Dysthymia

Hi, I'm looking for a good programmer to help me with a survey I'm putting together for a grad school project. I need to have it available both online and offline (for the internet-challenged) and to make it as user-friendly as possible. Here's a transcript of a conversation I had with someone (who can't help me right now b/c of other commitments but was nice to offer suggestions) so as to not retype everything:

Please contact me at knotsck@yahoo.com if you are an experienced programmer and can help me with this project in a time crunch (for a fee, of course).

1. The survey has multiple pages so I would need to carry the hidden data from page to page until the end (I have over 200 questions so it is quite extensive).

This shouldn't be that hard to do using session variables. You could also write the responses to a DB after each page rather than trying to pass them all from page to page to page and writing at the end. This way if something goes wrong, you don't loose all the data.

2. One section of the study requires that when respondents answer and hit enter, the options to proceed or stop will be given to them. Also, on this section, I will need to give them immediate feedback if they get the item wrong .... (example, the word "INCORRECT" should appear in big red bold letters if they answer wrong). Not sure how to provide this immediate feedback based on responses.

This can be accomplished via javascript, but that does expose the "correct" script in the page. If it okay for it to take them to a new page after each response, then you could handle this with PHP.


3. On the section above, respondents will only have 15 second to answer the question. I will need to set up a "timer/clock" type thing on the page that is visible to respondents so they know how much time they have left. At the end of the 15 seconds, they will be sent automatically to the next item.

This will probably require either a java applet or a snippet of javascript code to display the countdown timer. From a psychological test taking point of view, keep in mind the potential increase in anxiety that having a timer creates. Is it absolutely necessary to limit responses to each item to 15 seconds? Is that sufficient time for the respondent to answer?

4. At various points in the study, I would like to give respondents the option to continue or save their data and come back to it later. It is quite long so people may not want to sit through it all in one shot.


Good plan, though you'll need to make it clear at the outset which sections they can stop in the middle of. By saving to a DB after each page is submitted, their past results could still be linked by their logon info.

5. I would like to provide a password/username or some sort of method that prevents any random person from filling out the survey. In other words, I'd like folks to fill out by invitation only and to have a way to identify who's filling it out (through the use of password or ID# only as names will not be used in the survey itself)

This is definitely possible and would be a good thing to do. Ideally, I'd suggest creating a table a DB that contains a list of email addresses and randomly generated participant ID#s. You can then have a script that runs to send out the announcement to each person along with a link containing their ID#. During their first login, you can have them create a username/password that will be easier for them to remember on future visits, in case they loose their ID# and/or delete the original link you sent them. You can also have it check the ID# to make sure that the same ID# isn't used twice.

6. I'd like to set something up where the likelihood of a respondent submitting the form more than once is reducted and a way that respondents can only participate in the study one time.

linking the ID to email address is one way to do this... basically, if they don't have the right ID#/original password combo supplied in the link you send, they can't get in... and each combo can only be used once. Placing cookies on the machine is another option, but can create problems if it is possible that multiple people might use the same machine to complete the survey.

7. Most of all, I need to set up a way to reliably get the data and, if at all possible, to have it sent in a way that will be easy to transfer/ import to SPSS.

Not a problem... just have a script to pull the data from the DB and output it as comma delimited.. you can then save the output. The READ ASCII DATA option in SPSS can handle the importing of the data without a problem... you could even have the script add the variable names at the beginning, though creating a separate SPSS syntax file to create the variable names, labels, value labels, etc. might be a better option to save yourself time later on going through and having to enter all that info.