create an ID from demograpic data provided on a survey to match pre and post responders spss 14.0

i'm doing a survey with juvenile detention detainees and want to develop and extract an ID number from the survey responses without the responders suspecting an ID is being created on them

sneaky pete stuff--paul
PaulStuartAsked:
Who is Participating?
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

PaulStuartAuthor Commented:
id number creation from survey using spss 14.0
8080_DiverCommented:
Well, it is going to somewhat depend on what data you are collecting and, absent that information, I can only suggest that you figure out some unique set of data that will occur in each of the surveys for any given individual and then use that to develop the ID.

Can you provide more information on what data will be gathered?

If you are having the survey's printed, you might be able to have a unique ID printed at the bottom of each page as though it were a form number and then, assuming you could make sure you give the same "form number" to the same person each time, you could use that for your ID.
richdiesalProfessorCommented:
Hi Paul -

The ethics of what you're suggesting aside, there is no automated way to do what you're describing in SPSS 14.  There are things you can do, however.

Here is the basic technique:
1) For each demographic value, assign a unique base-2 number to that value.  
For example, if you had two demographic variables, race and gender, you could code it like this:
1 - female
2 - male
4 - white
8 - asian
16 - latino
(etc)

2) Use SPSS's "recode into different values" function to create entries for each of these - for example, you'd have a "female" variable (worth 1 if they were female, 0 if they weren't), and an "asian" variable (worth 8 if they were asian, worth 0 if they weren't).  

3) Add all of these values together.

The reason you use base-2 values is that no sum of lower values can equal a larger value.  For example, a white male is a 6, a latino person who did not report their gender a 16 - no other combination of demographics can produce each value.  Technically speaking, there are other values you could use (because a person probably cannot be simultaneously white and asian, for example) but for your purposes, sequential base-2 values all the way up will be fine (if you have lots of values, you could have something attached to .

4) Use the Add Variables procedure to match cases based on this ID number.

Depending on your dataset, there might be multiple entries with identical ID numbers - this will be because two (or more) people have identical demographics.  There is nothing you can do about this - these cases will be unidentifiable.

I also want to note that it is impossible to ever be sure that you have the original people matched to the new ones, even if you do go about this procedure.  There is always a possibility, for example, that there are two people with identical demographics - person 1 completes the pre-test while person 2 completes the post-test.  You would never know this.  The only sure-fire way to avoid this problem is to code the surveys with their identities somehow (there are many ways to do this where it is not obvious to the survey-taker that you are doing this, both online or on paper, but from the sound of it, you already have the data).

You should also be aware that survey takers are not 100% consistent in their responses to demographic items across administrations (and I suspect juvenile detainees may be even more erratic).

Hope that helps!

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
8080_DiverCommented:
To build on richdiesal's response, If you use the technique described but you need to have a unique identifier for a given individual, you could incorporate some additional questions that would provide a greater probability of achieving uniqueness.  For instance, if you identify Juvinile Detainees as those being of age 18 or less, then you could obtain their birth month (01 through 12) and their age (01 through 18) and append those to the value that richdiesal's process provides.

Mike McCrackenSenior ConsultantCommented:
This question has been classified as abandoned and is being closed as part of the Cleanup Program. See my comment at the end of the question for more details.
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
Databases

From novice to tech pro — start learning today.