why use 999 for blank survey data in SPSS?

[ehb23]

I have cleaned up my survey data (Likert 1-5) for analysis in SPSS. Some survey items were left blank by respondents. I left these cells blank and ran a factor analysis. A stats savey friend changed all the blanks to 999 and got different results. (she may have done other things differently too)

Is it necessary to change blank data cells to 999. Shouldn't SPSS be able to deal with that?

Thanks

[vinaitheerthan]

Hi,

Sources of Missing values in the research can be
1. No response
2. Not applicable
3. Answer out of range
4. Typo
You should always examine the data and decide how to deal with this missing data. Sometimes you don't know why the respondent didn't answer the question.

You will need to code the missing answers to differentiate the reason for missing.

1) "." - This is called system-missing code.
2) 9 - doesn’t know
3) 8 – does not apply

Missing value codes must be values that are not possible values that the respondent would have used.

The number of digits in the missing value is based upon your variable. If you want code missing data in the age of respondents the maximum number of digits are 3.So you can use ‘999’ to code the missing value as this not a possible value for the rest of the responses of the variable age

Regards,
Re.Vinaitheerthan

[ehb23]

Thank you for the detailed reply... i went into spss and found how to do it. I have heard it is better to do this rather than just leave the cell blank.