mix up tablerows

erynj
New Member

Posts: 9

mix up tablerows Apr 19, 2020 4:43:26 GMT

Quote

Post by erynj on Apr 19, 2020 4:43:26 GMT

Hello,

I am trying to design an experiment which has a block that proceeds as such {picture1 -- picture2 --response.}. This is a short experiment so there are only 8 possibilities for picture 1 and another 8 for picture 2. So I have a table with a column for picture 1 (rows a - h) and a column for picture 2 (rows a - h). I realized that when the block is run, it uses one table row at a time, meaning that picture (a) from the picture 1 column will always be presented before picture (a) from the picture 2 column. (If I am understanding this?)

Is there a way to randomize these so that (b) from picture 1 might be presented before (d) in picture 2, etc.? Or should I create all the different possibilities as rows in a table even though I only need 8 trials?

I hope that makes sense! Thank you!

PsyToolkit
Administrator

Posts: 545

mix up tablerows Apr 19, 2020 7:50:34 GMT matia likes this

Quote

Post by PsyToolkit on Apr 19, 2020 7:50:34 GMT

1) Indeed, you can only use 1 row per trial, and only one table per task

2) Creating all 64 (8*8) possibilities is the easiest way to deal with this, indeed. There is actually a tool in PsyToolkit that helps you with quickly creating such tables: Click "table" in the left "create" menu, and read the guidance, that might help)

3) Alternatively, you could code it without tables. An example of how to select stimuli randomly can be seen in the Nback experiment (also in others, but this is the one I can think of): www.psytoolkit.org/experiment-library/nback2.html

Note that all bitmaps in the bitmap section are actually simply numbered 1, 2, 3, etc (from the computer's point of view -- it knows the "names" of bitmaps, but internally, it simply replaces these with numbers starting from 1). So the two examples below do the same. Look at the example below, in which there are 3 bitmaps, a smiley, a tree, and a house. First, you get the easy conventional way of coding:

bitmaps

my_smiley

tree

house

task

show bitmap tree

Now you can do the same as follows:

bitmaps
my_smiley
tree
house

task
show bitmap 2

The reason is that the "tree" word is actually just replaced by the computer with a number. You can use this to randomly choose one of the 3 bitmaps as follows. In the example below, you set a new variable called "$my_random" to a random value between 1 and 3. Then after that, one of the three bitmaps is show without using a table.

bitmaps
my_smiley
tree
house

task

set $my_random 1 3

show bitmap $my_random

Now, you can make this even more sophisticated by choosing stimuli from different lists. Say, for example, that in your specific study, you want picture 1 from the first list of 3 pictures in your "bitmaps" and picture 2 from the second list of 3 pictures, you can do that as well. For easy of coding, let's say the first three pictures are pictures of types of vehicles (that is, pictures 1-3) and the second are pictures of types of buildings (4-6):

bitmaps
bus

truck

tractor

house

flat

tower

task

keys x m

set $my_random_vehicle 1 3

set $my_random_building 4 6

show bitmap $my_random_vehicle -200 0

show bitmap $my_random_building 200 0

readkey 1 5000

save $my_random_vehicle $my_random_building RT STATUS

You can even make this more sophisticated by not allowing the same number as in a previous trial and so on.

Thus, with this way of coding you get a lot more flexibility.

Generally, the only downside of this type of coding is that when you change the list of your bitmaps, you need to make sure the numbers still match the order as used in the bitmaps section.

In principle, you could also write the random line above as follows: set $my_random_vehicle bus tractor

because, for example, "bus" to the code is just meaning "1", given that it is the first bitmap listed in the bitmaps section, but I think that may be a bit more difficult to understand when reading back the code later.

Last Edit: Apr 19, 2020 8:05:30 GMT by PsyToolkit

erynj New Member Posts: 9	mix up tablerows Apr 19, 2020 17:19:00 GMT Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by erynj on Apr 19, 2020 17:19:00 GMT Thank you so much for taking the time to explain this. It's exactly what I needed to see.

O
Guest

mix up tablerows Apr 20, 2020 12:16:57 GMT

Quote

Post by O on Apr 20, 2020 12:16:57 GMT

Hello,

I think I have a similar problem as erynj. I would like to design an AMP experiment. A trial looks like follows: prime - blank - target - mask - rating. It is important that both prime and target are presented randomly. I have tried to do this as described above. However, when I set the variables I added "random" before the numbers (it didn't work before):

set $my_random_prime random 1 60

set $my_random_target random 61 120

So far the experiment works. But it is important that no prime and no target is shown repeatedly (they should only be presented once)- how do I do that? (I tried "all_before_repeat" in the tasklist, but there were still repetitions)

Thanks a lot!

PsyToolkit Administrator Posts: 545	mix up tablerows Apr 20, 2020 13:30:21 GMT Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by PsyToolkit on Apr 20, 2020 13:30:21 GMT Yes, in order to do that you can build in a while loop. I will write a quick lesson and post it later today!

PsyToolkit
Administrator

Posts: 545

mix up tablerows Apr 20, 2020 14:26:09 GMT matia likes this

Quote

Post by PsyToolkit on Apr 20, 2020 14:26:09 GMT

I have now added a new lesson to explain all of this. Could you please check it and let me know if that answers all your questions?

www.psytoolkit.org/lessons/no_table.html

If you like PsyToolkit, don't forget to subscribe to it on Twitter to get the latest updates and to like it via Facebook:

twitter.com/psytoolkit

www.facebook.com/pages/PsyToolkit/152198261589443?ref=hl

Last Edit: Apr 20, 2020 14:26:28 GMT by PsyToolkit

erynj
New Member

Posts: 9

mix up tablerows Apr 20, 2020 19:30:05 GMT

Quote

Post by erynj on Apr 20, 2020 19:30:05 GMT

I ended up with the same issue as the above question, in that I didn't want any of the images to repeat. The tutorial is very helpful to see, thanks for adding it!

If you want to make it so that an image never repeats at all (not just consecutively), is the solution to keep adding code to the while loop(s) to make it check for all previous trials?

Thanks! (subscribing on social media now.)

O
Guest

mix up tablerows Apr 20, 2020 20:05:18 GMT

Quote

Post by O on Apr 20, 2020 20:05:18 GMT

Thank you very much!
But if I understand the while loop (as explained in the lesson) correctly, it only prevents a stimulus from being repeated in two subsequent trials.
I'm afraid I need it a little more complicated: in my experiment, each of the 60 primes and 60 targets should only occur once, in other words, once a stimulus has been presented, it should not be presented again during the rest of the experiment.
In the end, all primes and targets should have been displayed once, and the combination of prime and target should be random (that is why I cannot use the table + "all_before_repeat" option, which would otherwhise do the job). I hope this is somehow possible?

PsyToolkit Administrator Posts: 545	mix up tablerows Apr 20, 2020 20:51:01 GMT Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by PsyToolkit on Apr 20, 2020 20:51:01 GMT That is indeed impossible to code, I am afraid, unless you use tables.

Post by erynj on Apr 19, 2020 4:43:26 GMT

Post by PsyToolkit on Apr 19, 2020 7:50:34 GMT

Post by erynj on Apr 19, 2020 17:19:00 GMT

Post by O on Apr 20, 2020 12:16:57 GMT

Post by PsyToolkit on Apr 20, 2020 13:30:21 GMT

Post by PsyToolkit on Apr 20, 2020 14:26:09 GMT

Post by erynj on Apr 20, 2020 19:30:05 GMT

Post by O on Apr 20, 2020 20:05:18 GMT

Post by PsyToolkit on Apr 20, 2020 20:51:01 GMT