOo
OF
or
oc
OOF
Pro be

(continues on the following page)

 CASE 27 HOUSING PRICES II-7

BEDS HEAT GARAGE AGE PRICE

SQ_FT BATHS STYLE FIRE BASEMENT SCHOOL
2375 3 0 am 1 ao 000
it i
2809 6 a) 195 .000
, oi2 ng a oe 000
816 19 0 61 AOU,
1008 ue 0 63 -500
SEZ 16 1 66. 950
1008 a7 1 68 .000
838 oe 0 68. 694
1008 ve Al 69° 000
£720 al il 70 aa
1242 2 0 WE .000
O12 18 il Dee 900
4316 8 i s)he 000
1490 aS it 88. 000
1118 16 a 88. BIS
1278 alle a 89 soa]
1490 10 it 89 .700
1418 14 il 90 . 000
1250 Ly) i 90 . 800
1881 o i: oF -500
1355 Y i 25 000
1518 Ze 0 O82 000
1606 ANS dl 99 500
2124 14 1 LOVE 000
2099 ils3 ol is nlo0
Aeae 9 a 125 000
1608 13 il SZ 000
1320 13 il 16. 000
1680 7 1 79 -900
1434 12 1 yack 900
1664 if 1 Save 000
13/9 8 al Silas 158
1634 10 il 87 - 200
1400 i il ous. 500
1664 16 a Oils 500
1548 8 il 92 5:00
2040 15 1 93, 5 OC
L792 3 1 94. 000
1865 A i Siok 700
1833 7 il 97, - 400
16297 3 il 98 »900
2198 eZ 1 SOR 000
17.07 We u 92 .000
2096 i A 102 -400
2004 13 ul 104 . 000
1818 KRW
BS
WKB
KUBO
WK
RP
FPWWBWWWPSS
FE
8S
2
"OPO
OO)
OS
ONO
10
Ore
OC
ee
BS
CS
OC
9S
TO
OO
Oe
a
OO
OS PEP
PPP
BPP
PER
PPEHPO
PEEP
EP ND
DUD
NI
NH
WOH
SOY
dO
|W
NIN
NO
NY
NNO
NO
NM
NM
DO
NH
WON
DH
NN
DN
ND
NNN
HN
G EPP
EE
EHP
PEPE Ang
RBWWWWWdwWwWwWunuUPPWdPPWWPPPWWP
W
DNDN 1 107 . 000
PRPPPPRPRPRPEPPRPPORPRPRPRPO
PPP
PEP
NNNNNNNNNNNNNNNNNNNPRPBRPRPRP
WWWWWWWWWWWPWWNYNNDWWNYNWWWNHNWNHNNW
CON
TO
OS
CO
FO
FS PORPOOFR
GGG
o

(concludes on the following page)

II-8 CASE 27 HOUSING PRICES—I

BEDS HEAT GARAGE AGE PRICE

SOaET BATHS STYLE BASEMENT FIRE SCHOOL
$935,055 5 0 2 2 i = Sain 7 900
107s a
23 Sipe 3 5 0 Z 2 0 Ly 000
1035
TOT hou 85 s 0 2 2 it 9 OR 000
BoA Ue 3 0 ? Z al 2 500
ci sie
2072. 2 e4 5 0 2 Y le 11 V2.000
ZAQS irs 3 0 2 2 i) a Pe 900
2060 4 4 0 2 Zz i g eo 600
2064 7 <3 3 0 iz 2 ab 5 500
114.
2S eed 4 ) 2 Z. x 8 414 .700
ZA OES 4 0 2 z Hl Bal iTS. 000
1984 4 3 io 2 2 il lt ae 000
2130 pea 3 0 2 a at 20 ad 000
2049 = 4 5 0 2 a a RS) Ze 000
ZOD FES 3 0 2 3 i 6 U2: 000
2264 BT Ser) 2 3 at v 130% 000
22604 73 3 0 Z 3 A 10 Lo 000
25:62 0 £4 3 0 2 2 i 6 TZ 000
22162 ee iS 0 Z Z i a 132 500
23 53°9 6A 5 0 2 2 al 5 a3 “SU
2293.03 3 0 2 S 7: 2 144 500
2282 4 5 0 2 2 al 9 150 ue
2542 4 3 0 2 2 it 6 153i 800
2380 4 3 0 2 2 il 6 Sie 000
2505 4 3 0 Z Z 1 7 156 20.0
2804 4 3 0 Z S 1 al EB OZ 000
PP
PRPPP
PRP
RP BE
PRPPPRPO
RPPPPPPRPPEPEPPRPRP
 28
HOUSING PRICES—II

You currently own a home in Eastville, Oregon, and want to put your house on the mar-
ket. You’re not in any particular hurry to get rid of the house, and would like to try sell-
ing it yourself. One way to determine a reasonable asking price of a house is to call one
or more real estate agents and seek their advice. Another is to hire an appraiser—this
approach would cost several hundred dollars. You’ ve been wondering if there might be
an easier and cheaper way to understand what determines selling prices in the area.
On your daily after-dinner stroll through the neighborhood, you’ve passed several
houses on the market. At first, you hoped that you could get a feel for the market by
simply studying these few houses. But the more you think about this problem, the more
confused you get. You’re pretty sure that bigger homes sell for more, but you don’t
know how much more. Also, it seems as if fireplaces would be a desirable amenity in
Oregon; your house doesn’t have a fireplace, and you’re not sure how much to lower
the asking price because of this. You’ ve also heard that location is the most important
thing in the real estate market, and you wonder if prices tend to vary by school district,
one indication of the quality of the neighborhood.
A friend has collected some information for you about selling prices and other char-
acteristics of houses sold within the last few months in your neighborhood. These in-
clude square feet, number of bedrooms and bathrooms, type of heating, and school dis-
trict. This data set is shown below. At first, you decided that your problem was
solved—you would only need to find a house just like yours in the data set, and use its
price as your guess.
Of course, it’s never that easy. Although you know the specific characteristics of
your house, such as square feet, number of bedrooms and bathrooms, and so forth, not

These data were used by Ellen L. Chilikas, David S. Abelson, and Brian R. Landry to price a house
owned by one of the group members as part of a student project. The name of the suburb from which
the data were taken has been changed.

lI-9
 IlI-10 CASE 28 HOUSING PRICES—II

one of the houses in the data set is a perfect match for your own home. Some of the
houses in the data set come close to yours, but every comparable house is different from
yours in at least three aspects. You’re wondering if there might not be another way to
use the data to come up with an estimate.
The data are contained in the file named HOUSES, which includes information on
108 houses. In this data set, SQ_FT is the variable measuring the total square feet in
the house. BEDS and BATHS are number of bedrooms and bathrooms, respectively.
HEAT and STYLE are categorical variables. HEAT takes on the value of 0 for gas
forced air heating and 1 for electric heat. STYLE is the architectural style of the home:
0 indicates a trilevel, 1 indicates a two-story house, and 2 indicates that the house isa
ranch-styled home. GARAGE is the number of cars that might fit in the garage, al-
though you don’t know whether the garage is attached. AGE is the age of the house in
years. FIRE indicates the presence (FIRE = 1) or absence (FIRE = 0) of a fireplace
and BASEMENT indicates the presence (BASEMENT = 1) or absence (BASEMENT
= (0) of a basement. PRICE is the selling price of the house in thousands of dollars and
SCHOOL is the school district (0 = Eastville school district; 1 = Apple Valley school
district). All else the same, Apple Valley is viewed as being the more desirable of the
two school districts.
Use these data to uncover the important determinants of selling prices of houses in
your neighborhood. Then prepare a description of how your findings might be used as
a general method for estimating the selling price of any house in your neighborhood,
such as yours.

DATA SET
BEDS HEAT GARAGE AGE PRICE
SORT BATHS STYLE BASEMENT Pi SCHOOL
2d:Cee me z 0 0 i 1 a af 59.900 1
NOLO haces Z it 0 z 0 3 i. 64.000 0
1296 44 2 0 0 2 i 1 0 66.500 0
132Oa 4 0 0 2 il ates: 1 66.500 0
14 Om s a 0 0 iL 0 6 a 66.900 0
te See 2 0 0 es al ia} 1 68.000 0
LEhOSue S Z 0) 0 DB, 1 20 0 68.500 0
Ld ge Site A 3 0 0 2 i Le 0 69.000 0
1794 4 Zz 0 0 2 il 18 0 2. 950 0
LAG4, 3 2 0 0 2 0 ples a FL Og 0
1725 <3 2 0 0 2 i 3 0 42.692 il
SG ig weed 2 0 0 1 0 8 ua 2a Ou 0
1996 4 Z 0 0) 2 0 3 1 Pom ee il
1764 4 2 i 0 eZ i aS a 76.000 0
LLG. 13 a 0 0 2 0 8 0 76.000 1
LO A “a 0 0 2 O: is i 77.500 0
1392" 3 a 0 0 2 il 8 ib 79..900 a
1664, 3 - 0 0 2 0 aL 1 79.900 0
Uae Ss 2 0 0 2 1 14 0 81.000 L
l/s ee) 3 0 0 a 0 18 i 82.800 0
GouOil a 3 1. 0 2 i aS i 84.900 0
1664 3 2 0 0 2 0 9 1 85.000 0

(continues on the next page)

 CASE 28 HOUSING PRICES—iI|_ II—11

BEDS HEAT GARAGE AGE PRICE

SQ_FT BATHS STYLE BASEMENT FIRE
SCHOOL
1973 3 0 13 0 86. 000 1
1384 5 89 2260
1431 i 89 900
1950 23 90% 000
1452 4 oa 000
1829 LO 92 439
1652 if 94. 646
1516 10 oT aoe
1998 ln 98. 100
1984 g 98 149
1840 9 105% 000
1823 3 ash oe 000
2150 12 ae 900
2096 9 3 000
2575 le 124 000
2375 il T3386 000
2809 6 UNS -000
912 9 Sie 000
816 a9 Gite 500
1008 al/ 63 500
912 16 66. 950
1008 a 63.5 000
838 Pal: 68. 694
1008 19 Oo 000
1120 ta LOG 452
1242 2Z 1S .000
912 18 Os 900
1316 8 83: 000
1490 13 88. 000
1118 16 88. 879
1278 ol 89 347
1490 10 89. 700
1418 14 90 .000
1250 a 20 800
1881 3 Sales 500
1355 7 95%. 000
1518 22 23% 000
1606 3 re500
2134 14 LOE. 000
2099 13 Lis 750
2232 2 Zoe 000
1608 13 132 000
1320 13 76. 000
1680 i V9 2200
1434 2 aoe 900
1664 7 Sor 000
1379 PR
PRWWP
WW
RPP
WWWWWHUPW S)}
(SS)
J.
SS)
ea
aye
Ke
ene)
orotate
Koh
(oe)
(=) EEE
WWWWWWWWWDNYWDYDY
(eeeaen(=)
(©)
(eye)
ere)
sen)
e=)
tela)
fey
(=)
(eee)
lowe
le
(ey
ers) NY HEOHOHO8 LE
NONNNNNWNHNNNNNNNNNNNNN
NNWWNHNWWWNHWNNWWNHYNNWNHYHNNWNHWDHNNNHN
DY
WD
WPWWOWw
WUOPRPWPB
UKPRUOWP
KR
WP 87% 158
PRPRPORPRPRPRPRPRPERPRPRPRPRPRPR
EPP
PEP
PEE PRPRPPRPRPRPRPRPRPORPR
HE
HBP
HBP
PEPE
PHP Oo
PRPPPPR
NNNNNPRPRPRPRPRPRPRPEPRPEPRPEPEPPRPERPERPE
DOODDDODODOOCDCOCOCOCGCOCO

(concludes
on the next page)
Il-12 CASE28 HOUSING PRICES—I

BEDS HEAT GARAGE AGE PRICE

SOSET BATHS STYLE BASEMENT FIRE SCHOOL
1634003 3 0 2 2 et 510 A 87.200 1
L400" 33 3 0 2 2 0 t i 917, 500 1
1664 4 4 0 2 Y: i LG i 91;. 500 i
1548 3 3 0 Z 2 i 8 fi 925500 1
2040 4 5 0 2 2 i 15 i 932500 0
1792 4 S 0 Zz 2 I ) ai 94.000 £
1865" 3 S 0 2 2 i 12 il 96.700 0
Le2sars S 0 2 2 1 a Linens 97.400 if
E20 TES 3 0 2 2 a 2 i. 92.500 1
Z2LIBe v4 3 0) 2 2 i 12 i 952-000 0
DOW eS 5 0 2 2 d lee 1 99.000 0
2090RSs 3 0 2 2 it 7 Le pb L02 ¢400 1
2004 4 3 0 2 2 Hl 13 1 «1042000 0
1818 4 5 0 2 2 al £9 L = LOvAOUo 1
HIBS tS Sees 10 2 2 2 ie. Ly 14072900 z
25817 5 0 Ze 2 0 17 7 “109.090 1
TOT ere 3 0 2 a a 9 i Uigd £62006 u
24h a) 5 0 2 2 i 9 Lotpbloed500 i
20/274 3 0 2 2 i dd Tp aki, 008 i
ZA0DS% 63 eS 0 2 2 ft ia i. LIZ SCG 0
2060 4 4 0 2 2 q 9 Letts 680 =
20647 53 5 0 2 2 a 5 tae i.
Zi4A8) $5 4 0 Z 2 t 8 Ls tle 700 -
2110 BES 4 0 2 a a ad 1 1152000 1
1984 4 es 1 2 2 1 14 tne 121-9080 1
ALOE £4 3 0 2 Z a 20 1 121.000 1
2049 4 8 0 2 2 L LS t= 125.000 A
2035 <3 5 0) 2 3 1 6 1 - 126008 =
Z2A64 £3 5 0 2 3 1 a i” 130,000 £
2264 3 3 0 Z 3 1 10 1 ~“t31.000 a
2362 4 3 0) 2 a a 6 E132 000 0
2262055 3 0 2 2 a i i 132.500 1
20 Smell 5 0 2 2 it 5 i) 13820508 0
22985 53 3 0 2 5 - 2 1 144.500 1
Za o2e 4 s 0 z 2 at 9 i 150.593 z
2542 4 3 0 a Z 1 6 1 152.800 1
2380 4 3 0 2 2 Z 6 195.808 1
2505 4 3 0 2 Z 1 iL 1. 156.2900 is
2804 4 3 0 2 3 Zt 7 1 192.000
 Jae
BOOKS

The retail book market has become very competitive. These days bookstores offer more
than just books: magazines, international and domestic newspapers, notecards, sta-
tionery, maps, books on tape, reading lights, and gifts are just some of the items found
on bookstore shelves. Large retailers such as Barnes & Noble are able to offer low
prices and large inventories to their customers, much to the chagrin of local bookstore
owners. The large chain bookstores have recently added coffee shops or delicatessens
to entice buyers to stay and browse.
A client of the consulting firm for which you work is considering investment into the
retail book sales industry, and has hired your firm to undertake a financial study. Your
job is to examine the data in the file named BOOKS. This data set contains general fi-
nancial characteristics of five retail bookstores (Standard Industrial Classification [SIC]
code 5942). The firms were chosen to include different-sized business entities: two large
corporations (Barnes & Noble and Borders), two medium size firms (Books-A-Million
and Crown) and, for comparison, one of the smaller stores—Village Green Bookstore.
Barnes & Noble, in addition to selling hardcover and paperback books, publishes
books for exclusive sale by Barnes & Noble outlets. The firm also operates a retail book
mail-order business. Borders Group, Inc. is a holding company which owns and oper-
ates book “superstores” and smaller stores such as those found in shopping malls. Bor-
ders bookstores offer prerecorded music and videotapes, and claim to specialize in
hard-to-find items. Borders is one of the largest firms in its SIC group, employing al-
most 23,000 people in 1997.
Books-A-Million is a relatively new corporation offering the full line of hardcover
and softcover books and magazines. Books-A-Million also distributes books to whole-
sale clubs, supermarkets, department stores, and mass merchandisers. Crown Books

These data and company descriptions come from various company reports.

ll-13
 Il-14. CASE29 BOOKS

offers the usual range of books, audiotapes, videotapes, and newspapers, and also some
computer software. Headquartered in Columbus, Ohio, Crown has recently experi-
enced financial difficulties: net income for Crown was — $860,000 in 1997.
Village Bookstore, out of Rochester, New York, sells much more than books. Ac-
cording to corporate reports, Village offers “books, newspapers, magazines, periodi-
cals, greeting cards, boxed stationery, writing papers, books-on-tape, selective video
tape rentals and multimedia products, humorous and traditional gift items and food
items such as fresh and packaged desserts, gourmet chocolates, coffees and ice creams
as well as other food items; and oo educational retail ee which sell chil-
dren’s educational toys and products.”
The data set contains the longest time-series available for each of these five firms.
Incomplete financial information is indicated with an asterisk. Use these data to un-
dertake a study of this industry. What do the historical trends have to say about these
firms? Use an appropriate statistical method to characterize net operating assets of
these firms. What factors influence net operating assets? How? Prepare a summary of
these matters for presentation at next week’s meeting with your client.
~

DEFINITIONS
Firm 1 = Barnes & Noble, Inc.
2 = Books-A-Million, Inc.
3 = Borders Group, Inc.
4 = Crown Books Corp.
5 = Village Green Bookstore, Inc.
Year Year, 19XX
PPE Property, plant, and equipment - Total (net), in millions of $
Assets Current assets - Total, in millions of $
Liab Current liabilities - Total, in millions of $
OpAssets Net operating assets - Total, in millions of $

DATA SET
Firm Year PPE Assets Liab OpAssets
‘ly 91 Sanaa SOiGresas 33:2 5350 57/933
Al oF, M2, O38 SSiley gashCMS} 3645739 64.747
al 93 PS Specs) ANNISY AMIS ENO GH BS) 2507 662
1 94 OOraow, 5a ons 4 363.2939 3722590
ih 95 PRES AS SYS) Gnmcwee 457.896 BOSE Cie:
at 96 CMS stoi) 848.188 621.688 546.380
il G7 AleyeAl I/ 7/24 866.955 654.263 647.466
2 oa. Se ooo SiO Rous cs *
2 92 56 ee Sea x oi
2 93 ORES) 57.239 24.455 Ad137
2 94 Deo og ooo 282 Ol BORO
y O15 SBoncio4 Zao 29 eee 1327595
2 96 AO 53 AZ OMONOE AQ rewle Sa LA SrelOS
zy 7 OSs ean is ANSE) 6 7-55) SOO 156.252
3 94 OTOL SISO 439.700 290.400
3 95 ZAAR ZO” 816.200 558 A010 502 22010
(concludes on the following page)
 CASE 29 BOOKS H-15

Assets Liab OpAssets

-500 740 .200 542. 900 440. 800
-200 846 - 400 634. 500 Ou 100
541 85 «032 31% ar ou 396
os 96. 243 3A, 588 68. 943
OeL 100; 562 21 631 Hake 022
oOo 2 O97 49 ~495 87 2239
a 149 432 2 622 o7), oa
018 147 ~ 809 BS3 492 106 2355
ee 180 oeL0 6c 981 110 ao,
004 182 MORL LOS. aS} Og, mA
hoe 150 9/4 - 606 O67) T30
ovo 141 -983 coXI OnE) 82 eA)
-462 NO 200 e651
PowWNA NO BEARS

eo ou eee: wii 23316

-496 060 tae naa
499 ooo ~061 ne
fo U5 394 002 .907
2615 004 lh) oe
~670 «815 304 melee:
Seon . 760 478 og)
-934 aa SREB FOSS
WwW
SPW
PPPS
PRR oS
Oo
oe Ook
AnnnnininninwiPP
Eee WAN
DN
WWW
OP oa OF EF .042
FNHNNN
WOOOwW PWRPENPRPRPED
23910
 , ; + ;
win bibaint n> ag pests,

calientemh traoO iam acs

roe, ea (Pea 1a! GOES ye ,wae

=
7 Lei) Talks BOO rt ihe 1 Ke
£68,845 - ot ero yt Ew ate, Vil
¥ te ae: asa gestaea
; 2 ETB: ie ape bagel nd af Ks eabaahic am
ieee "arive ee s iris han iteney Tak a
Seer = if allen eel Vac a uP eddie.rte #
2 en. aif HG) iE oO eoi PS -
' a See “pi? adi
— laxvadent iit}: pe nufe oe? ter
Na eta iae fin Re “ "aks ba a ai
epeese Ke oa a é wie wenaed.tevied m2 ihn “ac Te a Sig a
_ BRAS jan
ip. FAR ey be yc hin
kh wing <1 , dimers — ee,Fieger ssw
* ean! pihrvias {c me ged Wy sxe Mealy >
ez « 7 * fs
Y,
aa ae Wg

Yan
* '
‘ifh Ne

.ci,? _

oR er wi
i
7 e*b :

‘tere: 6 ae
Coe

pd

birt sty Mi £
‘~~ : -
 30
ELISA PLATE
UNIFORMITY

ELISAs, Inc., manufactures plates used by hospital laboratories to test for the presence
of antibodies to viruses such as lupus or HIV. The specific test process, called the
ELISA method, consists of coating the bottom of plastic wells (like little test tubes)
with a substance that contains molecules with which the specific antibodies will react.
Serums from the patient are placed in the well, and if they contain the specific anti-
bodies, those antibodies will bind to the bottom of the well, too. The test produces a
blue color if antibodies are present. The amount of blueness depends on the number of
antibodies present. The degree of blueness is then measured by a machine that produces
readings of “optical densities,” the final value of interest, on a scale of 0.000 to 3.000.
A low reading on this scale indicates low numbers of antibodies present.
The plastic wells are manufactured in lots of 500 plates. Each plate contains 96
wells, arranged in a pattern of 8 rows and 12 columns, as shown below:

Column number

Injector1 -> O
Injector 2 -> O
Injector 3 -> O
Injector 4 -> O
Injector 5 -> O
Injector 6 -> 0
Injector 7 -> O
Injector 8 -> X O1O O70)
O10}
O10,
O10 O10
©
O1O1O}O1O
1OlOlo
OlOlO CMOS(©)
@)(S)
@)(@)(@)
@)
(9)
G(@)
GOVE
OO 17010 Ore,
O10
O1OlO
O
QiOlO)O7O
OO000000
ClOlOsO1O.OCL OLOTOlO1O1O
OO ClOLOlO1OO
CO

This case describes a real situation, though the name of the company has been changed. (The
ELISA process is real.) Information and data about the situation are used here with permission. They
were provided by a team of students, including Susilo Hertanto, Michael Tigges, and several others.
ll-17
 I-18 CASE30 ELISA PLATE UNIFORMITY

The machine that fills the wells with the substance to which the antibodies bind has
a single bank of eight injectors, which correspond to the rows in the diagram above.
The machine first fills the eight wells in column 1 simultaneously: each cell is filled by
a squirt of solution from the injector directly over it. The bank of injectors then moves
to column 2 and fills those eight wells simultaneously, and so forth. After filling the
wells in column 12, the entire bank of injectors returns to the positions over column 1,
in preparation for the next plate. Differences.in the resulting wells, then, may be at-
tributed to variations between the columns and/or variations between the injectors, as
well as to variations in the other parts of the total measurement process.
In principle, if the wells are filled with identical serums (from the same patient) and
if the resulting optical densities are measured by using the approved method, the re-
sults should not vary. In practice, of course, they do.
ELISAs, Inc., has provided the data below, which consists of the readings for 95 ob-
servations on each of five plates taken sequentially from a single lot. The same patient’s
serums were used throughout. The specific plates tested were designed to detect anti-
bodies of the disease lupus. Note that for each plate, the well in row 8, column 1
(marked with an X in the diagram), is used as a control. Its value in the data below is
“missing.” ELISAs, Inc., wants to know how well (.e., how uniformly) its process
works and to understand any ways in which it might be improved.
Based on the data below (also available in file ELISA), how would you advise
ELISAs, Inc.?

DEFINITIONS
PLATE Plate number (1 through 5)
ROW Row number (injector number, 1 through 8)
COL Column number (1 through 12)
OPTDEN Optical density (0.000 through 3.000)

DATA SET
Oo ROW
PLATE COL OPTDEN OBS PLATE ROW COL OPTDEN
RR1 al 1.421 18 1 2 3 1.338
2 a 1.446 403) 1 3 3 1.208
3 a 1.614 20 al 4 a 1.084
4 1 Lod 8 a i 5 3 ioio3
5 a 1.452 22 ut 6 3 15 20
6 i 1. 190 23 i a 3 Looe
is oe eo od 24 i 8 3 1.389
oyrnuewne
w 8 2 x 25 a 1 4 1.466
il 2 1. 266 26 1 2 a 1.298
2 Z Lea 0 27 di 3 4 1.288
3 Z 1206 28 1 4 4 1.165
4 2 dco 29 I 2 4 15295
5 2 ds 4d 30 dL 6 4 RG
6 2 1.002 Sab a f 4 1.523
i? 2 17207 32 di 8 4 i Gad
8 Z 129 33 2 aL 5 1.454
PRPRPPP
PPP
OW
YRDUBRWNHHR
HEHE 1 3 1.804
PRPPPPPPPPPPPHPHEP 34 i. 2 > 15356
(continues on the following page)
 CASE 30 ELISA PLATE UNIFORMITY I-19

OBS PLATE ROW| CoOL OPTDEN OBS PLATE ROW COL OPTDEN
35 | £525 83 i alta a ae UD
36 .392 84 aA 586
Sp .483 85 tel) om
38 W
FP
~
Ou . 633 86 ale ia B
oo .744 87 av. i675
40 .550 88 ial . 889
41 .578 89 eZ 742
42 .370 90 ue 806
43 .788 od UZ. - 803
44 .442 Oe eZ S558
45 3388 93 2, W2B2
46 .786 94 eZ 568
47 .844 OD de. 505
48 APIS 96 eZ Be
49 .503 oi, fe 610
50 .604 98 «7 OL
5A San. 99 -946
SL .500 100 AOS
5S .586 RO w096
54 .803 102 aU6d
Do .679 dk08; 665
HEPOPPRP
PEE
P
PEPE
56 .013 104 *

57 .645 105 LA'S

5S .695 106 .340
59) £891 107 OM
60 .628 0S ae
61 .840 109 160
62 .611 TO 093
63 .847 Aa «aoe
64 .802 2 <a
65 Seay ils 243
66 .814 114 . LBs
67 .875 Lal's ~248
68 Te LLG ae
69 .424 lee and
70 1001 118 209
71 1)
SO:
IO
OO
(OO.
SS
XO
Oy
ON
ON
OF
OV
0)
~7
<=]
<2
ss)
SF
=)
OL
OT
4OO
W
CO .885 Tig Le
We, .652 120 ~245
TS .714 WA LO?
74 rea. M4 SILAS
75 .957 125 .246
76 .884 124 «2kO
Va. .762 eS) .074
78 .873 126 ~ 126
ag 758 27] 20M
80 .708 128 os
81 £569 dra) FPoOowHnuP
DHVP
OAH
PP
OWAHAAVU
WARD
EF
OAAU
WNP
WN
FP
UF
OH
&O
W iS
Gis
wmwowWOWw
eases
Sas
OM
Nww
Oo
SS
s 020
82 PPR
PRPPPPPPPP
~I
OO
FWNE
UU 6G6HS
OCOOoCG .799
ooo PEPER
PEPE
EPP
EP
EEE
PEPER
ERE
PRPPPPPPPEPEP SiO)
EPENEPERPEPRPEPRPEPEEPEPEPEP
EBPRPPRPRPPPRPEP BE
PPP
PEPEHRPHREPPRPRPRPRPRPRPRPERPRPERPRPRPRPERPEERPERPRPRPRPRPRPERERPHRPRPRERE
FE dO O1
NODNMNNDNMNNNNNNNNNNN
OAAOPWNHNPOAKHDUUBPWNHPWOANKDUPWNKFPWOATIAUYPW
NF 083
PRPPPPPP

(continues on the following page)

 II-20 CASE30 ELISA PLATE UNIFORMITY

OBS PLATE ROW COL OPTDEN OBS PLATE ROW COL OPTDEN
132 2 3 5 Af OG Ih, z 3 5, a 1.184
132 2 4 5 1.148 180 me 4 tel 1.400
133 2 5 5 LO cS as 2 5 at 1220
134 2 6 5 1.184 182 2 6 ai L276
135 yy 7 5 15298 183 Z 7 i 113350
136 2 8 5 1.487 184 2 nd oy 1.476
139 ez t 6 1.148 185 2 i. i? 1239)
130 2 2 6 Dl's 186 a 2 a2 152
So 2 3 6 Ass O69 187 2 3 12 12233
140 Z 4 6 Ieee 188 2 4° 12 1.284
141 Za 5 6 dG L39 2 5 12 LZ
142 2 6 6 it ALSO 19:0 z 6 12. 152310
143 2 X 6 ie3.66 Ove Z i 12 1.4363
144 Z 8 6 1.416 192 2 8 12 1.366
145 2 al Z ley is 3 1 1. 53k
146 2 2 v dy 228 194 3 2 J 1.946
147 2 3 7 1.226 195 3 3 a 1.662
148 2 4 i i262 L936 3 4 1 L198
149 2 5 a 1.040 Lou, 3 5 1 1 Dao
150 Z 6 7 1 3s 198 3 6 a 12504
52 2 a 7 A eo nk 1:9 3 a i 1.415
52 2 8 vi 1.625 200 3 8 i *
153 2 di 8 al ela 201 3 1 2 1525
154 Zz 2 8 1g, 202 3 2 2 1 oS
159 2 3 8 be 203 3 3 2 1.23e
156 2 4 8 1.260 204 3 4 2 Lota
157 2 5 8 Lee. 205 3 5 2 de eee
158 a 6 8 Pes 206 3 6 Z i228
159 2 7 8 1.419 207 3 7 2 1.019
160 2 8 8 1.547 208 3 8 2 0.943
161 2 cf 9 ine! 209 3 1 3 1eoks
162 2 2 2 TP Gi02 210 3 2 3 Lt.582
163 2 6 9 J slO9 2a S 3 3 L538
164 2 4 9 1.543 242 3 = 3 Sr
165 2 5 9 259 218 3 5 5 1.444
166 2 6 2 1.423 214 3 6 3 206
L167 2 i 9 1.448 2a5 3 7 3 1.467
168 a 8 Gg ESers 216 3 8 3 1.338
169 2 2 10 L320 ane 3 2 4 +99
10 Z 2 10 2 eo 218 2 2 4 1.407
ie 2 3 10 1.269 249 3 3 4 bi 267
V4 2 4 10 Eo SG 220) , 5 4 4 LL
173 2 5 10 1.95, ed 3 5 4 Loess
174 2 6 10 L305 wee 3 6 4 a ae
175 Z 7 10 1.434 4223 3 7 4 1.252
176 2 8 10 1.439 224 3 8 4 1.290
ye Z a Ly Zag aap 3 1 a 2s
178 2 Zz ri Le. ae2 226 3 2 2 £330
(continues on the following page)
 CASE 30 ELISAPLATE UNIFORMITY II-21

OBS PLATE ROW _ COL OPTDEN OBS PLATE ROW COL OPTDEN
227 3 3 5 1.425 275 3 3 ie 1.668
228 3 4 5 £23201 276 3 4 11 1.584
229 3 5 5 1.268 277 3 5 14 4 ge
230 3 6 5 1.279 278 3 6 11 1.446
231 3 7 5 1.344 279 3 7 Li 1.684
232 3 8 5 1.272 280 3 8 1 1.284
233 3 1 6 1.1052 281 3 4 12 2.057
234 3 2 6 147 282 3 2 12 1.801
235 3 3 6 1.465 283 3 3 1 1.562
236 3 4 6 1.509 284 3 4 12 2.054
239 3 5 6 1. 380 285 3 5 12 1.636
238 3 6 6 1.338 286 3 6 12 1.769
239 3 7 6 1.435 287 3 7 12 1.759
240 3 8 6 1.446 288 3 8 12 4.922
241 3 4 7 1.437 289 4 1 1 1.425
242 3 2 7 1.574 290 4 2 1 1.867
243 3 3 7 1.503 291 4 3 1 1.625
244 3 4 7 1.722 292 4 4 1 1.460
245 3 5 7 1.506 293 4 5 1 1.491
246 3 6 7 1.394 294 4 6 1 12222
247 3 7 7 1.446 295 4 7 i 1. Bea
248 3 8 7 1.695 296 4 8 1 *
249 3 i 8 1.493 297 4 1 2 1.416
250 3 2 8 1.535 298 4 a 2 1.192
251 3 3 8 4459, 299 4 3 2 1.246
252 3 4 8 1.592 300 4 4 2 1. 180
253 3 5 8 1.565 301 4 5 2 1.283
254 3 6 8 1.47% 302 4 6 2 1. ee
255 3 7 8 1.423 303 4 7 2 1.156
256 3 8 8 1.552 304 4 8 2 1.295
257 3 1 9 1.514 305 4 1 3 1.185
258 3 2 9 1.554 306 4 2 3 1.149
259 3 3 9 1.527 307 4 3 3 1, 17
260 3 4 9 1.761 308 4 4 3 1.198
261 3 5 9 1.553 309 4 5 3 1.163
262 3 6 9 1.618 310 4 6 3 1.096
263 3 a 9 1.623 3aa) 4 i 3 1.268
264 3 8 9 1.503 312 4 8 3 1.278
265 3 1 10 134 313 4 1 4 ior
266 3 2 10 1.591 314 4 2 4 0.968
267 3 3 10 1.650 315 4 3 4 1.306
268 3 4 10 1.517 316 4 4 4 1.364
269 3 5 10 1.435 317 4 5 4 1.200
270 3 6 10 1.607 318 4 6 4 1, 29S
291 3 7 10 1.338 319 4 7 4 1.304
272 3 8 10 1.636 320 4 8 4 1.286
273 3 i a 1.449 2a. 4 d 5 1.338
274 3 2 View a1 58e: 322 4 2 5a 1. 2e7,
(continues on the following page)
 I-22 CASE30 ELISA PLATE UNIFORMITY

OBS PLATE ROW CoOL OPTDEN OBS PLATE ROW COL OPTDEN
323 4 3 5 1.198 37a 4 3 Di 1.530
324 4 A 5 1.<379 372 4 4 a 1 $439
325 4 5 5 1-442 373 4 5 11 1.491
326 4 6 5 1.524 374 4 6 1 1.418
Bou 4 7 5 1.458 375 4 a 11 1.646
328 4 8 5 14375 376 4 8 11 1.536
329 4 1 6 1.480 377 4 a 12 12793
330 4 2 6 1.340 378 4 o 5s. 1 2742
331 4 3 6 1.280 379 4 oh 12 T2090
332 4 4 6 1.508 380 4 4 2 1.591
333 4 5 6 1.399 381 4 5 12 1.992
334 4 6 6 Pe5i 382 4 6 12 1.815
335 4 7 6 1.608 383 4 Fp 12 1.692
336 4 8 6 1.3001 384 4 8 12 1.380
337 4 Zl 7 1.296 385 5 1 1 1.794
338 4 2 7 1.279 386 5 2 1 22190
339 4 3 a 1.426 387 5 3 1 1.981
340 4 4 7 1.338 388 5 4 1 2.348
341 4 5 7 1.O32 389 5 5 iL 2 2014:
342 4 6 7 e200 390 5 6 1 2.024
343 4 7 7 Vea T 391 5 7 ie 2.000
344 4 8 Pi 17403 392 5 8 1 *
345 4 i 8 oer 393 5 1 2 1-943
346 4 2 8 1.693 394 5 2 2 15022
347 4 3 8 1.504 395 5 3 2 1.683
348 A A 8 1.414 396 5 4 2 1.908
349 A 5 8 1.528 397 5 5 2 1.690
350 4 6 8 1.378 398 5 6 2 1.610
351 4 q} 8 1.680 399 5 7 2 1.656
352 4 8 8 17293 400 5 8 2 1.684
353 4 1 9 1.596 401 5 Z 3 1.669
354 A 2 9 1.646 402 5 2 3 L502
355 4 3 9 13647 403 5 3 3 1752
356 4 4 9 17728 404 5 4 3 1.689
357 4 5 9 1.601 405 5 5 3 1.538
358 4 6 9 1.654 406 5 6 3 1.356
359 4 7 9 1.689 407 5 7 3 1.597
360 4 8 9 1.369 408 5 8 3 1.796
361 4 iE 10 1.384 409 5 1 4 1 £320
362 4 2 10 L3728 410 5 2 4 1278S
363 4 3 10 1.569 411 5 3 4 1.821
364 A 4 10 1.654 AID? 5 4 4 12728
365 4 5 10 1.599 413 5 5 4 1.615
366 4 6 10 1.607 414 5 6 4 1.426
367 4 y 10 1.597 415 5 7 4 1.610
368 4 8 10 1.611 416 5 8 4 12752
369 4 iN 11 1.592 417 5 i 5 1.709
370 4 oy et 1.408 418 5 2 5 17708
(concludes on the following page)
 CASE 30 ELISA PLATE UNIFORMITY I-23

OBS PLATE ROW eon OPTDEN OBS PLATE ROW COL OPTDEN
419 Oo -465 450 5 NO i POO2
420 =f Ou 451 2g
421 O24 452 992
422 Sw 453 sel
423 Taso 454 -936
424 eh? 455 ENO)
ORNS
NG)
ROP 21.09
425 - 366 456 mes
426 -654 457 wield
427 -402 458 10 073
428 -642 459 10 668
429 540 460 10 eae
430 - 680 461 10 . 803
431 904 462 10 070
432 ~ 980 463 10 - 190
433 vt 92 464 10 pesheill
434 oD 465 fal oo
435 2165 466 11 Ome,
436 So a2 467 alfa 51S,
437 - 696 468 alba TODS
438 .894 469 alta 824
439 093 470 An: AOS,
440 sa OF 471 alr) £022
441 - 148 472 del 5 BE
442 Ooi 473 dZ 21
443 tt 474 AZ 034
444 028 475 ee -246
445 aol Io 476 2. TO
446 «J 16 477 12 625
447 peo al 478 ily 253
448 041 479 2 one.
449 One
Oh
On
caisson
Oi
Os
On)
Om)
Oa
OT
er
on
Om
Ol
Oi
On
Or
sO W
FB
U1
Ys
WO
FE
PWN
HAY HEHE PE 480
FPoOIHAHOHPWNHPRPOTAUPWNPFPWOTN
00:
0G"
100500.
00)
00100
GOr
NO
oO
Gr
Ul
Oi
OT
CI
OF
CO)
Oy
01,
0107
6G),
0)
J]
SJ
=)
SJ
ST
SS Gal
Wale
Gn
al
Orson)
Oa
Gr
On
Oa)
sO)
1G)
OT
Ol
Oy
Om
yOT
Oi
Oren
One
Ole W
KS
MN de APasic
NNNPPNNPPE
NPP
PEN
HEB
BEPNNPPNFPRPENNPRPPPPPEPEPEPPEPEPEPEPERPEPRP
Oi) UPWNHPOATIAHAUBPWNP
OIA
 <

ron inc

“A
oe

=
@
ah a2 'Oars
a)
2
A
a
Yo
or
wv
yfi
1
a ;:
' 2, ~
aaa
¢
the
= i 4 20
=f
7
ee af

an 7 4 Ps
- 7 2

yeee.2>.4 fs. &

A
wv
et

bi, .
=
\1
mw
‘fr:
ne
i
wt
@ta
he
id
i

L.
¥
outs
Pe
Ay
hed,
+

oa
te
ay
_
~ e
:
RNa
-

«
™

ef L

he
ai

iv

Pte
Ping
5
? ®

ee
bubee
 OT
VIOLENCE ON TV—I

As the general manager of KTDS, the NBC affiliate in Tidusville, Oklahoma, Chris has
a range of responsibilities, including programming, personnel, advertising, and public
relations. His least favorite activity is responding to customer complaints. Unfortu-
nately, there’s been an unusually large number of complaints in the past few months
from viewers and advertisers alike.
Most of the recent comments are objections to the level of violence on KTDS pro-
grams. Chris is sensitive to this issue because he has observed a gradual increase in vi-
olence on TV over the past 20 years. Chris really prefers the old-time movies in which
dirty deeds were neatly sanitized and violent crimes occurred behind the scenes. He is
sympathetic to the recent callers.
Nonetheless, he’s in a tight spot. Chris knows that small, vocal groups do not nec-
essarily represent the population at large. People who feel strongly about an issue are
likely to speak out, while those who are content tend to remain silent. While the recent
callers have denounced the level of violence on KTDS shows, Chris knows he must un-
derstand and serve all of the KTDS viewers. ;
Chris has a suspicion about the source of the recent calls. Four months ago, a flam-
boyant politician announced his candidacy for mayor. This candidate has received a
great deal of air time on the local news due, in part, to his impassioned outbursts. Some
people love him, others despise him, but almost everyone tunes in to the evening news
in hopes of catching the latest controversy. One continuing theme of his platform is vi-
olence in America in general and violence on KTDS in particular. Over the past four
months, the candidate has suggested that those opposed to violence in the media “let

This case is based on an idea of Ann Lee Bailey, an MBA student at the University of Colorado at
Denver.

Il-25
I-26 CASE31 VIOLENCE ON TVv—I

their voices be heard.” Chris suspects that this fellow has inspired a large portion of the
recent complaints to KTDS.
The situation has become so volatile that Chris is convinced he needs to take some
kind of action. To sort out the hype from any real concerns of his constituents, Chris
has commissioned you to help him get some structured feedback. He would like to un-
dertake a survey (and has the staff to do the legwork), but he would like advice on the
specifics. In particular, he needs to know:
¢ What questions should be asked, and what is the best way to ask them?
¢ How should survey participants be chosen and how many should be chosen?
¢ What sort of confidentiality guarantees should be made to participants, if any?
¢ What is the best way to contact potential participants?
¢ Should the station offer some kind of free gift to induce higher participation rates?
If so, what? :
¢ Should one survey be used to address the concerns of the viewers and another sur-
vey to appease the station’s advertisers? If so, how should the surveys differ?
Prepare a report for Chris that addresses these and any other issues you think he
should be aware of. Include in your report a copy of the questionnaire.
 we
OVERDUE BILLS

Quick Stab Collection Agency (QSCA) collects bills in an eastern town. The company
specializes in small accounts and avoids risky collections, such as those in which the
debtor tends to be chronically late in payments or is known to be hostile.
The business can be very profitable. OSCA buys the rights to collect debts from their
original owners at a substantial discount. For example, QSCA might pay $10 for the
right to collect a $60 debt. QCSA takes the risk of not collecting the debt at all, of
course, but often a single official-looking letter yields full or nearly full payment, par-
ticularly for small debts.
Profitability at QSCA depends critically on the number of days to collect the pay-
ment and on the size of the bill, as well as on the discount rate offered.
Arandom sample of accounts closed out during the months of January through June
yielded the data set below (and in file OVERDUE). Write a brief memo to QSCA man-
agement advising them on the relationship, if any, between size of bill and number of
days to collect. In this data set, the variable DAYS is the number of days to collect the
payment, and BILL is the amount of the overdue bill in dollars, while TYPE =1 for
residential accounts and 0 for commercial accounts.

ll-27
 I-28 CASE32 OVERDUE BILLS

DATA SET
DAYS unebl Lbess TYEE DAYS, (BILL. . TYPE DAYS BILL, TYPE
41 pa gL i 302 aL 83 98 0
60 202 0 60 197 0 53 220 0
86 19 0 47 288 0 69 150 0
81 OF. 0 26 150 1 48 273 i.
Oe 201 iL Hed 158, 0 25 146 z
90 50 0 Bo Ze 0 60 210 0
94 46 0 ou 158 1 42 210 ‘ll
83 95 0 30 149 1 36 205 1
84 100 0 70 154 0 50 302 0
19 140 0 34 180 1 68 187 0
47 299 0 38 205 * Ze 95 ii
33 187 1 42 220 1 ae 46 is
47 264 ik 29 162 if 44 301 0
69 180 0 83 a7 0 47 289 0
9. oie 1 50 3174 1 19 98 i
36 WES) i 49 250 0 67 teks, 0
30 154 i o> LoS 1 3 149 0
39 310 0 16 80 - of 70 0
63 205 0 43 225 © 82 90 0
17 110 1 oh 310 al 63 Zu 0
85 nS 0 aa. Lng 0 74 153 0
oa 100 af 74 150 0 24 150 1
49 20% i 67 201 0 92 80 0
83 25 0 22 OF sh 65 146 0
13 v5 t 53 ano 0 99 60 0
16 79 a 5 90 iL 47 288 i
53 240 0 Dad. 220 0 51 264 0
40 9%) i 10 50 i 39 2i1 Zz
47 Si 0 80 110 0 2/ 140 dL
48 Zug i 47 289 a 44 250 1
70 162 0 1s) 70 1 55 199 1
43 240 x 11 60 i, 6 ao 1
 dd
THE BRAZING
OPERATION

In one of the phases in manufacturing a precision flow meter, a vacuum brazing process
is used to attach hardware to the flow meter tube. The process consists of placing the
“fixtured” assembly in a sealed vacuum furnace, pumping a hard vacuum on the fur-
nace, slowly heating the assemblies to approximately 2050 degrees Fahrenheit, and
then rapidly cooling them down to the ambient temperature.
The company wants the assembly to have a metallurgical structure that resists corro-
sion. One aspect of such astructure is a good grain size, a measure of the size of the ac-
tual crystals in the crystalline structure of the flow wall. The grain size is measured by a
standard specification called the grain count, a numerical reading between 1 and 10. The
larger the grain count, the finer the crystalline structure (and the more corrosion resistant).
The following data (and in file BRAZING) were gathered from process control
charts. The run numbers are not consecutive because a variety of operations, not just
the brazing operation, are performed in the furnace.
Based on the data beginning on the next page, what recommendations do you make
concerning the process?

DEFINITIONS
RUN Run number
TEMPAVE Average maximum temperature (degrees Fahrenheit)
TIMEAVE Average time at that temperature (minutes)
VACLEVEL Vacuum level (in torr units; higher numbers indicate a poorer
vacuum)
COOLRATE Cooling rate, in degrees Fahrenheit per minute
GRNCOUNT Grain count, in standard units, computed from a sample of the
resulting product

David R. Berge provided the data on which this case is based and described the manufacturing situation.

ll-29
 II-30 CASE 33 THE BRAZING OPERATION

DATA SET
RUN TEMPAVE TIMEAVE COOLRATE VACLEVEL GRNCOUNT
326 2045. ase 342
327 2046. 12s 352
232 2046. 2% 331
355 2047. Dae 304
336 2044. 12. Gt
Soi 2047. iZe 347
341 2046. ae 391
346 2047. 12 354
348 2046. i 310
349 2046. LZ 345
Bo 2046. are 305
353 2046. eis 349
399 2047. hee 314
358 2046. t2e 363
361 2045. ae 354
365 2042. 12% 346
367 2048. Thee 302
369 2044. tow 361
Sk 2046. iz. 326
373 2050. Le 389
B15 20502 2H 390
377 20502 te 336
801 2048. are 306
3:35 2049. 126 353
339 2048. a2. 357
391 2049. 12. 310
394 2049. LD 363
397 2050: dis 330
399 2050. Lt 351
403 2049. ars 392
407 2048. 2s 337
412 2:10:50, 2 326
415 2050. it
2m 363
417 2049. Las 391
419 2048. 12. 385
421 2048. ieee 347
425 2048. 12. 372
427 2050. ZK 398
429 2049. re au
430 20ers ve. 352.
433 2046. ar. 350 —
437 2046. i2e 342
440 2048. 2% 338
449 2002. U2n 354
453 Zo Ur tas 355
456 2048. UW DONATI
WMO
WO
OC
OEF tga KL
WPBDWOWHAOKRNAIOFPNDW 311
NHDUAWONONAAHAKRFPWUAHAKFE
RB
DS
HDAYWAAHAWAIAIKDATANPUAHAWAWAHADWKPWAIK
WIDP DN
UW
TV
SP
WUHDWAHAWAIWANT
Bh
or]
WT
DKO
SSN"
iO)
DO
BLO
iO
OS
WO
WOR
~)]
Al
CO
PB
Om)
Oo
oO)
te
lo
RP
0)
~1
=I
WN
A
ON
BOW
©
Of
We HDE
DNA
UAHPUAOA
KH
WO
OV
GY
UI
0)
~)1
O&O
OD
=!
AHD
OO)
WDM
HN
bWH
OD
VP
AWA
YOKF
DW
ON
OrRPWPR
WOHNNUWONRAON

(concludes
on the next page)
 CASE 33 THE BRAZING OPERATION ll-31

RUN TEMPAVE TIMEAVE COOLRATE VACLEVEL GRNCOUNT

458 2050. 2 4 315 5 of
463 2050" dees 348
471 20505 ae 352
» 473 2050. Ae 360
477 2049. IZ 353
481 2047. ges 374
486 2048. ONS)
ee Lei UI
NAN
OD 347 NAhPNAN
UA
PSODHD
&ODORU1
OF
oy
ov
UTOUW
OO
WNWON
a a ese
a
/ ar ;

iden Z ar endOCy ‘S a
sae
i a.ee
2v6US
salad os a tS 12Vei IGk ne

2 Lb. TeiPaviere GIMEAVER SE 20

as, wept 4 %. - rar.t
Te , O88
aid ocee - 2
eee, mae ::e an +1308. 0. at- ms
247. 00#€ 4 sf “THAT BOS 6.
pied Gate -82.4.8.47 meebo. 7
ravesee, 22.) Ber abate, Bees oe a a

/
aia
=
fae

mls
0
: *
ne
tos a)
as
oe
oak
‘2
-

)
La

=Ly
ik
7
»

7
= 7 7s
ow
ye
=
o
a
Pes.
a
lon ma

Cen
xe
7
y ae 7

i
1
7
A
‘7
JA-
 a o4
LEADING INDICATORS

Decision makers need reliable forecasts of key business activities to efficiently run an
organization. Failure to anticipate future demand for a product can lead to foregone
revenue, customer dissatisfaction, or lost market share. On the other hand, overesti-
mation of future demand may lead to surplus inventory, with associated costs. Accu-
rate projections of all costs of doing business, from personnel to capital, are needed to
prepare budgets for an upcoming fiscal year. Although predictions about future busi-
ness needs rarely turn out to be perfectly reliable, all firms do some kind of forecast-
ing to assist in planning.
Some businesses rely on leading indicators for preparation of forecasts. Leading in-
dicators are current levels of economic activity believed to foreshadow the direction
and magnitude of future activities. For instance, since durable goods such as refrigera-
tors and cars are time-consuming and costly to produce, manufacturers are more likely
to step up production only when future demand is expected to be strong. Demand for
large-ticket items seems to depend on consumers’ confidence in their economic future,
since consumers can postpone purchases of durable goods when times are hard and ex-
pected to remain so. Consumer confidence is one of the many leading indicators
watched closely by business executives.
Inventories, wholesale trade, shipments of durable goods, housing starts, and the
producer price index are other important leading indicators. Economic indicators are
collected and published by organizations such as the Federal Reserve Board, the Bu-
reau of the Census, and Dismal Sciences. See, for instance, www.dismal.com or
www.census.gov/briefrm/esbr/www/brief.html.
Leading indicators can help prepare company forecasts if the firm can isolate which
indicators apply to their industry. For instance, we might expect that mortgage appli-
cations and the interest rate on mortgages might foreshadow future activities in the
construction industry.

II-33
 II-34 CASE 34 LEADING INDICATORS

Can you prepare a model to forecast the need for new orders for materials and sup-
plies in the construction industry using the leading indicators available in the data in
the file named LEADING? The definitions below include citations from the World
Wide Web that will assist you in updating and revising the data set. Make use of ex-
planatory material about these data on the web if necessary.

DEFINITIONS
Yr.Month Year and month of the time series; 91.01 is January 1991, and so
forth .
Pop U.S. resident population in thousands (source: U.S. Census Bureau,
Current Population Reports
www.census.gov/main/www/subjects.html)
PermitSA Building permits (seasonally adjusted annual rate in thousands) for
new, privately owned housing units; the data are based on 17,000
permit-issuing places through 1993, and 19,000 places beginning in
..1994 (source: U.S. Census Bureau, Housing Units Authorized by
Building Permits, C 40
www.census.gov/const/www/index.html)
PermitNA Building permits (unadjusted monthly estimates in thousands) for
new, privately owned housing units; the data are based on 17,000
permit-issuing places through 1993, and 19,000 places beginning in
1994 (source: U.S. Census Bureau, Housing Units Authorized by
Building Permits, C 40
www.census.gov/const/www/index.html)
Const_SA New orders for construction materials and supplies in the U.S., in
millions of dollars, seasonally adjusted (source: U.S. Census Bu-
reau, Current Industrial Reports
www.census.gov/ftp/pub/indicator/www/m3/index.htm)
Const_NA New orders for construction materials and supplies in the U.S., in
millions of dollars, not seasonally adjusted (source: U.S. Census
Bureau, Current Industrial Reports
www.census.gov/ftp/pub/indicator/www/m3/index.htm)
MortRate Interest rate on 30-year, fixed rate, conventional mortgage loans
(source: Federal Home Loan Mortgage Corporation
www.bog.frb.fed.us)

DATA SET
Pop Const_SA PermitNA
Yr.Month PermitSA Const_NA MortRate
Vike Oik ~ABHOGO 786 14043 LAS A0 5vs5 9.64
OV .0e aoUo eS 853 1316, Sano Sale 54.1 Sia
FiO as LOwe one isos 13826 Soa, O50
91.04 251246 916 14199 14333 I059 9.49
OO Dara Do ae oa 14385 14735 23853 hee
OOO ee ADs ooue 964 14389 HGS 23.2 9.62
FIN eee 973 15293 14845 Ones 200
iT US | e20eors 944 14566 Loe Om 84.9 9.24

(continues on the following page)

 CASE 34 LEADING INDICATORS lI-35

Pop Const_SA PermitNA

Yr.Month PermitSA Const_NA MortRate
el 09 252661 974 14645 15637 80. oe
91 4 ALG) 252932 991 14761 15445 89% . 86
OAR 17 2o3 19% 984 14553 14173 67% au
oi bz 253385 1061 14457 13229 68. OU)
923 O01 253589 HOW 14790 12964 67. -43
92% 02 253780 1146 14653 14132 Was .76
927 03 22.3293 1082 15085 15384 SS, 94
225 04 254238 1054 15305 15462 103% 282
92% 05 254488 1056 15185 15582 oF, Or
928 06 254749 OSH 15860 17198 TOS! OL
92% 07 255011 1089 15102 14611 TOL ois
92. 08 255304 L075 15275 16052 93s Ie
92. 09 259 04. 1114 15864 16913 99% 4
92° 10 255843 dale2 15649 16351 OT: .09
92. TL 256082 plelalas 15523 15164 Ue aoe
92 12 256297 ALLS 15961 14572 Da mae
93° O01 296512 Lay 15892 13926 Ol. -02
93. 02 256700 1148 16766 16290 on - 68
93F 03 256889 1056 16307 16628 99°: 220
os} 04 257100 1104 16248 1GSW2 107. -47
23 205 251318 Tne? 16566 17005 LOT: -47
93 OG 251551 LAL 16284 17676 Ls -42
o3% 07 PS MEREJe ala 16079 15521 104. aoe
938 08 258056 1230 16305 17214 Ee wed
93% 09 258307 teal 16639 LTU22 eae ae
93% 10 258555 eu 16660 17428 107. -o3
oBF qe 258766 1357 7135 16858 937 = 16
930 ee 258965 1461 17426 oe. 1007 wad
94. 01 259159 1392 72351 15180 80. .06
94. 02 wha
yeieib 1279 17421 16994 Salm io
oAe 03 259494 Be 18476 18745 T26= - 68
942 04 209023 nowy 17985 18016 20 oe
94. 05 2599S 1383 No Wa 18444 oe - 60
94. 06 260140 Pee 18201 19589 33. -40
94. 07 260372 1347 18143 17424 iia, 61
94. 08 260610 1386 18342 Toa See oa
94. 09 260853 1426 18384 LOSES LAT -64
94. 10 261087 1401 18747 13596 Tie 225
94. TT 261297 1358 19049 18766 100. alee
94. 12 261496 1420 19350 tO. 94. -20
95% 01 261687 1293 OTD) 17560 aoe mee
oS 02 261838 i A82 19474 18948 Sale -o3
95 OS 262024 5 19030 ooo 3k -46
aa 04 262251 1243 18548 18016 LAO soe
gon 05 262451 1243 18807 18444 IZ aes 6
oo% 06 262661 1275 18529 19539 12o% Soy,
9s 07 262890 1935) 18902 17424 LG6.. WON
UN
AOD
BRP
HO
PWAINDP761
AWAITIWEHEHRPEOBRP
NWODAWDNWMOUWOUWO
AAI
AWAADAADAADAA
IANA
ANYAIAINA
WADA
AA

(concludes on the following page)

II-86 CASE 34 LEADING INDICATORS

Pop Const_SA PermitNA

Yr.Month PermitSA Const_NA MortRate
957088 1263123 1368 19382 19355 132.8 7 986
952098263366 1405 19105 19515 124 6 7.64
952210 4263609 1384 19169 19596 1214, 7.48
9O5el1FF 263804 1450 18926 18766 107.8 7.38
959226 6263988 1487 18730eed* 1768156 97.4 7320
96.201. . 264162 1357-3 18927 16807 S173 Daw
6 hO2e% 264295 L407 18639 18115 94.9 7.08
96.03 264485 1423 18676 18862 219.9 Tebe
96.04 264701 1459 19265 19209 139.4 F293
96.05 264878 1452 19652 20161 14065 8.07
96.06 265072 1415 19653 21093 130. 5 8.32
96.07 265284 a4 57) 20269 19538 135a7 8225
96.08 265498 1423 20053 22237 129.8 8.00
9620981 265723 1399 20195 21493 121.0 8.23
96 F10SF 265943 1349 20057 21159 L23 27 A=I2
O61 PEE 2661310 13,9; 19903 19603 100.7 Pew
96el27 £266373 1405 20264 18500 99.2 7.60
9723015 266490 1395 20471 181223 88.9 ae 2
9780281 266620 1438 20507 19749 92.8 T.65
9720381266789 1457 20421 20301) 120.0% 790
97.04 266993 1442 21342 21498 Sy se 8.14
97 a0ST T26717% AA B.D 20613 21094 T3128 1208
975062 267368 1402 20896 22442 131.6 7.69
O7 S07 et 267575 1414 21574 21058 13225 7.50
O7 2087) 267792 i3i9'7, 20873 21970 ea? 7.48
97.09 268014 1460 21203 22052 134.6 7243
S97 BrOe! 2682314 1476 * x 134.0 *
 dO
INTERRUPTING TINA

Tina is a graduate student adviser in the College of Business at the University of Michi-
gan at Trenton. Her job is sometimes hectic, but she likes its flexibility. She sets her
own schedule, which can vary from week to week if needed, as long as she gets the job
done and works an average of 20 hours per week.
Tina is one of five advisers in the College of Business. Advisers have a myriad of
job responsibilities that include counseling students, processing admissions, maintain-
ing and updating student records, responding to requests from professors, and provid-
ing information to prospective students. Each adviser is trained to respond to all of
these needs but also has a specific area of expertise. Tina is solely responsible for ver-
ifying student eligibility for graduation. She has to pull each student file to check
whether all required courses have been completed and the student has a minimum
grade point average of 3.0. This is a time-consuming process—about 200 students ap-
ply for graduation each semester.
Tina’s office is located in the middle of the advising center, which is in the middle
of the College of Business. She suspects that a quieter office in a corner somewhere
would eliminate an awful lot of the noise, confusion, and extraneous requests. On the
other hand, she needs to be available to answer quick questions for her colleagues and
students. Tina is the senior student adviser with a great deal of knowledge. She is also
an expert in the Student Information Computer System (SICS), which handles all of the
student-related administrative record keeping. Often she’s the only one on the floor
who knows the ins and outs of the SICS.
Student advising is a big part of her job. Tina can schedule appointments in advance,
or she might see students on a walk-in basis. It’s tough to know from day to day how

These data were collected and provided to us by T. J. Gaub. The name of the institution has been
changed.

I-37
 I-38 CASE35 INTERRUPTING TINA

many students will need to be served on a walk-in basis or how long their appointments
will last. Nonetheless, the college has recently instituted a total quality management
(TQM) program, which requires an eye toward customer (in this case, student) satis-
faction. This means that students must now be seen on demand. Tina also advises stu-
dents over the phone; advising sessions by phone can be either prescheduled or un-
scheduled.
In addition to these chores, Tina interacts with her supervisor, answers questions for
professors, trains newly hired advisers, processes administrative paperwork, and an-
swers phone calls.
Tina had more difficulty than usual last semester completing all of the graduation
checks. It wasn’t that there were more students graduating last semester. The problem
seemed to be an unusually large number of interruptions. Even though the job calls for
her do a variety of tasks beyond just graduation checks, things seem to be getting out
of hand! Tina’s supervisor is unable to pay any overtime, so the only solution for this
semester is to identify and eliminate interruptions where possible.
To study this problem in more detail, Tina decided to perform an experiment. She
created a worksheet on which she recorded every incident that took her away from
graduation checks over a six-week period. The worksheet includes the time she clocked
in and out, the type of incident, and the starting and ending time of the incident. The
worksheet is shown below. At the end of the day, Tina calculated the length of each in-
cident in minutes and categorized each incident by general type:
=
SICS problems
=
Work with coworkers
= Work with supervisor
= Answer phone (unscheduled)
= Clerical
= Student advising: walk-in
= Student advising: scheduled appointment
= Meetings
= Student advising by phone: scheduled
<7“
TaAtmmoanwy
= Other
As the weeks went on, Tina’s coworkers asked what she was doing with the clip-
board and worksheet. Tina was mysteriously unresponsive to their questions.
Tina has asked you to analyze and interpret the data for her. Use whatever statisti- _
cal technique you feel is appropriate to prepare a report.

WORKSHEET

ue aaa START END LENGTH

DATE INTERRUPTION TIME TIME (mins) TYPE

10/12 del 37450 162815\0 computer eee? diles53

problem

|counsel Rick
Rick | tee
clans isles
ee

peers
is pial 09 ae 10
Betty?
proof memo Ziel) 12:4
(continues on the A ieeepage)
 CASE 35 INTERRUPTING TINA 1|I—39

TIME “TIME START END LENGTH

DATE IN OUT INTERRUPTION TIME TIME (mins)

phone calls ABA) ere 3

finde tide A esers} £250 Oey

10/13 Sse 4:55 check address Si 2 i

message Sg
= Br29

>

4:36
| |
10/14 1:45 On :00 question about ISSO als

Se ee ee
answering
machine aoe

eae eae
file

Sa eee | |
Betey

[sites erotics[20s |ane] a |a

a seallaesibad vc aaa |PoSE |
a eS a Signature
elee
rojae | 4:5 ||5250
5750 asi6 é
=i walk-in a 27 ale

4
Bo
10/19 10740 fosaOM [en ae es
phone 10:47 py

appointment
|
fe
tok phone call Lis23

Fser | phone call 11:38 ALA tS, Face

On

(continues on the following page)

=a eee
II-40 CASE35 INTERRUPTING TINA

DATE IN ouT INTERRUPTION | TIME | TIME (mins) | TYPE

|walk-in |12:05 (par ae

phone call 1237 2

10/27 3.30. |) EE Oo 2:39

FORGO 2:43 Dee
password

computer BO Beak 16
problem

i172 HORSO RAS Snes Osa: 10238 a

problem

Mark steals HOS deOtrAal: Si

the computer ;

check on relics 5 L200 5 E

professor

2 :10
phone eee S40 7 ae
appointment

clea /g3 3:00 5 QQ

11/4 LOSS
2:00
oO)
oO

(concludes on the following page)

 CASE 35 INTERRUPTINGTINA Il—41

TIME
IN
TIME

eo
OUT

6730
INTERRUPTION

appointment
START
TIME

4:05
S(=3G
TIME

4:05
4:30
(mins)

25
ye
TYPE

F
G
eres appointment 5305

training B35
Shelly

11/9 9:45 eZ ee) appointment O50

ay 330

2:00 personal call Bd) ai 10

87515)

4:10 1 Ol
AS 5)

525 Nh
bo

a3 7/11 11:00 a1) 2230 a) oO

3:06 3:08
ery walkin 4:24 4:26 Nw]
OF
Oi
Ol
oO

aie [aera [os [wetiein as Pane[2

QQ]
a}
4

P| |_| nevetornteare[2230 |aaa [as

ty

oO

al Ol

BES Use[past Ie|evar

ewe Svein crow [as
omilee)

fp [vaticin [a | 4:45 |]
WwW
= Ons
P] Callies
 . - ‘ + -
oT i flees K _ . : 7
78
AE ite taeae . - a " —— a ;

co. bbe, Bal Lib 4 4 ie ES

‘ ee ’ oo

jommnigea
as ~ gi
p.298eee et4 P- .
.
- / ensign <n nls >
-4@
ahi
<o

i ment ne ee
ei,
an
OteeA
if =
tt. urs.
al Bed:
4
«2
- poh m6 ‘ }

Nd nail
£ URS Saf
eg epee oe 1
a nn eee apy Fim Sel ae
mn roeds n= /
wead a

Se

ne—_.
:
eS
ee q
iar allel Me” praia
ae pe — apee. aie
2 tal 8 le sh mere all
ft wan te v4
ss 4 >
ie z i ', i‘ 2 € _*
Tt. lapel ereeompren ns 4 2 ym A

jo

rite sent = 3) : Rl iam

piece 5 ; Ta ts
H 1. es ;
As aw =< 3 ao
tl—
we
ce li 8a ot re >

me lh me a
 36
VIOLENCE ON TV—II

As the general manager of KTDS, the NBC affiliate in Tidusville, Oklahoma, Chris has
a range of responsibilities that include programming, personnel, advertising, and pub-
lic relations. His least favorite activity is responding to customer complaints. Unfortu-
nately, there’s been an unusually large number of complaints in the past few months
from viewers and advertisers alike.
Most of the recent comments are objections to the level of violence on KTDS pro-
grams. Chris is sensitive to this issue because he has observed a gradual increase in vi-
olence on TV over the past 20 years. Chris really prefers the old-time movies in which
dirty deeds were neatly sanitized and violent crimes occurred behind the scenes. He is
sympathetic to the recent callers.
Nonetheless, he’s in a tight spot. Chris knows that small, vocal groups do not nec-
essarily represent the population at large. People who feel strongly about an issue are
likely to speak out, while those who are content tend to remain silent. While the recent
callers have denounced the level of violence on KTDS shows, Chris knows he must un-
derstand and serve all of the KTDS viewers.
Chris has a suspicion about the source of the recent calls. Four months ago, a flamboy-
ant politician announced his candidacy for mayor. This candidate has received a great deal
of air time on the local news due, in part, to his impassioned outbursts. Some people love
him, others despise him, but almost everyone tunes in to the evening news in hopes of catch-
ing the latest controversy. One continuing theme of his platform is violence in America in
general and violence on KTDS in particular. Over the past four months, the candidate has
suggested that those opposed to violence in the media “let their voices be heard.” Chris sus-
pects that this fellow has inspired a large portion of the recent complaints to KTDS.

The survey design and data set compilation were undertaken by Ann Lee Bailey, an MBA student
at the University of Colorado at Denver. The scenario has been altered to preserve the anonymity of
the survey respondents.
Il-43
 \I-44 CASE36 VIOLENCE ON TV—t

Chris needs to sort all of this out. To understand the views of all KTDS patrons, he
has commissioned an independent consulting firm to devise and administer a survey.
The survey results are in, and he wants you to analyze the data (which reside in the file
named VIOLENCE). Below is a copy of the questionnaire.
The survey was done over a three-week period in October. Two hundred phone calls
were made and 106 people declined the invitation to participate in the survey. Of the 94
participating respondents, two people were offended by the question of income and re-
fused to answer that particular question. Nonresponses are indicated in the data set by
an asterisk. A random selection of phone numbers from the Tidusville phone book was
used to select the sample.
What is Tidusville’s perception of the level of violence on KTDS? Does one’s per-
ception of violence on TV vary with gender, age, marital status, income, or education?
Do parents with children at home have a different tolerance for violerice than those
without children at home? Do viewers who spendalot of time watching TV become
desensitized to violence? Chris needs a report as soon as possible.

SURVEY

weGender (0) (1)

male female

. Age (1) (2) (4) (5)

under 20 20-30 31-40 41-50 over 50
. Marital status (0) (1)
married Single or
divorced

Do you have children at home? (0) (1)

yes no

. Household (1) (2 (3) (4)

Income under $20,000 $40,000 over
$20,000 -40,000 -60,000 $60,000

. Education (1) (2) (3) (4)

high some college graduate
school college graduate school

. How many hours per week do you watch TV?

(1) (2) (3) (4) (5) (6) (7)

Oil, 8-14 aS el 2228 ADs3e 36-42 43 or
: more
In your opinion, how violent are most TV programs?
ri) (2) (3) (4) (5)
much too somewhat violent a little not very
violent too violent violent violent
 CASE 36 VIOLENCE ON TV—II I-45

DATA SET
AGE CHILDREN SCHOOL VIOLENT
GENDER MARRIED INCOME HOURS

(concludes on the next page)

II-46 CASE36 VIOLENCE ON TV—I

AGE CHILDREN SCHOOL VIOLENT

GENDER MARRIED INCOME HOURS
 aol
COMMERCIAL PROPERTY

In early 1992, the Arapahoe Assessor’s office made local headlines. Arapahoe County, ad-
jacent to Denver County in Colorado, had just undergone its annual property assessments.
Many people thought the assessed values were much too low. If this were true, it would
have implications for funding of schools, the perceived value of the properties and prop-
erty tax revenues for the county. The Arapahoe County Assessor was in the center of all of
this—he was responsible for the low assessed values and he owned one of the properties.
Is there any way to determine whether the assessments were fair? To understand
how buildings might be assessed, you have collected information on recent selling
prices of commercial property in Arapahoe County. Of course, the selling price need
not equal the assessed value of the property, but in most jurisdictions the assessed value
is supposed to be proportional to the fair market value of the property. Selling price is
an obvious measure of that value.
Understanding the determinants of selling prices of commercial property is difficult.
Private residences typically range in size from several hundred to several thousand
square feet, but commercial property can be the corner drug store or the high-rise
downtown. The expected use of the commercial property may influence its selling
price; buildings designed to be restaurants, warehouses, or barbershops probably sell
for different amounts. Like residential property, though, it might be expected that older
commercial property sells for less, all else the same.
The data file, named COMMPROP, contains data on selling prices and physical
characteristics of 75 commercial properties sold in Arapahoe County from July 1988
to June 1990. Do these data suggest which factors determine selling prices of

These data were compiled and analyzed by George Carr, a tax appraiser and consultant for the Col-
orado Division of Property taxation; Rochelle Hrigora Hass; Lisa E. Pieper, Research Associate; and
Mark Truitt. The data come from “Arapahoe County 1992 Property Assessment Study Final Report,”
by Thimgan & Associates, Inc., conducted for the Colorado Legislative Council, Denver, Colorado,
September 15, 1992, p. 110.
I|-47
 I-48 CASE37 COMMERCIAL PROPERTY

commercial property and, hence, which factors should be included in a fair assessment
process? Is there seasonality to selling prices of commercial property? If so, how might
this seasonality influence the ability to use selling price as a proxy for assessed value?
Prepare a report of your findings.

DEFINITIONS
CLASS Primary use of the property:
0 = Merchandising
1 = Lodging ‘
2 = Offices
3 = Special purpose
4 = Warehousing
5 = Multi-use
6 = Manufacturing
LANDSIZE_ Size of the land parcel in square feet
TOWNSHIP Location by township within Arapahoe County:
0 = Greenwood Village
1 = Aurora
2 = Englewood
3 = Littleton
4 = Unincorporated
SQFEET Number of square feet in the building
FLOORS Number of floors in the building
DESIGN Specific property characteristics impacting its use:
0 = Store/Retail
1 = Hotel
2 = General office
3 = Medical office
4 = Restaurant
53 = Garage/Auto
6 = Warehouse
7 = Industrial
8 = Other
9 = Commercial condominiums
QUALCON Quality of construction as assessed by an independent
audit firm:
0 = Fair
1 = Average
2 = Good
BUILT Year the building was built (19XX)
SYEAR Year the building was sold (19XX)
SMONTH Month the building was sold (1 = January, etc.)
PRICE Selling price of the property in dollars
 CASE 37 COMMERCIAL PROPERTY ll-49

DATA SET
LANDSIZE SQFEET DESIGN BUILT SMONTH
CLASS TOWNSHIP FLOORS QUALCON SYEAR PRICE
6 24393 2 LIOOUPe Lt 7 0 72 Se wc 295000
coe S235 2 2400 1 7 0 Lo roe 2 120000
6 35912 2 2s6oR #1 d 0 8 So 3 395000
6 28000 2 11So x1 “ 0 TA 9 6 280000
4 18229 2 6885 1 6 0 ve]. 90 2 195000
6 14840 2 LOOCOre 1 gi 0 ime aa) 2 205000
6 13160 2 S550" *1 a 0 1a, 99 4 201000
6 43065 2 10240 1 7 0 13 Bee i2 350000
4 21587 2 LTZcOr 1 6 0 78. 89 <10 300000
4 45000 2 25224) 1 6 0 J 8. 39 2 425000
4 17480 2 G00) 1. 6 0 £5 69 8 207500
2 6250 2 so00" 1 ie 0 80. 89> 10 230000
1 48250 Z 43228 3 dt . 72 89 “14 600000
0 15000 i TESOr 1 0 0 42 90 250000
3 25000 ah 3864 1 4 i. HO) BS Se Ls: 200000
3 41136 a 3478 1 5 au oes ad 350000
3 37635 i 3544 1 4 1 60 90 Z 217000
5 20000 1 5640 1 0 e 78 89 ci 262500
0 13583 al 2400 1 0 iv thie he) 1 241250
2 79933 a 38924 3 Z af 82° 69 e127 1300000
LeA267767 LMS T5265 t7 1 2 82, 8es9 3) 237700000
0 17550 i 5040 1 0 0 70 atS9 Get? 182250
0 9828 4 2490 2 2 0 65589 2 80000
2 49648 t SH Grae 2 0 PWS SS 3 400000
S 60364 a 8641 2 4 1 iD S89 6 618000
2 1024 - oS DUS 9 1 80790 4 92500
a 1024 i 1408 2 9 i 80. 90 i 85000
2 52 af 1440 2 9 1 Sai s88. "a2 99500
2 LiUS2 i. 1440 2 9 iI 84°89 “ail 87000
2 £152 1 1408 2 9 - 83. 90 4 83000
2 io 2 cl 1408 2 9 it 83 90 5 84000
3 233438 1 44562 1 0 0 86° 89 8 2100000
Z 11644 Z S81if 1 g e 84 90 6 102500
2 2055 a. 9352, 2 2 L 80 89 3 759000
2 54320 a 52 Lore 2 1 S35 eoa 4 1450000
4 30381 i 12980) 2 6 0 81 88 7 300000
2 5502 : 1625.31 9 ae Bee 5 58100
4 28070 1 8918, 1 6 1 85-569 saeZ 175000
4 102148 1 116037 6 i 86 90 6 285000
5) 107157 if 42a 0 0 a. = 850000
0’ 106252 i S5S365 1 0) 1 18") 89> *i2 OT S008
2 4800 i, ALTO. 1 2 1 81 89 741 165000
2 6300 fl Oven 2 A Sl .89 g 350000
4 44012 4 12843 2 6 1 S289 8 400000

(concludes on the next page)

II-50 CASE37 COMMERCIAL PROPERTY

LANDSIZE SQFEET DESIGN BUILT SMONTH

CLASS TOWNSHIP FLOORS QUALCON SYEAR PRICE
4 41528 4 8400 1 Siiue HO BAee oo i. 232000
5 23460 0 5480 1 0 di (ae 59 8 310000
2 104548 0 59854 4 2 ab 74 88 8 3350000
2 S7991 0 14764 1 2 1 65 90 4 810000
2 85410 0 10492 1 a i 68 88 10 600000
Ze 193321 0 7 TS 2ar8 2 ue 80 89 7. -3325906
3 174240 6 48887 3 3 2 85 89 1 1440000
2 191403 AW tlie tes 2 1 82 .89 12 4683300
2 87120 a 1917 492 2 Al 85. 89 9 652350
4 266108 4 811 OOmet 6 1 8290 5 2100000
4 6400 4 6292e4 2 1 80 88 41. 227000
4 8000 4 7880) :1 8 cL 12 8S 8 300000
4 4800 4 4704 1 2 iz 12790 2 189900
4 8000 4 B92 41 2 0 12) 39 ULES Z0CUD
5 20034~— 2 8352° 2 3 a bp oe 9 475000
3 81295 2 392A 66 3 1 Bie 1 700000
5 15625 zz 6084 1 0 ul 85 89 9 =6«4513000
0 4625 2 192 Tr2 0 0 36°53 89 7 85000
0 25000 2 6080 1 0 0 60 89 2 260000
4 24000 2 10260 1 6 0 83 88 7 550000
5 21978 2 6080 1 = al 88 89 8 600000
Oe 28828 3 OO 0 0 Comoe 3 90000
3 24700 3 S3.608 1 0 0 ta) Oo 2 300000
0 5625 2 4500 1 0 0 bay BY 1 105000
3 84040 3 31461 3 3 iL 66°89 it = 650000
0 4000 3 4489 1 0 0 60 89 4 105000
8 16800 3 S4/8- 1 3 1 Ji 88 8 215000
5 267300 3 89820 1 0 1 26 90 6 1900000
4 217800 3 74550 1 6 i 35 _ 90 5 1900000
0 24137 5 Spe 0 i 81 88 9 440000
by 91500 5 188508 1 0 i 78 $89 11 4640000
 38
AIDS RATES

Our intuitive impressions of rates are often erratic or wrong, particularly about issues
that are sensitive, emotional, or scary. For example, ask 10 or 20 people to estimate
the rate of acquired immunodeficiency syndrome (AIDS) cases in the United States,
first for males, and then for females, and record their answers in units of, say, cases
per 1000 people. Prepare an appropriate summary of their responses. Record your
own estimates, too.
Now use the World Wide Web to find the latest data from, say, the Centers for Dis-
ease Control and Prevention, about AIDS rates in the United States, and prepare an ap-
propriate summary of that data. How accurate were the opinions of the people in your
sample? How consistent?

lI-51
 ak
io
.|

é / oAe 3 : Hy GE OeG
Pe: she ae ee ae a. | 17 an “Se
oo aS Dae ee : 4: o ; es

Ie
¢ ee

ee
om >

Be 5 ces ah BOOK 42743

“2 Pe, ee eae ; ; M
4 ie oe = au id eile ‘| i 4 > i 2 oS 7c ’ eum

7 . et rf ai 7 *, , q BY 3 : » ; an : ee

ote | es ea B25 9 OVER eS UTR Gi aS

b he 4 e233. 3 1 os . : Fi Pa ea 2

ean ey ; 2oR i: eal het gh.*> 42° Fie

chy . 5 * ah hi. a te ie ; r = Jd Le “ 1h i Ee
- . a - oe * by td * a . 7 a , a ’x 3 . at 3

ya hinavalencnie§ erst bites reaftssie,‘heh enpleningmi-avbiaad 8

piies.0} sigqosy 0-7. ¥ ine Paabars 3VE sre He dahon’, nvaiheesasa Set © or ae
a.
Afysijnl adi ni usenit (2tA's bingy? ganckl Aeanni iri poe To apivgds 85
ass
Ang Ao zian ae ett Srey twser bru ablehes. wh woe bar gal 20b-dert #S. 3
etre eeenogest tah area faith: eter af ouagert aa dy OOOT og z
es. 00 vesaadtetey Givers . «
nehiar A Mae petit Vea] sti beasy ew.obi. SQA onebeagO Te
<sie
a
3 c a*
. BSiageng ink teva b ent tS
SA) 96 asvat Cod,
wae Rae natioro®t bas lA Irie = ee
a attdomeee st pase pleithce 36 mh (abih belt Yo! ae
eadoH an)ne ste Te + Qpaeleaos es
07 eta
es 3 7

. fg ee ov : ah OPar -
fi,
o,oa
 dQ
SOFTWARE
SERVICE CALLS

“What can you do?” asked Sam. “The number of service calls we get in any week or
month jumps around so much, it’s no wonder that sometimes we seem to be over-
staffed, while at other times we are completely swamped and can’t get to all the re-
ported problems anywhere near as quickly as we ought to.” Sam knew his supervisor
wouldn’t like his explanation much, but he still felt it was an accurate description of
the problem he faced.
Sam managed a department in the national service center for a software develop-
ment company. As part of the customer service and warranty operation, the national
service center staffed a hotline, to which customers could report discrepancies or prob-
lems. Customers could file formal, written trouble reports (STR), but Sam was con-
cerned about telephoned trouble reports, not STRs.
His department received direct service requests by phone (SCALLs) for one specific
product in the firm’s catalog, a fairly large package for computer-aided design and
drafting. The package had been sold for 18 months. His problem was to have the ap-
propriate number of service representatives on hand to provide decent response time to
customers. “Decent” was not defined in any specific terms (at least not yet), but Sam
agreed with his supervisor that performance lately had certainly been unsatisfactory, at
least in spurts. Sam knew that the allocation of service representatives to different prod-
ucts could in principle be changed from time to time. The difficulty was how to know
in advance how many would be needed for which product.
Sam stared at the numbers on his desk and remembered the tone in his supervisor’s
voice. He had to figure out something. The numbers on his desk are also available in

This case describes a real business situation. The data are used here with the permission of the students
who originally provided them to us.

lI-53
 I-54 +=CASE39 SOFTWARE SERVICE CALLS

file SERVCALL. Q is the quantity of the product sold in each month, and MONTH is
a count of calendar months elapsed since the product’s introduction. The other vari-
ables are defined above.

DATA SET
MONTH Q STR SCALL
il 20 28 76
2 rg 18 63
S 28 10 19 :
4 oe) D3 73
) 2 47 89
6 Pai 41 oF
Wl 4 BS 68
8 ial 34 18
9 45 66 85
10 ae 28 88
alee o 63 aS
ole 14 58 68
Le 10 40 70
14 8 43 84
eS Zt 41 85
16 Q 30 Wiz
iy, ce 32 74
©O CO BS 72
 40
TITANIC SHIPPING
COMPANY

Titanic Shipping Company depends on its nationwide system of computer terminals for
entering orders, tracking shipments, and other routine functions. Some 1500 terminals
and personal computers in more than 70 locations are connected throughastate-of-the-
art telecommunications network to central computing facilities in Newark, New Jersey.
Thanks to good backup procedures and steady improvements in communications tech-
nology, the system has become reliable, and failures (in the strict sense of the term)
have been rare.
The performance of the system has troubled the users, though. At any point during
a normal day, several hundred users may be entering or retrieving information about
the company’s shipments or updating customer records. When response time for these
transactions deteriorates, users can’t work effectively. Recently, they’ve been letting
the computer center manager know their feelings—forcefully.
The computer center manager understands their concerns, but isn’t sure how to re-
spond. To some extent, of course, improving response time requires things like new fa-
cilities or better performance tuning of their systems, both of which take money. That
can be negotiated in the normal company budget process. Right now, he is more con-
cerned about losing his credibility with the users he is supposed to serve. Several years
ago, the users’ representatives agreed that response times of two seconds would be ad-
equate. Since then, the computer center has tracked average response time, as measured
by an internal tool in the computer’s operating system. According to this measure, per-
formance has been generally satisfactory, at least for large portions of the time. Still,
the users complain that the system is not meeting their needs. The manager realizes that
part of the problem is one of “translation”: the users don’t know how to specify exactly

The company name and the specific data for this case are simulated, but the situation corresponds
closely to a number of real situations with which we are familiar.

I-55
 I-56 CASE 40 TITANIC SHIPPING COMPANY

how the computer center ought to measure response time (at least not in any precise,
quantifiable terms), but when performance is bad, they know it, and they complain. As
long as the computer center continues to track performance by using a measure that
(apparently) is unrelated to the users’ perceptions of their needs, relations between the
users and computer center management will probably continue to deteriorate.
During the last quarter the computer center has developed a system that asks each
user to rate the system performance as “OK” or “Bad” before signing off. Thirteen
weeks’ worth of data from the 10:00-11:00 hour for each workday are given below and
in file TITANIC. The 10:00 hour is usually a busy one, and generates many of the user
complaints. Data from two holidays have been omitted. ‘
The manager wonders if these data can provide some insight that will allow him to
develop a better measure of response time. He really wants to know two things:
¢ How should he measure response time?
¢ What target or standard performance should he use?
The measurements in the data set come from the software in the operating system
and could be obtained routinely if it were helpful to do so. In setting the standard, the
manager wants to use a value that corresponds as closely as possible to the users’ per-
ceptions, so that when the computer center’s own measurements indicate acceptable
performance, the users will agree. Can you advise him?

DEFINITIONS
AVERESP Average response time during that hour (seconds)
RESPS50 50th percentile of the response times during that hour (seconds)
RESP75 75th percentile of the response times during that hour (seconds)
RESP90 90th percentile of the response times during that hour (seconds)
RESP95 95th percentile of the response times during that hour (seconds)
NBAD Number of users who rated performance “bad” when signing off
NOK Number of users who rated performance “OK” when signing off
 CASE 40 TITANIC SHIPPING COMPANY I-57

DATA SET
DAY RESP50 RESP90 NBAD
AVERESP RESP75 RESP95 NOK
2 eS FORE? ioe 944 De 426 Bee
aes wo? ped .008 oe ak. 2, 276
ENDS ree | sao? -734 882 144 153
2 O38 eo) ose 230 2003 ul)
a oD al al - 682 764 104 od
- 380 seo ie Os) -228 oe, 343
486 .694 384 EE 06 338
UFPWNEF
AANA 049 oD 039 034 wed 326
5 Al 27 US wo22 2293 147 syayll
se -734 <oOF -482 wle2 33)
«266 «/ 85 oes 948 05:9 343
-090 -488 . 888 350 ody, 314 OWNF
WF

OES ~o1L9 - 686 -044 «079 346

3029 oft© -556 504 7210 £5 247
-465 .080 028 342 ao ke 5 283
-144 00 083 7090 2012 SL de3
Sead sod 928 a2 «68 2 281
Bis ey) ato. ~618 2639 364 5) 261
- 800 - O11 043 13:6 .708 356
ae Ouh 029 so27 430 S02 315 2S
food 7 oo! 27 479 wa B28 10
22s 2085 OS 2097 7696 B De
LOZ ood 7590 “Ise 743 310 35
ore . 807 r50 934 ogg 343
2916 ~ 061 OOS - 603 aoa 258
-754 ~053 . 188 2815 Om BD
pDDZ 2768 19.99 -ioo aoe 341
oo 085 .061 wo O92 340 18
2076 ee eles) . 868 .974 B55
eS 207 -421 7900 re 209 43
o> 089 090 vols 2oOl) 355
PoUS 186 al = 7 10 oll 128 169
7100 .374 - 662 203 “562 S12
wos 10910 760 950 043 Se
- 880 086 ~O37 662 - 663 B53
OS aiede 2052 432 444 ou 258
682 - 346 - 900 A192 -436 306
e902 2785 od 493 269 4 INST]
~045 898 ord 841 Oost 4 267
194 009 2963 0G -OgL 20 265
306 696 Zoe 421 020 334
eo we ied Syl aes 474 432 7 286
- 016 oo 724 147 .060 348 i
oad7 668 poh <a] 2185 530 9
SP
RP
RR
FPP CoN
Se)
Re eOF
SOON
Nae
DOs
eS
NN
NNN
Ne
NN
Nee
i
iOe
Oe nao
OP PW
FN ~ 202 ar
NOWWNWREFRFPFREN
NPWNWWWWWWNNDN
PWWWWWWWWWWNNNNNNNNNNRFPRPRPRPRPRPRRP
WO
OO
WAIADUBPWNPFPOWWOHANAHAUPWNF
WONANIAUPWNRFPOWO
OPWNFOWO PAI- oO
DANIAN
WW
UB
OF
WUNDWUNWNON
WW
PP 185
RPPRPOORPENNEFPEHENENNEFPNOHBENFPENDOOPRRPORP
PPP OPAPP
WWUKPRPWUHOU ho
WDAWHREAKPWAIAKHAANANAWAAAN
PWAHAHDPWHAWAAHAUANHIUNAUNIIN
WPRPWWPRPWWPRBP
SP

(concludes
on the next page)
I-58 CASE 40 TITANIC SHIPPING COMPANY

DAY RESP50 RESP90 NBAD

AVERESP RESP75 RESP95 NOK
46 ee 617 7.83 2355 864 6005 2Ad 82
477 -440 aris -455 1699 -405 seo
48 ~741 402 soak 574 AVOF 425
49 028 7309 a3) se92 -776 aT
50 218 -464 008 2956 -401 303 Zio
Sil a6 Oo 4699 so 15 -508 2 265
pe 192 924 2057 409 249 soe
53 216 .898 «168 -904 n/ DO aes PB
54 aoe aoe O23 28.6 -416 303
5D) 142 GS Low abe) 1997. 336 14
56 eo “aL 96 148 290 648 64 230
oT Sealed 704 486 )4 ae) 333
58 -464 30k aD .054 BOT 529. 20
50 souk L2G i186
NPPRPRPPORPRORPRRPREE Ou 882 358
60 7396 ? oO 1OD3 ~0 a6 2 L6 SD 1 238
61 876 2030 . 060 628 2501 350
62 451 iS eas 468 sh? 318 18
63 FD~924
FEN
NEP
NPE
FPNONPNNNEPPed-446 DO
W
D
WYN
WP
WNP743 OW
KBP
KEK nO 2)
NOwWPPRPWWNYN NW
PNA942
FUNDNAANWAKRNAN
WY
HAI
DWDWAU 316
MnyoWornnN
 AT
CAMPAIGN STATISTICS

“That’s good!” said the congressman. “We can use that to show the voters just what
happens when our opponents stay in office for four years.” He was referring to data that
showed that the effective income tax rate had increased from 14.1% to 15.2% of ad-
justed gross income during the period 1974 to 1978.
“What I’d like, though,” he went on, “is to break it down into smaller categories,
like, say, $10,000 to $14,999. That way, we can kill two birds with one stone. First, we
can choose the category that looks the worst for our opponents. Second, the people in
that category will know that we’ re talking about them, not just some ‘average’ or some-
one a lot richer or poorer than they are. See what you can work up for me!”
With that, he was off to a campaign appearance. The data he referred to are given
below. Provide the breakdown he wants.

TABLE
TOTAL INCOME AND TOTAL TAX (IN THOUSANDS OF DOLLARS)

Adjusted gross income 1974 income 1974 tax 1978 income 1978 tax

Under $5,000 41,651,643 2,244,467 19,879,622 689,318

$5,000-$9,999 146,400,740 13,646,348 122,853,315 8,819,461
$10,000-$14,999 192,688,922 21,449,597 171,858,024 17,155,758
$15,000-$99,999 470,010,790 75,038,230 865,037,814 137,860,951
$100,000 or more 29,427,152 11,311,672 62,806,159 24,051,698
Total 880,179,247 123,690,314 1,242,434,934 188,577,186

This case is based on real data, originally reported by C. H. Wagner in The American Statistician,
Volume 36, No. 1, February 1982, pp. 46-48. Reprinted with permission from The American Statis-
tician. Copyright 1982 by the American Statistical Association. All rights reserved.
II-59
II-60 CASE 41 CAMPAIGN STATISTICS

The effective tax rate for a given category is the tax paid divided by the total income
reported in that category. For example, for 1974 the effective tax rate in total was
123,690,314/880,179,247=.141 and in 1978 the corresponding number was
188,577, 186/1,242,434,934=.152. These are the numbers to which the congressman
was referring before he left for the campaign trail.
 42
HOSPITAL CHARGES

When administrative duty’s to be done, the hospital manager’s lot is not a happy one.
For example:
¢ Physicians have control over a substantial fraction of hospital costs, including the
length of time a patient spends in the hospital and which drugs and other supplies are
used in the patient’s treatment.
¢ On the other hand, the physicians are usually not employees of the hospital, so
while the hospital is responsible for managing the costs, its administration has rather
less control over those costs than it does in many other business settings. Hospitals do
generally grant “admitting privileges” to individual physicians and review their per-
formance on a periodic basis, but short of rescinding those privileges (or threatening to
do so), the hospital has little direct influence over many expenditures.
¢ Further, the hospital’s revenues are determined largely by the patient’s insurance
coverage, which may pay either a more or less fixed amount based upon the “Diagno-
sis Related Group” (DRG, a classification system used by many federal and other pro-
grams), or a percentage of the bill (in more “classical” kinds of insurance schemes). To
what extent differences in insurance coverage may or may not influence the patient’s
treatment or quality of care is a matter of occasional controversy.
Being sure that analyses of physician performance data are fair and balanced is im-
portant in any such performance review. The 289 patient records in the data for this case
(available in file HOSPITAL) are for normal deliveries of babies. DAYS is the number
of days the patient spent in the hospital, CHRGS is the total expenses charged to that
patient, PHYS is a code identifying the physician, and PAYOR indicates the type of in-
surance the patient carried: 1 for managed care insurance and 0 for commercial insur-

These data are real, but the name of the hospital and of the person who collected them are confi-
dential.
Il-61
 I-62 CASE42 HOSPITAL CHARGES

ance. Other things being equal, do there appear to be any substantial differences in
charges among the various physicians? Write a brief report summarizing your findings.

DATA SET
DAYS CHRGS PHYS PAYOR DAYS CHRGS PHYS PAYOR
2 1701 10 0 ‘2 929 nS 1
2 3500 4 a 3 3464 14 1
2 3165 10 0 2 SAS 10 0
a PISS 14 i Z 3034, 14 4.
2 2607 2 it 2 3716 eZ i.
z 2944 i iu 2 Zio 6 1
4 5063 2 ih 3 3430 6 a
2 23.96 13 a 3 3041 6 A
4 4903 2 it 2 2247 iL 1
2 2280 iW di 3 ASD 2 1
2 2959 a i 2 3050 12 iz
3 3585 4 a. Z 2199 eZ, L
2 S073 13 iH iz Za a0 { 1
Z. 2880 14 I 2 2620 2 ifs
2 2358 10 0 3 3369 13 ale
3 3418 2 1 2 3090 i 1
3 3604 a 0 4 6340 id 1
3 3480 6 th ve 2900 10 iL
2 3306 10 0 3 5.109 a i
3 3047 4 1 3 3503 4 As
2 Any, ie 1 2 ZO SY. 12 al
2 2439 4 1 2 2941 14 1
3 2647 ae L 2 2668 14 z
Z 1874 6 1 z 2138 2 1
1 2439 10 ik 2 2227 2 0
2 2424 14 1 2 2838 10 )
2 2450 a 1 2 2681 2 0
2 2198 14 0 3 3945 f 1
2 3355 10 1 2 2026 o 0
2 2609 0 3 3160 10 1
2 2324 Z L 2 S270 2 0
2 Ba.077 v2 i 3 4146 6 x
2 2649 ve 0 2 2150 Li id
3 3375 10 Hl 2 2435 i i
a 3906 6 il 2 2343 10 1
3 4049 14 i 3 3982 2 i
z 2564 13 a 2 2285 10 2
2 2745 at a 2 3001 14 cL
3 1254 6 al 2 1864 a 0
Z 2955 Z iL 3 4711 ee au
2 2334 i. i 3 3283 2 0
Zz 2809 Wi 1 2 2592 a 2

(continues on the following page)

 CASE 42 HOSPITAL CHARGES _ II-63

DAYS CHRGS PHYS PAYOR DAYS CHRGS PHYS PAYOR

a. 2081 ee i 2 2804 2 0
1 e791 14 i 2 2801 13 0
2 S729 2 a 2 2638 7 i
2 2a) 10 i. 2 2263 10 0
3 4510 ae 1 2 2449 oa 0
2 2017 “ Tk 2 2473 1 0
2 3218 14 a 5 2864 6 1
3 a392 a a 2 2534 lal a
2 1854 4 ) 2 Zoo 10 Al
14 14898 2 eh 2 3768 2 a
3 3819 2 Xa 2 3194 10 0
3 2644 4 1 1 2153 12 0
3 2551 LS 0 il 1840 io 0
3 3492 vale i 2 2522 2 a
2 2666 i 0 i 2145 12 i.
3 4248 2 i Os 2403 6 al
3 2500 4 1 S 3287 2 i
2 S242 10 ul zZ A24S 6 i.
3 ZOO 7 1 Z 3182 7, 0
2 2761 14 p 2 2954 10 0
2 1905 2 i 2 2158 TL 0
3 3394 10 1 2 3455 12 0
2 3663 10 a 2 2307 14 1
3 3059 6 dl 3 2798 10 -
a 2528 10 1 2 2408 Ae x
eC 2823 z a e 2468 aLab 1
6 6710 Akal A 3 3081 e i
2 2543 a3 a 2 2218 y iz
3 2785 2 0 3 2133 ee 0
2 2589 IP 0 4 5499 13 1
2 3183 at 1 2 2250 2 a
1 1806 10 1 2 ZI L0 tha 0
Zz Sap 10 0 il 2220 17 0
2 3204 7 i 2 2176 10 0
2 2921 2 0 a 2048 2 al
3 4933 2 0) 2 2566 a i
2 4722 13 a 2 2450 et O=
3 5633 13 a 2 2582 AL i
2 2780 10 0 2 2612 7 0
2 2798 4 1 2 2683 6 al
3 2066 7 i 2 INSEXT) 14 0
3 2494 id il 2 3202 10 a
3 2724 10 0 3 Slo 10 il
2 Zee i 0 2 2833 12 1.
A dhs. 2 1 2 3458 10 0
2 ao29 4 a ut 2789 7 A
Z 3094 14 0 2 L590 i) cL

(continues on the following page)

I-64 CASE 42 HOSPITAL CHARGES

DAYS CHRGS PHYS PAYOR DAYS CHRGS PHYS PAYOR

3 3868 a 1 2 3032 ig 0
Me SU Ieh a2 2 3 2091 4 0
2 3109 10 0 il 2886 LO -
a 2097 4 0 t 3381 2 0
2 S182 aS il a 21D2 10 0
2 3034 6 2 1 367 & 13 1
ah 2a 14 a 2 2840 13 1
i) 2178 y dl 2 3105 13 0
3 4185 ae 2 y 2288 .° id ‘
Z 1567 10 a 2 3128 14 0
2 2347 14 1 2 Zand a 0
a 2268 10 1 2 2602 10 0
3 3617 2 0 2 2604 14 1
2 2858 10 A 2 1023 a2 0
2 2542 10 i 2 1899 10 0
2 2665 + 7 ft 2 2334 10 1
2 2436 6 2 eC 2540 2 a
Z 292 2 7 3 3636 6 £
Z 21D ul 0 2 4206 Z 0
3 aLSe 12 0 y 2414 12 1
2 2308 4 0) 2 2206 13 Zz
e 3800 i 0 2 2907 2 1
2 2242 iz zl a 3027 14 1
u 2314 2 a 2 3050 14 1
al 1960 6 0 2 LORS a 1
B 4183 oi 1 2 2692 10 1
a 1663 4 1 3 2564 di 0
” 2070 7 0 2 3623 10 1
3 2153 6 1 2 2741 13 1
5 22109 6 0 2 2888 2 i
3 3218 10 i 2 Se00 6 0
2 2419 3 0 3 2873 Zi il
3 3289 LL 0 2 2582 10 1
2 De, 4 ol 2 2640 2 0
1 2741 s 0 i 2068 14 1
3 3454 10 i a 2123 10 0
2 3848 10 i 2 3480 14 +
y Ange 7 0 2 4475 6 0
2 2594 v 1 Z 2802 14 1
2 231hke 6 1 2 2898 = i
i 902 A 0 3 3708 10 0
2 295 Z 1 A 2078 10 0
a 3230 6 i: 2 2570 ia i
2 Zi2k 2 0 3 3689 14 a
S 2687 2 iM 2 4674 12 0
i 1848 14 i 2 2534 14 0
3 2891 4 0 2 2197 7 0
Z 2698 iz 0 2 J139 12 0
(concludes on the following page)
 CASE 42 HOSPITAL CHARGES I-65

DAYS CHRGS PHYS PAYOR

3 3826 a 0
2067 10
2840 Zi
2465 2
1947 ee
2267) dl
2840 10
chal 6
S137 2
ZI95 2
2563 10
1924 4
2062 6
2663 13
2184 z
Nt
bo
tombon
BONE
se 3064 10 FN
CT
Si
I
SG
Sa)
 e a

pels a ye, 28s PA eta '

yee,
Cal Ta
“1 2>°
Jee)
13 OE Si
t
jue
- at i = ay a ee Gt
7s — = - a ; 2x8
ul
oe
2s
é

we. a 7 #12) * .* ~
A oe it) c
he . a... « RA4® a ol |» OF
<=. aee® a : >
' ~ sn
an
et
oJ
au
©ww
Se
€F~ Ley 0 6
[aye +
0 € on
ea
aad
ra, y

1
te

A's
ad aon~

~~ i 9 »e
bers
POCE
ead
22 TR c

54 *< G +. na ca3e _

Oe
2 :
ey he
a>
BA
L340
; ‘a : &.,
or 4 , : »
a 5674
se :
ei
«®

an Ly} a : , . eine fa aa
2 4 4 Bs é onieh i
a dae 3907 ’ io 7

2’ weal . a 3a? a6-

ee 7 7. Veda See
cs | ¥27 q ist} [wes
7 ism . a). Saad
aa. ‘ ; M4 ii? |.
: - 4 i? - =
ea 19 @0 = oo
ay f Po aia 47° ta 24 S=
- —_ . —Ns yes
. 4°. =
7 *

¢ le

ae a
‘
7 gs
_

: rs ie? '
‘ é
° ; oe 4 i om

_ 7 .
7. aa 7 : "
ot , ;
7 7, j
7 ®

® 4
> : Da
s 7 Se

>

- :
- 4 4
- » & _

d ‘@
‘ ~ . -

i
“ » A , » Byars 5:
La
es i
a a
me ee
; a
 43
COAL TESTING

Coal is a heterogeneous bulk commodity, typically sold under contracted terms that in-
clude bonuses and penalties associated with various quality parameters and impurity
levels of the coal. In many cases, predicting and controlling the quality of the coal ac-
curately can mean the difference between economic success and failure for a coal mine.
The heating value is a quality of particular interest. It is the basis of payment in most
coal supply contracts. In the United States, the heating value of coal is measured in
British thermal units per pound (BTU/Ib), by using a national standard test procedure
(ASTM D-2015). ASTM standards also indicate that repeated measurements should
give values within 100 BTU/Ib.
Normally, the only way to determine heating value is to send a carefully prepared
sample off to a laboratory and wait for the results. This can lead to disputes, if the cus-
tomer and the supplier don’t agree on which lab to use or if the results from different
laboratories give drastically different results.
In an effort to determine the extent of any intra- or interlaboratory variation in
heating value measurements, one coal company prepared 50 samples of coal from a
common area (so that results were expected to be very close). They shipped the sam-
ples in two batches to five reputable laboratories. Each batch consisted of five sam-
ples sent to each lab; the two batches were sent at different times. The samples were
drawn randomly, packed in water (to minimize oxidation effects) and sent in sealed
plastic containers. The labs were not aware that an experiment was being run, and
the coal company hoped that the results would thus reflect the labs’ normal operat-
ing procedures.

This case is based on areal situation, described to us by H. Erik Scherer, P. E., Mining Engineer
and Consultant. He also provided the data.
\|-67
 li-68 CASE 43 COAL TESTING

The labs made several measurements, but only those for heating value are given
below. What do you think the coal company should conclude about the perfor-
mance of the laboratories on its list, at least as far as heating value measurements
are concerned?
The data are available in file COALLAB. The variables are: LABNO, the labora-
tory number; BATCH, the batch number (0 or 1); and BTU, the BTUs per pound as re-
ported by the laboratory.

DATA SET
OBS LABNO BATCH BTU
al i 0 11655
2 a 0 11690
3 i 0 11654
4 i 0 11653
5 al 0 Li66>
6 BO ie a LiL 7oe
y/ il al 11676
8 il dl bee
9 al il deal
10 il il 11684
ile 2 0 11665
12 , 0 11618
13 2 0) LAS 2
14 2 0 11686
SS 2 0 P25
16 i 1 11602
17 2 1 11582
18 2 1 11764
ws 2 1 Li613
20 2 il 11637
2K 3 0 11586
De 3 0 11588
25 3 0 Jo
24 3 0 tig6
25 3 0 Lio22
26 5 i 11428
2a] 3 Al tie 29
28 3 al 11488
29 5S 1 11632
30 3 1 11478
ad 4 0 L165.
32 4 0) 11780
33 4 0 11663
34 4 0 11769
35 4 0 11800
36 4 i 11630
CW 4 i LiS5e87

(concludes on the following page)

 CASE 43 COAL TESTING I-69

OBS LABNO BATCHBTU

38 4 ahTi567
39 TES3
40 11630
41 11591
42 11605
43 11549
44 i590
45 TL6O9
46 L739
47 Lied
48 11768
49 11768
50 & 11794
PrPPRPPOGDOOCOFRP
AMMnaAnnnnnn
 ™ mie: - one ee a

OB Fora te [
pie re a

tpi alent f NI, VR

too Wa saa erate
maasioe (ALihe fete
heTi ape ae enggre ‘
y Ste stimeeraed? oo 2 ho
Phe hale ioe 4 peste wees dete ak ee
Hilynuma ATI Bee port SOLS ae ed
eeis
ie ie » anton fos comin
onere
eee, die) OO
a fF ma CETLE ; =
HAR een TORAH:
1 pared
; a vt o> WATTLE
2 ‘ perce
: as | : ‘
. }
i yi I nw a
2 a

%
>

i rs 4

a
‘—

7 :
Ss ~.

= ot

i io ‘
é ~

at

a= 7 cs
r a

as ‘ :
a J
4 ‘ :
- ) i ih

. 4

=~
 44
PERFORMANCE
EVALUATIONS

Ann, an independent contractor who specializes in human resources issues, has been
asked by the XYZ company to undertake an independent study of its employee per-
formance evaluation practices. XYZ, a nonprofit organization, is a professional review
organization in the health care industry. Most of its revenue comes from government
contracts for Medicare and Medicaid utilization reviews.
The employee performance appraisal involves a fairly complex mix of subjective
and objective scoring. Each evaluator fills out a form that covers 11 categories of em-
ployee performance, such as adaptability, accountability, communication and decision-
making skills, sensitivity to cost-effectiveness issues within the organization, atten-
dance, and training and development of other employees (see Exhibit A).
Nonsupervisory personnel are evaluated on 8 of the 11 categories; supervisory and
management employees are rated on all 11 categories. The same form is used for all
employees, although the weight applied to each category varies with the employee’s
grade and responsibilities, as shown in Exhibit B. Weights are assigned in such a way
that any employee can receive a final score anywhere from 0 to 200. Higher scores
mean better performance.
Each employee has an evaluation initiated on the 12-month anniversary of his or her
hire date, so evaluations go on at XYZ throughout the year. Any employee who is a new
hire, has been transferred between departments, has been promoted, or has been de-
moted is considered to be on probation. These employees are given three-month eval-
uations in addition to the annual evaluations.

These data are real, although, because they deal with personnel issues, the name of the company
has been disguised. Claudia Roberts Salvano collected the data set and provided it to us. Claudia re-
ceived an M. S. degree in Management in 1993 and was the Manager of Human Resources at the time
that she collected the data.

I-71
 I-72 CASE 44 PERFORMANCE EVALUATIONS

EXHIBIT A EMPLOYEE PERFORMANCE APPRAISAL FORM

ANNUAL PERFORMANCE EVALUATION -

Date Evaluation Due: Time on Job: Employment Date:

Employee: Title:
Evaluated by: Title:
Reviewed by: Title:

RATING KEY
eae aeae ee re ee
RATING POOR(P) FAIR(F) COMPETENT(CP) COMMENDABLE(C) DISTINGUISHED
nc ce ana A BIE er ee DaeNO EPO a
INDIVIDUAL RATING P= (P+ E= "E+ CEE CEs C- C+ D-—
CATEGORY POINTS 1 2 3 4 5 6 7 8 9
a

TOTAL POINTS 0-50 51-70 71-90 91-110 111-130 131-150 151-170 171-190 191-200

Raise % 0 Osean 1o3 2% 3% 4% 5 A% 5 %% 6 4%

DESCRIPTION OF CATEGORY
Poor Performance does not meet requirement of position.
Fair Performance adequate most of the time. Marginal performer in one or more areas.
Competent Responsibilities are handled competently. Meets all requirements of position.
Commendable Knowledge and mastery of position performance marked by high quality.
Distinguished Performance superior in all aspects. Quality leaves nothing to be desired.

Employee Manager Approved by

Date: Date: Date:

Employee Agrees Employee Does Not Agree

 CASE 44 PERFORMANCE EVALUATIONS _ II—73

EXHIBIT A = (continued)

Annual
Three-month

EMPLOYEE EVALUATION

Employee Name Due Date

1. ( ) INITIATIVE/ADAPTABILITY/ACCOUNTABILITY Rating__ Pts___

Demonstrates ability to work with others in organization. Adjusts to changed work assignments or conditions, accepts
guidance and special assignments. Demonstrates capabilities for establishing work priorities and begins tasks on own.
Fulfills the full range of accountabilities indicated in his/her job description.

COMMENTS:

COMMUNICATION Rating___ Pts___

Clearly, concisely, and persuasively expresses ideas and instructions verbally and in writing and understands the ideas
and instructions of others.
COMMENTS:

DECISION MAKING Rating_ Pts___

Demonstrates the ability to evaluate and make proper decisions within a reasonable time frame. Evaluates data,
reaches sound conclusions, makes appropriate recommendations, and follows through with timely decisions.

COMMENTS:

COST EFFECTIVENESS RatinGemeatots aa

Demonstrates and promotes an awareness of the importance of cost effectiveness and reduces expenses (through
unnecessary labor and time, wasted supplies, unneeded equipment, etc.) and reduces costs when and where practi-
cable.

COMMENTS:

QUANTITY OF TASKS Rating___ Pts___

Performs day to day assigned activities and special projects, within prescribed standards and ina timely fashion.

COMMENTS:
 I-74 CASE 44 PERFORMANCE EVALUATIONS

EXHIBIT A (continued)

6.
() QUALITY OF TASKS Rating__ Pts___

Produces high quality work with minimal error rate.

COMMENTS; —.___ eee

ne
ee er eee

e
pee ee S ee

ee a ee

ee ee ee

tA) ADHERENCE TO CORPORATE POLICY AND DECISIONS Rating__Pts___

Understands the rationale used in decisions, accepts the final decision, and is able to properly convey the decisions
to others when appropriate. Adheres to all Corporate Policies outlined in the Personnel Manual and related policy
memos.
COMMENTS:
e
ee

8. ( ) ATTENDANCE Ratngee isa

Arrives at work on time, returns from breaks and lunch on time, schedules vacation time in advance, schedules
Doctor and Dentist appointments so as to use the least amount of work time. The number of hours of sick leave
used is a major component for this category.

COMMENTS:

9.( ) PLANNING AND ORGANIZING Rating____ Pts___

Effectively schedules and coordinates the flow of work. Organizes and utilizes his/her people, materials, and
equipment. Organizes and utilizes his/her own time, meets established deadlines. Demonstrates ability to prepare
annual budgets and to operate within approved budgets.

COMMENTS:

10.( ) MANAGING, CONTROLLING, AND DELEGATING Rating___ Pts___

Leads and motivates people to effective group and individual effort. Directs, reviews, and checks work of subordinates.
Establishes and maintains standards of performance. Takes corrective action when necessary. Defines and delegates
authority to others.

COMMENTS:
 CASE 44 PERFORMANCE EVALUATIONS — II—75

EXHIBIT A = (concluded)

11. (| ) DEVELOPMENT OF OTHERS Rating___ Pts___

Encourages and assists individuals, through training and other methods, to acquire knowledge and skills essential for
effective job performance. Selects individuals with potential for promotion and assists them to qualify for advancement
with assurance that equal employment opportunity corporate goals are met.

OWEN SMe ee ee roe

 II-76 CASE 44 PERFORMANCE EVALUATIONS

EXHIBITB PERFORMANCE EVALUATION CATEGORY WEIGHTS

Category Director Manager RN Technical Secretarial Clerical Medical Director

Managing, Controlling/
and Delegating

Accurate performance appraisals are important. They provide feedback on an em-

ployee’s contribution to the company, and the annual appraisal score is used to deter-
mine merit raises. When this appraisal system was introduced several years ago, eval-
uators attended extensive training sessions designed to ensure as much consistency and
fairness as possible. Evaluators are also provided with detailed guidelines (Exhibit C).
In spite of this, employee complaints continue to surface. Ann’s job is to undercover
any non-work-related biases in the evaluation process.
Concerns expressed by the employees at XYZ include the following.
¢ Evaluators are occasionally late with their evaluations. Rumors at the company sug-
gest that the later the evaluation, the lower the score, as the manager may be attempting to
postpone a controversial or hostile discussion with the employee about poor performance.
¢ There may be interdepartmental biases. Some managers may be attempting to raise
their own department’s status and salaries by artificially inflating the evaluation scores.
e Scores may be bumped up a few points to push a favored employee into the next
higher raise category.
e Three-month probationary evaluations are considered unnecessary by the em-
ployees, because three months are generally not enough time to adjust to a new posi-
tion or to reach the peak level of efficiency in a new job.
¢ There are rumors that the company is attempting to force early retirement by giv-
ing the long-time employees lower scores than others.
Ann has a sample of 144 employee evaluations conducted over the past calendar
year; 10 variables were collected from each employee record as described below. The
first 125 observations in the data set refer to employees who were still employed at
XYZ at the end of the year; the remaining group left the company at some point dur-
ing the year. The data available from these evaluations are contained in the file PER-
FEVAL. What should Ann’s recommendations be?
 CASE 44 PERFORMANCE EVALUATIONS _ II—/77

EXHIBITC DETAILED EVALUATION GUIDELINES

EVALUATION GUIDELINES
This guide has been prepared to help you conduct the performance appraisal. The objectives are twofold:

1. General guidance (ideas to keep in mind when appraising performance).

2. “How-to's” of preparing the evaluation and conducting the session.

. BE FAIR AND IMPARTIAL

Try to disregard bias about an employee. Concern yourself only with the facts and concrete instances of employee
performance. Bias and prejudice may exist, but if you are aware of them, you will be in a good position to control them and
their effects. A positive attempt must be made to be as fair as possible to the employee and to XYZ.

. APPRAISAL PERFORMANCE
This is your role as a manager. Appraise performance against each accountability. It is not fair to base an evaluation, for
example, on a person’s educational attainments.

. PERFORMANCE—NOT POTENTIAL
Accurate assessment of individual potential is important; however, the concern here is performance of the past year. Rate
only the performance and do not rate what you think his/her potential is.

. ACHIEVEMENT—NOT PROGRESS
Performance is rated in achieving results during a fixed time period. Success should be measured against previously
agreed upon objectives.

. APPRAISE PERFORMANCE FOR THE WHOLE YEAR

One month of commendable does not offset 11 months of satisfactory.

. RATE EACH ACCOUNTABILITY INDEPENDENTLY

Do not let your performance rating on one area affect your rating against others. Try to rate each aspect of the evaluation
independently.

. CROSS RATINGS
Usually, evaluations are completed by the employee’s manager. In some cases, another manager within XYZ may have
important insights regarding the employee’s performance; for example, if the employee worked under another manager for
a good part of the year, it would be necessary to seek the views of the previous manager. Keep in mind, the goal is to have
a complete, objective, and fair appraisal of performance.

. BE A CONSCIENTIOUS RATER
Give a favorable rating only when the individual has merited it. Forget about giving people a “break.” Often, this is difficult for
us to accomplish. We want to be fair and often this means giving the benefit of the doubt. If we continue to do this over a pe-
riod of time, we will find the standards we have set have strongly been lowered.
By lowering standards, XYZ’s chances of attaining organization goals are weakened. Additionally, we hurt the individuals
being evaluated by asking them to produce less to reach the same level. Be strict and fair to both parties. These two goals
do not necessarily conflict.

. MORE FREQUENT APPRAISALS

If an employee’s performance needs improvement, establish a time frame for corrective action and re-evaluation. Discuss
performance expectations, actions the employee can take to meet those expectations, and a specific time when you will
re-evaluate the employee’s performance.
 II-78 CASE 44 PERFORMANCE EVALUATIONS

DEFINITIONS
SCORE Evaluation score in points _
LAG Difference between the due date of the evaluation by the manager
and the day on which the results were discussed with the em-
ployee, in days
EMP@EOY Years with the company at the end of the year
POS @EOY Years in the current position at the end of the year
TYPE Type of evaluation (0 = 3 month evaluation, 1 = annual evaluation)
GRADE Employee’s grade scale: 1—7 are clerical or technical, 8—9 are
skilled employees, 10-12 are supervisory,‘13—17 are management
DEPT Employee’s department number
EDUC Employee’s education level: 12 is high school grad, 13 is one year
training beyond high school (e.g., secretarial training), 14 is two
years beyond high school (i.e., Associate’s degree, nursing
diploma), 16 is Bachelor’s degree, 18 is Master’s degree, 20 is PhD
EMP@EVAL __ Years with company at time of evaluation
POS @EVAL Years in the current position at the time of the evaluation

DATA SET
LAG POS@EOY GRADE EMP@EVAL TYPE
SCORE EMP@EOY DEPT EDUC POS@EVAL
184 cl Woe Wes) ANe 8 ale Pees) Te M2 i
155) 9 20 ee a OS HES a3 1.820 02 al
v3 2 opt Ol ae ON) tee 5:5 153 16 Ors 10 OTs 0
nine eeu Vo3O Oss@ SS 4 a3 OR 2510 Om23 0
102 iL) Seite Chg Vial 2 eZ deo 4.07 1
178 4 ORG5a so OS en Oe 3 eee S00 Son a
1S 2 eS ee eS ea S 16 ZO Ze 1
128 Ls OsedO vOeaOn WZ 8 16 O2290 0229 0
aS 35) AIS Bell Gz 2 2 eel eG) 2500 af
SS = 2 Ae Oe eeal ea 16 AWeePAEH0, eee 0
158 le Ble OA as 5 10 le? 4.130 Ins03 dl.
125 0 Cnt94 Oo eo 8 16 6.000 6.00 al
lass 2 oss OSs) /s 8 14 le 50 0.26 0
2 le CS Sm ORs 73 al? 18 6.290 OE22 0
172 a Ey Tiss | Shes), /h24 8 16 SE080 39903 A.
inS)S 49 MeTO” AUS TO)” Mee! 8 Tay Os 67 Or eas Hl!
an 14 Cre Ono) ae SS 4 dbs} DRooU 2.90 i
LoS 7 Taeshe sta, Taye 10 18 Vaal ae OF ‘l
128 ae Oe516mm => Oneal 2 PS, Ome 0 OS 0
ea 62 SRA2 Se 42) eel 10 16 Sree0 ated 1
134 45 Ors Ossie 62 8 16 ORS 70 On3 7 0
126 3, OE i MR er tet 2 se Oe 0.05 0
3a: 16 4.24 4.24 41 10 18 4.050 4.05 1
ASG =—2 OS S6ss OOK SG" 55 2 12 0.240 0.24 0
155 8 Oe Sih MOSS vo cAtih algal 16 Oee27/€ O27 0
125 Ay 0.42 0.42 40 3 16 0.300 0.30 0

(continues on the following page)

 CASE 44 PERFORMANCE EVALUATIONS lI-79

LAG POS@EOY GRADE EMP@EVAL TPs

SCORE EMP@EOY DEPT EDUC POS@EVAL
128 PhAl Oss 0.89 He. 14 Oe Sale OFesel:
164 Al, Oi)Me An OND 64 14 4.010 4.01
, 90 aul OVA O74 2 2 ORB ZO ORB2
146 oO Ai Ot Oe me 13 14 3.470 12,99
a59 Lg 425.0728 OD a 20 05.950 O30
da a2 4. 80% 24,80 74 16 4.040 4.04
129 36 S542 3..42 62 14 35 OE Srua0
149 el Ao 63 16
DNOonMDOoORE
O © 2.340 ibs(0)4
182 =) TioMel ais oe Te. iG 6.980 67.98
ide 42 3 1612.38 62 14 Be. ho U Zi
37 oi: O24 eae WAL 16 6.470 Or22
Ld. 30 i 3 PN iS) 62 aAr2, Zee O) ZrO
144 34 1 ala NOLS DS V2 it AOC OF,
163 28 ee dso" 61 14 OS de OS
165 6. 09e 2 Oka 74 16 6...750 4.98
128 ila 6, OZ Re D9 We a O. Lae br 0B.
133 le29 ha 1.29 63 14 OO 5 02
162 aio) Bin)OME 252 64 14 BF S10 399
£19 Oi. Om 0. SS 65 16 WCAG O27
L3il 34 OSS WsSh7/ 74 de 0.350 C235
149 62 Avaaa) UL BD PZ 2awOO 0.42
Leh 24 Os Jae 0. 41 2 0.320 OZ
136 13 eS Alay eleal- 63 14 1.040 1.04
150 i C2560 RO. 26 41 14 O...310 Open
135 0.8) 90.81 7S — 14 Ora 0 Oar
AG: 64 17.24 17.24 62 ale ae Oe Waren
ID ag Ul350e 0 38 ee ig Oe SOG) ORS0
SS DES.OMEED IO.O 65 he 5.000 5. 0€
L2y le set Bw cay 62 aS) ORIG Oreo
61 Sy) 4212 Os 20 62 14 4.790 Voss
nS 1% oo 16as 0. oo =) 3. 253.) O23
LAL 28 109 E07 hk 5S 14 2390 Oost
160 —14 1.42 181-42 WZ 14 ORS 6G ORS6
oy coal 2 vOo (i209 eZ. aa 14 (eS oO 3. O00
155 2, C0aN 2.00 72, 14 AsOlO 2601
124 ae Sse” at V3 bh 16 ee 30 OsA5
141 12 142, Ske) « dee ves 14 OF 960 0.28
124 Ls WVCoe a4 6L 14 1.040 OA
31 79 AL 76 ee 47716 63 14 4.220 ATRL
bd A SOMA 209 73 eZ. 8690 ZrO
135 10 ASOLES 0. 56 62 bh 14 BE 100 OE28
164 6 1, 04067 64 64 14 7.600 EO
eA 0.46 0.46 62 14 Os 2He O27,
Aaat Ae Stee OO. 73 bk 16 4.670 O26
ives a Spl LY 6.50 61 ce OSU ore
TPE 27 oie Pare 74 OO!
WING
GoW)
1S
Ul
CO
C08
&))
Or
16)
DSN)
WSF
89
100)
Gd)
83)
Oo)"
Coco)
We
WE
@ 14 20 10 AN: aSSolo
SooS
Eo
|ooo.
oo
Soi
lj
eee
es
oe
SiS
oe
oo
SS
po

(continues on the following page)

I-80 CASE 44 PEFORMANCE EVALUATIONS

LAG POS@EOY GRADE EMP@EVAL TYPE

SCORE EMP@EOY DEPT EDUC ' POS@EVAL
128 15 0-940) 0.94 62 S.A, LO 0.290 O ao 0
23 —6 0.63.00 Ob mmo 2 8 Gs 0.240 0.24 0
2S2, ee 0.83:55,0288,47.0 6 16 02240 Ona al
Ai 3 34 8.73 Ee OatSeem Os 8 14 8.100 8.10 1
126 36 Oe 90 28 0.54.2 3 d 2 Ue O35 0
120 8 PERLE VE Se Ty 8 14 2.020 O22,0 0
2a 30 0.9 Geae0s 26.65 5 IZ On 330 01.333 0
di 12 0. S2me OSDZa) LA 8 14 0.280 0228 0
i229 58 Ol D6 a e.O.59'6 aie Gol! 3 14 0.410 0.41 0
132 2 Ds De 2c we ak 10 16 eI Oe eet a
75 10 sO s5 086.6 s5U ane 10 16, 2610 [92 0G COG z
7 6. el 2.)GSal2. Oe 13 4 L200 eZ uf
155 20 T SO Rec Vues Hh3 es 16 Dogs eee) 1:
SZ =2 3520 ee 32 Came 8 14 A Dee 1
Ud Dyrene OL 002 fe b2 5 2 5s deo) 0.26 0
d2 18 OF 94Re 0.9oem oe 8 14 L300 L230 iL
170 4 OpaZtS 4256 OZ 8 16 5.020, vag. 1
159 as 4.80 4.80 65 8 16 4.040 4.04 I
124 a OS65 BE0S 65907 2 8 14 0.300 0239 0
174 ==ih 3.44 3.44 62 8 16 2980 Se ceo i
20 9 2. 4480 2 bS 7a 13 16 1 580 0.28 0
139 0 96g 0.09 3855 10 14 1.320 0.24 0
139 0 Zo1OTe 2a)Oe ake a 2 2.006 2.00 1
160 5 i G9 eatsa eu a 14 18 i330 0.99 is
133 =8 2. SOmeO DS ORL 8 14 4.990. 47:99 1
147 14 LOO ae ho o61 64 8 16 .D6Q — 2256 1
126 15 0. 52a 0. 52.061 8 18 0.290 0.29 0
ii = 6.787a0.28.5 64 Lo 18 6.740 0.24 0
140 16 QC. 30am 0.74 a t2 e -s 0.520 0.29 0
124 60 t.33ee 0.595 G62 10 14 4,919 0.42 0
104 10 i. SO Slt. SO CUe2 2 14 4.030 i035 1
126 30 Aw ORO S27, dS 8 14 2.840 0.33 0
140 20 5.05%) 0.98, 644 4 12 4.430 0.30 0
106 7 SriSat OS, eet 9 12 7. 8E0 oes 0
138 7 Vets av Ocaoa Te 4 12 0.270 Ong 0
13.0 5 2.25.00. 80, %55 C 12 1.630 S27 0
L05 35 O32e" 0s28. 072 8 16 0.350 0.35 0
143 2) Ble2alas Bi eS ae 16 14.030..14.03 ui
L314 27 2.94) 02,69. 955 10 18 2.320 #07 i
LOS SLO CELIO 26e5 0. 28ye74 7 24 S21 2220 0.22 0
17d =7 5. oS ane 08.2 10 14 4.380 #43 1
152 22 2 61h 22225 55 5 i 2.500 “2.06 at
7. 15 DC meD. caeu 64 3 12 2.20 5.04 -
152 8 L964 42.96. 96s 2 sia i 020 #02 1
115 69 QO. 46590252872 3 12 0.580 0.44 0
80 8 0.48 0.48 41 4 12 6.270 0.27 0

(concludes on the following page)

 CASE 44 PEFORMANCE EVALUATIONS I-81

LAG POS@EOY GRADE EMP@EVAL MEE,

SCORE EMP@EOY DEPT EDUC POS@EVAL
apn Aas, Vis 41 16 2060 Oy 2)
134 Pago» Oa, 61 14 dee Or
pod Se 1. oon 0.74 55 LS teow O26
37 ene OS 55 18 Loa Ors 0
29 Ae OlSie 62 als 2.047 Ones
154 16.eo Lo. ou Al 14 A GrsS'9Oto 13.25
sll DitOO lS: Hell 14 B ke.Q erage
104 De Oe Vol. ae 55 de2 ayaa ils Oat
rot Ota ye eZ. 14 0.240 0.24
eae Ocot es. Oo d 62 14 0. $50 eis)
a. Le ple iS. 79 65 14 Ue SAOe Ae!
90 Leo ere 74 14 0.280 0.28
123 Al ame epee) 74 14 ORO O027
Oa 8.25 4.49 62 16 Te Oo So
40 OS CO S16 62 ie 0.480 0.48
ey 2.80 1 2280 74 14 ZAO 2002
WZ el ml 2a 74. 14 ROO Oe
ro 1.84 1.84 V2 18 0.960 0.96
OS L3G LOS 55 ie, IO. SAW G5]
106 2 US) Oboe aS 14 1.680 ig a7
116 Oy Smelt.S 41 We 0.280 O28
al Zi aD oe 62 16 Zr0.6 0 Ze10S
143 ev! Se 65 16 Up ee eee,
90 0,65 9 90:65 74 14 0.260 ORG
a3 Oe 2 41 ee O30 eel
fey Ona? 0142 GAs wobBNM
oO
mW
Wom
W
© 16 0.280 Or8 [FS
FS
12a
(
(2
SSeS
ea
Leas)
She)
[Se
Bee
aS
ST
 2 . a ie be

darian Sa ygeake
7 7 gaGltece eC ean
oe zs. eX,
p itt, =! ih
~s - “Ov 7 Ne ae 7

ae jen \ Ury:. 4
we
' ie . veh. 0 i: Gs 4 é 4

fees B= FY GOs Seen bh

- 7 Bet eae iy f
cy 7 oe J wo 7 & =% ef i,

ie Gib orn; t
$4.0 AAT
‘SUS F neil / if oe :
5 | ee,

Pa aU

1a ie: e
 uw 49
AMCHEM’S TAX CREDIT

As Harold Davies sat down to finish work on Amchem’s tax return, he knew most of
the details were available in the huge pile of files on his desk. There was one part,
though, where he wasn’t sure how to approach things—the sales tax on gas production.
Amchem manufactured chemicals in a plant in the Denver metropolitan area. A re-
cent change in the Colorado tax law provided that all nonproduction use of power was
subject to state sales tax. Harold knew that much of Amchem’s use of natural gas was
production-related. The problem was: how much? Public Service Company of Col-
orado bills and records revealed only the total gas consumption, not its eventual pur-
pose. Harold knew that some part of that consumption went to heat the research facil-
ities and other nonproduction and administrative buildings. That part of it, of course,
depended on the weather, among other things, and would be subject to sales tax under
the new law.
Harold suspected that most of the gas consumption went into the production
processes for two groups of chemicals Amchem produced, labeled generically Group
X and Group Y. The chemicals in Group X were fired in electric furnaces that use lit-
tle gas (what little was used was used in other parts of the production process), while
those in Group Y were fired in gas furnaces and used (Harold suspected) quite a lot
more gas.
Harold felt that as long as he had a rational, defensible basis for his estimate of how
much gas consumption was associated with the production process, he could get a
credit of some kind for sales taxes paid on the total amount, and he felt it would be a
substantial amount. Looking back over some of the files he’d requested, he found he

This case is based on areal business situation, though the names of the company and its products
have been changed. The data were provided by David W. Howell.

I-83
 I-84 CASE45 AMCHEM’S TAX CREDIT

had information and some quantities of presumed interest, for each month (MONTH)
of the last five years (YEAR), as follows.
¢ Gas consumption (GAS): in hundred cubic feet (CCF), from gas company
records.
¢ Heating degree days (DEGDAY): a measure of cold weather, obtained from the
local weather bureau. The amount of gas required to heat buildings is expected to be
related to this measure; that’s why the weather bureau tracks it and provides it.
° X production: production of chemicals in group X, from Amchem’s monthly
records, in thousands of pounds.
¢ Y production: production of chemicals in group Y, from Amchem’s monthly
records, in thousands of pounds.
The data are available in file AMCHEM.
Harold knew that all these factors somehow contributed to overall gas consumption
at Amchem. He figured other factors such as annual plant shutdowns in summer (“turn-
around months”) might affect things, too, but their effects were presumably repre-
sented by lower’production during those months. One particularly extended shutdown
in July of year 5 might have distorted things a bit, though.
As it got later in the afternoon, Harold wondered if there were some way to organize
the information coherently. What he needed was an estimated amount of gas con-
sumption for production use for year 5 that he could defend rationally. It would be even
nicer if the method could be made routine, so it could be used in following years. Of
course, it would also be possible to simply not claim the credit, but that was unappeal-
ing for any number of reasons.

DATA SET
YEAR MONTH DEGDAY GAS Xx M
il al 1206 TBID0) yolk 3 Carma dao
a 2 942 LOST AS Beet 2A. 0
al 3 657 66408... 1.54, 023.2
il 4 421 D0 29A.. p13 199.9
Al 5 220 DOS BB ALSO. 2.69
1 6 55 45292 eg Su gaol.9
fl i 0 S714AS * (i360 Lae
i 8 6 43.03.34) del 2igid 8.7
i 9 246 34696 903 8.6
nt 10 460 53204 9.16 Bred
al 11. 1017 SELOOG, 0.Oy es
ul 12 1070 48270 0.0 5.9
2 wv TE a0 TLL 4.5 oe
2 2 780 60922 Fob 1S.8
2 3 457 D9 Dee ppllrcrd pele dy<an
Z 4 359 44813 Oi.Sed wha?
2 5 226 46704 T oSiyrk 9 5h
2 6 i Seana TOON gen
2 7 0 Syopo wh) OS. ee icnd
2 8 2 39746 OiOram Score

(concludes
on the next page)
 CASE 45 AMCHEM’S TAX CREDIT II-85

YEAR MONTH DEGDAY GAS Xx A

2 a 222 50892 Orerke (92
2 10 Soi ASTI? Aisa 20.5
2 pt oe SLaLG CLO57 (16.9
2 ez 1104 Sy Sa Peo. 15.5
2 i 1038 HAO Me Sa oe ocd
3 2 770 69414) 4 U3'7 35.4
3 S aa. Voge Lae oa O
2 4 338 S57 Si ie t36
3 5 Zao Sireodies Ue Gk
3 6 36 s2o02 “lac7 ~A0.6
3 a 10 50987 14.1 46.2
3 8 SS) 36257 isi emetic
a 9 186 Ai DON USO Zoe8
3 10 493 C0697, V1358 Sia
3 ep US OHO Sde lel S228
3 12 1214 POOR3) CEO 23,4
4 z 1243 S76300 15200 <2 ivs9
4 Z 892 OS252) 15.0 e273.5
4 3 815 Visi D2 2 SeeskeOU SEl eiS
4 4 453 52700) 16nd — 347.2
4 5 203 OSA ler OmmS SLO
4 6 1S 5305Opel ity 3 Os
4 uy 3} AL154) 34) 26.4
4 8 6 543530 UBS 3 Du68,
+ 9 154 By] 4)5\Ome or SS ee)
4 10 412 I23824" A853) 34:56
4 dea: 774 TO 0S eel Sop Ot
4 AZ $63 YAOI ES Sie EE BS
5 i: 942 6/563 eS 9.4
5) 2 1216 So TAO. 1 Aae0 | 249
5 3 OS), S20 54 13 cl 50)
5 = 453 Gulia9 eeeeS 8.8
5 5 An OIG lf Ue.8
5 6 101 AT DTG se desiO BS 9)
5 7 0 MES 2 36 8.6
5 8 4 48714 CrOmmmeleo ae,
5 g 186 BO ON ae Oru
5 10 498 OA Oe Ore
5 11 710 58444 12.6 16.6
5 ee, 1233 6295Ae Od 5 13 et
 tei FACIE ant MMA
2ARRAN” Anite ve Pl i eeae
. Sb ge Sle ee ae >
: rn iy = a as 7 oa tee
; > . . _ “Teh Je Ee -Z

hetjeaeiebiiicdeel pee
na bash VI cf, ae
F 2 zy
S ~ ;

i : a ak = se
te
en! - + hieseni ight 2 "i, te
Fi MOONS ato Hf bic, oe st
“9 Hatters ,
‘i hil © ee 7 a
i foe Vee
“a e mae 2 : , Se ‘ he aae

o riaseesta , Re ary co ~~ - a
Le ‘ geoy forte ayy A daw coe ide , nt
1 Of t 4 7

. GOHIRE M
we (et 20 Voile OSE o fe ;
oe Han tGs dail] (GAM criticl nately pried &
— ai Avttifad th) Quits ~§E4GA > ONO& 2one@ (1d oe
‘7 Pho, bat tebe eee
decmcindd Padibie’ ! ehigior aE de gk
) eaateet bop pier A ONS doris, Sky pepticiolegly
mau ry 441: dS OE Ob as ® ib Mita eit
Yi yts >
al fa! ea ppl ine, 1aeARS onindereg I hove dave
Ue &1) etiog) V = TN hd ahede! Eat teil |

iaarhe © tps’ réeny Sean LAO K ipony 8 Ont le crated ineSecret TR

a ey i Oar osha Weak! hale rhe. pr e G0\ vaahs!
innounans fe
> Mate, BRAS Dit fapee io Sidoly ox Cole ee ahh eat e
= : Wit pod aly ‘Maes repre ho
% . : 4 :
: 5.4 i » ? { d 4 + 2 - ;
7 an 2 Z ( ..% erate i : ‘ * :
7
oer . > Mf a pGkcec a OE : .
7 i. 7
7
y : ‘ ‘o
a
we
7 a” FRAG 7. 4 > Fess) Ja of
_ >
- me fevSy coy =e ai é
< : = sf
¢ é =T @ - ; sf LA a é 7
- ‘ bax j ee >
7
_—
o
oe 2 od
a)
_ ~ rh ; ’ ave . . é
- é % @ eed}
> i
ee
4 » 7 €
ft ; ~ “
=e OP 4 Bish * be: «
ay ‘
7 74 “4 1 I Z = 3 &
-.
>. & a mt — }) i % é
. 5 4 | fh - i 2
-
:
: sa 2 BETA : = 2
i } hc a pep mG
7 1 fer tic oe et. Be
S ax, @ eG
J [ c a we j i ‘J “ % ne 2

ae
- 448 - , «9

5 a9 ih * As

5 ~~ - all we
® ay 9 han

ia s:2 ad +
at. 2
ne — _
a ; .
~ i
i as Cae
 46
NATURAL GAS PRICES

Your company produces natural gas and needs astatistical model to provide short-term
forecasts of natural gas prices. Of course, forecasting the price of natural gas is difficult,
because many different factors influence price. The demand for natural gas from the res-
idential, commercial, and industrial sectors of the United States is a major factor. Nat-
ural gas prices may be seasonal, because, for instance, home heating requirements are
higher in the winter months. Prices will also reflect the costs of producing natural gas.
In addition, natural gas prices tend to mirror prices of related energy products such as
coal and oil. Gas prices are also influenced by embargoes, production levels in other
countries, domestic regulations, and other national and international political factors.
You have agreed to do an initial analysis of trends in natural gas prices. To accom-
plish this, you have been given monthly data on natural gas prices from January of 1984
to December of 1991. The numbers represent average wellhead price measured in dol-
lars per thousand cubic feet. The data are in the file named GASPRICE.
The Energy Information Administration (EIA) maintains a website of the most current
data on natural gas prices. This site is constantly updated as earlier published values are
replaced with actual prices, and revisions are made to historical data. The latest data may
be located by reviewing the EIA home page www.eia.doe. gov/emeu/mer/contents.html.
The accuracy of your forecasts may depend on having the most current data available.
Construct a model or other statistical representation of natural gas prices and esti-
mate prices for the first three months beyond the last month in your data. Prepare a re-
port on your methodology and forecasts for presentation at a meeting next week.

Earlier data in this case were collected and analyzed by Thomas G. Funk. He found the data in the
Historical Monthly Energy Review 1973-1988 and the Monthly Energy Review, both by the Energy
Information Administration, Department of Energy. Subsequent updates were located in the website
indicated in the case.
I-87
 I-88 CASE 46 NATURAL GAS PRICES

DATA SET
Read across the row:

Oi al O7 64 - 6:7 70 . 68 .69 BMay” - 63 a0

se .64 aac .62 64 £53 soo 51 47 42 5a,
316 a8 720 226 6 00. ~ 96, .85 30 aed 218
sg ~ Ld » 16 74 a3 ce Og 65 ~65 . 66 - 63
56 sod 64 ie 0
hoe woe 84 2 #0 59 “2 sD wOG
02 Be 08 he «16 209 99 . 81 .69 220 ~O1 269
05 61 oe nS 66 92 aa .85 ~
5D 49 eg
48 49 coe -56 6 94 04 96 02 49 350
48 ~43 34 ~43 Soo 82 39 .00 aie 26 ao
42 Sod 02 23D 84 92 70 43 OF “03 93
.00 06 ve -98 299 04 HO 02 203 PLS - 86
aI6 ~82 0 - 00 105 eA 190 94 «0 «85 -98
Oe 48 -47 B24 OD 58 -43 43 Aue 54 62
.84 5 tis
eer
er
Se
SN 2o9 oS) .08 eh .08 25 <0 94
- 85 PRPRPPPP
PENN
PRPPNNFPRPRP
NNN
NPPRPNRPRPRPRPRPEHE . 60 ee.
“210 WwW OD OD NPPNPRPRPERER
pal’ WNHORPPENFPRPERPHRPHRPENNND LD weWN
eS04
&OMA
SSS 16 ND
PNUPENPRPRPRPERENND . 10
NNPPNNFPRPRPPRP
NNPPNNNEFPRPRPRPRPEPEND
PNPRPPRPRPPRPRPRPENND
ES
 47
WESTAR BEVERAGE SALES

The Westar hotel, a 250-room business hotel located in a resort area of the Colorado
Rockies, has two lounges on its premises: the Pour Nous Lounge, a pub-style lounge
that seats about 50, and Gaieté, a 200-seat nightclub.
Jane Bartley is a staff analyst for Westar’s corporate parent, charged (among other
things) with predicting and analyzing beverage sales (liquor, beer, and wine, primarily)
for the chain’s business hotels. In fact, her deadline for producing some sort of forecast
for Westar beverage sales was the end of the week. After years in the business, Jane had
seen a lot of approaches to forecasting beverage sales, but none that struck her as par-
ticularly helpful. Most of them used only room occupancy as a basis for forecasting.
Not that it was unreasonable to consider room occupancy: clearly, the number of
people staying in the hotel might well influence beverage sales. Further, hotel occu-
pancy was (at least to some extent) predictable, as people made advance reservations,
so it could be used for forecasting. In the case of the Westar, though, Jane suspected
other things were involved, and the existing corporate forecasting methods (such as
they were) made no allowance for them. Specifically Jane wondered if season, promo-
tional food offered during happy hour, and the presence or absence of a piano player in
the Pour Nous Lounge might be factors to consider. She was able to gather data for the
most recent 30 months on some relevant variables (described on the next page). The
data are available in file WESTAR.
The promotional food had been discontinued after month 12 and the piano player after
month 13 as the previous management attempted to cut costs. Data from months 14 and
15 were unavailable because a change in ownership about then had led to some changes
in record-keeping techniques (and the subsequent unavailability of certain records).

This case is based upon areal situation, described to us by Gail M. Derby, who provided the data.
The names of the hotel chain and the lounges have been disguised.

lI-89
 I-90 CASE47 WESTAR BEVERAGE SALES

Jane suspected that with these data she could develop a better understanding of the
factors that really influenced beverage sales at the Westar and perhaps even influence
her department to consider some new and better ways to do the forecasting.

DEFINITIONS
MONTH Month number (month 1 was a January)
BEVTOTAL Total beverage sales (in $), pub and nightclub
BEVPUB Beverage sales (in $) for the Pour Nous Lounge
BEVNITE Beverage sales (in $) for the Gaieté night club
FOODPUB Amount spent on food (in $) for the Pour Nous Lounge
FOODNITE Amount spent on food (in $) for the Gaieté night club
ROOMS Total room occupancy for the Westar

DATA SET
BEVTOTAL BEVNITE FOODPUB ROOMS
MONTH BEVPUB FOODTOT FOODNITE
i 24443 7254 ESO BEAT) Zao 2107 4154
2 22924 72313 15686 2341 200 2141 4457
3 2750 8150 USS SAL 3559 230 S123 OUR
4 ZIZS 2 8081 AOL Bue) Gs 210 2586 4456
5 21447 7OnL 14436 2508 220 2230m aS
6 20404 6076 14328 2674 220 2454 5383
rf ESO Ei 8001 11050 2381 210 2S Ot)
8 1927315 8636 10599 2 ial 2:30 21747 95030
g) 21498 9064 12434 1624 220 1404 5372
10 26619 8760 IN. 8159 S52 210 1642 4889
ata 25600 71794 17806 1676 Zi) TA5On 4295
ale, BOSS Saya 22202 2706 220 2486 3425
13 23483 8186 Lou, ZA TT 0 ZAG S393
1 4 * * * * * * *

all 5 * * * * * * *

16 26004 7703 18301 2465 0 2465 .4395

le, 19624 6764 12860 2183 0 2183 4524
18 20459 6524 139135 3589 0 3589 5044
Ug) 20956 4209 16747 3010 0 3010 5174
20 29754 7084 22670 32115 0 S27 5eeosG
Bik 29098 Spal 2 3ulhlal 3225 0 8225) 47/06
ae 30559 Don9 24680 3448 0) 3448 4259
23 BOs 5882 PAS MENS 3.8315 0 3835 4284
24 33:833 4440 29393 3785 0 3785 3364
25 30095 6935 23160 3198 0 3193 4250
26 28163 4421 23742 AG5i, 0 2657 ee as42
aul 34051 5561 28490 3054 0 3054 5006
28 BOn39 6033 27106 293.0 0 2930 4761
AS) 29468 5180 24288 wGiou) 0 1657 3860
30 28896 6222 22674 2407 0 2407 5300
 48
PRICE-FIXING CASES

Price-fixing (when two or more companies conspire to fix the price of a product or ser-
vice rather than let it be determined by normal market forces) violates a number of fed-
eral laws in the United States, in particular, section 2 of the Sherman Act. Findings of
guilty in such cases can lead to severe fines and other criminal penalties for the firms
involved. One factor that influences the number of such cases in any given year is the
actual number of price-fixing conspiracies, of course, but others include the zeal with
which the Department of Justice prosecutes such cases and how it interprets both the
law and the evidence. In a sense, then, we may think of the number of cases of price-
fixing as the output of a complex process, a process by which society spawns such con-
spiracies and the government investigates and prosecutes them.
The data below are the number of price-fixing cases for each year from 1961 to
1987. This covers the administrations of Kennedy and Johnson (Democratic), Nixon
and Ford (Republican), Carter (Democratic), and Reagan (Republican). The data in-
clude only those cases that resulted in either guilty findings or pleas of nolo contendere
(which can be considered equivalent to guilty pleas for our purposes here). Each case
is listed under the year in which an indictment was returned.
Do these data support the claim that the Reagan administration (1981-1988)
adopted a more lenient policy with regard to prosecuting price-fixing cases than had
been the case in previous administrations?

This case is based on data provided by E. W. Eckard, derived from the reports in the Commerce
Clearing House Trade Regulation Reports.
I-91
 II-92 CASE48 PRICE-FIXING CASES

DATA SET
YEAR CASES YEAR YEAR. CASES YEAR CASES
PY Gal 1968 1975 10 1982
1962 1969 1976 et 1983
1963 1970 1977 ce 1984
1964 1974 1978 1985
1965 LOW 1979 1986
1966 1973 1980 14 1987 Wo
wo
NImn

1967 WO
SP 1974
NuPN 1981
 49
PROPERTY CRIMES

You’ ve been asked to undertake a study of the determinants of property crimes in the
United States for presentation at an upcoming meeting of business executives and com-
munity leaders. Your task is to provide evidence, for or against, common perceptions
about property crime. Are crime rates higher in urban than rural areas? Does unem-
ployment or education level contribute to property crime rates? How about public as-
sistance? What other factors relate to property crimes? The file names PCRIMES may
help answer some of these questions. Use these data to prepare a report on the charac-
teristics of and determinants of property crimes in the United States.

DEFINITIONS
CRIMES Property crime rate per hundred thousand inhabitants (property
crimes include burglary, larceny, theft, and motor vehicle theft);
calculated as # of property crimes committed divided by total
population/100,000
PINCOME Per capita income for each state
DROPOUT High school drop out rate (%, 1987)
PRECIP Average precipitation in inches in the major city in each state
over 1951-80

The data for this case come from a variety of U.S. government sources, including: the 1988 Uni-
form Crime Reports, Federal Bureau of Investigation; the Office of Research and Statistics, Social
Security Administration; the Commerce Department, Bureau of Economic Analysis; the National
Center for Education Statistics, U.S. Department of Education; the Bureau of the Census, Department
of Commerce and Geography Division; the Labor Department, Bureau of Labor Statistics; and the
National Climatic Data Center, U.S. Department of Commerce. The data set was originally collected
by Louis J. Moritz, an operations manager.

I-93
 I-94 CASE 49 PROPERTY CRIMES

PUBAID Percentage of public aid recipients (1987)

DENSITY Population/total square miles |
KIDS Public aid for families with children, dollars per family
UNEMPLOY Percentage of unemployed workers
URBAN Percentage of the residents living in urban areas
STATE 1 = Alabama 26 = Montana
2 = Alaska ‘27 = Nebraska
3 = Arizona 28 = Nevada
4 = Arkansas 29 = New Hampshire
5 = California 30 = New Jersey
6 = Colorado 31 = New Mexico
7 = Connecticut 32 = New York
8 = Delaware 33 = North Carolina
9 = Florida 34 = North Dakota
10 = Georgia 35 = Ohio
11 = Hawaii 36 = Oklahoma
«12 = Idaho 37 = Oregon
13 = Illinois 38 = Pennsylvania
14 = Indiana 39 = Rhode Island
15 = Iowa 40 = South Carolina
16 = Kansas 41 = South Dakota
17 = Kentucky 42 = Tennessee
18 = Louisiana 43 = Texas
19 = Maine 44 = Utah
20 = Maryland 45 = Vermont
21 = Massachusetts 46 = Virginia
22 = Michigan 47 = Washington
23 = Minnesota 48 = West Virginia
24 = Mississippi 49 = Wisconsin
25 = Missouri 50 = Wyoming

DATA SET
CRIMES DROPOUT DENSITY PRECIP URBAN .
STATE PINCOME PUBATID KIDS UNEMPLOY
le OOS il? 60453 05 Or. S10 ep mele eyOare dee 60.0
DAS Ste Opens See Ore 4.3 ORS MoS ae oo nr Oe 64.3
SiO Corl eS iS7/ eo Ol0 3.5 HOS ANS a al 683 Soc
AS FS) Orel lle ere Dig A GrOe 1) ere2 Wed 5S
Seo OS nino5 sme alec OFC Onl Dye Srl amlehieS Die OSs
Ono Obs / C4 ly eA Steel, SOF aoa lions 6.4 80.6
1 VAG427.2 2277/6 2138 AOS One Ome ae. SFOR a Gas
8 4347-4 17699 2920 Ao S4ieh 26F 2148 Sr FOC
SSS EMG >AGae Gr Al 22 1 ALO SS a5 84.6
IO EXONS hg 2 TA SENO) Siva) Or LOS rae 5 me Cuero bio 62.4
sl Olean) eae) ANS TAOS Csi Sh SS Sisk 36.5
WP SUS CA MAS 2S Zieh daar Ome lets 7 Sno 54.0
ae Sa ORR e Galle eZ 2) (SRE SR SOO 3 3 6.8 Sans
VA SAHOO yA nalAl S70 Sa G2 Gas ) ae! 64.2

(concludes on the next page)

 CASE 49 PROPERTY CRIMES _ II-95

-CRIMES DROPOUT DENSITY PRECIP URBAN

STATE PINCOME PUBAID KIDS UNEMPLOY
PS S019 14764 13.54 Dan) 5076 19348 33.6 4,5a Se 56
eo aha 15905 1579 3163 S005 336225 16 Bae «=GIT
ap eevee) AZ795. 932.1 Vale 93.9 204 43.6 Too 30.8
te OV4353 12193" 7384 8°.'8 SOTO SOs OoOe), T0527 67
ie -342063 14976 212 Cine 33.49 37054355 See A AU e5
e0 AS97.9 19316 23.5 See ee mes Zen eo Ae er CONS
el ea7 is 2070L* 26.0 D2 MP S2ce 586) A378 Divi l SES
@a 342.7 16387 28.6 B55) —162..2) 481 310) 2:0) OLS
Moe fee ero’ — Uae 4.6 Diol, oa te Oa ALOy 68.8
eareaeo?.,6 LO992 34.4 Lisl §55.5 .119 5258 8.4 4753
eo) 4292.2 15492" 23:,.9 DD PaO 2042-3349 Died 6 S7o
26° 4144.0 12670 15.5 4.5 Sao 862, ol iA 653° 85279
275 2866,8 15189 13.3 Sie 2059, 320. 3053 Ba Oat
28 S56/2.5 17440 18.7 220 OF One) mamma en ies
2OM Seo. 119016 “254 ei eZ Oe Bae oO Zk | 2a
20 R47 12.5 21882 2053 B08 103.3)OF eS Oat Wade OOS. Fee
Bie S046. AAS 326.8 Ou 26 a2 Sone Or WMedted:
Sear oe 2 0P 92 99 3373 8.00 378.0 523 39753 A. 8456
33 4360.4 14128 30.9 DOO eeeeee Sa.0 (Azo
342665.2, 12720 11.6 Sat 956. 3o0, 31554 4.8 48.8
35. 419352 15485 20.8 Tsp Zoe. 2298" 3728 Gol WSs
36) 915456 13269 24.2 4.8 BT wn 3 D\0) 09 ba Oy 25
SI eo LAGS2 29.2 4.0 23.8 348.3 [4 58 O19
So e2odA) 12168 18.9 6.1 -267,4 347 40.0 5 a0
39 4007).5 6793 28.0 6.0 940.9 “450 .41.9 3g JS dome
AQ 4671.2 12/64 - 32.2 OrSee A oe Some orlen© 4.5 D4)
AME ZA OY Sr LA) 5 Sie. ora 9 he ZI0S 5 39 ee Ope
BO SI 30. 13009) 324.8 654. 116.9 155. 4325 DO otes sOU DA
43 136 D.1. 14640 34.1 4.5 64-3: 169° 42...) eS UIES
A535 57, 9 L201. Ny Boe IMO Ay Vs ESS) 4.9 84.4
AS e098. 215382. 17.3 yo 60,1 (469 337 2c0 S28
46.0538.) )25 17/6040 23.4 AO. eS ero oD ie 2 BO Cy)
Ai me646.0 160569 4 21.9 Nae 69.9 443 38.6 Oiedug ho
ASP eg LOT eV L658. 2Aad Si W168. 238,40 07, OF a 3 Gis
49 3757.6 15444 16.3 a Bo.) ATS. (30.9 4.3 642
DOL esos eel 38. 620 aa Je Aeo 3.03. 703-3 Oi3g Oe ecial
 : ; ae
ine 7 me): ea tid
ae
. Beil.” Bakang veRTCR AL Daa He MORE a
eS. an
; ie ; * { ~ } oy ® i
= iw ; E \

1 re

| of idm ax: envi YT ToMuaer er : . : mi or

eek r ieee. alone 2 Sees Pl vig dane epic lc hart
Ge Be Buk * roe e : on onoa
os
- "8. ‘i ss. WINGMPLO Yoe
eee staid er
7 < a G2. oe, eo: BR AR <,&¢ yi , Mew ees
mee ee watt or SH mee Tees
; 7 a, ea ‘OTe ‘e, hE

ee Ce “Sysaver Lines ’
iy Oth O82 ¥.Sey Orem, 10r0E pipetics
Guth £es 6.kar SOME Leet? a
Sar rel sg.sc (eeomeeer- cata c eee Me”
@.te eff 2.22 PEt RECO S Tetee =
SB eptien
» Bat HOT 9.20 §ebei2 Li -gBagh Sigealimi, -
Kt sae ce | Ce sued OCA Orem Pees 0 ae
Coe cE «0c eergicr secre er ee Ae
£5" ats 8 € eteiretids « Oberroe Roe
a.ef TOP T.o8. PP poe 2s bene a
Bich Tce 8. tfotis aiiieeysc o3$ Pepagigio! “Te
See \ast-¥. teeta peaset oohpade felted = 9p
Ret mta:.0 gee leeiiew ch 8 oie og pat areitien oo
aw #2 ctr [re ciesoe beenyen gery ing ree ee
bal OSc 2.e - oorhopy- o¢ier eae eee
Bit, 446) v2 Veer ouppog ekAe
239d et” C=
yt a ee ee ee ee ee | 32 7
DAE O° 4.05. Sh rewveet seept tee re
0.08 tht bad ticeMenphboemstcre Sy ‘sa
Fk 62h 2.080 “tow rind -_ eget eokleges. eee
SAS aed _t shi st x Tiegh. haf cS a2 eet Gee a
Rot Str ML p.2 - Seco erler) rah tes eer a
xe | £86 221 6.345 Seca ei Roe esp 7:
tS) Giks. CAL £068, 7.8 £80 “Gbabl £2080 oes bis
. #4 ea tab cht 2.0% a a. Ret Logt 2, 2b! tegen
Bene) t1th *36 1.00 A i.Ti~. CHGl S-MCOk, Se es
an : Ree. Te ice Sees eee an
: 4 be
eer
CORO RS a WHO a RE

ee| eea aes

==
era se
4 SEO) OU. 8h Maaa sh o. cur
ca i. 258 nbe ay 9 Cane ReA
e

j
FAN, a4pA - FUR sO8 2572 Oo
~ a>
» ® 4
ye SORE e : : 7 v2 4,
a
ee
5 ~ i iS
A> A ' . ‘
ic® ; > ay
se wae

’ q
» i . ae a
a> {a
cs 44
- 5 7 . 7

Boos:
wa
@ weFg
3°Nel,
Od
-~.
=
 00
ROCKVILLE PHYSICIANS’
GROUP

In late November, Dr. Eaton sat down to review the results of a study of cesarean sec-
tions (C-sections) at the Rockville Physicians’ Group (RPG). He knew there were in-
teresting findings in the data gathered over the past year, and he wanted to understand
those findings. For example, he had discovered that the C-section rates differed dra-
matically among the various RPG physicians, from about 14% of all births to about
30%. He also knew, though, that how he summarized and communicated those find-
ings to the obstetrics/gynecology (OBG) staff at a meeting scheduled for mid-January
was probably just as important as the results themselves, at least in terms of getting
them to “buy in” to the ideas of quality monitoring and measurement that he wanted to
establish at RPG.
RPG was founded in the early 1950s as one of the earliest group practices in its ge-
ographic region. Located 30 miles from a major city, Rockville was at that time per-
haps best known as an attractive suburb and as the home of a state university. Since
then, it had grown enormously to about 100,000 residents, plus more than 10,000 stu-
dents. Other industrial and governmental offices were located in or near Rockville and
a major hi-tech firm had its laboratories in a university-sponsored research park near
town. Thanks to its attractive scenery, generally liberal attitudes, and a far-seeing (ag-
gressive?) no-growth policy, Rockville had become an affluent (exclusive?) city, and
RPG had prospered along with it. It was easily the largest such practice in the Rockville
area and had contracts and arrangements with most major insurance and reimburse-
ment agencies, various health maintenance organizations, and others. In all, 50 physi-
cians worked at the center. Total other staff ran about 185.
Dr. Eaton had joined RPG more than 30 years earlier, after leaving the armed services,
and practiced internal medicine there until recently, when he retired from active practice

This care is based on a real situation, though the names of the organization and the individuals
have been changed. The situation was described for us by John S. Avery, M. D.
I-97
I-98 CASE50 ROCKVILLE PHYSICIANS’ GROUP

and took an (approximately) half-time job as RPG’s medical director. As one of his first
major forays into administration, he had played a major role, including negotiations with
contractors and various regulatory agencies, in the design and construction of a building
expansion for the group. He had been involved in contract talks with some of the insur-
ance and other reimbursement agencies, and his interest in computers and data manage-
ment had grown in the last few years. He had discovered he enjoyed administration, and
felt he’d been reasonably successful at it. ;
There had been a quality assurance committee for some years at RPG, chaired by
one of the physicians. Historically, topics for study had been selected in a more-or-less
democratic manner by the physicians involved, but those topics had tended to be ad-
ministrative rather than medical (studies of waiting times, telephone systems, and so
on). Dr. Eaton felt the Group needed to be systematically involved in quality monitor-
ing of the medical aspects of their operations, beyond the usual kinds of case review
sessions they routinely held. In his words, he was
hoping to build into our quality assurance efforts . . . these [attitudes and philosophies]:
¢ Focus on the center of a quality/frequency bell curve, moving the curve to the right,
rather than on the detection of instances of poor care, lopping off the left tail of the curve.
¢ Encouraging self-examination, becoming aware of what our patterns of practice actu-
ally were. Stepping back and looking at ourselves. Escaping habits of which we were
scarcely conscious.
¢ Encouraging a research attitude, being quantitative and systematic in looking at our-
selves. Creating the possibility of publication.
¢ Having the courage to be independent; learning from ourselves, not being bound by the
received opinions of the day.
I saw this as a way to enhance the fun of practice as well as the quality of care, to lessen the
insidious onset of staleness and routine that you see in all professions.

Accordingly, the ground rules of the committee were changed to require that the top-
ics be restricted to the study of the patterns of care and the use of the results in medical
decision making. This year, after some discussion with various people involved, the com-
mittee decided to focus on the OBG department, and particularly the study of C-sections.
Dr. Eaton set up a small team working under his supervision to gather and process data.
The report from this study was to be presented at the mid-January meeting.
Cesarean section rates in the United States have been running about 20% to 25% of »
births in recent years. This rate is a concern to many in the medical community, both
because of the costs involved and because it is dramatically higher (double, in many
cases) than the rate in the rest of the developed world. There is perhaps some evidence
that it is influenced by fear of clinical failure and/or malpractice claims, but Dr. Eaton
didn’t know specifically whether that or any other identifiable influence was at work
in the case of the medical center, whose rates had been around the national averages in
recent years.
Working with Mary Spiegel, an assistant, he identified all births, including C-sections
and vaginal deliveries, from October 1 to September 30. From the computer system used
for administrative and billing operations, they were able to identify the following infor-
mation for all the cases.
¢ Chart number.
¢ Procedure code.
* Doctor number (six different physicians were involved).
¢ Date of service.
 CASE 50 ROCKVILLE PHYSICIANS’ GROUP _II-99

¢. Charge.
* Financial code (type of insurer, or none).
¢ Diagnosis code.
,In addition, they identified a sample consisting of all C-sections (178 of them) and 176
randomly chosen vaginal deliveries for more detailed analysis. For the sample of 354
patients, charts were pulled and the following further information was recorded.
¢ Maternal age.
° Parity.
¢ Gravida status.
* Immediate and underlying reasons for the C-section.
¢ Potential reasons for a C-section (this refers to cases in which a C-section was not
performed, but where some mention was made of a condition potentially leading to a
C-section).
¢ Time in labor.
¢ Apgar-1 and Apgar-5 Scores (a measure of general condition at age one minute
and five minutes for the baby).
¢ Gestational age.
¢ Birth weight.
¢ Maternal death (Y/N).
¢ Fetal death (Y/N).
While it was a time-consuming exercise, there had been no particular difficulties in
obtaining the information, at least none that Mary (with occasional consultation with Dr.
Eaton) had been unable to resolve. The process had been greatly facilitated by the exis-
tence of fairly standard (nationwide) forms and coding schemes for OBG procedures.
The combined data had been entered into a database package for a personal com-
puter. A sample page of direct printout from the program is attached (Exhibit A). Us-
ing the database package, it was possible to select certain subsets of the data based on
user-specified criteria. For example, it was possible to ask to look at all those (and only
those) cases in which labor time was greater than two hours and the Apgar-1 score was
less than 5, if such a combination was considered interesting. It was also possible to ex-
tract data from the data base package and “send” it to spreadsheet programs or graph-
ics packages.
The Quality Assurance and Utilization Management (QAUM) committee consisted
of the RPG’s administrator, nursing supervisor, the head nurse of the area involved
(OBG, in this case), and the area physicians. All were expected to have reviewed the
agenda and materials for the meeting beforehand and everyone was expected to com-
ment on the materials or the topic generally. At the meeting scheduled for mid-January,
the topic would be the results of the C-section study. Dr. Eaton’s original goals for the
study had been basically to provide information to the staff, help them understand what
the data suggested, elicit their suggestions for changes that might be made, and obtain
their support for the monitoring and quality activities he hoped to continue at RPG.
Now armed with the results of the C-section study, he wondered how best to present
them to the OBG staff.
 II-100 CASE50 ROCKVILLE PHYSICIANS’ GROUP

EXHIBIT A SAMPLE PAGE OF PRINTOUT

100% indexed 44632 Records indexed

- display structure
Structure for database: C: :
Number of data records: 1017
Date of last update s 03/14/91
Field Field Name Type Width
1 CHART Character
2 PROC_CODE Character
3 DOCTOR_NUM Character
Numeric
Numeric
AV ID: Numeric
IAG_CODE Character
ASS Character
Numeric
Numeric
Numeric
REE Numeric
*# Total ## : W
-ONNNAKUNNNSUH

Command Line 3<Cs>s sRec: 1/1017

Enter a dBASE III PLUS command.

Record# CHART PROC_CODE. DOCTOR_NUM AGE PARITY GRAVIDA DIAG_CODE CLASS BIRTH_W
T GEST_AGE REASON_1 REASON_2
313560 59410 103 33 1 V222 2
2 40 0 o-
Z 351602 59500 103 39 2 V222 1
Ss 39 11 18
222600 59410 _@ 34 V222
Ss 50 0
fA 307185 59500 >N 6612
6 40 22
302185_59500_—
5S oO p 6612
0 oO o
- 0 N V222
0 0 0
pt 349135 59500 V222
0 40 24
@—249135-59500— @° 7684
0 o
-9—350727594 10- V222
0 0
Jf 349645 59410 V222
Ss 42 °
fi 335661 59410 @ V222
40 °
-
Ss
42340728 59410 Pp @ V222
0 ° °
1334283289410 @° V222
0 re) °
14—255875-59410_. V222
0 ° °
YS 342472 59410 -@ V222
8 37 )
336306 59410 V222
4 40 ° 09
(O72
0%
60
2Oe*
OF
Ol
iD
Om
£0
6)
SOs
0
@°

JY 293233 59500 V222

2 39 14. NQ
 *

ri
IG
_
- = Vee,
- - ol~
_ ‘oe
 Hi
i Irwin
Wyte TTL

Irwin/McGraw-Hill Y,
A Division of The McGraw-Hill Companies