Intermediate IBM SPSS
Advanced Statistical
Techniques for Difference
Questions
Pawel Skuza
Statistical Consultant
eResearch@Flinders / Central Library
Please note that the workshop is aimed to be a brief
introduction to the topic and this PowerPoint is
primarily designed to support the flow of the
workshop. It cannot be seen as either an exclusive
or exhaustive resource on the statistical concepts
which are introduced in this course. You are
encouraged to refer to peer-reviewed books or
papers that are listed throughout the presentation.
It is acknowledged that a number of slides have
been adapted from presentations produced by the
previous statistical consultant (Kylie Lange) and a
colleague with whom I worked with in the past (Dr
Kelvin Gregory).
Pawel Skuza 2013
Statistical
Consulting
Website
http://www.flinders.
edu.au/library/rese
arch/eresearch/stati
stics-consulting/
or go to Flinders
University Website
A-Z
Index S
Statistical
Consultant
Introductory Level
Introduction to IBM SPSS
Introduction to Statistical Analysis
IBM SPSS - Intermediate Level
Understanding Your Data (Descriptive
Statistics, Graphs and Custom Tables)
Correlation and Multiple Regression
Logistic Regression and Survival
Analysis
Basic Statistical Techniques for
Difference Questions
Advanced Statistical Techniques
for Difference Questions
Longitudinal Data Analysis Repeated Measures ANOVA
Categorical Data Analysis
IBM SPSS - Advanced Level
Structural Equation Modelling using Amos
Linear Mixed Models
Longitudinal Data Analysis - Mixed and
Latent Variable Growth Curve Models
Scale Development
Complex Sample Survey Design / ABS and
FaHCSIA Confidentialised Datasets
Pawel Skuza 2013
(1) How to
check?
(Examples with APA Style)
SPSS
Statistical Package for the Social Sciences
PASW
Predictive Analytics Software
IBM SPSS Statistics
Pawel Skuza 2013
IBM SPSS Statistics Base
IBM SPSS Regression
IBM SPSS Advanced Statistics
IBM SPSS Complex Samples
IBM SPSS Categories
IBM SPSS Exact Tests
IBM SPSS Missing Values
IBM SPSS Forecasting
IBM SPSS Custom Tables
IBM SPSS Conjoint
IBM SPSS Statistics Programmability Extension
and AMOS
(2) How to cite?
In late 2009 SPSS Inc. was taken over by IBM Company
and the software changed its official name twice over the
period of one year. From SPSS it was relabelled to PASW
(Predictive Analytics Software) and later to IBM SPSS.
Consequently, there may be books, online resources, etc.
that use either of those different names but in fact refer to
the same software.
Flinders University has licence for number of IBM SPSS products
(versions 19, 20, 21) covering following modules:
START SOFTWARE
HELP ABOUT
SPSS Inc. Released 2007. SPSS for Windows,
Version 16.0. Chicago, SPSS Inc.
SPSS Inc. Released 2008. SPSS Statistics for
Windows, Version 17.0. Chicago: SPSS Inc.
SPSS Inc. Released 2009. PASW Statistics for
Windows, Version 18.0. Chicago: SPSS Inc.
IBM Corp. Released 2010. IBM SPSS Statistics
for Windows, Version 19.0. Armonk, NY: IBM
Corp.
IBM Corp. Released 2011. IBM SPSS Statistics
for Windows, Version 20.0. Armonk, NY: IBM
Corp.
IBM Corp. Released 2012. IBM SPSS Statistics
for Windows, Version 21.0. Armonk, NY: IBM
Corp.
??? SPSS / PASW / IBM SPSS ???
IBM SPSS on Flinders University
SPSS / PASW / IBM SPSS
Pawel Skuza 2013
For details explaining various modes of obtaining access to the
software go to
Pawel Skuza 2013
http://www.flinders.edu.au/library/research/eresearch/statistics-consulting/spsslicenses-and-technical-support/licenses-for-university-and-home.cfm
Pawel Skuza 2013
Levels of Measurement and Measurement Scales
EXAMPLES:
Ratio Data
Differences between
measurements, true
zero exists
Height, Age, Weekly
Food Spending
Temperature in Celsius,
Standardized exam score
Interval Data
Differences between
measurements but no
true zero
Ordinal Data
Ordered Categories
(rankings, order, or scaling)
Service quality rating,
Student letter grades
Nominal Data
Categories (no ordering
or direction)
Marital status, Type of car
owned, Gender/Sex
MEASUREMENT
Parametric
Statistics
Nonparametric
Statistics
Dependent Variable
Approximates Normal
(Scale) Data and
Assumptions Not
Markedly Violated
Dependent Variable
Clearly Ordinal Data or
the Assumptions Are
Markedly Violated
Dependent Variable is
Nominal or
(dichotomous) Data
Compare
Example 1
Figure 4.11 from Dancey, C. P., & Reidy, J. (2004). Statistics without maths
for psychology : using SPSS for Windows (3rd ed.). New York: Prentice
Hall.
Example 2
Table from Pallant, J. (2007). SPSS Survival Manual : A step by step guide
to data analysis using SPSS for Windows (SPSS Version 15) (3rd ed.).
Maidenhead, Berkshire. U.K. ; New York, NY: Open University Press.
Example 3
Flowchart from http://gjyp.nl/marta/Flowchart%20(English).pdf
Similar ones in other resources
Pawel Skuza 2013
Selection of an Appropriate Inferential Statistics for Basic, Two Variable Difference
Questions or Hypotheses PART 2
Level of Measurement
of Dependent Variable
Selection of statistical methods
Pawel Skuza 2013
Selection of an Appropriate Inferential Statistics for Basic, Two Variable Difference
Questions or Hypotheses PART 2
One Factor or Independent Variable with 3 or
More Categories or Levels /Groups /Samples
Level of Measurement
of Dependent Variable
Independent
Samples or
Groups (Between)
Repeated Measures
or Related Samples
(Within)
Means
ONE-WAY ANOVA
GLM REPEATED
MEASURES ANOVA
Parametric
Statistics
Mean
Ranks
KRUSKALWALLIS H TEST
FRIEDMAN TEST
Nonparametric
Statistics
Counts
CHI-SQUARE
SIGNIFICANCE
TEST
COCHRAN Q TEST
Adapted from (Leech, Barrett, & Morgan, 2008, p. 74)
Pawel Skuza 2013
Compare
One Factor or Independent Variable with 3 or
More Categories or Levels /Groups /Samples
Independent
Samples or
Groups (Between)
Repeated Measures
or Related Samples
(Within)
Dependent Variable
Approximates Normal
(Scale) Data and
Assumptions Not
Markedly Violated
Dependent Variable
Clearly Ordinal Data or
the Assumptions Are
Markedly Violated
Means
Analyze Compare
Means One-Way
ANOVA
Analyze General
Linear Model
Repeated Measures
Mean
Ranks
Analyze
Nonparametric Tests
k Independent
Samples
Analyze
Nonparametric Tests
k Related Samples
Dependent Variable is
Nominal or
(dichotomous) Data
Counts
Analyze
Descriptive Statistics
Crosstabs
Analyze
Nonparametric Tests
k Related Samples
Adapted from (Leech, Barrett, & Morgan, 2008, p. 74)
Pawel Skuza 2013
One-Way Analysis of Variance
Kruskal-Wallis H test
Reproduced from (Leech, Barrett, & Morgan, 2008, p. 81)
Pawel Skuza 2013
Pawel Skuza 2013
ANOVA Terminology
ANOVA Terminology Balanced Design
Factor
Factor
Treatment
The independent variable
The treatment condition
A categorical variable
Sometimes called the cause
Levels
The categories of the factor
The treatment levels
One
Two
Three
X11
X21
X31
X12
X22
X32
X13
X23
X33
Xn1
Xn2
Xn3
X1
X2
X3
s12
s22
s32
Treatment
levels
Replicates
and their
scores on a
dependent
variable
Replicate
The subjects
Balanced
The same number of replicates (subjects) in each treatment level
Unbalanced
Different number of replicates (subjects) in each treatment
condition
Pawel Skuza 2013
Hypotheses of One-Way ANOVA
Note the type of
statistics we have
here
Pawel Skuza 2013
Hypotheses of One-Way ANOVA
Null hypothesis
Alternative hypothesis
All population means are equal
i.e., no treatment effect (no variation in means among
groups)
H o : 1 2 ... J
At least one population mean is different
i.e., there is a treatment effect
Does not mean that all population means are different
(some pairs may be the same)
It just means that at least one mean is statistically different
from the other means
Note that this hypothesis is saying that the subjects are
really drawn from the same population
So, this hypothesis is saying that at least one mean
is drawn from a different population
And there is no treatment effect
Pawel Skuza 2013
Hypotheses of One-Way ANOVA
Pawel Skuza 2013
Partitioning the Variation
Variation is a measure of spread
At least one
mean is
different
from the
other means
Of distance from a mean
ANOVA is concerned with variation
And in particular partitioning the observed variation
Total variation can be split into two parts
SST=SSB+SSW
Where
SST = the total sums of squares (the total variation from the
overall or grand mean)
SSB = the sum of squares among or between groups (the
variation from each groups mean from the overall mean)
SSW= the sum of the variation within each group
Pawel Skuza 2013
Pawel Skuza 2013
Total Sum of Squares
Total Sum of Squares
SST = SSB + SSW
SST ( X 11 X ) 2 ( X 12 X ) 2 ... ( X ni n j X ) 2
Response ,
X
nj
SST ( X ij X ) 2
j 1 i 1
Where:
SST = Total sum of squares
J = number of groups (levels or treatments)
nj = number of observations in group j
Xij = ith observation from group j
Group 1
Group 2
X = grand mean (mean of all data values)
Group 3
Pawel Skuza 2013
Among-Group Variation
Pawel Skuza 2013
Among-Group Variation
SST = SSB + SSW
SS B n1 ( x1 x ) 2 n2 ( x2 x ) 2 ... nJ ( xJ x ) 2
Response, X
SS B n j ( X j X ) 2
j 1
Where:
X3
X1
X2
SSB = Sum of squares among groups
J = number of groups
nj = sample size from group j
Xj = sample mean from group j
X = grand mean (mean of all data values)
Group 1
Group 2
Group 3
Pawel Skuza 2013
Among-Group Variation
Pawel Skuza 2013
Within-Group Variation
SS B n j ( X j X ) 2
SSW ( x11 X 1 ) 2 ...( X 12 X 2 ) 2 ...( X ni n j X J ) 2
j 1
Variation Due to
Differences Among Groups
MSB
SSB
J 1
Response, X
Mean Square Among =
SSB/degrees of freedom
X1
Group 1
Pawel Skuza 2013
Group 2
X2
X3
Group 3
Pawel Skuza 2013
Within-Group Variation
Within-Group Variation
SST = SSB + SSW
J
nj
SSW ( X ij Xj ) 2
nj
SSW ( X ij X j ) 2
j 1 i 1
j 1 i 1
Where:
SSW = Sum of squares within groups
J = number of groups
SSW
NJ
MSW
Summing the variation
within each group and
then adding over all
groups
Mean Square Within =
nj = sample size from group j
SSW/degrees of freedom
Xj = sample mean from group j
Xij = ith observation in group j
j
Pawel Skuza 2013
Pawel Skuza 2013
Partition of Total Variation
Obtaining the Mean Squares
Total Variation (SST)
SS B
MS B
J 1
MSW
SSW
NJ
MST
SST
N 1
Variation Due
to
Factor
(SSB)
Variation Due
to
Random Sampling
(SSW)
df=J-1
df=N-J
Mean Variation
Between Groups
(SSB)
Mean Variation
Within Groups
(SSC)
Pawel Skuza 2013
Pawel Skuza 2013
Partition of Total Variation
Three sum of squares or
squared deviations serve as
the backbone of ANOVA
Total sum of squares: overall, how
far are the observations from the
overall (grand) mean
Sum of squares between
For a balanced design: how far is
the group mean from the overall
mean
One-Way ANOVA Table
SStotal X ij X
J
SS B n j X j X
SSW X ij X j
Sum of squares within
For each group: how far are the
observations within that group
from that groups mean
Source of
Variation
SS
df
Among
Groups
SSB
J-1
Within
Groups
SSW
N-J
Total
SST =
SSB+SSW
N-1
MS
(Variance)
MSB =
MSW =
F ratio
SSB
F=
J-1
MSB
MSW
SSW
N-J
J = number of groups
N = sum of the sample sizes from all groups
df = degrees of freedom
Pawel Skuza 2013
Pawel Skuza 2013
One-Way ANOVA - F Test Statistic
H0: 1= 2 = = c
H1: At least two population means are different
Test statistic
MS B
MSW
The F statistic is the ratio of the among
estimate of variance and the within estimate of
variance
The ratio must always be positive
df1 = J -1 will typically be small
df2 = N-J will typically be large
Decision Rule:
Reject H0 if F > FU,
otherwise do not
reject H0
MSB is mean squares among groups
MSW is mean squares within groups
Degrees of freedom
df1 = J 1
df2 = N-J
Interpreting One-Way ANOVA
F Statistic
(J = number of groups)
(N = sum of sample sizes from all populations)
= .05
0
Do not
reject H0
Pawel Skuza 2013
Assumptions
Reject H0
FU
Pawel Skuza 2013
If assumptions are not met:
Interval data
Independent samples
Multivariate normality
Homogeneity of variances
Absence of outliers
MORE IN
More of a problem if have small sample
sizes and/or different sample sizes
Constant variance typically more important
than normality
May influence the p-value in either direction
Garson, G. D. (2012). Univariate GLM, ANOVA, & ANCOVA.
Asheboro, NC: Statistical Associates Publishers.
http://www.statisticalassociates.com/glm_univariate.htm
OR SIMILAR RESOURCES
Pawel Skuza 2013
Assumptions
Pawel Skuza 2013
Effect Size
Chiarotti, F. (2004). Detecting assumption violations in mixed-model analysis
of variance. Annali dell'Istituto Superiore di Sanita, 40(2), 165-171.
Lix, L. M., Keselman, J. C., & Keselman, H. J. (1996). Consequences of
assumption violations revisited: A quantitative review of alternatives to the
one-way analysis of variance F test. Review of Educational Research, 66(4),
579-619.
McGuinness, K. A. (2002). Of rowing boats, ocean liners and tests of the
ANOVA homogeneity of variance assumption. Austral Ecology, 27(6), 681688.
Schmider, E., Ziegler, M., Danay, E., Beyer, L., & Bhner, M. (2010). Is It
Really Robust? Reinvestigating the Robustness of ANOVA Against
Violations of the Normal Distribution Assumption [10.1027/16142241/a000016]. Methodology: European Journal of Research Methods for
the Behavioral and Social Sciences, 6(4), 147-151.
Yang, H., & Huck, S. W. (2010). The Importance of Attending to Underlying
Statistical Assumptions. Newborn and Infant Nursing Reviews, 10(1), 44-49.
Pawel Skuza 2013
Sometimes called magnitude of effect
A measure of the degree to which variability among
observations can be attributed to treatments
Measures of effect size in ANOVA are measures
of the degree of association between and effect
(e.g., a main effect, an interaction, a linear
contrast) and the dependent variable.
They can be thought of as the correlation
between an effect and the dependent variable.
If the value of the measure of association is
squared it can be interpreted as the proportion
of variance in the dependent variable that is
attributable to each effect.
Pawel Skuza 2013
Commonly Used Measures of Effect Size
Four of the commonly used measures of effect size in AVOVA
are
Eta squared
partial Eta squared
omega squared
Intraclass correlation
Eta squared and partial Eta squared are estimates of the
degree of association for the sample
Omega squared and the intraclass correlation are estimates of
the degree of association in the population
SPSS for Windows displays the partial Eta squared when you
check the display effect size option
p2
Eta Squared
Eta squared is the proportion of the total
variance that is attributed to an effect
It is calculated as the ratio of the effect
variance (SSeffect) to the total variance
(SStotal)
ssB
ssT
I
Pawel Skuza 2013
Eta Squared
In our example it would be (4716/5836)=0.8080
Pawel Skuza 2013
Eta Squared
Very easy to calculate
But it is a biased statistic
ssB 4716
0.8080
ssT 5836
This means that schools (the treatment) account for over 80%
of the variation in student achievement
Tends to overestimate the true value in the
population
Better to use the omega-squared
ANOVA
Distance
Between Groups
Within Groups
Total
Sum of
Squares
4716.400
1119.600
5836.000
df
2
12
14
Mean Square
2358.200
93.300
F
25.275
Sig.
.000
Pawel Skuza 2013
Omega Squared
Pawel Skuza 2013
Partial Eta Squared
The formula for the omega squared
statistic is
p2
The partial Eta squared is the proportion of the
the effect + error variance that is attributable to
the effect
The formula differs from the Eta squared formula
in that the denominator includes the SSeffect
plus the SSerror rather than the SStotal
where k=numbers of group,
n=number of participants
Pawel Skuza 2013
Pawel Skuza 2013
Exercise 1
Is there a difference in average students
mathematics achievement among four groups
established by fathers occupational status?
Comparing Means from More Than Two Populations
If there are J groups, then the number of
pair-wise comparisons is J(J-1)/2
! Assumptions
So, if the number of groups was 5, then there
would be ten pairwise comparisons
Exercise_1_a.sav & Exercise_1_b.sav &
Exercise_1_c.sav
Simplified data from PISA 2003 Study Australia &
Indonesia
(The Programme for International Students Assessment)
http://www.pisa.oecd.org
The probability of a Type I error is a single pairwise
comparison is
And the probability of one or more type I errors within
the set is greater than
If C is the number of comparisons, then the actual error
rate is
p 1 1
Pawel Skuza 2013
Probability of a Type I Error
Probability of Type I Error (=0.05)
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1
11
16
21
26
31
36
41
46
51
56
Pawel Skuza 2013
Post Hoc Tests
When a significant F has been obtained and the
factor (or treatment) has more than two levels,
post hoc tests can be used to determine which
particular groups differ significantly from one
another
Multiple comparisons procedures are methods for
identifying differences among the group means
once the hypothesis of overall equality has been
rejected
NUmber of pairwise comparisons
Pawel Skuza 2013
Differences Between Post Hoc Tests
Post hoc tests differ is in the degree to which
they control experimentwise error or
familywise error
Familywise error refers to refers to the probability
that a family of tests will produce any Type I errors
For our example in which we looked at whether school
(00008, 00009, 00018) was associated with
mathematics achievement, the family of tests would
refer to those that test the differences among the
schools
Pawel Skuza 2013
Pawel Skuza 2013
Differences Between Post Hoc Tests
The greater the degree to which a post hoc test
controls for familywise error, the more
conservative it is said to be
All other things being equal, the more
conservative a post hoc test is, the less power it
has
Making decisions about which type of post hoc
test to use often involves a trade-off between
Type I error and power
Pawel Skuza 2013
Bender, R., & Lange, S. (2001). Adjusting
for multiple testing - When and how?
Journal of Clinical Epidemiology, 54(4),
343-349.
Page 374 from Field, A. P. (2009).
Discovering statistics using SPSS : (and
sex, drugs and rock 'n' roll) (3rd ed.). Los
Angeles: SAGE Publications.
Pawel Skuza 2013
Exercise 2
Is there a difference in average time students allocate
on mathematics among groups established by
students self expected occupational status?
! Assumptions
Exercise_2.sav
Simplified data from PISA 2003 Study Australia &
Indonesia
(The Programme for International Students Assessment)
http://www.pisa.oecd.org
Pawel Skuza 2013
Analysis of covariance
Analysis of covariance
Pawel Skuza 2013
Applications
an extension of ANOVA in which main
effects and interactions are assessed
on DV scores after the DV has been
adjusted for by the DVs relationship
with one or more Covariates (CVs)
Pawel Skuza 2013
Analysis of Covariance
Major applications
Increase test sensitivity (main effects and
interactions) by using the CV(s) to account for
more of the error variance therefore making
the error term smaller
Adjust DV scores to what they would be if
everyone scored the same on the CV(s)
This second application is used often in nonexperimental situations where subjects cannot be
randomly assigned
Pawel Skuza 2013
Combination of both categorical (factors)
and continuous (covariates) predictors
of a continuous dependent variable
Covariates
Confounders: nuisance variables that are
(linearly) related to the dependent variable
First remove the effect of the confounder(s)
and then test the effect of the factors on the
dependent variable
Pawel Skuza 2013
Analysis of Covariance
Default ANCOVA model:
All interactions between factors (as in
ANOVA)
Main effects of covariates
No interactions between factors and
covariates
No interactions between covariates
Assumptions
Typical ANOVA assumptions still hold
In addition there are few more:
(a) Linearity is assumed that each CV has
a linear relationship with the DV and other
CVs
Pawel Skuza 2013
Pawel Skuza 2013
(b) Homogeneity of Regression
Assumptions
(b) Homogeneity of Regression
the slope of the line predicting the DV from
the CV should be the same for each level
of the IV.
In other words the regression coefficient
(B) relating a CV to the DV should be the
same for each group.
Pawel Skuza 2013
Assumptions
(c) Reliability of Covariates
it is assumed that each CV is measured
without error (this is unrealistic).
Pawel Skuza 2013
Exercise 3
Is there a difference in average students
mathematics achievement among groups established
by fathers occupational status while controlling for
disciplinary climate in the classroom?
! Assumptions
Exercise_3.sav
MORE ABOUT ASSUMPTIONS IN
Garson, G. D. (2012). Univariate GLM, ANOVA, &
ANCOVA. Asheboro, NC: Statistical Associates
Publishers.
http://www.statisticalassociates.com/glm_univariate.htm
OR SIMILAR RESOURCES
Pawel Skuza 2013
Simplified data from PISA 2003 Study Australia &
Indonesia
(The Programme for International Students Assessment)
http://www.pisa.oecd.org
Pawel Skuza 2013
10
The Design: An Example
Suppose we had
Assignment to three schools as one factor
And gender (sex) as another factor
We now have a much more complex ANOVA
design
Effect of school
Effect of gender
Effect of any interaction between school and gender
Two-way ANOVA
An interaction example would be when girls prefer one
school over another because of some feature of that school
Pawel Skuza 2013
The Design: An Example
Treatment: Schools
Treatment: Gender
Female
Male
7/10/2013
One
Two
Three
X111
X211
X311
X121
X221
X321
X131
X231
X331
Xn11
X112
X212
X312
Xn21
X122
X222
X322
Xn31
X132
X232
X332
Xn12
Xn22
63 X
n32
Pawel Skuza 2013
Factorial Design:Two-Way ANOVA
Examines the effect of
Two factors of interest on the dependent
variable
e.g., Percent carbonation and line speed on
soft drink bottling process
Interaction between the different levels
of these two factors
e.g., Does the effect of one particular
carbonation level depend on which level the
line speed is set?
Pawel Skuza 2013
Pawel Skuza 2013
Two-Way ANOVA Sources of Variation
Two-Way ANOVA
Two Factors of interest: A and B
Assumptions
r = number of levels of factor A
Similar to One-way ANOVA
c = number of levels of factor B
MORE IN
Garson, G. D. (2012). Univariate GLM, ANOVA, &
ANCOVA. Asheboro, NC: Statistical Associates
Publishers.
http://www.statisticalassociates.com/glm_univariate.htm
OR SIMILAR RESOURCES
Pawel Skuza 2013
n = number of replications for each cell
n = total number of observations in all cells
(n = rcn)
Xijk = value of the kth observation of level i of
factor A and level j of factor B
Pawel Skuza 2013
11
Examples:
Interaction vs. No Interaction
Two-Way ANOVA Sources of Variation
SST
Total Variation
n-1
SSB
Factor B Variation
c1
SSAB
Variation due to interaction (r 1)(c 1)
between A and B
SSE
Random variation (Error)
rc(n 1)
No interaction:
Interaction is present:
Factor B Level
1
Factor B Level
3
Factor B Level
2
Factor B Level
1
Mean Response
SSA
Factor A Variation
Degrees
of
Freedom:
r1
Mean Response
SST = SSA + SSB + SSAB + SSE
Factor A Levels
Factor B Level
2
Factor B Level
3
Factor A Levels
Pawel Skuza 2013
Interaction
Pawel Skuza 2013
Main effects
If interaction is present then main effects can no
longer be interpreted
Decomposing significant interactions
Profile plots
Parallel lines => no interaction
Simple main effects
Pair-wise comparisons of the levels of A, within each
level of B
Specify reduced set of comparisons of interest if
possible
Significant main effect => not all levels
of the factor have the same response
If more than 2 levels, post-hoc tests are
needed to determine which levels are
different to which others
Adjustments for multiple comparisons
All pairwise tests (Bonferroni, Tukey etc)
Comparison to control (Dunnett)
Pawel Skuza 2013
Higher-way ANOVA
n-way => n factors
Full factorial model => all main effects,
plus all 2-way, 3-way, , n-way
interaction terms
If higher-order terms are non-significant
may want to remove from model to give
greater power to estimates of the lowerorder terms
Pawel Skuza 2013
Exercise 4
Is there a difference in average students reading
achievement among males and females who are
from two different types of family structure?
! Assumptions
Exercise_4.sav
Simplified data from PISA 2003 Study Australia &
Indonesia
(The Programme for International Students Assessment)
http://www.pisa.oecd.org
Pawel Skuza 2013
Pawel Skuza 2013
12
SPSS BOOKS (Hard copies)
Chapters 7,8, 10 in Allen, Peter James, & Bennett, Kellie. (2012). SPSS
statistics : a practical guide : version 20. South Melbourne, Vic.: Cengage
Learning Australia.
Chapters 19, 25 in Argyrous, George. (2011). Statistics for research : with a
guide to SPSS (3rd ed.). Los Angeles: Sage.
!!! Chapters 5 & 6 in Landau, Sabine, & Everitt, Brian. (2004). A handbook
of statistical analyses using SPSS. Boca Raton: Chapman & Hall/CRC.
SPSS BOOKS (Online copies)
Hard copies and online versions
!!! Chapters 18 & 19 in Pallant, Julie. (2010). SPSS survival manual a step
by step guide to data analysis using SPSS (4th ed.). Maidenhead: Open
University Press/McGraw-Hill.
Chapter 10 in Morgan, George A. (2011). IBM SPSS for introductory
statistics : use and interpretation (4th ed.). New York: Routledge.
!!! Chapters 8 in Leech, Nancy L., Barrett, Karen Caplovitz, Morgan,
George A., & Leech, Nancy L. (2011). IBM SPSS for intermediate statistics
: use and interpretation (4th ed.). New York: Routledge.
!!! Chapters 7,8 in Kinnear, Paul R., & Gray, Colin D. (2009). PASW
statistics 17 made simple (replaces SPSS statistics 17). London ; New York:
Psychology Press.
Online versions
!!! Chapters 10,11,12 in Field, Andy P. (2009). Discovering statistics using
SPSS : (and sex, drugs and rock 'n' roll) (3rd ed.). Los Angeles: SAGE
Publications.
Chapter 9 in Bryman, Alan, & Cramer, Duncan. (2011). Quantitative data
analysis with IBM SPSS 17, 18 & 19 : a guide for social scientists. Hove ;
New York: Routledge.
Chapters 9 & 19, 22 in Noruis, Marija J. (2008). SPSS 16.0 [or later
versions] Statistical Procedures Companion. Upper Saddle River, NJ:
Prentice Hall.
Chapters 10,11,13 & 14 in Larson-Hall, Jenifer. (2010). A guide to doing
statistics in second language research using SPSS
Pawel Skuza 2013
BOOKS More Theoretical Level
Tabachnick, B. G., & Fidell, L. S. (2007 or later). Using
multivariate statistics (5th ed.). Boston: Pearson/Allyn
& Bacon
Quinn, G. P., & Keough, M. J. (2002). Experimental
design and data analysis for biologists. Cambridge:
Cambridge University Press
Whitlock, M., & Schluter, D. (2009). The analysis of
biological data. Greenwood Village, Colo.: Roberts
!!! Levine, Gustav, Page, Melanie C., Braver, Sanford
L., & MacKinnon, David Peter. (2003). Levine's guide
to SPSS for analysis of variance (2nd ed.). Mahwah,
N.J.: L. Erlbaum
Pawel Skuza 2013
SPSS Online tutorials and resources
(!!! Please keep in mind that usually online resources are not academically peer
reviewed. Despite many of them being of high quality as well as being very useful
from educational point of view, they shouldnt be treated as a completely reliable
and academically sound references)
- Statnotes: Topics in Multivariate Analysis, by G. David Garson
http://www.statisticalassociates.com/
- UCLA Institute for Digital Research and Education - SPSS Starter Kit
http://www.ats.ucla.edu/stat/spss/sk/default.htm
- Getting Started with SPSS for Windows by John Samuel, Indiana
University
http://www.indiana.edu/~statmath/stat/spss/win/index.html
- Companion Website for the 3rd edition of Discovering Statistics Using
SPSS by Andy Field
http://www.uk.sagepub.com/field3e/SPSSFlashmovieslect.htm
- SPSS for Windows and Amos tutorials by Information Technology
Services, University of Texas
http://ssc.utexas.edu/software/software-tutorials#SPSS
- Journey in Survey Research by John Hall
http://surveyresearch.weebly.com/index.html
Pawel Skuza 2013
SPSS Help and Resources
SPSS has a range of help options
available
Topics
Used to find specific information
Tutorial
Find illustrated, step-by-step
instructions for the basic features
Case studies
Hands-on examples of various types of
statistical procedures
Statistics coach
To help you find the procedure you
want to use
And manuals available online http://www01.ibm.com/support/docview.wss?u
id=swg27021213
Pawel Skuza 2013
SPSS Help and Resources
Online SPSS FORUMS
(!!! Please keep in mind that usually online resources are not academically peer
reviewed. Despite many of them being of high quality as well as being very useful
from educational point of view, they shouldnt be treated as a completely reliable
and academically sound references.
!!! Suggestions / Guidance found on forums should be especially
treated very doubtfully, yet they may point to more reliable academic
resources and be somewhat of help.
that is endorsed by IBM SPSS
http://www.listserv.uga.edu/archives/spssx-l.html
Other forums
http://groups.google.com/group/comp.soft-sys.stat.spss/topics?gvc=2
http://www.spssforum.com/
Pawel Skuza 2013
Pawel Skuza 2013
13
THANK YOU
Please provide us with your feedback by
completing the short survey.
Pawel Skuza 2013
14
