Possible to run 4-6 trials of the MST on one subject

[Daniel Callow]

Hello,

I am a graduate student at the University of Maryland College Park and I am interested in using the MST in an upcoming pilot study. I will be looking at the effects of acute exercise on hippocampal-related tasks in older adults. 

I would like to administer the MST 4 (or if possible 6) times to each participant (before and after each condition). Some of these tasks would be performed close together, probably no more than 40 minutes apart. From what I understand I could use sets 1-6 for this? It is my understanding that there is not a strong training effect for the tasks and that these 6 sets use different pictures and have similar difficulty? I would, of course, counterbalance the order of the sets between participants, however, I wanted to get some input on whether this task would be suitable for a study design like this.

Thank you for making this task so accessible and for any advice or input you can provide!

Best,

Daniel Callow

PhD Student, Neuroscience and Cognitive Science

Exercise for Brain Health Lab

University of Maryland, College Park

ddcc...@gmail.com

443-254-6298

[Shauna Stark]

Right. We have tried to balance Set 1-6 for difficulty and have minimal overlap in stimuli between sets, minimizing interference between them. We have tested different sets of the MST with just a few minutes delay and not observed a practice effect, although typically our delays are much longer in our intervention studies (1-2 weeks). So, yes, I would would advocate the MST for the type of design you are proposing. If you do end up implementing it, I hope you'll post to the group here if you do observe any practice effects with the brief intervening delay, but we have not observed one.

Hope that helped!

Shauna

[Daniel Callow]

Hello Dr. Stark, thank you for the reply!

When you say there is minimal overlap does that mean there are repeats of some of the images between the different sets? Or that some of the images may be related/similar to objects in other sets?

Have any of your previous papers or work stated that you don’t see a training affect (curious because it might be useful for me to cite when defending my use of the task) or is it primarily anecdotal?

Finally, is there any optimal order to implement the 6 sets (I.e. use 2 and 4 together and 1 and 3 together on the same day?). Based on your previous response and my understanding I would assume not, but I just want to check!

Thank you for the thorough response. I will certainly make sure to notify the group if I notice any training effects.

Best
Daniel

[Shauna Stark]

We have worked hard to balance the sets and have minimal overlap, but it seems like there are still a few items that may overlap by name, but not the exact same images, between sets (e.g. motorcycles). In order to replace them, we need to figure out the similarity score for new pairs, which we are working on. In the meantime, you should be able to randomize the sets in any way - I have no reason to believe that Set 1 and 3 interfere with each other specifically, etc. You can find our immediate test-retest data in Stark et al, (2015, Behavioral Neuroscience) in the Supplementary Materials, which I have attached here.

Best,

Shauna