Thread Tools  
OlgaLazareva
Junior Member
 
Join Date: Nov 2013
OlgaLazareva is on a distinguished road Posts: 4
Correction trials, Posted November 11th, 2013 #1
I am creating an experiment in which I want the participants to repeat the trial if they made an incorrect response until they get it right (correction trial).

I can handle it through the rule "if incorrect response, then restart the trial", which is fine. At the same time, though, I would like to count first presentations only to determine whether a behavioral criterion has been reached. For example, if a participants responded correctly on 8 consecutive trials (first presentations only), then the program moves on to the next block. In addition, I need to record in the data file whether it is the first presentation or not. I think this can be accomplished through the parameters but so far I was not able to figure out how to do that.
Reply With Quote
MonikaT
Senior Member
 
Join Date: Mar 2008
MonikaT is on a distinguished road Posts: 1,279
Posted November 12th, 2013 #2
Attached is a sample experiment that demonstrates how this is done in SuperLab 5 using the parameters feature. Also, the steps are written out below.
1. Create a counter parameter, in the sample it's named "consecutive successes." It should have an initial value of 0.
2. Add a further action to your existing rule. The existing rule you described as being "if incorrect response, then restart the trial." This rule needs to have two actions:
  • The first is to set parameter "consecutive successes" to -1 (negative one)
  • Then to restart the trial
3. Now, create a second rule that states, "if correct response, then increment parameter by 1"
4. The third and final rule is: "if parameter > 7, then exit block"
5. Within the Block Editor for the block that contains these trials, go to the Looping tab, and choose "Keep presenting trials in this block forever"

One important point about using "repeat trial" in a rule, the placement of "repeat trial" in the first rule is significant. This is an action that causes a jump from one point in the experiment's execution to another point.

Whenever a rule's action executes a jump, then all remaining rules attached to that trial are skipped. So for each incorrect response, only the first rule is executed. Then, for correct responses, we end up executing rules 2 and 3.
Attached Files
File Type: sl5 t_1868.sl5 (6.7 KB, 10 views)
Reply With Quote
Thread Tools
Posting Rules
Forum Jump:
Copyright Cedrus Corporation, 1991-2014.
All times are GMT -7. The time now is 01:29 AM.