[solved] Open Sesame Tutorials

Roger

Hi Folks,

I'm working to have our undergraduate psychology students begin to use OpenSesame as part of their research methods courses. The current tutorial is a good start, but the task is very challenging for people new to this area, so I've started creating a "newbie" tutorial. Please let me know if you have any suggestions, comments, or would like to be involved. :-)

Cheers,
Roger

sebastiaan

Hi Roger,

Another tutorial would be immensely helpful. I don't have any specific suggestions, but I noticed that the following aspects generally cause the most confusion, and may thus warrant some extra attention in the tutorial.

• The loop - sequence structure can be a bit confusing, especially when there are multiple layers.

• The prepare - run strategy can also be a bit confusing. For example, providing feedback through a sketchpad generally doesn't work, because the sketchpad is prepared in advance (ie., before a keypress has been collected).

Just out of curiosity: What type of task will you use for the tutorial?

Regards,
Sebastiaan

sebastiaan

Come to think of it: It may be best to design the tutorial around the upcoming 0.25 (the differences with 0.24 are minor, but especially for novices minor differences may throw them off). I hope to have a Mac OS test-package of 0.25-pre10 soon, in which case I'll let you know.

Roger

My first attempt was walking students through building an experiment - it didn't go very well (excessive cognitive overload).

On my second attempt I provided the students with a "big picture" description of a Human Factors study (a modified version of the affordances example). Then they each participated in the experiment. Finally, we examined the reaction time output data in R.

One challenge was deciphering the CSV output file. It took me a little bit to realize that the leftmost columns were running totals, not the more typical trail-by-trail data. Also, the output file has 49 columns, but we were only interested in 4. Would there be a relatively easy way for the output to be structured so the "important" (e.g., correct, response time, etc.) data is located in one place (e.g., leftmost columns) while the "housekeeping" output (e.g., background color, looping logs, etc.) placed somewhere else (e.g., rightmost columns)?

Okay, I'll work on the tutorial with the Mac version of 0.25.

sebastiaan

Would there be a relatively easy way for the output to be structured so the "important" (e.g., correct, response time, etc.) data is located in one place (e.g., leftmost columns) while the "housekeeping" output (e.g., background color, looping logs, etc.) placed somewhere else (e.g., rightmost columns)?

Haha, well, no, I'd like to stick to an alphabetical order, otherwise it will become impossible to find the relevant data! But you can select in the experiment which variables you want to log (in the logger item). By default, it logs everything, kind of as a safeguard against forgetting important variables. But you can also select only a few variables, to make the data-file easier to process.

Roger

Ah, I didn't realize that I could select the variables to log - that's the perfect solution! :-)