Data Screening: Peek-a-Boo!

Preparing SPSS datafile for analysis is not an easy task. This is in fact the most critical stage because I have to use full concentration and be as thorough as possible throughout the whole process. Phewww...!

Although I am thankful that I can skip the data entry stage (because I paid someone/so called RA to help me with that daunting task during my fieldwork), I still have to go through the whole process of checking and screening my survey data. So, it took me nearly three months (Since 16 Nov 2009 - January 2010) just to do the screening, the checking, identifying 'errors', going back to the individual questionnaire to correct these errors, and entering the correct response into the SPSS datasheet. Here are the specific work and date that I did which I recorde in my SPSS log book:

1. 16 Nov 2009 - checking the codebook
2. 17 Nov 2009 - checking for errors: Mailed survey data (Section A)
3. 18 Nov 2009 - cont. checking for errors (Section B)
4. 19 Nov 2009 - working with online survey data: recode from excel into spss
5. 1 Dec 2009 - working with mailed survey data: checking interviewees' responses
6. 2 Dec 2009 - locate BIG mistake in spss datafile: 36 redundant ID (data entry)
7. 3 Dec 2009 - re-checking for errors in mailed survey data
8. 9 Dec 2009 - data entry 6 newly arrived questionnaires (ID364 - 369) - consultation with Dr Sharil
9. 10 Dec 2009 - data screening: online survey data
10. 15 Dec 2009 - compute total scores, compute missing values.
11. 16 Dec 2009 - working with mailed survey data: follow-up with 7 respondents to ask for their responses to account for missing values
12. 21 Dec 2009 - data entry: got one more questionnaire via mail. Total mailed survey: 370. Then, proceed with checking for outliers and preliminary analyses
13. 22 Dec 2009 - Read JP and run visual displays of data.
14. 24 dec 2009 - working with onliine data: total collected was 268 but afer the cleaning process, I got 141.
15. 30 Dec 2009 - Run preliminary analyses and reliability analysis using online data
16. 31 Dec 2009 - Finally got my 'pure-clean' data: Mailed (369) and online (139) equal 508!
17. 7 Jan 2010 - Read Kline (2005) chapter 3: Data preparation and screening and re-run data screening again to double check accuracy (mailed data).
18. 8 Jan 2010 - re-run data screening with online data
19. 15 Jan 2010 - Practice stats analyses with 3 data files: mailed, online and combined.
20. 19 Jan 2010 - run SPSS tutorial
21. 21 Jan 2010 - Analysing open-ended comment boxes in Section A (manually!!!)

The biggest challenge that I encountered when I found that there were 'redundancies' in the participants' ID. There were about more than 30 cases that I had to check and double check. This means that I had to go back to the source - my collected questionnaires - to locate the 'right response' and the 'right case' to match those already enetered into SPSS datafile.

Overall, data screening was a very time-consuming process. I thank you so much Julie Pallant for guiding me in a step-by-step manner during the process of data screening and checking. I recommend others to use Julie Pallant book to learn the whole process of preparing, analysing, and reporting quantitative data for their research!