|
Group: Forum Members
Posts: 3,
Visits: 2
|
Hello community,
I've got a problem here. We've been recently using Inquisit for an IPT (Initial-preference Task).
There is a methodical mistake though; the questions have not been labeled properly (radiobutton, scale from 1 to 7)
Example:
<radiobuttons iptL> / caption = " L" / options = ("1", "", "", "", "", "", "7") / orientation = horizontal / required = false </radiobuttons>
As you can see, we've labeled only the extreme options, thus we find only "1", "7" and a major amount of free spaces in the Inquisit data.
The data simply MUST have been recorded somehow, the current problem is that we have no clue how to differenciate the options 2 to 5.
Is there a way to make use of our data or could you say the experiment is screwed?
Greetings, V.
|
|
Group: Administrators
Posts: 13K,
Visits: 108K
|
> The data simply MUST have been recorded somehow, the current problem is that we have no clue how to differenciate the options 2 to 5.
Well, Inquisit, I'm sure, has recorded what you instructed it to record. And that, unfortunately, was to record the exact same "value" -- "" -- for options 2 through 5. There simply is no way to differentiate between those given your setup.
For the record, there would have been a proper way to do this, and that's by using the /optionvalues attribute as in
<radiobuttons q1>
/ options = ("1", "", "", "", "", "", "7")
/ orientation = horizontal
</radiobuttons>
<radiobuttons q2>
/ options = ("1", "", "", "", "", "", "7")
/ optionvalues = ("1", "2", "3", "4", "5", "6", "7")
/ orientation = horizontal
</radiobuttons>
<surveypage mypage>
/ questions = [1=q1; 2=q2]
</surveypage>
<survey mysurvey>
/ pages = [1=mypage]
</survey>
which I recommend you run to see the difference.
|
|
Group: Forum Members
Posts: 3,
Visits: 2
|
Thank you, Dave! This is pretty helpful.
Is there a way to save the previous data...? I mean, the data is there in a way.
Greetings, V.
|
|
Group: Administrators
Posts: 13K,
Visits: 108K
|
> I mean, the data is there in a way.
What exactly do you mean by that? As already said and (hopefully) illustrated by the example code I posted, there is *no distinguishable data* for options 2 to 5. It's not possible to "recover" something that was never recorded to begin with, unfortunately.
|
|
Group: Forum Members
Posts: 3,
Visits: 2
|
Okay, I hoped there was some way.
Thanks for your help, Dave
Greetings, V.
|
|
Group: Administrators
Posts: 13K,
Visits: 108K
|
I'm sorry this happened to you. One recommendation for the future, if I may (this isn't to add insult to injury): Proper testing prior to live data collection probably would have caught the mistake / prevented this from happening.
|