BUSINESS STATISTICS A Textbook for B.Com, Semester Ill, Paper DSC- 3.1 & B.Com. (Hons.), Semester IV, Paper DSC- 4.1 As per latest UGCF-2022 Pattern Based on the National Education Policy - 2020 Dr. J. K. Thukral Former Associate Professor Department of Mathematics Shri Ram College of Commerce University of Delhi - 110007 JKT PUBLICATIONS Delhi Scanned with CamScanner‘A hts reserved No ato this bok ay be wpebued, dard 7 Real SGA fanart in any fom by any ras. electons mecants! pace nan tverwse, witout te prio writen permission om De pubtanon Maiog: Te cing an unastorisd ctn reatin fo weep wrk may ut bm ied ‘ain for damages and emia! prosecutor, meetin Spectat Hote :Phaccopy of Kerang of eductoe! boots witha hewmen pemieson of publisher i legal end against Copyright Act. SGenaral: Whe every otha baen made fo presen ute ifomaten ands oor, te uth isnot sponsible forthe consequences of any ation taken on tw bose ane po ©Dr. J. Thukral Published by : JKT Publications Second Revised Edition : July 2024 ISBN : 978-81-965747-1-0 Price : 7599 Printed in India SYLLABUS BUSINESS STATISTICS B.COM. SEMESTER, PAPER DSC - 3.1 B.COM, (Hons.) SEMESTER-IV, PAPER DSC 4.1 LATEST UGCF - 2022 PATTERN UNIT I: Descriptive Statistics Measures of Ceatal Tender y Cones and roperies af averages inci ‘median end mode, estat mean, ‘Measures of Dispersion: An overview of age, quar deviation an mean deviation ‘Standard deviation; Variance ad coeticient of varation ‘Monens: Compusion end significance; Skewnes; Kurcis UNIT IL; Probability and Probability Distributions ‘Theory and Approaches to Probbily: Probability Theorems: Adan md Muliglicaton; Conon ‘robabilty and Bayes" Theorem. ‘Expectation ad varinceof random variable, DisseteProbbilty Distbuons : Binomial and Poison (Properties ant Applicators) ‘Nonnal Discibcon: Properties of Nonnal Curve; Compusion of Pobailities and Applicators UNIT KI: Simple Correlation and Regression Analysis Conelaton Analysis: Meaning and yes of eotelaten; Conlin Vs Cason ; esos Cooter: of Coneaton (eraptation and properies; Probable and Standard errs; Rankcoreton Regeason Arasis: Precip of ast squares an reyestoa ines; Regression eguxins an eration Propertis of repesion coer; Relationship beween Coan ard Regression soclfcere Standard Enorof Estimate. UNIT IV: Time Series Analysis ‘Time eres data Componens of time series; Acie and Multiplicative modes ‘Trend analysis: Firing of tend using principe of least squares linear and second Shifting of Ong and Conversion of annul linear vend equation w qutely Vioe-vers, UNIT V; Index Numbers Meaning and wes of index uber Construction of inex numbers: Methods of Lazperes, Paasche and Fisher's Wea index Construction ané Us of Consus Price Indices; BSE SENSE, and NSE NIFTY. ce rarbe'a vy basis and ‘Scanned with CamScanner"asd op fb 60 ay Be rapa, ond 9 ai lon oF Mei oy tom by a8) eats, OHO, Pecan Robey tin OF a cle rater pomaton fa he pas tuning: Te dag fo Uitones ath ono coy wok yr bot Sr mage fd cil psnen Spec! nt Pty caring of war Dots tod he wn pein of puch Bon onda pri At Caner a wary ft as ben mae pres aubalomaon a i ar be iris pet fre eoserines ary acon en oe ns of fs ok © Dr. J. K. Thukral Published by : JKT Publications Second Revised Edition : July 2024 ISBN : 978-81-965747-1-0 Price : 599 Printed in India SYLLABUS BUSINESS STATISTICS B.COM. SEMESTERIII, PAPER DSC - 9.1 B.COM. (Hons.) SEMESTER.IV, PAPER DSC 4.1 LATEST UGCR - 2022 PATTERN UNITI: Deseriptive Statistics ‘Messre of Central Tendency: Concepts and properies of everages including with mean, ‘media and mode ‘Measure of Dispersion : An overview of range, quartile deviation and meen deviation ‘Standard evnto ;Vaciance and epecien of variation ‘Moras: Computation and signifcane ; Skewees ;Kictosis UNIT 11: Probability and Probability Distributions ‘Theory and Approaches to Probability: Probability Theorems: Adon and Maliplicaton; Condition! probability and Bayes Theorem. “Expectation ad variance ofa random vrsbe Distete Probability Dishibtions -Binomial and Pisson Prope and Applicaics) ‘Noma Distibution: Properties ofNoral Curve; Computation of Probabilities and Applications, UNIT LIT: Simple Correlation and Regression Analysis (Contin Analysis: Meaning andiypes of corelstion; Camelia Vi Causation; Pearson's Cuicert ‘of Coneaton (computation and proper; Probable an Standard errors; Rank corre ation Regesion Anji: Pick oft sures and repress ine; Region exitionsand cation: Peopartis of regesion coefficients; Relitionship between Conelatioe and Regesion coticens + Standard Err af Estimate, UNIT IV: Time Series Analysis ‘Tioeseres ata: Components of ime series ;Adtve and Maliplicative mode's ‘Trend analysis: ining of tend using principle o” east squares linear and second-desseenarbo‘a, Shihing of Origa and Conversion of annual linear tend equation to qunrtety ‘mith basis aad UNIT V: Index Numbers Meaning and uses of index numbers Construction of lndex sumbers: Methods of Laspeyres, Paasche and Fisher's eal indo CConstustonand Uiity of Consumer PriceIndices; BSE SENSEX, and NSE NITY. ‘Scanned with CamScanner1 STATISTICS: AROVERVIEW LLL Meaning of Statistics 1.2 Pumetions of Statistics 41.3 Limitations of Statistics 14 Collection of Data LS Classification of Data 1.6 Tabulation of Data 17 Frequency Distribution 1.8 Graphic Presentations of Frequency Distributions ‘MEASURES OF CENTRAL TENDENCY 2.1 Introduction 2.2 Average 23 Requisites of 2 Good Average 24 Arithmetic Mean 2:5 Properties of Arithmetic Mean 26 Combined Mean 2.7 Merits and Demerits of Arithmetic Mean 2.8 Weighted Arithmetic Mean 29 Median 2.10 Merits end Demerits of Median . 2.11 Other Partition Values : Quartiles, Deciles and Percentiles 2.12 Graphic Method of Computing Medien and other Partition Values 2.13 Mode 2.14 Graphie Method of Locating Mode 2.15 Merits and Demerits of Mode 2.16 Geometric Mean 2.17 Weighted Geometric Mean 2.18 Merits and Demerits of Geometric Mean 2.19 Harmonic Mean 2.20 Weighted Harmonic Mean 2.21 Relationship Among AM, GM and HIM aa-227 2a 215 2.19 220 223 231 231 aaa 282 268 2.69 an 279 280 284 290 293 coments (ex 2.22 Merits and Demerits of Harmonic Mean 293 2.29 Choice of an average 294 3. MEASURES OF DISPERSION 31-355 3.1 Introduetion, 31 3.2 Dispersion. 32 33 Measures of Dispersion 32 34 Properties of @ Good Meesure of Dispersion 3a 35 Range 33 36 Merits and Demerit of Range 34 3.7 Quartile Deviate. or Semi Inter quartile Range 34 3.8 Merits and Demerite of Quartile Deviation 37 © 39 Mean Deviation or Average Deviation 39 3.10 Nerits and Demerits of Mean Deviation aie 3.11 Standard Deviation 3.12 Combined Standard Desiation 3.13 Coefficient of Variation 323 3.16 Mathematical Properties of Standard Deviati ass 3.15 Merits and Demerts of Standard Deviation as9 4 MOMENTS, SKEWNESS AND KURTOSIS 41-469 4.1 Inveduction 4 42 Central Moments Moments about the Mean, aa 43 Raw moments - moments about an arbitrary point 34 44 Relationship Between Central and Raw Moments az 48 Relationship Between Central Moments and Moments about Origin 4.11 4.6 Skewnese as 4.7 Measures of Skewness 448 48 Karl Pearson's Coefficient of Skewness 47 4.9 Bowley’s Coefficient of Skewness 438 4.10 Kélly’s Coefficient of Skewness 448 4.11 Moment Coefficient of Skewness 449 4.12 Kurtosis 483 4.13 Moment Coefficient of kurtosis 453 4.14 Dispersion, Skesmess and Kurtosis Compared 465 5 PROBABILITY THEORY 51-561 5.1 Introduction si 52 Permutations and Combinations 5a ‘Scanned with CamScannersins Station 5:3 Basic Terminology of Probability 53 5.4 Approaches to Probability Se 55 Addition Thecrem of Probability on 5.6 Canditional Probability en 57 Independent Bvents nes 5.8 Multiplication Theorem of Probability 522 59 Bayes! Theorem: a DISCRETE PROBABILITY DISTRIBUTIOKS 61-647 6. Introduction i 6.2 Random Variable 64 623 Discrete Probability Distribution 62 64 Expected Value of a Discrete Random Varisble 65 69 Binomial Distribution 62 6.6 Derivation of Binomial Distribution 613 67 Constants of Binomial Distribution 620 6.8 iting of Binomial Distribution 628 69 Poisson Distribution 620 6.10 The Poisson Distribution a8 a Limiting Case of the Binomial Distbution 6x 6.11 Constanta ofthe Poisson Distribution ce 6.12 Fitting of Poisson Distribution eat NORMAL DISTRIBUTION 7a-7:90 7.1 Introduction ™ 712 Probabiity Density Function ofa Continuous Random Variable 72 73 Normal Distribution 72 714 Properties of Normal Distribution 73 175 Standard Normal Distribution 18 7.6 Area Under the Standard Normal Curve 18 77 Avea Under any Normal Curve na (CORRELATION ANALYSIS— oan 811 Introduction " 82 Correlation : Some Definitions 2 83 Types of Correlation — 8 84 Correlation and Causation a 25 The Correlation Coeficient ts £86 Methods of Studying Correlation . 87 Scatter Diagram Method cones 8.8 Covariance ad 8.9 Karl Pearson's Coefficient ‘of Correlation a. 8.10 Computing the Correlation Cocfficient ao 8.11 Coefficient of Correlation for Bi requenc 5 or Baris Pe — 8.12 Probable Error of Coefficient of Correlation 7 Pateeuione 828 8.13 Coefficient of Determination a ‘8.14 Spearman's Coefficient of Rank Correlation Ke 8.15 Computing the Rank Correlation Coefficient, ee 8.16 Merits and Demerits of Spea ‘Rant co rma’ ask Colon ‘9. REGRESSION ANALYSIS aaa es 9.1 Introduction “ 9.2 Meaning of Regression o ‘9.3 Lines of Regression - The Least Squares ‘Approach - 94 Regent Catia ~ Some formne 3 95 Papertes of RepetionCaticate 4 9.6 Sundar Ee Eimnte sa 9.7 rarest Cotes or Bras se reued Pegi Dion 98 Ddersee men Cran nd Regeeee oa 10 TME sBnuns ANALTetS 0 a ; 1646 ‘10.2 Time Series - Some Definitions iy 10.3 Analysis of Time Series waa n 102 102 103 108 10.38 108 coven dua Pad ations onthe Bquons INDEX NUMBERS re 104 Components of a Time Series 105 Mathematical Models for a Time Series 10.6 Methods of Measuring Trend 10.7 Shifting the Trend Origin aa 11.70 HL Introduction thd 11.2 Index Number - Some Definitions na 11.8 Characteristics of Index Numbers, na 11.4 Types of Index Numbers u3 11.5 Uses of Index Numbers ng 11.6 Problems in the Construction of Index Numbers ns 11.7 Methods of Constructing Index Numbers us 11.8 Quantity Index Numbers 11.26 ‘Scanned with CamScanner‘Business Statistics vad 1130 11.9 Value index numbers ua 11.10 Test af Consistency or Adequacy mat 11.11 Consumer Price index "a 11.12 Stock Market Indices hae 111.13 Limitations of index Numbers APPENDICES ALAS ‘Append: Table Ata eon sas : . c1-c12, ‘AppendicC: Bxamination Papers ———_¢iEn—____ STATISTICS : An Overview OBJECTIVES ‘ler studying the matenaln this chapter; you should be able to 2 Understand the meaning of statistics. Distinguish between descripti statistics and wferential stansties Descnbe various chivacteristios of statistics ‘Know important functions of statistics, Idersfy tmtavons of statist, Know diferent categories of data requved for statistical analysis Understand the meaning, objectives akd entena of classyficanen and tabulation, Describe various terms relating to frequenty distnbuton Distngush between discrete and continuous frequency sistnbunors Present frequency distrbusans graphically hy vanous methods such as fastogram, frequency polygons, etc ocooce woo 1.1 MEANING OF STATISTICS ‘The word ‘statistics is emmonly used in two ways. [n the first way, "statistics" is Used in plural sense meaning numerical {ecta or data. When used in plural sense, the study of statistics i also termed as “Descriptive Statistics". Descriptive statistics includes the procedures for collecting, classifying, summarizing and presenting data ‘Alcnowiedge ofthis aspectof statistics enables us to evaluate critically the information Presented in reports, artkes, ete. Today, with the development of probability theory, Statistics is also used in a broader and singular sense, When uscd in singular sense, the word ‘satstcs" refers to statistical methods that not enly deseribe important be features ofthe data but methods that allow us to proceed beyand the collected data ‘Scanned with CamScannerau into the area of decision making through generalizations and predictions. When Used in singular sense, the study of statitics is also termed as “Inferential Statistics We give below some definitions of statistics used in plural sense acre x frm Da us aren alas akrtgean evee sic ema es ee ore ea acta ee eet a tn dante ilar Sosa a nonin domafecedtamonateen yagi * gaeies semen ndueeesy 4 Deewana cpr sth nome indy Sa i dt te amen sa ee rat asic aim de sca lt er * se — Prof. Horace Secrist onan ii det aia ste ee vec Ima be empha ate ein a oped ere a emis tatran eo Cham tin nce tienen tes sorcerer, seers tae Basen om ae clear mag reine fu mandoremtarenn Rea pe entation Sera tnaeen cena eae ence ee cate Soave wapegeaent eet er pean mainly Somme se seers cise cae rae Sete unless they are supported by numbers. For example, statements like, “India is, ree pee reenact ag seems ete es nie ae ears ee ee roe tart tn a na Sl sera deme rar nen ened ‘Statics :An Overview is 4. Statistics are enumerated or estimated according to reasonable standard of accuracy. Enumeration simply means counting the sctual nuraber in the ddata, such as the mumber of male teachers in a school. However, whenever itis impossible or impractical to observe an entire aet of abservations, then the data ‘re estimated by using the powerful techaiques of sampling and estimation But the estimated valuea will not be as precise and accurate ao the actual values, The degree of accuracy of the estimated valves wil depend on the nature and purpose of the enquiry. 5. Statistics are collected in a systematic manner. Tae data must be collected in a systematic manner. Toe data collected in a haphazard manner may lead to erroneous conclusions 6. Statistics are collected for a predetermined purpose. The purpose and objective of collecting the data must be clearly defined, decided and determined prior to data collection. This would facilitate the collection of proper and relevant data. For'example, the data on the agricultural production, in itself does not serve any purpose unless we know the regions, commodities and periods for hich they are required. 7. Statistics must be placed in relation to each other. The statistical data that are callected should be comparable with each other. Ror example, the data collected on the prices of different commodities in a departmental store will not be considered statistics. However, prices of one commodity in different departmental stores constitute statistical data since chese prices are comparable. ‘We now give some definitions of statistics used in singular sense. Statistics au Statistical Mothods - Some Definitions 1, Sates may bcled te aceon of uring —Botey Ses ay iy cafe he ins ofa —Beatey Sates the cero he meceurnent scaler, egrdedase sae wraltenenfesiaers howls 4. Statistics is the science of estimates and probabilities. — Boddington 8 The since of statis is he meted of fdang elective, nan! or sal thenamenen om herent bned ron he wos tcnuneranor cece ae” ang Statics is he Sclerbe whch deals Uh the maha of cling, classe ‘esoing compart and interpaing rameriod dats vow neigh og ‘pherefemay “Setigman ‘Statistics may be defied es the science of collection, presentation, analysis and ‘eprtatin of orca dat, croton and Goedon Satin i a methodofdecton making he ace unary one Ds sumer data neler i glance ‘Scanned with CamScannerrf ushese Satates 19. The science and tof handing aggregate of facts observing, enumeration, recording, classifying and theratse systematically treting them, ~Hartow In the light of above definitions, itis natural to ask the following question Is Statistics a Sclence or An Art or Both? Inorder to answer the above question, itis necessary to know the meaning of cence and art The word ‘scienor flere to a syatematizes body of knowledge I tas cause and elfect rationship to find out generalisations which are called "awe" of that scence. Statistica is also a science but only in'a Limited eente va specialized branch of mnowledge. tie ot similar to exact sciences such us Physics, Chemis, etc In the word of Croxton and Cowden, “Satsies maybe defined asthe scene of collection, presrtatin analysis and terrain of numeric dota ‘Statistics isan art ls. “Act” is that branch of knowledge which tll us the best possible methods wo as to meet certain specifications. A statistician le an artet ‘il fil in his ob ihe does not posses the requisite sl, experience and patience ‘hile using statistical toaafor any problem. Inthe words of Baslow ‘Sait is oth ‘science and an ar of handing aggregate of facts - observing, enumeration recording, classifying and otherwise systematically eating hem* 1.2 FUNCTIONS OF STATISTICS 1m this section we shall describe aome of the most important functions of statistics. 1L. Precisenees and Definiteness, Statistical techniques enable us to present facts in dear, precise and definite form. 2. Condensation. With the help of eatstical tools a mass of data can be condensed into a few presentable, understandable and significant figures. For example, the complex data may be presented as totals, averages, percentages, etc. ‘3. Comparison. Cozparison is also one of the importaft functions of statistics, ‘The absolute figures themselves do not convey any sipulicant meaning, I is their comparison that help us draw conclusions. Statistical devices like averages, ratios, percentages, graphs, diagrams, etc. are the tools that can be ‘employed forthe purpose of compariscn, 4, Formulation and Testing of Hypotheses, Statistical methods are extremely ‘useful in formulating and testing hypothesis for the purpose of establishing & relationship between two-or more variables. For example, the degree of ‘association between price and supply, or between amount of rainfall and yield of ‘rop, or betwee advertising expenses and sales cin be obtained using statistical ‘ole Sinilary, hypothesis ike whether or not advertising campaign is effective 1 eae Snreay of epi can be etd by appropriate ata to 5, Prediction. Statistical techniques can be used to analyse the past ‘predicting some future trend. For example, the demand ofa particular product {for the next year can be predicted by knowing the demand (or that product in the ‘past and the current market trends and possible changes in the factors thet affect the demand. Statstis Tan Goariow 1s economic, business and other policies at state, national or global ievel Fos oeder, import-export, social welfare, wages, ete. are formulates on the busis of 23 LUMITATIONS oF starisics The ld satis, hog, wide eed in lon ery phere, ye ‘2 mm kataton. Some ofthese intron ae deere age 7 1. Statistics des nat deat wth indtdua gure oxyde wth aggregate of tcis er Gis, Sng or naed gues cance! cee tae he rank sg igre iting wo the hog oft sdk & hes a oe 5 cate nls the he students these Case F eight of ether students the ene class se 2 Sins a tse tae Stes sty te fate which ae capable of rng stated in numbers or aang, Oentinng esomenen sich a hones intligece, ena, poner ne eae seid icy in stan unes the nes ae cape is ct 4 Beatson aus are not. law arent enact Cone onan werner ‘Statutes can be missed Sits ia te ise. Th ble tobe misused Therein famous fatement "he gz done bu the lars en gue’ eh vet cmisua cf sats Thesainel esl can be mane dtcior ee according to one’s own interests, Peal a 5 Statistical nerpreatons require high de stand re high depres of skit and of the subject. In order to get meaningful results, ws 7 | exact. Unlike the laws of natural sciences, statistical quently, the conchision based on them are truc only nomiedge and understanding of the statistical devices can propely hones statistic) dat. The statscal results placed in the hands of inexperienced to Talacious and misieading ingr,“Siis rite cy fom one ne ‘can make a God or a devil as one please”, ance fomsetshoe 1.4 COLLECTION OF DATA 4 this section we shall discuss how data are ‘seaningful and readily comprehensible form ‘Scanned with CamScanner on OOP BGOVOFINAGQDANNGQ GAL16 7 Bisineas Sales {s tocollect the dats. The purpose of cllection of data is to coleet evidence foc reaching sarod and cessation ta rim. Dac ‘be classiSed es ether primary oF ‘secondary, Primary and Secondary data Primary Data. Primary data are original data which are collected fer the fist tne by the inveatigator for the purpose ofa specific statistical investigation. fn the words cease, Willet and Simone, "Data orginally collected in the process of investigations ‘re known as primary data” Primary data are collected by the investigator himself using methods such as Personal Trterview, Telephone Interview, Mail Questionnaire, Door-to-daor survey, focus grou Rterviews and so or. The key poist about primary cata is that it i unique to the investigator and unless iti published by him, no ont else has access 19 i. ‘The menit of primary data is that itis direct information uncontaminated by being ce resited through another source. Primary data can be relied on because You raernnereit come fom and what was done tot ike coking someting yours Cae hat wert nto it The tain disadvantage of primary data is that itis ‘expensive and difficult to acquire, ‘Secondary Data, Secondary data are those data that have already been cletied Of eerie purposes other than the solution of the problem ia hand. According Sand tate “Sarondary cata are thse whick are alrecdy in eristence, and ih Me aan af some other purpore than the anerng of he question ina” kn he a cot ental, "Data collected by other persons are elled secondary data” Tere worth mentioning ere thatthe deta are primary tthe sore who calest® 2 process them forthe frst time and become secondary sopseane elpe who Uses rea ors owa porpe. The dstnaion between pmyary end sander £0) ienly of degre of detachment of erginal ource and nat of nt, In the words of is only of ls Cpe estincion between primary and secondary dale ree of danta can be cellested from external sources tuch a8 magazines, riticles, T.V., radio, Interset, and ¢0 02. aererey econdary dat stat ti ble o chraer end ees Tra prtpary data. The problem is that often the eimby, arcuracy and interity of secondary data is uncertain. 1.5 CLASSIFICATION OF DATA 6: aw cate 65 data re ect in aay stasicl event ase termed a am 38 i daca ote enon orarined ory ej ued OS DS we eral pm. Caenton and bin fea ore ac nd era cation ie emis 10 tabsaton an reparts e renee paper presen of wee a ean pthc prorat eangng tain SOUPSOE CAE STNG AST, Cheatin Fn pres ye fas coming te ata Aarne ' | americas rab netian, cma 17 etal ine grape eas cree of crrancing Wings (thr actly or Goer ase ms anes ring thas revembanoes fin en induce” 9f etvues that may sist arorgst @ vers of In the words of tacit, “Clea tee wade of ert ash res of ara da ae bnarelated pata -eummon characterises, or separating them ito differen: ‘Thus by way of classifica Faas yf ization vtuminous heterogeneous date comet inte 1.5.1 Objectives of Classification ‘The main ebjectves of classification are it Homie omen a eae ef dete: ‘Statistical data collected during the course of Sine we mac ht tpi w areas a efi ans the ral sierliemee ‘of the figures, unless they are resem acai ia Fal ite popes Sars ingen Senne range Lacnyproarti qe these ures are lasnifid according to sex, marital aoe tod ditindady die daaobat tert eeadygepe “i Cee ten prepare the data for tabulation, Only clasiied data can be presented in Bei intros senna a oe a 2 fate compton Cisne an ca api — = on houscholds classified on the basis of age, religion, education, sar ety cin cea wet cag es Ss od reson es eo study the relationship. Classification en: one: ring out caus relationship, if there is any in the data, abi as lees 5 1.5.2 Requisites of a Good Classification Although ro hard and fat ‘hough re hard and fat rls can be ad down for clasiteason he lowing rer] pines may be oben fora god darieatin Bxhaustive. Classification must be exhaus at [crue cusicaion must be ebausie in the sent Unt each and ery 2 tually Becluste.Thec! cna ate “ ‘The classes must not overlap, ‘«., each item of data should Sultabillty, The classical Sutil The sasiiaton snl cut to the object of imei Far gmp tn avenigon cored fo enter nen rains wars, lb an use sytem on he Dan 3. ‘Scanned with CamScannerra a esis Stes 4, Stab, The baie pri fection shuld be retined tev She the ring ot inrenr n th ube mas wl ea {Flay A good classfeatio shouldbe le being adjusted to changed & Romogentety Te tes consign shouldbe bomogencous 1 Unambigity. Te asicaton shouldnt ea a any abigay eaten ‘per of Classation. Data canbe clase’ onthe bain fle flowin a 1. Geographical, La, areasise or region wise. 2. Chvonolgia, ., with respect to occurrence of tine. 3. Qualaive, Le, by character or by trite 4 Quanaiate ie, by magnitudes. : Geographical Classifteation. When the classification is done on the basia of sroeraphical (oc regional) diferences such as states, ces, diate sores ene, {called geographical classicaon alooknown as span assieaton| For cass the production of wheat in diferent states of the Country may be classiled ate basis of geographical classifcation. Siaularly, the navonel income of diferet countries is an example of geographical clesication Chronological Cassifteation. When the data are cassie on th hase ome i is termed a chonologial lasication (ls know a tengo clasfeton) Tike say bein hours, daya, weeks, months, years, et For emmple the flowing data regarding sales ofa fms an example ofchroneogcal classification. Year 1999 2000 amo aoea 003 aooeomns ‘Sales fx '0009) 2 8 2 7 2 © % Silay, the data representing poption of India fr diferent years at shown ‘low isan example of chronlopenl clasication, Yer 11951 1961 3972 1981 aL 201 Pepuetonofindinfromes) «5 357 BS SHS ORS OS 1027, ‘Statistical data arranged chronologically constitute what is called a Time Series, (Qualzatie Clasefiction. When the deta are classed acceding to some nen- measurable characteristics uch ax beauty, honeay, sex, occupation, nations, ‘employment, Iteracy, ee, the classification in termed a5 quate or deseiptive. fn ‘Tuatatveclarsiication data are caanited according tothe presence or absence of 1 particular azibute alitative classification can be simple (dichotomous) or manifold. When the rer eens neh enol {wo classes are formed, one possesting the attribute and the other not possessing the attribute. This type of classification [s called almple (or dichotomous) . ith respect to one For example the clasication of population in a certain city with reepeat to 0 atte, say sex, in dichotomous ia nature as shown below Stakes an Ouniew 7 Female eapect to wo attributes, say sex and be classified with respect to ‘ex into Male Similarly, if the claseention js dene with ‘employment. Then the population mey fret, "Males! and Females’, Each of these classes may then further be classified into ‘Employed’ and ‘Unemployed’ on the basis of atinbute ‘employment’, and as such population is classified into four classes, vz, (9 Male employed, Male unemployed, (i Remade employed, and (i4 Female unesployed, Stl. the classiieation may be further extended ty considering other attsibutes euch as tmarital status’ ete te (ype ofelassiication, where twocr more atibutes are considered and several clases ‘reformed, is called 2 manifold classification as shown in the following diagras: Pn f + Nase ele + aap me oie vats canbe whi nabs Nateee Unetiied Pg ia Quantitative Classification. When the data are classified on the basis of a phenomenon whic is capable of quantitative meanurement such as heighe weight, age, come, expenditure, sales ec, itis termed as quantitative classification tes quantitative phenomenon under study is known a8 a vanable (or variate a hee (his clasafcaton fs also sometimes called dassfteaton by vacubies Usually. variables are denoted by capital letters auch as X, Y and Zand the values that these variables asnume may be dented by corresponding lower cae lesa ‘eample 1 Ut X dence the number of chien i aly, then Xs avaiable ‘that can assume any of the values 0, 1, 2, ‘ sna (2 Soppose we measure the heights ofall the students in a class, I denotes the ight oa teat in that an, then Xs vane which ean ake op ne There ace to types of variables, vi, inerete Definition. A dseretesariabe isa vrabe tat can assume only afte nut of values or if its values can be placed in one-to- ate 2 For example, the number of st fents in a class, the the number of accidents on the umber of chldren in a road, the number of . et typing ertors per paze, ete are ‘Scanned with CamScanner ee OO On110 Business Satistics some examples of discrete variables. Note that a discrete variable represents data ‘hat can actually be counted. Definition. 4 continuous varlableis a variable that can assume any value (integral ‘as well as fractional) in some specified interval or range. for example, the heights and weights ofthe students in a class represent continuous variables, Note that a continuous variable represents data thet are measured 1.6 TABULATION OF DATA ane cet ae cr ra oe rca ere ame ena er ae wr ago Pe er ares So eos al oer pn Tiny ne ee as sa ec ed ae aoe ae ee ety an ee a em te a chert ees suit sufficient explanatory and qualifying words, phrases and statements in the form of sneer yg a meno ne cae Rong ae characteristics. . ‘Another definition of tabulation given by Prof. H.Secrist empl oe te ean fi a ra “Tables are a means of recording in permanent form the lysis eS rr yb oe In the words of W-A. Spurr and C.P, Bonin! “A statistical table is a dassification of a chs a Cec oe ‘data in the form of a table so that it -_ ee easily ur ne rt oon tear, wan ells pe ue a characteristics. tion 1.6.1 Objectives of Tabulat , “Te main objectives of tabulation are described as flows © statistical table is to 1 aa rmstify Complex Data, The main objective of sans! ‘Stasis: An Overview un Sunmaie amass of mune inmaton and to ree oe maton ant proctitis tebe tent he pues reich oben Ae ne 4 complex mats of information nore less tnd ney 2. To Factlitate Comparision, {t facilitate int fom I elites quick compartion of ait ‘shown in, rows and columns, = insane Seed 9. Folsonomie Space lrncnsary reins aera blatin whch ‘ess in economy of space without saci quay and untines ean 4 To Depict Trend. Data condensed in the form ofa tb in the form ofa table cevea the pate oe Lend which erin cannot beundersacd na dsp orn af ence 5. Todelp Reference. Tabled data are god for reference purposes easily identified, epesimniteen Ds To Detect Error and Omissions inthe Data te some vital emissions are detected, 7, To racttat Statistteal Processing, tisony iter cassticaton and tabulation tht the data becomes renal preening The eacuatn of seer prion, et can be undertaken ol rte daa as eo Cueughane preliminary stage of classification and tabulation. . 8 To Clarify the Characteristics of Data. concise taba fora dearly events the characterstin of data and highlights is eigaifcant features. 1.6.2 Besential Pasts of a Table A god statistical table conn al that is required in as smal a space as postle without any les of cary. The peparetion ofa good statiseal tates an art Toe purpose of tabulation must always be kept in mind before preparing a good statistical {abl The folowing ae the easel parts ofa goed ets table 1. Table Number. A table should always be numbered for easy dentiction and reference fn future, The ‘ale numbers placed atthe tp ofthe table 2 Tile, Every table must bear tle. Te tile ofa table should be concise, lear and selferplanatry I's customary ta place the ie a the topo the table, eadnote. The nlormation n parenthetes following the tite sella heat. Teemplain some general characters azout the data th the eader shld know inorder tose abd interpret the able propel. 4. Captions. Captions ace the headings for vertical columns. They ar used to entity the fqures in the eokuns. 5. Stub and Stub head. The column on the left ide ofthe tabi is ele the atu and the niu lisngs, whe ident the contents of te ows and ies ae refered to as stb items. The beading above the stub items i ealleé the atubheed 6 Footnote. Fstnotes are often given athe botom athe abe o point apart irregularity, omision or charecteriatc ia the ale, Footates ae generally numbered or lettered, when there are several of them. If there ar oly ot oF {wo asterisks and eyabole euch a can be aed to dence them. only after tabulation that a ‘Scanned with CamScanner1a Business States 7, Source Note. Ifthe data have teen collected from a secendery source, itis advisable that the source of the data be revealed to give the proper credit to the original source. . Body. The body of the tabe is composed ofthe gures, The figures in the body ‘correspond to classification unde: captions and stub heads. 9. Ruling and Spacing In order to give neat and clean appearance ofthe table, there should be proper ruling and spacing. Horizontal rulings are seldom west in the body ofthe table: However totals may be separated from the main body of {he table by a herizontal ine. Vertical ines may be used to separate the eure Unit of Measurement. The unit of measurement showd always be stated slongeith the table, ift is uniform throughout the tale f diferent une have been adopted, these may be stated slong with ‘stubs’ er ‘captions’. ‘TABLE 1.1. Format ofa Blank Table ‘Title Headnote ifany) ———— es net [eerie [thet [eet | rot | | 10, Sates: An Ovenew TABLE 1.2. Coffee Drinking Habits of Town A And 1 Percentage) Teun A ewe Ariba Maes [ Fenates [Tota | tates | Fenaive [Toor ‘Coffe einkare = [40-35 [4s | 5-30] as To wa | Ss 225 | Nesfestintes | 20 | $6220 |ito-15 | 20 | as-15 | 10-20 sa5-_ | sss =30 |e Tear To-4 [40 [too | 35 00-35} 00 260 ss | | Note. The bold Sgures are given figures. The other fi appropriate additions or subtractions, Beample 3 In 2000, out of total of 1750 workers ofa factory, 1200 were members of ‘attade union. The niznber of women employees was 200, of which 175 cid ne bien tog trade union. fn 2002 the number of union workers increased to 1980 ol wack, the mumber of non-union workers fell down to igures are obtained using, Solution. The given information has been represented in Tabular formas shown in Table 13 ‘TABLE 1.3, Comparative Study of the Membership of ‘Teade Union in « Factory in 2000, 2002 and 2008 t Years 2000 2002 08 ] we rade nian | Wis [Peete] Tae Was | Pence] Tat aes] Foret tows fine |S hme [ime | as fee a Pes tennooer [35 | we [ae |'we | [ome [S| PS ee eamge ney on tenin amen seni | DE LT coun Touma: Pemales were 40%, Total coflee drinkers were 45% and Male non-coflee drinkers were 20%. a TounB: Males were 85% Male non-calfee drinkers were 90% and Female coffee drinkers were 15%. Represent this data in a tabular form. ‘Solution. The tabular sepresentation ofthe given data is shown in Table 1.2. ‘Note The bold Ggures are given Ogures, The other Spures are obtained using ‘appropriate addons or subtractions. 1.7 FREQUENCY DISTRIBUTION A frequency distribution is « convenient vay of presenting a large mass of ata ia tabular form by grouping the data. ‘There are two types of frequency distributions, wz. discrete and continuous Discrete Frequency Distribution (or Ungrouped Data). A eguencyditibution ‘epresented by « dveete variable i called a scrote roquasty diswibaceee be data ofa discrete variable fy tabular all posible values chat a diserete varabe can take on, alone corresponding fequencies ‘Scanned with CamScanner ear earOKO4FTAVVAAN an a14 I Xis diserete variable that can telce on values x; x. Busine Sais 1 xy with the corresponding ‘Statice: An Overview ‘TABLE 1.6, Frequency Distribution for the: frequencies f, fy, «+» fqthen the frequency distribution of Xie given by Wage 30-34 3 x x 4 * 35-39 5 Preqeney hi & & 40-44 2 ‘ s5-49 8 for example, Table 1.4 shows a discrete frequency distrition. fc pa a ‘TABLE 1.4, Frequency Distribution of umber of Chiliren t 55-59 i 7 —s ; 60-64 2 enteral ; 0 8 | Class Interval, An interval defining a class is called a class interval (or simply ; a {dase}, Each clas intervals identified bythe range of values assigned tothe clase - 2 8 | Forexample, in Table 1.6 the range of values of first clase 30 ~34 from 20 through 2 3 {38 Inclusive; che range of values of second class interval is from 35 through 99 2 {elusive and ro on. & a | Class Limits, The smallest and largest values that define a gven dass interval are Fetal refered to as its class limits, The smaller numbe: is called ‘he lower clas limit Jn. A frequency distribution represented by = and the larger marber is cli the upper clans limit. For example, for the clase cris Fey itn nan ort. | dhe mt a th on int Ps muse Siiiceweleraiacommciereaaementmcurenay | Revi ecient west Teen Seren ee aera | emtedicierame ont ey cia Oe ee | classes. These are (1) Jnclusive Method, and (2) Exclusive Method. } He, Table L.§ shows the eaatinuous frequency distribution ofthe heights of} 4. Inctustoe Method. Grouping of data into diffrent classes such that both the aera cess eee an Rel en = 7 acencrenasate called grouping by “inclusive method”. Here, the grouping of data into different ‘TABLE 1.5. Frequency Distribution of Holghts o' lasses is done in such a way that the upper class limit of one class does not ‘Number of Students fRrequency) coincide with the lower class limit of the next class. For example, the marks eighe Cetinetr) Z obtained by students ina class may be grouped a8 0-9, 10- 19, 20-29, ... 155-156 % «where the class interval 0-9 inladea all values from Oto 9 (bth inchucivg), 187-158 < the clas interval 10 19 inciudes all values from 10 to 19, and soon. Thus, for 159-160 2 instance, fa stent has secured 10 mark, he ie included inthe class 10-19 11-162 * 2, Bxclusioe Method. Grouping of data into diferent classes such thatthe upper nescies class Umit of one ciass is same as the lower class limt of the next clase ‘eens 2 a called grouping by “exclusive method” Here, it should be understood that the Tee 3 upper-class limit of any class is excluded from that class. For example, if the I 1. is obtained by students ina class are grouped at 0 10, 10-20, 20-30, sent Tol 50 » then the class interval 0 ~ 10 includes all values which are greater than LF ee eit to 0 tt less than 10, the cine interval 10-20 include all values } 4 Which are greater than or equal to 10 bot less than 20, and soon. Thus sta Some Useful Definitions ich are relevant to x continuous frequency | stdent has secured 10 mar, he wl be included inthe clas 10-20. some definitions which ae relevant w 4 6 ied He a a ee one shown Tate 16 Tale Leica Guan buon | Se ‘of 60st ts in a class, recorded to the nearest kilogram, | the weight fa group of 60 etuent i i ‘Scanned with CamScanner116 Business Sates Class Boundaries. The css boundailes are the values halfway between the upper lass lmitof one elass ane the lower elas init ofthe next cass, Each clase lug an ‘upper and lower clase boundaries. The class boundaries are obiined to aso fl the tp that might exist between classes. Uda the gap between the upper clas limit of one clans and the lowe clas imit of next class the clase boundaries for any class are given by Lime cota = over da int = Uppers tery uppercase nt + Fee hefeguccyceubuion gen TD 16 tee ee he wpe clase tetas starce: anc gae eee ae ee tert ers Sbtrecng as vlue rm ahr tate snteangg a ore ‘wall upper class limits, we obtain the class boundaries for each cla Table 1.7. 8 shown in. ‘TABLE 1.7. Frequency Distribution for the Weights of 60 Students (Class interval l Class Boundaries NoofSudens | 30-94 Ds a5, > | 35-39 345-298, 5 40-44 395-445 2 aids 445-495 ® S034 495-545 « 559 s45-595, 6 0-64 sos-45 | 2 rie place than ice hat case boundaries are always caved out 0 one more decal pl CESSES cbecrvean. the claw nite are expremed In the units uted fr Be.Stnug: Tu cane boundaries, however, ar eresed unite that ae amar than the nse in he memurenens or emmp Wf be de Li ee Teele umber uns, then te cas boda ar expend in Ser or ithe eave expressed intent unt then the nda e preset mundredie of uta v0 on Gace widih. The diftrence between the upper ard lower le boundaries of « Showin ese he Sasouih ress siny tmabe 17a le ier ar of qual wi nat, the with fac as iS. Clase Feqoenay The aumber of eberations that ona parslar dain ale Sree ee Peratnay ans denoted oy the eter 7. The cs fequney on the las Fem ractonr a percentage ofthe a equeny. We fer t thi va a oe froquonay of a cass. Area frequency distin presente te in terms of fractions or percentages. The relative frequency distri fer the dain Table 61s pen a Table L8 a een l | po sates “An Ooeriew TABLE 1.8, Relative Frequency Distribution Class ‘Class ‘Class Fre aie | ce ma | Sia z , | oe ae | See 5 2 | as om | eee z : | (Clase Marks. The value of the midpoint ofthe class Goundaries or class limite fa class intervals called the class mark or class midpoint. Thue __ lower ass limit + upper clase limit (Class mark is De upper class nit lower ease boundary + 2 For data in Table 1.8, the clase marke are 22,97, 42,47, 52, 57 ard 62. cass boundary Note may be noted that the class mask is used a the representative value of 2 kdven class: or example, for the first clas interval 30 = 36, the clase mark is 32 Which is used as a representative value forall data incladed in that class Open-ended Class interoal. class interval with an unspecified upper or lowe limit is called an open-ended class interval. For example, classes such as salary less than %10,000; age above 65 years; marks Jess than 10, ete. are“open-ended” classes since one ofthe class limit is not speed in each of them, Cumulative Frequency Distribution. The cumulative frequency ean be classed into two types: "eas than" cumulative frequency, f) "more thar’ cunulative frequen, () “Less than” Cumulative Frequency. The number of observations which are less than the upper class boundary of given clas interval is called the “Less ‘than” cumulative frequency of that class. Thus the “ess than” cumidative frequency of given clas interval is the sum ofthe frequency of the eves ciass and the frequencies of all classes having a clase mark less thaa that 2f the ven class, (ty “More than” Cumulative Frequency, The sumber of observations which are ireater than the lower class boundary of a given class interval is called the fmore than® cumulative frequency of that elass, Thus the “more then” cumulative frequency of a given clase interval is the sum of the frequency of the gives lass and the frequencies ofall classes having a than that of the given class, table listing cumulative frequencies is called a cumulative frequency distribute ‘The cumulative frequency distribution for Uhe data given in Table 16 is shown in Table 1.9. lass mark greater ‘Scanned with CamScanner aaawanAANNy gy o18 ‘TABLE 1.9. Cumulative Frquency Distribution “Cass Class Cas] Cm oR tera | soundates | ere |" | ecco | mene) w-% [5-5 |=] 3 3 | areas ass |aes-ass | ao | os | sesea | saeses wm |xs-ms | a | a | een» | wraee as [as-as | ce | ie |aoriesas | mcies a so |asses | 2 | we |aeruee | semen ss |ss-os | 9 | 6 | sees | aero [oom |ass-os | @ | 2 | strane | 2 ‘Example 4. Conveat the flowing isubuton nto more than” equendj Ase ibuin Seetiy wages ess han” no @ mw No of workers 2 we mm {Beth Une B.Com. 197) Solution. To obtain the “more than’ frequency distribution, we shall first convert it into an ordinary frequency distribution as shown in Table 1.10 (a, ‘The required ‘more than’ frequency distribution can now easily be cbtained from Table 1.10 (a) as shown in Table 1.10 (bi ‘TABLE 1.10 (a) ‘TABLE 1.10 (5) Ordinary Frequency Distribution “More Than’ Frequency distribution Weekly Wages Wo. of workers Weekiy wages Te. ofuorkers fo smorethan o- 2 a 0 Tova = aor 20- 40 92-41-51 » 109+ 51~ 160 40- © 1ss-sa-64 | | “ 45+64= 109 60. © is4-ise=38 | o 7438-48 80-100 201-198= 7 © z 1.8 GRAPHIC PRESENTATIONS OF FREQUENCY DISTRIBUTIONS In this section we shall discuss how to present frequency distributions graphically. ‘The information provide bya regency distor in tabular frm eae © rasp if presented graphicaly. Graphs present data in a two - dimensional pict Ep the tarontal ao we show the ales ofthe vara (he characters We we ‘measuring). On the vertical axis, we mark the frequencies ofthe classes shown on. the horizontal axis. False Base Line. One of the fundamental rules of constructing graphs is that the ‘vertical scale must start with zero even ifthe lowest ¥-Value (dependent variable) ‘associated with any X-Value (independent variable) is far from zero. Wren the gap between zero and smallest value of the dependent variable ¥is large, then the vertical scale is broken and the space between the = origin ‘O"and the minimum value (or some convenient valuc near that) of tae dependent variables omitted by drawing +8228 horizontal lines above the baseline, called the false base tine (see ig 1.2), Tae 3120 scale along Yoasis Is then framed § according. The main advantage of using & false bas ini that magnesia ‘uctuations present in the data so that they are clearly visible o the cease: Moreover, # it also economizes the space A frequency istribution can be presented graphically in 92 1004 spas cok mee ae 000 aor ao0a as Box neta 1, Histograms (2. Frequency Polygons: Liege Semen nme theorem teprn nteneee e ‘A nistogram isa series of rectangles, each proportianal in U the classes were unequal and the width ofthe intervals differed amoag the clases, then the eight of the bar for each class would be proportional tothe frequency 3s, where the frequency density of a class is obtained by dividing the clans frequency ofthat clase with the corresponding width ofthe clase. That i _ Glass Frequen: Prowiemry dtasily = aot ths dese aaa ‘The height of the bar fora class, in such acase, aay be ten asthe product of the corresponding frequency density and the width of the class having amaleat nee Note. It may be noted that histograms cannot be constructed for distributions with open-end classes, Example 5, Draw a histogram to represent the following data Day wages (2 0-10 10-20 20-30 30-40 No.of workers ) frequeney 40-80 50-60 woos ‘Scanned with CamScanner120 Business Statistic sotution. The histogram representing the given frequency distribution has been shown in Fig. 1.3. On the horizontal axis, we have shown the class intervals. On the vertical exis, we have marked the frequencies ofthe classes shown cn the horizontal abi, Sinee the class intervals are of equal width, the height of the rectangle for ach class corresponds to the frequency of that class. 20 7 2 Be fia as 6 2 ° O18 a0 w wo ius hes 713 ‘Beample 6, Draw a histogram to represent the flowing Srequency distribution: Mats 010 1020 M40 4059 S060 €0-70 7050 50-00 NoofSudents: 4 6 WW i Ww WO Bw § Solution. Since the class intervals are of unequal width, therefore the height of the rectangle for each class is proportional to the frequency density ofthat class. Infect, the height of the rectangle for each class corresponds to the product of the corresponding frequency density and the width of the claés having smallest size. Taat is, height of a rectangle = frequency density x width of the smallest lass Computation of Frequency Density Giassinteraal| Class wid] Frequency | Frequenay Density] Heighto/te reaunglo a a a 1e- 6+ 8 aio 0-10 wo fF cr 4 10-20 0 6 0 6 20-40 a u o7 1 40-50 10 6 16 16 50-6 0 irs 14 4 60-70 0 0 10 10 70-90 » 6 08 90-100 10 5 os 5 Stohr An Overview tar ‘he histogram representing the give fequency distribution in shoun in ig, 4 iB Be 2 . fo 5 al 2 ; : 010 20 30 40 80 «70 80 90 100 aks Mes ote. Ifthe midalues of various classes are 8e8 are given {hese must first be converted into classes as is shown inthe fllowing eearnple ‘Bxample 7. Draw a histogram to represent the jllowing data Md value (9 MS 5 SS ass ts Frequency 3 6 6 7 4 2 in place of clase intervals then. Set Scene we gin lc we td eet ter The ence even ry toons ain be sa 10/2 5 cach mil-vdue gute coveapndng eee ea sabe cate te nace Can cle oe Ea ee i inogan reeening Ie pees tec eee oy Frequency ohio 120 130 140 190160 Ho Gin ter rete ‘Scanned with CamScanner