Building digit span test in Superlab

Hi,

I wonder if it’s feasible to build the digit span test in Superlab. If I want to use stimulus list, is it possible to tell SuperLab to pull various number of event items at different conditions? Is there a template to start with?

Thanks
Min

We don’t have a template for a digit spin test.

What are the different conditions?

Sorry for the late response! I think I meant “pull various number of event items according to the feedbacks”. We plan to do the digit span test in an adaptive procedure. Starting from two random digits, if subjects repeat them correctly, then present 3 random digits, and move to 4 digits if they repeat correctly. But if they get wrong, we move to the shorter strings until they repeat correctly again. We stop testing when we find the first digit span that they have two consecutive wrong.

I think the first step for me is to know how to tell SuperLab to pull a certain number of random digit (in audio) from the stimulus list every time I run the trial. I have 9 (1~9) individual audio files in the list. Could you please help?

Thanks a lot!

When you say “if subjects repeat them correctly”, how is the subject responding? can they do it by typing digits 1 to 9 on the keyboard? Can you post the audio files?

digits audio files_1

Yes, you are right, the subject is responding by typing digits 1-9 on the keyboard. Here are the audio files.

Thanks a lot!

eight.wav (82.9 KB)

five.wav (105 KB)

four.wav (88.3 KB)

nine.wav (109 KB)

one.wav (74.2 KB)

audio file_2

Sorry I should have zipped them.

Thanks

seven.wav (106 KB)

six.wav (118 KB)

three.wav (81.5 KB)

two.wav (73.7 KB)

You can certainly build a digit span test with SuperLab, even one that’s adaptive.

Attached is an experiment with such a design.

experiment.zip (472 KB)

The key to this experiment is to use text parameters. One to keep track of what digits have randomly been presented and another to store participant input.

You can then compare these two in order to check for correctness and end condition, i.e. participant erring twice in a row.