Collecting touch responses on Android 4.2.2 Tablet

sbhimsaria

Hi,

I am trying to build the an Implicit Association Test using the graphical interface as I don't know how to use Python. Basically, I have used the German-British example from Open Sesame as a base to build my own experiment. I plan to customize it further once I have the basics in place.

I replaced the keyboard response to a touch response as I will be running this experiment on android tablets. I defined the number of rows as 1 and columns as 2. If I am not wrong, this should divide the screen into a left and right responses with center coordinates (0,0).

Subsequently, I have two sketchpads for correct and incorrect responses.

The problem is that when I run the experiment, for any response (left or right), the next screen is always the incorrect response sketchpad.

I am not sure what I am doing wrong. It would be great if someone can point me in the right direction.

Here is the screen shot of what the program looks like: http://img.cogsci.nl/?q=56dabee23406e

Thanks,

Shruti

Josh

Hi Shruti,

Your problem may be in the logic of the code that defines when your sketchpads need to be shown. Even without python scripting, you'll have to define some 'if this-then that' rules somewhere, because you only want some things to be shown in specific situations. If you click on the sequence item that your response sketchpads are placed in, you'll see these sketchpads in a list with so-called run-if statements. Did you (try to) define here under which circumstances the sketchpads need to be shown? (e.g.: run-if [correct] = 1).

Cheers,

Josh

sbhimsaria

Hi Josh,

Thank you for your response. To answer your questions, yes I have defined the run if statements as [correct] = 1 for correct and [correct] = 0 as incorrect. If I run the same with a keyboard response, it runs without any error. However, with the touch response it shows the error as I had mentioned before.

Thanks,

Shruti

Josh

Hi Shruti,

Have you also defined what the correct response should be every trial?
You can leave the Correct response statement in the touch_response item empty; rather, you could create a correct_response variable in the loop item where you indicate the correct response (1 or 2) in each condition. It may well be that you've already done this, but we got to know for sure of course

Cheers,

Josh

sbhimsaria

Hi Josh,

I have a simple task in the IAT where the pictures of some public/private organizations will appear on screen and I want my respondents to click on the left side of the screen for every public image and right for a every private image. For every response, I have a subsequent sketchpad item showing a red/green dot for incorrect/correct response.

The response mapping in the beginning of the experiment is as follows;

if var.subject_parity == 'odd':
var.public_response = 'left'
var.private_response = 'right'
else:
var.private_response = 'left'
var.public_response = 'right'

var.positive_response = 'left'
var.negative_response = 'right'

This is followed by the task loop which has the following variables;
category(public/private)
correct_response([public_response]/[private_response])
image (has public/private images)

The following sequence contains a sketchpad which has the image in the center which should record the left and the right response. Depending on the response, I want the next sketchpad item to appear where i have used the conditions [correct] = 1 for correct and [correct] = 0 as incorrect.

So the sequence has;

1. Activity sketchpad
2. Touch Response
3. correct feedback sketchpad
4. incorrect feedback sketchpad
5. logger

What do you think is that I am doing wrong?

Thanks,

Shruti

Josh

Hi Shruti,

Here's the deal: the touch_response item will return a variable response that is set either to 1 or to 2, in accordance with the areas of the touch screen. So you should indicate for each trial whether the correct response is 1 or 2; 'left' or 'right' doesn't work.

Cheers,

Josh

sbhimsaria

Hi Josh,

Thank you so much. That worked!

I was also making a very silly mistake where I had spelled 'correct' wrongly in the variable correct_response.

Another question, I would like to put a limit of 75 percent accuracy on the practice exercises. This means that if the accuracy rate of the respondent is below 75%, he will have to repeat the whole sequence. How can this be done?

Thanks,

Shruti

eduard

Hi Shruti,

You can put your sequence into a loop that is repeated a very high number of time (often enough, so that you are sure the loop is never finished during on experimental session). Then, you can check the box Show advanced options in the loop item. You should see a Break if field now. And there you write something like [avg_accuracy]>75.

Of course this requires, that there is a variable called avg_accuracy. You probably want to initialize it with a low value (e.g. 0) to make sure the loop is executed at least once. Then after the last trial of your sequence you have to calculate the accuracy and update the current value of avg_accuracy.

Does this make sense?

Eduard

car_mic

Hi,

I am new to OpenSesame and I wanted to ask you a related issue. I am using the German-British IAT example from Open Sesame and I want to create a touch screen version.

I have defined touch_1 (touch response) with the number of columns to 2 and rows to 1 (instead of the key_board response). However, in the response_touch_1, I always get 1 (even if I click on the right side of the screen). I was expecting to get 2 when touching the right part of the screen and 1 when touching the left part.

I am not sure what I am doing wrong.

Thanks,

Michela

Daniel

Hi Michela,
Even though your problem sounds similar to the one discussed in this thread, it is maybe better to post it in a new, separate thread. It will get picked up better and will be better searchable for when someone has the same problem as you in the future.