Item stack

Lorraine

Hi,

I wrote an experiment a few weeks ago. It could run. But recently, the experiment cannot run anymore (no warning, just freeze, so I have to abort the experiment). The detail said:

I use Mac, and after I updated the software, the problem is still there. Could you please maybe let me know what should I do?

eduard

Hi Lorraine,

that sounds like a bug. Could you try to reproduce the error? Like building an experiment that is as simple as possible, but still shows this unwanted behaviour? Then could you share the experiment here?

Oh, and you updated to Opensesame 3.3.6?

Thanks,

Eduard

Lorraine

Hi, Eduard,

Yes, I am using the latest version.

I also try to run an old experiment and it worked. So, probably it is something related to my code. Could you maybe help me check it?

To blink or not blink.osexp

eduard

Hi Lorraine,

The code runs fine on my machine (Ubuntu 18, with Opensesame 3.3.6), so it probably really is a bug related to the Mac version of Opensesame. Have you tried it with all backends or only with legacy?

In this case, I would indeed report it as a bug on the github page, unless @Daniel or @sebastiaan are already aware of this problem?

Eduard

sebastiaan

Hi @Lorraine ,

The experiment looks fine to me. The error message indicates that you were able to abort the experiment by pressing Escape and then Q, which means that it didn't really freeze (otherwise this wouldn't have been possible either). Could you describe in a bit more detail what exactly happens?

— Sebastiaan

Lorraine

Hi, Sebastiaan and Eduad,

Thank you very much. It works on my computer now, somehow...

Now, I have a new problem. I would like to add some buttons as a multiple choice response. However, I wish the text on the button could be changed based on the previous variable. Is that possible to define that in the form base plugin or use python under opensesame environment? If no, could you please recommend me how to make that work?

Best,

Lorraine

lvanderlinden

Hi @Lorraine,

Because the text on the buttons is defined online, you will need to create the content of the display in the Run phase of your experiment.

I see two options:

1. You could use a feedback item (instead of a form_base item) to create the buttons, followed by a mouse_response item to collect the click. (For the difference between feedback items and sketchpad items, see: https://osdoc.cogsci.nl/3.3/manual/stimuli/visual/)
2. Alternatively, you could create your form using Python inline scripting and place your code in the Run-phase tab of the inline_script item. For more information, see: https://osdoc.cogsci.nl/3.2/manual/forms/custom/#creating-forms-using-python-inline-script

For more information on OpenSesame's Prepare-Run, see this article:

• https://osdoc.cogsci.nl/3.2/manual/prepare-run/

I hope this helps. If you want more specific advise, please try to describe the behaviour you want to achieve as precisely as possible.

Cheers,

Lotje

Lorraine

Hi

Yes, I have tried to draw buttons using the python inline script (your second suggestion). However, I am still lost how to add dynamic text on the button in this way.

In my experiment, I randomly selected two letter stimuli from 8 letters. In the response page, there are 4 buttons. One button displays the two stimuli, while three buttons display unrelated letters. Irrelevant letters are also selected from those 8 letters. The button displaying the correct letters is also randomly selected.

Also, I wonder how the response could be recorded in this process?

Thank you for your help!

Best,

Lorraine

lvanderlinden

Hi @Lorraine

I made an example script for what I think you are trying to achieve (see comments for explanation). Does that help?

6761.osexp

Cheers,

Lotje

Lorraine

Hi, Lotje,

Thank you very much for your code! It helps a lot.

I changed a little bit to fit my experiment.

It is weird that sometimes this code works, sometimes it reports an error as you can see above. In the trial that works, the irrelevant_pair can be printed, whereas when it reports an error, it cannot print the pair.

It is really confusing. Thank you for your help. Hope to hear from you soon.

eduard

Hi Lorraine,

Can you try to print var.irrelevant_pair before assigning it to var.text1 - var.text6? I would be surprised if this assignment would work when irrelevant pair doesn't have as many items as you have var.text variables (so 6). You can also share your experiment if you wish.

Eduard

Lorraine

To blink or not blink.osexp

Hi,

It's got the same error when printing it beforehand. So I uploaded the experiment.

Best,

Lanting

eduard

The problem is that your construction with if's and elif's does not cover all possibilities of target letters. So, if none of the 4 cases occurs (so, T1 is not S or D, and T2 is not F and G), then var.irrelevant_pairs, will only have one item. Then, when you try to assign it to the six values, you get that error. So what you need to do is to make sure that var.irrelevant_pair will always have six items before you assign its values to the six text variables, probably by manipulating your if-elif-else structure to cover all possibilities.

Fyi, changing the order of the print statement and that assignment won't prevent the error message, but it does print out the value of var.irrelevant_pair, in all cases, so that you can see in which cases the error occurs. For example, in my case it occurred when the var.T1 = 'F'.

Hope this clears things up.

Eduard

Lorraine

Hi,

Thank you very much! Now the experiment works on my computer.

Now, due to the corona, I would like to run the experiment online. Since I am not familiar with the Javascript, I have two questions.

First, how could I set a variable in the javascript item and then can be assessed in later pages (such as sketchpad)?

Second, is there any alternative way in javascript that functions as "canvas" in python? I was thinking display the stimuli(which is a stream of letters) on the sketchpad, will it work?

Thank you for your help

Best,

Lorraine

lvanderlinden

Hi @Lorraine ,

First, how could I set a variable in the javascript item and then can be assessed in later pages (such as sketchpad)?

In an inline_javascript item you can store variables in the vars object, like so:

vars.my_variable = "some_value"

Next, you can use the square-bracket syntax in the interface (such as a sketchpad) in the normal way. For more information, see:

• https://osdoc.cogsci.nl/3.3/manual/osweb/javascript/

Second, is there any alternative way in javascript that functions as "canvas" in python? I was thinking display the stimuli(which is a stream of letters) on the sketchpad, will it work?

No, unfortunately there is no canvas equivalent for OSWeb (yet). For your specific question, perhaps an additional loop-sequence construction, nested in your trial sequence, would work? Similar to this situation (see example script)?

• https://forum.cogsci.nl/discussion/comment/21993#Comment_21993

Good luck and let us know if you have any further questions!

Cheers,

Lotje

Lorraine

Hi,

Yes, I have successfully used "vars" to store a global variable.

My experiment aims to develop a stream of random numbers, which two letters (T1 and T2) will be randomly assigned (or take a random position) in this stream. I think there are too many combinations if I use an additional loop sequence construction to display this sequence on the sketchpad. Would you please teach me how to do that in an easier way?

Best,

Lorraine

eduard

Hi Lorraine,

Are the positions entirely random or are they limited to a certain range of the stream? Is the stream continous or will you have several trial with a stream each? Is timing critical? In other words is it an RVSP? If you can limit the number of possible combination somewhat, then I think a loop table together with efficient use of the prepare/run phase strategies could take you quite far. The most difficult thing is going to be response collection then (probably not possible to record response until the stream is over).

Thanks,

Eduard

Lorraine

HI,

Yes, it is an RSVP (Rapid Serial Visual Presentation paradigm). So the position will be limited to a certain range, however, the distractors (numbers) and the stimuli (letters) are all randomly selected. We also expected to have 1 to 7 lags (sequence is also random) between the two stimuli. We expect each trial will adopt different streams.

Could you please elaborate more about the loop table and the prepare/run strategy? How are they combined could solve my problem?

Best,

Lorraine

eduard

Hi Lorraine,

The attached experiment illustrates a bit what I had in mind. Does it make sense to you?

Of course you would have to add a little bit of flexibility, like how many sketchpads you want to show, what the distractors are going to be, etc. But I think, this flexibility is implementable with Osweb.

Eduard

rsvp.osexp

Lorraine

Hi,

Thanks, I had worked that out.

The newest problem is that I want to use a touch_response item. I have already linked it to a sketchpad item, but the mouse cannot be seen when running the program. And the experiment just stuck on that page, meaning that I cannot clink anything. Could you help me figure this out?

Best,

Lorraine

eduard

Hi Lorraine,

Maybe try a different backend? See here: https://osdoc.cogsci.nl/3.3/manual/backends/#mouse-cursor-visibility

Otherwise, you can attach your experiment and I can have a look.

Eduard

Lorraine

Hello,

I have figured that out, thanks!

Another problem is when I try to use "if else" statement in javascript, the program only displays my last statement condition. Could you please let me know what I did wrong?

lvanderlinden

Hi @Lorraine,

So it sounds like none your if-statements is ever met, and therefore the else statement is always executed? Did you try printing the variable response_touch_response to see whether it indeed has one of the intended variables?

console.log(vars.response_touch_response)

And did you try writing the values as strings?

if (vars.response_touch_response) == "36"{
// if block goes here
} 

etc.

If this doesn't help, could you upload your experiment here?

Cheers,

Lotje