[solved] Time pressure within experiments

fvg4

Hi all,

I am trying to insert a timer in my experiment using OpenSesame, and I went through some codes in python websites, however I couldn't find any solution for this.

To be more obvious, what I do need, is a timer that counts down the 'remaining time' to finish the task by the participants. Aim: putting time-pressure as a variable in the experiment.

Thanks to all!

Regards,
Francisco

eduard

Hi Francisco,

You can access a timer by calling self.time(). This will give you the current system time.
If you sample it twice and take its difference, you will get a duration of the interval between the 2 samples:

t0 = self.time()
self.sleep(1000)
t1 = self.time()
print t1-t0 # should print something like 1000

# a timer

timer_start = 10000 # 10 seconds
used_time = timer_start - (t1-t0)
while used_time > 0:
    t1 = self.time()
    used_time = timer_start - (t1-t0)
    # do other stuff

The only thing you have to make sure, is that the timer is updated while the participant is doing the task, because otherwise the entire while-loop structure is pointless.

Does this help?

Eduard

fvg4

Hi Edward,

I will try it and let you know. Thanks for your answer!!
Francisco

fvg4

Hi Edward,

I tried to do what you suggested. I think I miss some part.. I can call self-time() in a inline_script you mean? And what should I put in the prepare_phase and run_phase of my inline_script?

However, I don't know if that is going to make what I want. The timer will appear in the screen with this?

Sorry being so dummy..

Thanks a lot!
Francisco

Josh

Hi Francisco,

So the goal is to display a timer onscreen, while the participant is doing the experiment? (you didn't specify this before).
Yes, you can call self.time() in an inline script. Since displaying the timer is going to be an ongoing process during the task (updating the time onscreen every millisecond for example), this should go in the run phase. Beware though, that any given taskscreen is a static thing by default - meaning that there is no way to create and display a canvas (= screen) only once but still have an ongoing timer. That's why I would suggest displaying the seconds only (i.e., not milliseconds) so that you only have to refresh this static screen once every second. Further, it will definitely be easier if the task itself can be displayed on a single screen per trial as well (at least for as long as the timer should be active). A way to go about this (using an 8 seconds timer) would be:

   from openexp.keyboard import keyboard
   my_keyboard = keyboard(exp, keylist=['z', 'm']) # if it's a 2-button response task, say)

   task_done = False
   start_time = self.time()
   last_digit = "8"
   exp.canvas.text(last_digit)
   exp.canvas.show()

   while (task_done == False and timer_out > 0):
          timer = 8000 - (self.time() - start_time)
          current_digit = str(timer)[0]  # takes the current second
          if current_digit != last_digit:  # i.e. when current_digit becomes '7'
                last_digit = current_digit
                exp.canvas.clear()
                exp.canvas.text(last_digit)
                exp.canvas.show()
          response_given = my_keyboard.flush()
          if response_given == True:
                task_done == True

    if timer_out < 0:
           exp.canvas.clear()
           exp.canvas.text("Time out!")
           exp.canvas.show()

This would display an 8 on screen that ticks down to zero. The different screens displaying their respective digits are made on the go; an alternative way (that might make the experiment run more smoothly) is to make all those eight screens in the prepare phase. Then you only have to use the .show() command in the run phase.

Hope this helps.

Cheers,
Josh

fvg4

Hi Josh,

First of all. Thanks for your both answers. I was confused with the other discussion and thought it was "the same" as I wanted to do. However after your comments I realize it wasn't.

So ok, what you just mentioned in your comment (in this discussion) is what I do need to do.

I have a trial in my experiment with 10 different images. In each of them there is a correct answer (I am using Z and M as key_response). I am using an experiment_loop to run randomly the ten images (as the OpenSesame tutorial shows) with a [cr] for the correct response on the keyboard.

I need to get the timer onscreen when the images appear, and after the participant makes the decision it will appear another fixation_dot and another image with another timer too.

I tried to use the code you suggested, but probably I did it in the wrong way because I got an error when the experiment got to the trial. The code you wrote I putted it in the run_phase. Should I do anything in the command break_if?

I will send the appearance of my experiment. - file:///Users/FranciscoMarcus/Desktop/Captura%20de%20ecra%CC%83%202015-04-28,%20a%CC%80s%2000.59.01.png

Thank you very much indeed!

Kind regards,
Francisco

eduard

Hi Francisco,

If you bump into errors, it is always wise to post the error message together with your question. This makes our life a lot easier to find out what went wrong.

In case you copied Josh's code as he wrote it above, your problem is that the code is not ready to be run yet. I took the freedom to add some things that enable you to see the timer on screen and adapt it to your wishes. Also I couldn't resist to make a pretty clock timer. If you don't like it, or it slows down your experiment, for whatever reason, of course you can use the simple seconds timer as well.

from openexp.keyboard import keyboard
from openexp.canvas import canvas

my_keyboard = keyboard(exp, keylist=['z', 'm']) # if it's a 2-button response task, say)
cv = canvas(exp) # initialize a canvas

def timer2clock(timer):
    '''
    converts a plain integer representing the seconds left to a proper
    clock
    '''
    clock = str(timer//60000) + ':' + str(timer%60000)[:-3]
    # do the formatting
    if len(clock) == 2:
        clock = '0' + clock[:2] + '00' + clock[2:]
    elif len(clock) == 3:
        clock = '0' + clock[:2] + '0' + clock[2:]
    elif len(clock) == 4:
        clock = '0' + clock 
    return clock

# define parameters
task_done = False
start_time = self.time()
timer = 10000 # 2 minutes
rest_time = 10000
clock = timer2clock(timer)
# show initial screen
cv.text(clock)
cv.show()

# update screen
while (task_done == False and rest_time > 0):
    # update time
    cur_time = self.time()
    rest_time = timer - (cur_time - start_time)
    clock = timer2clock(rest_time)
    # draw to screen
    cv.clear()
    cv.text(clock)

    ##############
    # DRAW YOUR STIMULUS HERE
    ##############

    cv.show()
    # get response
    k,t = my_keyboard.get_key(timeout = 50)
    if k != None:
        task_done = True

# visualization of effects
if task_done:
    cv.clear()
    cv.text("Correct!")
    cv.show()
    self.sleep(1000)

if rest_time < 0:
    cv.clear()
    cv.text("Time out!")
    cv.show()
    self.sleep(1000)

fvg4

Hi Edward,

So it seems that your piece of code gets me the timer onscreen!!

However the timer gets first or after the stimulus (as I put it before or after my stimulus in the 'sequence').

I saw that you wrote a parte in the code where I can/should put my stimulus, but I already have the stimulus written inside the sequence (fixation_dot; sketchpad; key_response).

Also, my response should be with the [cr] that I have linked to my stimulus, which seems that with the timer I have to possible responses (my [cr] and the two options the code gives) which probably will mess my results and data.

That is how I am inserting the Timer on my experiment - file:///Users/FranciscoMarcus/Desktop/Captura%20de%20ecra%CC%83%202015-05-3,%20a%CC%80s%2017.44.42.png

Any advices how to get the timer with the stimulus onscreen at the same time?

Thanks a lot for your help!!

Cheers
Francisco

eduard

Hi Francisco,

First of all, the pictures you're posting are not accessible. Can you give it another try uploading them, if you think they would help solving your issues?

You're right. Using fixdot, sketchpad and keyboard_response is rather difficult to combine with the timer-inline_script. The only way that maybe would work, at least as far as I know, is finding out how to access the canvas which is created by the fixdot and the sketchpad item and manipulate them in your inline_script to contain the timer.

Another option, which seams easier to me, is not using these items at all, and draw your stimuli also in the inline_script. The way to do it is explained in thisdocumentation. So basically, you need to initialize a canvas, then draw a fixation dot on the canvas and finally your stimulus. You could do this in the prepare phase of your inline_script to save computation time. During the while loop you'll have to add the timer to the canvas, so that you can see both at the same time.

The response collection is a smaller issue. In the beginning of your inline_script you can access your [cr] variable to get the correct response (e.g. exp.get('cr') or so), then you change this line:

my_keyboard = keyboard(exp, keylist=['z', 'm'])

into a more generic way of creating your keyboard:

my_keyboard = keyboard(exp, keylist=correct_resp) # correct_resp needs to be defined first.

Does this help?

Eduard

fvg4

Hi Edward,

Thanks a lot!
I just got it.

Best wishes!!
Francisco

IgottaPauer

Hey everyone.

I have a similar task like Francisco.

I want my participants to work under time pressure. For that reason I want to show a 30 second timer on the top right corner of my sketchpad for every trial. I am not using keyboard but mouse_response and I only have one sketchpad (so no fixdot etc).

I tried to modify the code above to fit my aims, which are:

placing the countdown on my sketchpad in the top right corner (not in the middle)

using a mouse response (instead of keyboard)

and using the countdown only on one sketchpad.

but I just can't manage... like not even a little :(

Could anyone help my modify that code for my purpose?

Regards.

Charlotte