Millisecond Online Community

    I'm currently designing a survey with 16 radio button questions. Each of these questions is multi choice with five possiable answers. I have two questions regarding this ....

1) How do i assign a value to each answer (e.g. so if the subject clicks on the first answer that will be recorded as 1 etc...)

2) At the end of the survey is it possable to give participants different feedback bases on their average score across all questions (e.g. if they score on average 3 or more points they will get feedback X and if they score less than three point they will get feedback Z). If so, how?

Any help much appreciated!

Hi there,

(1) Assigning values to the response options is done via the '/optionvalues' attribute as in

<radiobuttons example>
/ caption="Do you know what I'm talking about?"
/ options=("Yes", "No", "Maybe")
/ optionvalues=("1", "2", "3")
</radiobuttons>

(2) Yes, this is possible. You will have to use the <values> element and some 'expressions' to keep track of the participants' scores and then assign the appropriate feedback with some 'if...' statements. For starters, you should read through those topics in the Inquisit help-file. I also recommend you to take a look at the Single-Target-IAT script available from the samples page as a very good example of how this can be done.

Best,
~Dave

P.S.: I've attached a very simplistic script that demonstrates the tracking and scoring of values in a survey. It has 4 <radiobuttons> elements with each 5 response options on a single page. The achieved average score is reported on the second page. This might give you a more concrete idea of how to get what you want.

Hi,

I have a similar question, sort of, and I'm having the devil's own time figuring how to go about doing it.  Here's what I want to do:

I have 21 items, questions really, on media usage.  In the instructions I'm asking respondents to indicate "how often you've used a particular news, information, or entertainment source in the previous 7 days."

For each item there are the same options. 5. Used 5 or more times in previous 7 days; 4. Used 3 or 4 times in the previous 7 days; 3. Used 1 or 2 times in the previous 7 days; 2. Have used service in the past, but in the last 7 days; 1. Have never used. 0. Don't know.

In a regular web survey, I would use something like a grid, with the items running down the side, the options running across the top, and radiobuttons next to each item.  This way, respondents don't have go through numerous pages.

At the the end of the survey respondents will go to a PictureIAT task.

I've been going through the documentation trying to figure this out but for someone with limited programming experience, it's a little daunting.

All help gratefully received,

Ben

Hey Ben,

although I'm having a hard time finding the similarities between the original question raised in this thread and yours, I'll do my best to help you... ;-)

Inquisit doesn't have any kind of matrix-type question layout, but with some not so pretty trickery you can mimick the page layout you're looking for in Inquisit.

I've attached a small file containing a single surveypage with three questions to show you one way to get what you want.

Hope this helps,
~Dave

I guess to a non-programmer like myself, layout is layout, radiobuttons are radiobuttons, and so on.  Due apologies if the question went on the wrong thread.

The file attached does the job wonderfully (looks pretty to me, again, a non-programmer).  Much appreciated.  Gracias, danke, thanks.

Ben

No need to apologise, Ben. The remark was solely meant to be humorous... I hope the script I attached to my previous post is self-explanatory enough. If you continue to have questions, come back and post them - either here or in a new thread.

Best,
~Dave

Hey Dave,

Might as well keep these questions in the original thread. 

The file you attached works fantastically well as long as the response labels across the top are 1 letter wide.  The challenge here is to have labels such as "5 or more times in the previous 7 days" ; "Used 3 or 4 times in the previous 7 days"; and so on... and have each of these wordwrap in a square that's about 8 spaces wide, while keeping all of them together across the top.

I figured out how to make the rest of the questionnaire line up, no problems there.  But I can't get the captions to work.

preciate the help,

Ben

PS. I'm attaching the script file (which you wrote, little or no mods there), and a MS Word file that has all 21 questions in a Matrix format.  Obviously I wouldn't want to have all 21 on a single screen, only 5 - 7 but I'm sure I can handle that. 

Any chance you could attach your script to this thread (instructions here)? That would make things easier to troubleshoot...

Best,
~Dave

... and the script file.  Since apparently I'm can't attach two files to the same post.

As I said, it's not pretty. I've attached a revised version - it all comes down to using a seperate <caption> element for each response option and then fiddling around with their screen positions until it's looking about right.

~Dave

Maybe it's not elegant, but it does the job.  I'll take it gladly!

Gracias, thanks, viehlen dank, grazie mille... and so on.

Hi again,

Im still strugling with how to do branching in a survey! (i've never done scripts before and my brain just doesn't work that way apartently!) Was wondering if anyone has a e.g. of a survey script that contains branching?

thanks muchly

See the attached file for an example based on the demographics sample included with Inquisit. In this survey, participants are directed to different follow-up pages (i.e. branching) based on a previous response expressing their political views. You can do the same thing based on any custom <values> element used to track participants' scores throughout the survey (see previous posts).

Hope this helps,
~Dave

I nested "survey page" elements into "block" elements which are nested in an "experiment".  I do not use the "survey" element at all. 

So, my survey is now technically an experiment, correct?  This means the way the data is written to file is different, correct? 

What element should I use to choose and arrange the responses that are written to file?  Or, should I?

Thanks

 

Pete R:

I nested "survey page" elements into "block" elements which are nested in an "experiment".  I do not use the "survey" element at all. So, my survey is now technically an experiment, correct?

In Inquisit's terminology your survey is an experiment. Correct.

Pete R:

This means the way the data is written to file is different, correct?

Correct. The <survey> element writes datafiles containing a single line per subject. The <expt> / <block> element writes datafiles containing multiple lines per subject.

Pete R:

What element should I use to choose and arrange the responses that are written to file? Or, should I?

Currently, data output for the <survey> element cannot be customised. As you are using the <expt> / <block> element instead, you can (re-)arrange data output via the <data> element. Do you really need this? Depends on what you're looking for... I'd recommend generating some test data (for example using Inquisit's built-in <monkey>) using Inquisit's default data recording scheme (i.e. without specifying a <data> element first). Then look at the output files, decide what -  if anything - needs to be rearranged and specify an appropriate <data> element.

Pete R:

Thanks

You're welcome.

~Dave