[open] Visual Analogue Scale

Vit27

Dear all,

I'm programming my recognition experiment with OpenSesame and I've just started to learn python.
I want to collect continuos confidence ratings for 3 different categories (labels) using visual analogue scales (probably->confident). How can I do that?

Josh

Hi,

Step one is to draw a canvas (similar to sketchpad, see http://osdoc.cogsci.nl/python/canvas/)

Step two is to collect a mouse response, see http://osdoc.cogsci.nl/python/mouse/

The mouse response contains a button as well as coordinates, and these coordinates can be used to retrieve your visual analogue scale score. You would just have to determine the coordinates of the scales beforehand, and then calculate to which % the X-coordinate corresponds.

At some point you'll also want to do some subtle things, such as making sure that a mouse response only counts when it was actually on a scale. Your response collecting command will therefore probably have to be in a while loop, from which you only break when the response was valid.

Good luck!

Josh

Vit27

Thanks a lot Josh!

I'm stacked at the first point. Do I have to create (in Prepare) a canvas for each object (in this case: 3 rectangles, 3 words, 3 lines and 6 words) and ask OpenSesame to show them in the run section, right?
It shows me the canvas sequentially, not all in the same time. Why?

eduard

Hi,

For all items that are static (that is don't change depending on what you're doing on the screen), you have to use a single canvas. For all other items that do change, you need to add another layer and update it regularly. So, the first canvas, you can indeed prepare in the prepare phase. However, the dynamic layer has to go into the run phase, because the things you're showing (e.g. current selection), will change depending on your behaviour.

Does this make sense?

Eduard

Vit27

Yes Eduard, it make sense.

but how can I do that? I've done it, but the rectangles are displayed sequentially not all in the same time. I mean first the first rectangle appear and immediatly disappear, then the second rectangle appears. The same happens to the last one.

How can I display something like this immage?

Edwin

If you want to show three of those sliders at the same time, have a look at this topic: http://www.cogsci.nl/forum/index.php?p=/discussion/827/solved-rating-slider-data-logging/p1

The solution there is copied below (this should be in the Run phase of an inline_script item):

# import everything we need
from openexp.canvas import canvas
from openexp.mouse import mouse

# create the instances we need
my_canvas = canvas(self.experiment)
my_mouse = mouse(self.experiment, timeout=20)

# slider dimensions
slider_w = 500
slider_h = 10
slider_x = self.get("width")/2 - slider_w/2
# three different y-coordinates, one for each slider
slider_y = [self.get("height")/4, 2*self.get("height")/4, 3*self.get("height")/4]

# variable names to store the slider fillings in
names = ["slider_1", "slider_2", "slider_3"]

# questions to put above the sliders
questions = ["question 1", "question 2", "question 3"]

# loop through all three slider positions, to draw the empty slider canvas
for i in range(len(slider_y)):
    # draw some text (this can be anything)
    my_canvas.text(questions[i], y=slider_y[i]-100)
    # draw the slider frame
    my_canvas.rect(slider_x, slider_y[i], slider_w, slider_h)

# add a text below the bottom slider, telling participants to click to accept
my_canvas.text("Click to accept ...", y=slider_y[-1]+100) 

# loop through all three sliders, waiting for input
for i in range(len(slider_y)):

    # keep showing the interactive slider until a click has been made
    clicked = False
    while not clicked:

        # determine the slider fill based on the mouse position
        pos, time = my_mouse.get_pos()
        x, y = pos
        slider_fill = min(slider_w, max(0, x-slider_x))

        # draw an empty rect within the slider frame, effectively resetting it
        my_canvas.rect(slider_x+1, slider_y[i]+1, slider_w-2, slider_h-2, fill=True, color=self.get("background"))
        # draw the slider fill
        my_canvas.rect(slider_x+1, slider_y[i]+1, slider_fill-2, slider_h-2, fill=True)
        # show the updated canvas
        my_canvas.show()

        # Poll the mouse for buttonclicks
        button, position, timestamp = my_mouse.get_click(timeout = 20)
        if button != None:
            clicked = True

    # save the slider's filling
    slider_percent = 100.0*slider_fill/slider_w     
    self.experiment.set(names[i], slider_percent)

Vit27

Thanks Edwin!
This is amazing and really useful and it works well! But what if I would to let participant chose to answer just to one of these three sliders, and not to each one sequencially? Should I add a mousetrap to let the mouse visible?

Vit27

Hi Edwin,
In my output I cannot find any variable with slider_percent. I ust say that I inserted a logger not immediatly after the inline script, but in the end of the trial.

Anyone can tell me why?
# save the slider's filling
slider_percent = 100.0*slider_fill/slider_w
self.experiment.set(names[i], slider_percent)