Is this possible and how?

Comments

• dk11

  November 2017

  p.s. I realized that the table where I explain Variable 1, Variable 2, and Condition got impaired after I posted it, this is what it shall look like:

  
• eduard

  November 2017

  Hi,

  That is a little more complicated alright. Not sure whether it is implementable with a sketchpad. It might, but certainly not easily. I recommend you look at inline_scripts and how to prsent stimuli and collect responses with python code (e.g. http://osdoc.cogsci.nl/3.1/manual/python/canvas/).

  You can browse this forum a little. What you are looking for are examples of how response collection can be done independently of image presentaiton. The implementation will probably involve a while loop, one or two canvasses and a mouse item.

  It is a little late right now, but next week I might ahve some time to tinker some code together.

  But you give it a try first.

  Good luck,
  Eduard

  
• dk11

  November 2017

  Dear Eduard,

  thank you very much for your advice. I've spent this weekend searching for examples of how response collection can be done independently of image presentaiton, and I was also researching different python codes and the "coroutines" feature, and I did not make much progress unfortuntely because I still have not found any indication of how what I want to do would be possible.

  To make it easier to understand what I want to do, I made the basic version of my experiment in OpenSesame, where people need to estimate 1s time intervals by clicking inside the circle in the middle of the screen while a specific word is being presented below the circle. This part is easy to do in OpenSesame.

  The tricky part is that I still cannot figure out how to add a dot at the top of the screen that is flashing at either irregular or regular intervals (depending on the block) and is independent of response selection and the presentation of seconds.

  I would really appreciate any help and pointers you or anyone else may have on this. My experiment is scheduled to run in exactly 2 weeks and I did not expect this would be so difficult to build when I constructed the plan for the experiment. Thanks a lot.
• eduard

  November 2017

  Hi,

  I made a little script. Hope this gets you started.

  Good luck,
  Eduard

  
• dk11

  November 2017

  Thank you, I really appreciate it.
• dk11

  November 2017

  Hi Eduard,

  thank you once again for creating the script.

  I extended the script you created and shaped it into a full experiment, and now there are only 3 things (I hope minor) that I still did not figure out how to achieve.

  1. It seems that "time = ms.get_click()" outputs only aggregate times across all 30 clicks per each stimulus in the practice or block loops, but I need response times for each of the 30 clicks of the counter, and I have not managed to figure out how to do that.
  2. When a loop starts and participants start clicking to estimate seconds, mouse cursor stops being visible, and yet it is very important for the experiment that the cursor is visible so that participants can click inside the circle in the middle of the screen. I tried setting double buffering in legacy to "no" and also tried psycho and expyriment back-ends but that does not seem to be the reason for mouse cursor invisibility. If the cursor cannot be made visible, I will need to use keyboard rather than mouse responses.
  3. In "random_canvas" inline_script, I used the random.randint function (see line 23 in the attached experiment) to make the appearance of the black dot random (so that the interval between each blink is random). This function, however, does not seem to have a desired effect. Only when xpyriment back-end is used it actually works properly, but using xpyriment has different issues, such as a long delay between when participants click the mouse and when this actually affects the counter of the screen. Do you perhaps know how to achieve this in legacy?

  I would be really grateful if you would know how to resolve these 3 issues, and that will be everything I need to make the experiment work.

  Thanks for your help, I appreciate it.
• dk11

  November 2017

  P.s. I resolved number 2 so I do not need help regarding that anymore but still cannot resolve the other 2 points.
• eduard

  November 2017

  Hi,

  1) time = ms.get_click()

  That is not correct. ms.get_click() returns the button that wa spressed, the current position and then the time from experiment start on. So the correct way to get the time is button, pos, time = ms.get_click().
  See here: http://osdoc.cogsci.nl/3.1/manual/python/mouse/#function-mouse46get95click404242resp95args41
  If you want a response time. You have to manually compute it. Use the clock.time() function to start the timer at some point (e.g. when stimulus presentation starts), and then take the difference of time of button press with the initial time. Does that make sense?

  3

  Mh, I see what you mean. I think you should not include the random.randint(200, 5000) in the if statement, because on every iteration (of which there are a lot!), it would have a different value, so that the logic would be messed up. Instead initialize it once before the while loop (e.g. interval = random.randint(200, 5000)) and then use interval in the if statement. Inside the body of the if statement, you can then update the variable, by calling interval = random.randint(200, 5000) again.

  I haven't tested it, but does this happen to solve your issue?

  Eudard

  
• dk11

  November 2017 edited November 2017

  Thank you Eduard,

  I am getting there.

  • To capture timing, I added the following line of code in the while loop: var.rtime=clock.time()-t0 (see below). The main problem now is that the variable gets overwritten in every new iteration of the loop, so in the end the reaction time that is captured is that of the entire loop. Sebastiaan explained here how to change variable name in every iteration of the loop, but I have not managed to implement it in my loop: http://forum.cogsci.nl/index.php?p=/discussion/2850/logging-variables-in-a-while-loop. Being able to do this would resolve all my issues

    t0 = cv.show()
    interval = random.randint(200, 5000)

    while counter < 30:
    left_button, pos, timestamp = ms.get_click() # samples a response on every iteration (waits for 3ms)

    if left_button == True: # only after a click do the updating
        var.rtime=clock.time()-t0
        cv.clear() # empty the current screen
        # draw the updates information on it
        cv.text(var.Stimulus,x=0,y = -300, font_size=200, font_family='Arial Black')
        cv.circle(0,0,32, penwidth=5)
        cv.text(counter,x= 0, y = -100)
        # present the new screen
        cv.show()
        counter += 1
    

  • To achieve random display of the dot, I added the codes you suggested as below but I still did not get the desire effect. What I would need is that the dot is always displayed for 400ms, but at random intervals (ranging from 200ms to 4000ms) while the loop is not finished. Do you maybe know how it would be possible to achieve this?

    while counter < 30:
    left_button, pos, timestamp = ms.get_click() # samples a response on every iteration (waits for 3ms)

    if left_button == True: # only after a click do the updating
        var.rtime=clock.time()-t0
        cv.clear() # empty the current screen
        # draw the updates information on it
        cv.text(var.Stimulus,x=0,y = -300, font_size=200, font_family='Arial Black')
        cv.circle(0,0,32, penwidth=5)
        cv.text(counter,x= 0, y = -100)
        # present the new screen
        cv.show()
        counter += 1
    if ((clock.time()-t0)%interval) < blinkrate: 
        interval = random.randint(200, 5000)
        cv.clear()
        cv.circle(0,200,30, fill= True)
        cv.text(var.Stimulus,x=0,y = -300, font_size=200, font_family='Arial Black')
        cv.circle(0,0,32, penwidth=5)
        cv.text(counter,x= 0, y = -100)
        cv.show()
    

  Thanks again for all your help. This is everything that is still left to be resolved for my experiment and after that I will need no more help. I am also attaching the abridged version of the experiment if needed.
• eduard

  November 2017

  Hi dario,

  Attached the (hopefully finished) experiment. I changed the inline_scripts and also the general structure a bit, primarily to make things easier. Now there is only a single trial loop and a single stimulus canvas inline_script. (The only difference between condition was the flicker rate, so I added a single if-else statement to the script, and everything was good). Also it is never a good idea to have multiple logger files. Always use a single one. (unless you have two entirely different tasks).

  Let me know if there are problems.

  Eduard

  
• dk11

  November 2017 edited November 2017

  Thanks, perfect! Now the only issue when I run the script is that 0 in the counter changes to 1 only after two clicks rather than after one click. If you maybe know what needs to be changed in the script you created for that no to happen everything will run with no mistakes.
• eduard

  November 2017

  Fixed. It was actually quite easy. You should try to figure these things out yourself. Make a backup of the experiment, and then play around with the order of lines, etc. and see how the behaviour changes.

  Good luck,
  Eduard

  
• dk11

  November 2017

  Thanks, I actually tried many different things and had no idea it was this small twitch that would make it work.