Millisecond Forums

Thanks very much for your reply! I do have Inquisit 4, so will look at the documentation for that attribute now (I found this thread through Googling my question, and didn't realise I was posting in
the Inquisit 3 forum... doh).

Part of this may be possible using Inquisit 4 Web, which allows a script to change the finish page via the <defaults> element's /finishpage attribute. Details are available in the documentation.

With Inquisit 3, I don't see a way to do this.

Thanks to everyone who has posted in this thread, I've found it very helpful!

Does anyone know whether it is possible to use the 'additional query parameters' process to specify finish parameters based on values calculated DURING the experiment?

I'm planning to calculate some overall feedback for my participants at the end of their session (using this process: http://www.millisecond.com/forums/Topic7543.aspx), and if it's possible to output this overall feedback score for each participant as a finish parameter, it would be very useful to me.

My participants are going to be completing the same experiment multiple times as part of a training study, and in an ideal world, I'd like to use their performance on the previous day to determine the difficulty of the experiment on the next day... I was thinking that using the overall feedback score as an input parameter for the next day's session may be one way of doing this. However, I am in way over my head here! If that's not possible, it would still be useful if I could output their feedback value, as it would allow me to track the value each day and encourage them to improve those scores (I won't have time to run data analysis in between each training session and calculate their accuracy manually).

I'm very new to Inquisit, so please excuse me if all of this is totally unrealistic. Any help would be very gratefully received!

You can use alphanumeric ids as subject id (i.e. the query parameter supplied by the survey). Then you will not be able to make condition assignment based on subject id, of course (a numerical id would be required to do that). Thus you need to set /groupassignment to random. There currently is no other way to capture query parameters in the data file other than using it as subject id. It might be worth talking to the survey folks to find out if their app is able to provide unique numerical ids.

Regards,

~Dave

Hi all,

I'm encountering a similar problem, but mine is that the information queried from the survey firm link is a string, which takes the place of subjectID in my datafile.  It then enters everyone's subjectID as '1111111' or something like it.

So, is there a way (a) to allow the query to capture alpha or alphanumeric strings?  And (b) since I don't know how those strings are generated initially, is there a way to then randomly generate another subject number, so I can randomly assign participants to experimental conditions using the <expt> command?

Thanks for any suggestions -

Penny

Thank you [:)] My selection scheme will be unfortunately more complicated so probably I will write to you again later, when I complete the script [;)]

Best greetings,

Kasia

Here's a latin square of order three:

  1 2 3
1 a b c
2 b c a
3 c a b

To run three <block> or <survey> elements a,b,c in one of the respective orders based on subject number / using <expt> and <variables>:

<expt>
/ blocks = [1=one; 2=two=; 3=three]
[...]
</expt>

<variables>
/ group = (1 of 3) (one=a; two=b; three=c)
/ group = (2 of 3) (one=b; two=c; three=a)
/ group = (3 of 3) (one=c; two=a; three=b)
</variables>

<survey a>
[...]
</survey>

<survey b>
[...]
</survey>

<survey c>
[...]
</survey>

Regards,

~Dave

What a relief..[:)]  Thank you!

Kasia

Yes, you can.

Hi there,

I am a little bit confused -  if I use Inquisit for conducting internet survey can I use the <variables> option in
<expt>  for latin square selection (depending on participant's number)?

Best,

Kasia