html5-img
1 / 22

Chapter 16 Starting the Data Analysis

Chapter 16 Starting the Data Analysis. Winston Jackson and Norine Verberg Methods: Doing Social Research, 4e. Analyzing Data with a Computer. Requires statistical software program Appendix A provides guidelines for using Statistical Package for the Social Sciences, known simply as SPSS

kane
Télécharger la présentation

Chapter 16 Starting the Data Analysis

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Chapter 16Starting the Data Analysis Winston Jackson and Norine Verberg Methods: Doing Social Research, 4e

  2. Analyzing Data with a Computer • Requires statistical software program • Appendix A provides guidelines for using Statistical Package for the Social Sciences, known simply as SPSS • See textbook for step-by step instructions on using SPSS for Windows (Versions 10 to 14) • Labs are available with this text to provide additional learning activities for students using SPSS (www.pearsoned.ca/text/jackson-methods) © 2007 Pearson Education Canada

  3. Rules for Data Entry A. Creating a raw data file (Syntax file in SPSS) • Give each case an ID number. • Number questionnaires or data-collection forms, starting with 001 (or 0001 if more than 1000 cases) on the top right-hand corner. • The ID number is used to link questionnaire with data in case errors are found • Code any uncoded questions • Any open-ended questions or occupational prestige scores that need to be looked up should be assigned values © 2007 Pearson Education Canada

  4. Rules for Data Entry (cont’d) • Do a column count • This will identify the position where each variable will be entered Variable Columns ID 1-3 Record # 4 Blank 5 Gender 6 Yr of Birth 7-8 © 2007 Pearson Education Canada

  5. Rules for Data Entry (cont’d) • Enter data with a partner • Fewer errors with help of another person • Note new page with blanks • Errors can be more easily identified if you leave a blank to mark a new page of the questionnaire • On Figure 16.7 (next slide), it is evident that there is an error on the fourth line • Can go back to case 004, and re-enter data from page 2 © 2007 Pearson Education Canada

  6. Figure 16.7 Data with Blanks Between Questionnaire Pages © 2007 Pearson Education Canada

  7. Rules for Data Entry (cont’d) • Simplify missing value codes • Where possible use a 9, 99, or 999 to indicate situations where a respondent refused to answer a question. • When something is left out but it means zero, use a zero. • In 9-point Likert scales use the 0 to indicate a missing question. © 2007 Pearson Education Canada

  8. Rules for Data Entry (cont’d) • Document research decisions • If a response is in doubt (two numbers circled, two answers when only one was asked for), flip a coin to determine which response will be taken • Circle in red and write your initials next to the decision © 2007 Pearson Education Canada

  9. Rules for Data Entry (cont’d) • Code for information not on questionnaire • When more than one person is coding un-coded data, it is a good idea to enter a code for each data entry person; in interviews, code the interviewers. If systematic differences occur one can then quickly identify whose questionnaires belong to which person • Use double data entry • Can be done to find errors © 2007 Pearson Education Canada

  10. Rules for Data Entry (cont’d) • Save the data • If entered in SPSS, save it as an .sps file • To do so, SPSS will ask if you want to save the file, when you click yes, it is saved as an .sps file • Check for errors • Check for errors prior to data analysis • E.g., using O (letter) instead of 0 (zero) © 2007 Pearson Education Canada

  11. Rules for Data Entry (cont’d) B. Point-and-Click Raw Data Entry • Click on File/New/Data • Move cursor to the cell you wish to start with (the cell will be highlighted) and enter the data for each variable, moving across the row • The value will show on the screen © 2007 Pearson Education Canada

  12. Rules for Data Entry (cont’d) • To insert a new case • Position the cursor on the case below where you want to insert the new case • On the toolbar, click Data/Insert Case • To insert a new variable • Position the cursor on the variable following the spot where you want to insert the new variable • On the toolbar, click on Data/Insert Variable • Enter data for all the cases © 2007 Pearson Education Canada

  13. Creating and Saving an SPSS.SAV File • For modest to large surveys: • Use Syntax Editor to enter commands to define variables, attach labels, and indicate missing value codes • To begin, click File/New/Syntax • To process commands, click on Run • See sample syntax file in Box 16.1 (next slide) • For small surveys (10–15 variables) • Use point-and-click direct-entry method • To begin, click File/New/Data (see pp. 431–434) © 2007 Pearson Education Canada

  14. Box 16.1 Sample Syntax Commands to Create an SPSS System File Title system file creation, Social Science Faculty Review, Winston. Data list file = ‘C:\oia\sscience\social science data.SPS’ / id 1-5 v1 8 v2 9 v3 10 v4 12 v4.a to v4.7 13-19 v5 20 v6 22 v8 23 v9.1 to v9.6 24-29 v10 to v12 30-32 v13 34 v16 35 v17 36 v18.1 to v18.5 37-41. variable labels id “identification number” [label up to 40 characters] /v1 “Year of Graduation” /v2 “Program” /v18.5 “Quiet Study Space”. value labels [each label limited to 20 characters] /v1 0 “1995 or earlier” 1 “1996” 2 “1997” 3 “1998” 4 “1999” 5 “2000” 6 “2001” 7 “2002” 8 “2003” /v2 1 “Major” 2 “Advanced Major” 3 “Honours” /v4.1 to v4.5 0 “No” 1 “Yes”. missing values v1, v2, v3, v4, v4.1 to v4,5, v5, v6, v8, v13 (9) /v4.5, v4.6, v9.1 to v9.6, v10 to v12, v16, v17, v18.1 to v18.5 (0). Frequencies var = v1. © 2007 Pearson Education Canada

  15. Steps in Analyzing Project Data • Run and print FREQUENCIES (nominal and ordinal variables) and DESCRIPTIVES (ratio variables) • Keep for future reference and reporting results • Use RELIABILITY and COMPUTE commands to construct indexes • Select appropriate procedures for analysis • May need to RECODE some variables (create a new variable by adding the letter “r” to the name of the original variable) © 2007 Pearson Education Canada

  16. Steps in Analyzing Data (cont’d) • Run analysis for intervening or source of spuriousness models (see Chapter 17) • Create summary tables (see Chapter 18) • Write your report using appropriate headings (see Chapter 18) © 2007 Pearson Education Canada

  17. The 3M Approach 3M is a way to remember how to decide which procedure to use to analyze relationships • Model: >X  >Y • Measurement: Nominal, Ordinal, Ratio • Method (choice depends on level of measurement; see Table 8.18, next slide): • CROSSTABS • MEANS • CORRELATIONS © 2007 Pearson Education Canada

  18. Appropriate Methods of Analysis by Level of Measurement © 2007 Pearson Education Canada

  19. Getting Errors and Warnings • If you are using the Syntax method of entering SPSS commands, expect to get errors and warnings when running SPSS jobs • SPSS has error checking routines that help you identify the error immediately © 2007 Pearson Education Canada

  20. Tips on Detecting Errors • Expect errors • Normal part of data analysis • Examine error and warning messages carefully • The character or symbol creating the problem will be listed • Make certain SPSS is accessing the necessary files • Fix first errors first • Fix the first error(s) and re-run the job; may correct all other errors or warnings © 2007 Pearson Education Canada

  21. Tips on Detecting Errors (cont’d) • Stuck? Re-enter the command line • Sometimes we cannot see an error (e.g., O (capital letter) instead of 0 (number) [as in v1O versus v10] • Just re-enter the line, and run again • Examine results on screen before sending to a printer • Sometimes we make mistakes that are obvious when we see them (meant to use a re-coded variable for a CROSSTAB, but used ratio level variable by mistake) © 2007 Pearson Education Canada

  22. Tips on Detecting Errors (cont’d) • Double check variable list • Make sure you are using the appropriate variable; this is particularly important when you have done recodes for variable • Check for a premature FINISH command • Forgetting to remove/move a FINISH command when editing a syntax file would result in the part of the job appearing before the finish command being run. • Make sure it’s at the end of the file. © 2007 Pearson Education Canada

More Related