Hi I am working with the latest demo version of Superlab 4 (4.0.3c), (have purchased but am awaiting my license number to be emailed).
I am setting up a training program in which the same two trials are presented multiple times until certain accuracy criteria are reached. I have been using a single block with these 2 trials in it and using macros at the block level to keep repeating these trials, e.g. present 12 trials and if accuracy is less than 10 out of 12 correct present another 6 trials etc. I’m using codes that are changed on the basis of correct or incorrect responses.
While the macro (see attached jpg) seems to happily count the 12 trials it seems that it will only use the last 2 trials when working out the accuracy (or however many physical trials there are in a block, e.g. 6 if there are 6 trials that are being repeated). That is, as long as I get the last 2 trials correct it will tell me I have 10 out of 12 correct, even if I got all the trials prior to that wrong.
If I physically have 12 trials in the block (i.e. made by duplicating my 2 trials 6 times) the macro runs fine.
So I’m wondering how Superlab counts the trials in a block if it can work out 12 trials have been presented but not use them in working out the accuracy. I suspect I may be misunderstanding something…
I’m a little confused by your screenshot. If you’re waiting for a license number, you would have to be running under Windows, but the screenshot looks almost like OS X… but not really. Are you running a custom skin on Windows?
Are you marking the trials with a code value? Is this how you are keeping track of accuracy? This is the method we generally recommend, but this method isn’t compatible with your implementation method. This isn’t a bug–it’s a design issue. When you repeat two trials to get twelve total, you still only have two trials. SuperLab keeps track of the number presented independently of the trials themselves. When it looks back, it sees that two trials have been presented twelve times. The number of trials with a given code value can’t be larger than the number of trials that exist–regardless of how many times they’ve been presented.
Instead of duplicating the trials, you could possibly point one of the events at a stimulus list with duplicate items. Then, when SuperLab runs the experiment, it will physically create as many trials as items in the stimulus list. Then your counts will appear as you are expecting them.
Thanks for clarifying how the correct responses are counted relative to the trials and repetitions of the trials in a block. That’s confirmed what I thought and I’ll change the way I try to do it, now that I understand. I’ll try your suggestion of using events with multiple instances of the same stimuli. Ta!
And yes, well spotted, I am using OSX, but with a custom skin. Danielle has told me she’ll email a link rather than a license number. And yes, I am using the April 19 4.0.3 version.
