Data analysis copy

• 1. DATA ANLYSIS BY ; Ms. KAVITHA. M. R

2. Introduction;Once the data are collected with theresearcher turns his focus of attention ontheir processing.A researcher has to make his plan for eachand every stage of the research process.As such, good researcher makes a perfectplan of processing and analysis of data. Tosome researchers data processing andanalysis is not a very serious activity. 3. Introduction;Data process is an intermediary stage of workbetween data collection and data analysis.The collected data through various tools suchas interview schedule, questionnaires, datasheets, and field notes contains a vast mass ofdata.And these collected information doesnt givesstraight away answers to research questions.The collected data is in raw materials, so itneeds processing. 4. Data processing consists of anumber of closely relatedoperations;EditingCoding and ClassificationTabulation 5. Once the data is collected, then theseaspects to be followed very importantly1. Checking the questionnaire Thorley.2. Sorting out and reducing information collected tomanageable proportions.3. Summarizing the data in tabular form.4. Analyzing facts .5. Interpreting the results or converting data into statements,propositions or conclusion which ultimately will answer theresearch questions. 6. Editing:Editing of data is a process of examining the collectedraw data to detect errors and omissions and tocorrect these when possible. a careful scrutiny of the completed questionnairesand the schedule. Editing is done to assure that the data are accurate,consistent with other facts gathered, uniformlyentered, as completed as possible and have been wellarranged to facilitate coding and tabulation. 7. The editing could be classified as;1. Field editing and2. central editing. 8. Field Editing;Field editing consists in the review of the reportingforms by the investigator for completing (translatingor rewriting) what the letter has written inabbreviated and or ill illegible form at the time ofrecording the respondents response. it is done as soon as possible after the interview,preferably on the very day or on the next day. field editing the investigator must restrain him andmust not correct errors of omission by simplyguessing what the informant would have said if thequestion had been asked. 9. Central Editing;Central editing takes place when all forms orschedules have been completed and returnedto the office. This can be done by individual in case ofsmall study, whereas in the larger studies ateam of editors will do the editing. 10. The editors would edit the inappropriateon missing replies; the editor cansometimes determine the proper answerby reviewing the other information in theschedule.Sometimes the respondents can becontacted for the clarifications. 11. Important aspects in the editing process are;1. Completeness;2. Accuracy;3. Uniformity; 12. Coding ;Coding process assigns numerals or othersymbols to the several responses of thedata set.In simple coding is translating answersinto numerical values or assigningsymbols/ numbers to the variouscategories of variables to be used indata analysis. 13. Coding is necessary for efficientanalysis and through it the severalreplies may be reduced to a smallnumber of classes which containthe critical information requiredfor analysis. 14. A. Numerical coding;Coding need not necessarily benumeric, it can also be alphabetic.Coding has to be compulsorilynumeric, when the variable is to besubject to further parametricanalysis 15. B. Alphabetic coding;A mere tabulation orfrequency count orgraphical representation ofthe variable may be givenan alphabetic coding. 16. C. Zero coding;A coding of zero has to be assigned carefully to avariable.In most of the cases when manual analysis is done, acode of 0 would imply a no response from therespondents.Hence, if a value of 0 is to be given to a specificresponse in the data sheet, it should not lead to thesame interpretation of no response.For ex; there will be a tendency to give a code of 0 toa no answer. However, if the respondent had notchosen yes or no, then a different coding than 0should be given in the data sheet. 17. QuestionnumberVariable/observationResponsecategoriescode1.1 Organization Private PtPublic PbGovernment Go3.4 Owner of vehicle yes 2No 14.2 VehicleperformanceExcellent 5Good 4Adequate 3Bad 2Worst 15.1 Age Up to 20 years 121-40 years 240-60 years 3Above 60 years 45.2 Occupation Salaried SProfessional PTechnical TBusiness BRetired RHouse wife H 18. Classification; The process of arranging data ingroups or classes on the basis ofcommon characteristics. Data having common characteristicsare placed in one class and in this wayentire data get divided into number ofgroups or classes. 19. Classification according to attributes :The data are classified on the basis of commoncharacteristics which can either be descriptive(such as literacy, sex, emotions, honesty) etc.A descriptive characteristic refers to qualitativephenomenon which cannot be measuredquantitatively; only their presence or absence inan individual item can be noticed.Data obtained this way on the basis of certainattributes are known as statics of attributes andtheir classification is said to be classificationaccording to attributes. 20. Classification according to class-interval:The numerical characteristics refer toquantitative phenomenon which can bemeasured through some stastical units.Data retaining to income, production, ageweight etc.Such data are called as stastics of variables andare classified on the basis of class intervals.For ex; persons whose income says are withinRs 202-Rs 400 can form one group, those whoseincomes are within Rs 401- Rs 600 can anothergroup. 21. In the case of open ended questionsto classify responses the researcherlooks for major characteristics of theresponses and puts accordingly. 22. In case of attitude scales a researcher has tokeep in mind the direction or weightage ofresponses.Forex; when responses strongly agree iscoded as 5 the subsequent codes would be inorder.Therefore, if there are responses like agreeundecided , disagree and stronglyDisagree, they have to be coded as 4,3,2, and1. 23. The matrix questions have to be coded takinginto consideration each cell as one variable.For ex; if the column of matrix representsemployment status, namely permanent andtemporary and the row represents employersor type of employer, namely, government andprivate, the first cell would represent a variablegovernment permanent .The second cell would represent governmenttemporary and so on. 24. Tabulation:Tabulation is the process of summarizing rawdata and displaying the same in compact form(i.e. in the form of statistical tables) for furtheranalysis. It involves counting the number ofcases falling into each of the categoriesidentified by the researcher.In a broader sense, tabulation is an orderlyarrangement of data in columns and rows. Thiscan be done manually or through computer. 25. Importance of Tabulation;It conserves space and reduceexplanatory and descriptive statement toa minimum.It facilitates the process of comparisonIt facilitates the summation of items andthe detection of errors and omissions.It provides a basis for various statisticalcomputations. 26. computerized data processing;When the sample size is large and orwhen the variables studied is vast andinter related data can be transcribed tothe computer for further easyprocessing. 27. Computerized tabulation is easy with the help ofsoftware packages. The input requirement will be thecolumn and row variables (EXCEL spread sheet).The software package then computes the number ofrecords in each of the row/column categories. Themost popular package is the Statical Package for SocialSciences (SPSS, Systat ).It is an integrated set of programs suitable for analysisof social science data.This package contains programs for a wide range ofoperations and analysis such as handling missing datarecording, variable information, simple descriptiveanalysis, cross tabulation, multivariate analysis andnon-parametric analysis 28. Example of Master Sheet / CHARTThe master sheets explain complete coding process of the collected data.RespondentsSl no.NoVariable labelsAge designationLevelofeducationMaritalstatusNature ofworkDuration ofworkWagesAttitude ofemployer1 A B D C D A A A2 B A D C C A A E3 C A D C E D E A4 D E D C E A A C5 E E A D B E C D 29. Some important points that can bekept in mind while preparing thecoding for computerized data entryare;Use a natural coding scheme:Avoid the blank space as a codingcategory.Do not use the , +, symbols . 30. Presentation andInterpretation ofData.Construction offrequency table: 31. Components of tables:The major components of a table are;HeadingTable numberTitle of the tableDesignation of unitsBodySub-head: heading of all rows or blocks of itemsBody head: headings of all columns or main captionsand their sub-captions.Field/body: the cells in rows and columns.Notations Footnotes, wherever applicableSource, wherever applicable 32. Types of tables;Uni-variate /One way tablesUni-variate analysis refers to tables, which givedata relating to one variable.Uni-variate tables, which are more commonlyknown as frequency distribution tables, howfrequently an item repeats.Further properties of distribution can be foundout by various measures of central tendencies. 33. ExampleThe below table showing awareness of the respondentsLevel ofawarenessDistribution of respondentsFrequencies %High 110 39.3Medium 106 37.9Low 64 21.8Total 280 100.0 34. Bi-variate /Two way table:Distribution in terms of two or more variables and the relationshipbetween two variables are shown in two- way tables.Example: table showing levels of awareness towards the Act and wage Differentialsof the Respondents.Awarenessabout theACTWage differentialsHigh Medium l Low TotalHigh 94(66.2) 99(10.5) 7(13.5) 110(39.3)Medium 37(26.1) 58 (67.4) 11(21.2) 106(37.9)Low 11(7.7) 86(30.7) 52(14.6) 280(100.0)Total 142 86 52 280(100.0) 35. Multi-variate Analysis:If the researcher is interested in assessing thejoint effect of three or more variables, he usesthe techniques of multi-variate analysis.The most common stastical technique used formulti-variate analysis is regression analysis.In the first step of multi-variable analysis, theresearcher has to obtain the correlationbetween the variables which are havingstastically significant correlation.These variables are put in the regressionanalysis. 36. Diagrammatic Representation:A diagram helps to understand the data easily. All stastical packages, MSExcel and Open Office offer a wide range of graphs. In case of qualitativedata most common graphs are bar charts and pie charts.The most commonly used graphic forms may be grouped into thefollowing categories;GraphsHistogram & Bar chartsPie chartPictogramPyramid Diagram 37. Graph:The graph offers a visual presentationof the results.The line graph can also providemarkers for each data point marker foreach variable can be different todistinguish it from other variables. Themarker also highlights the value at anyspecific point 38. 14121086420Category 1 Category 2 Category 3 Category 4Series 3Series 2Series 1 39. HistogramsHistograms and bar charts have two axes that intersect at right angles.One axis (horizontal one) is divided into intervals representing thevarious possible values of the variables. The other axis (vertical one) isdivided into intervals representing frequencies. Histogram depicts thedata graphically with vertical bars adjacent to each others.65432102 4 6 8 10 40. Bar chart:Bar charts are similar to histograms except thatthe bars are separated. Bar charts can be usedfor nominal level data to show differentvariables. Ex, diagram;765432101 2 3 4Series 3Series 2Series 1 41. Pie charts:Pie charts are circular charts divided into slices. Each slicerepresents a possible score and its size is proportional tooccurrence of those scores in the sample. The componentparts from the segments of the circle. The circle chart isusually a percentage chart. The data are converted topercentage of the total and the proportional segments;therefore give a clear picture of the relationship among thecomponents.1.41.23.2 8.21st Qtr2nd Qtr3rd Qtr4th Qtr 42. Pictogram:Each picture represents (man, animal, vehicles, crops,) acertain number and total number of pictures gives the totalnumber of events/ elements. Pictograms are an exciting wayto summaries data and are often used in newspapers.Examples for pictogram; 43. Pyramid diagram;Pyramid consists of many levels and presents one or twovariables. 44. RESEARCH REPORTThe final and very important step in a research study is towrite its report. The research report is a means forcommunication our research experiences to others andadding them to the fund of knowledge. 45. Functions of research report;It serve as a means for presenting the problem studied,methods and techniques used for collecting and analyzingdata, the findings , conclusions and recommendations in anorganized manner.It serves as a basic reference material for future use indeveloping research proposal in the same or related area.A report serves as a means for judging the quality of thecompleted research project.It is a means for evaluating the researchers ability andcompetence to do research.It provides factual base for formulating policies andstrategies relating to the subject matter studied.It provides systematic knowledge on problems and issuesanalyzed. 46. Different steps in writing report:Logical analysis of the subject matter;Preparation of the final out linePreparation of the rough draftRewriting and polishing of the roughdraftPreparation of the final bibliographyWriting the final draft 47. Planning report writing;The target audienceThe communication characteristic ofaudience.The intended purposes of the reportThe type of reportThe scope of the reportThe style of reportingThe format of reportOutline/time table of the report 48. BIBILOGRAPHY;Methodology of research in social sciences Dr.O.R. Krishnaswami, Himalayan publishing house,1999.Research for social workers ,An introduction tomethods, 2nd editionMargaret Alston , Wendy BowlesResearch Methodology Methods and techniques(Second Revised Edition 2004) & third edition 2014.C. R. Kothari, New Age International Publishers.Practice of social research social work perspectiveD.K.Lal DasRawat publication, 2004. 49. THANK YOU