ThunderingHERD
04-28-2007, 05:22 PM
[Irrelevant backstory]
I'm involved in research attempting to find links between certain endophenotypes consistent with schizophrenia and various SNPs of certain genes. Part of the process involves having the participants fill out a large # of questionnaires, some of which were originally intended to be orally administered by the experimenter. Early on I suggested that it would be much more efficient, and create much less variance on the experimenter adminstered stuff (particularly with regards to "sensitive" questions), to have all of the questionnaires filled out on a computer.
A few days later the prof emails out a link to a survey site that he's gotten demo account on, asking us to enter the questionnaires on the site. It quickly becomes apparent, however, that there are too many branches and contingencies in certain questionnaires to make this feasible. I searched for other survey sites and software and couldn't find anything that did exactly what we needed to do (unfortunately, I wasn't aware of MS Infopath at the time, which seems like it would work well). So I tell the professor that I had enough (rusty) programming knowledge to come up with something that would do exactly what we needed.
So right now I have the program working with 3 the three rather simple questionnaires. If you're interested, want to point out how much I suck at programming, or criticize my awful database design, you can download it at http://www.researchingen.com/files/question.zip (the menu is accessed with the "!" button and the password is "open"; "?"s provide help).
Anyway, I've run into some problems that have been causing me to put off working out the two remaining questionnaires--the ones that really needed to be computerized in the first place. The first problem is that they're just very poorly designed (from a pychometric standpoint), and it's discouraging to even look at them.
[Actual question]
The second problem is that, though I'm relatively competent at VB.NET/2005, I have no previous experience with database design or programming. I figured out enough to get through the first 3 questionnaires, as they were rather simple, but I have some issues with these final two.
I want to store as much of the raw data as possible, but also want to have it in a format where it's easy to compute a final score (we haven't actually decided on how to score it yet, which makes the whole thing more frustrating). For example, one question asks the participant what sports they played as a child, and presents a rather long list of checkboxes with various sports. Internally, the sports are split into three different categories. What would be the best way of storing this in a database? Creating a seperate field for each sport? Given that this sort of question comes up several times, this would create a massive number of fields. Should I create a seperate table for each question like this, and create a field for each response in that table? Or is there some better way to do this?
Any suggestions are appreciated.
I'm involved in research attempting to find links between certain endophenotypes consistent with schizophrenia and various SNPs of certain genes. Part of the process involves having the participants fill out a large # of questionnaires, some of which were originally intended to be orally administered by the experimenter. Early on I suggested that it would be much more efficient, and create much less variance on the experimenter adminstered stuff (particularly with regards to "sensitive" questions), to have all of the questionnaires filled out on a computer.
A few days later the prof emails out a link to a survey site that he's gotten demo account on, asking us to enter the questionnaires on the site. It quickly becomes apparent, however, that there are too many branches and contingencies in certain questionnaires to make this feasible. I searched for other survey sites and software and couldn't find anything that did exactly what we needed to do (unfortunately, I wasn't aware of MS Infopath at the time, which seems like it would work well). So I tell the professor that I had enough (rusty) programming knowledge to come up with something that would do exactly what we needed.
So right now I have the program working with 3 the three rather simple questionnaires. If you're interested, want to point out how much I suck at programming, or criticize my awful database design, you can download it at http://www.researchingen.com/files/question.zip (the menu is accessed with the "!" button and the password is "open"; "?"s provide help).
Anyway, I've run into some problems that have been causing me to put off working out the two remaining questionnaires--the ones that really needed to be computerized in the first place. The first problem is that they're just very poorly designed (from a pychometric standpoint), and it's discouraging to even look at them.
[Actual question]
The second problem is that, though I'm relatively competent at VB.NET/2005, I have no previous experience with database design or programming. I figured out enough to get through the first 3 questionnaires, as they were rather simple, but I have some issues with these final two.
I want to store as much of the raw data as possible, but also want to have it in a format where it's easy to compute a final score (we haven't actually decided on how to score it yet, which makes the whole thing more frustrating). For example, one question asks the participant what sports they played as a child, and presents a rather long list of checkboxes with various sports. Internally, the sports are split into three different categories. What would be the best way of storing this in a database? Creating a seperate field for each sport? Given that this sort of question comes up several times, this would create a massive number of fields. Should I create a seperate table for each question like this, and create a field for each response in that table? Or is there some better way to do this?
Any suggestions are appreciated.