Millisecond Forums

Problem running inquisit 3 program on inquisit 4

https://forums.millisecond.com/Topic14054.aspx

By GeorgeY - 8/24/2014

Hi,

We had a task that was programmed in inquisit 3 and was working normally. We have since upgraded to inquisit 4 but have realised that there has been an error in the stimulus presentation. Specifically, the 18 trials in the task is intended to run so that the participant is presented with them randomly without replacement. We have found instead that the participants have been presented with the trials randomly but some have received the same trials a number of times and other trials none at all (it sounds like it has been presenting randomly but with replacement).

The code is identical to what we were using in inquisit 3. We have tried the error now with a PC and Mac and the same thing happens. I am concerned that this same issue may have occured in other tasks we have been using but I have yet to check.

Could you please advise on whether this is a known issue? Should there be differences in how trials are presented when changing from inquisit 3 to 4? I am just guessing that it is related to the replacement section so I have pasted it below (the same problem happens with the baseline trials) but it could be something else. Could you advise on solutions to this problem?

Thanks- George

***selects the 6 Baseline pattern sets (indices 1-6) randomly without replacement
<counter selectindexB>
/ items = (1,2,3,4,5,6)
/ select = noreplace
/ selectionrate = trial
</counter>

***selects the 18 Experimental pattern  sets (indices 1-18) randomly without replacement
<counter selectindexE>
/ items = (1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18)
/ select = noreplace
/ selectionrate = trial
</counter>
By Dave - 8/24/2014

There is no known issue w/r/t sampling without replacement in Inquisit 3 vs Inquisit 4 -- it should (and according to my tests does) behave identically.

The <counter> elements you provided are insufficient to point to any issue that may exist in your code. Please provide the portions of the code that actually illustrates how they are used -- i.e., how often they are sampled from and when.