Number Memory Test


I want to set up a fairly simple number memory test experiment as follows but am having trouble working out how/ if I can do it in S/Lab 4.08:

I want to present a set of 5 numbers, no repeats, shown on the same screen in the form XXXXX, chosen from 0-9.

I then want to present 40 numbers chosen randomly from 0-9, repeats allowed, where the participant will need to press 1 if the number was in the list and 3 if the number wasn’t in the list. Data to be captured will be correct/ Incorrect response and reaction times.

Any help gratefully recieved!

Use a Trial Variable to display your numbers in the form of XXXXX. You will need to select Screen Location as the type of trial variable. Create your screen location levels for the 5 numbers you want to be displayed. Then, create 5 more positions that are not visible on screen. After you have 10 level positions created, go to the “Level Selection” tab. Under, “Use the following order:” select Random With No Replacement.

All 10 numbers need to be in a separate event. The first 9 events need to end “Immediately after the event is presented.” The last event should end after a correct response or a time limit.

In the Event Editor, for each of your 10 events, go to “Settings,” then “Position.” Under “The position of the picture is:” select Variable. Then, under “Select the variable” choose the Screen Location Trial Variable you created.

Under Presentation Option in the Event Editor:
Event 1: Erase Screen before presenting stimulus and Keep stimulus invisible is selected.
Event 2-9: Only have Keep Stimulus invisible selected.
Last event: Erase Screen before presenting stimulus and Keep stimulus invisible are NOT selected.


That’s great, thank you - I have set up that as you have outlined and I now have my instructions followed by a display of five numbers randomly selected from 0-9. Next I need to run 40 trials which randomly select from 0-9, compare with the 5 displayed numbers here and determine whether then number presented was in the original list. I’ve looked at Expression Editor and Macro and can’t find a way to compare the presented number with the original list - would I assign a code to each number to do this? I can understand how I would do this in Visual Basic type code - but doing it in Superlab is proving more tricky!

This can be done in SuperLab, but it’s a little involved. Before getting into details what exactly do you want to accomplish? Do you want to present feedback to the participant? What data do you need to gather from this?

Ok, as I said in the initial post I want to carry out a simple test of working memory - using 5 numbers in this case.

The participant is presented with the 5 numbers, randomly selected (no repeats and displayed XXXXX)from 0-9, which I have now programmed thanks to your help (see attached file). They are told to remember the numbers.

They are then presented with a number of trials - 40 in this case - where they are shown a number randomly selected from 0-9 (repeats allowed) and asked to press left mouse if the number was in the original list and right mouse if it wasn’t (or any other key combination). Participant will be given feedback if the selection was incorrect but it will just go to the next trial if correct. The feedback may not be neccesary - that is currently the subject of some discussion! No overall feedback will be given to the partcipant at the end of the experiment - just a “Thank you for taking part” message.

The Superlab data file will need to contain:
The original 5 numbers
Whether responses are correct or incorrect
Reaction time

I hope this is enough information - thanks for your help with this Monika

NumberMemorytest04.sl4 (12.6 KB)

It makes a big difference if you will need to present feedback. Have you come to a conclusion regarding this?

Yes, we will need feedback if the number is incorrect - however, if that makes the programming too difficult it could be lived without…

Hi Liza,

Monika ran your question by me and I told her that it can be done. Upon thinking about it some more, I came to realize that it cannot. Sorry for the confusion.

Hi Hisham,

Many thanks for your reply.

Just to be clear:

The storage of the 5 random numbers and the comparison of the further trials to them is what can’t be done? And this is regardless of whether we require feedback to be given to the particpants or not?

Displaying 5 out of 10 stimuli is not a problem. Having participants respond in the next 40 trials and storing their responses is not a problem either. But providing the feedback is.

Hi Hisham,

Ok, have spoken to the people in question and on consideration of what you say they have decided that feedback during the trials is not as important as they first thought! So if you could help me to program the 40 trials as specified that would be great!

Please see the attached experiment.

NumberMemorytest-Revised.sl4 (14.7 KB)

Monika, Thanks very much for that - It presents what is required on the screen but the data collected always has C as the error code - therefore it isn’t coding responses as incorrect and therefore doesn’t seem to be comparing the 40 presented numbers to the 5 initial presented numbers - is this the part that Superlab can’t do? When Hisham said this could be done but no feedback given I presumed he meant no feedback to the participants at the time rather than data collection…

This is what Hisham meant when he asked about feedback. You can get around this by creating an Excel macro that recodes the error code correctly. Unfortunately, we do not have enough in-house experience to help guide you through the process.

Thanks Monika… I’m pretty good witrh Excel, however, looking at the attached Superlab output file I can’t discern any way of working out what a/ the inital presented numbers were or b/ what the numbers presented for the participants to respond yes/ no to were… If I could do that then writng the excel macro to code the reponse yes/ no would be a simple process…

Or, did you mean that the experimenter would have to write down the presented numbers and then input those into Excel to sort out the correct/ incorrect responses?

Lizatest2.txt (2.65 KB)