Saving correct/incorrect textbox responses

[carlyleannah]

OS: Mac OS Catalina 10.5.7
PsychoPy version: v2020.2.10
Standalone – unsure

I am building an experiment that in which participants view a video and then type a response that matches the video.

The videos and responses are on a loop with a Conditions .xlsx file attached. The loopType is set to Random. To collect data on whether participants’ answers were correct or incorrect, I used the following code in the Code Component:

if practice_type:
    if practice_type ==corrAns:
        corr = 1
    else:
        corr = 0
    trial_prac.addData('correct', corr)

When I go to run the experiment and type responses in the textbox, they all show up as 0 (aka not correct) even when the answers are correct.

This is what the xlsx conditions file looks like:

And this is what the data output looks like. All the answers show up as 0 even though some are correct. (Capitalization doesn’t appear to matter as I did another run that matched the capitalization in the Conditions file.)

[Michael]

• What is practice_type?
• When does this code run (relative to the text box)?
• What is the result of inserting:

print(practice_type)
print(type(practice_type))
print(corrAns)

@carlyleannah

(Capitalization doesn’t appear to matter as I did another run that matched the capitalization in the Conditions file.)

It will matter after the current problem is solved - if subjects aren’t typing capitals, then the string comparison will always fail. To get around this, would suggest that you do something like this to force each side of the comparison to have the same case:

if practice_type.lower() == corrAns.lower():

[carlyleannah]

Thank you for getting back to me. I’m still learning coding and PsychoPy, so I appreciate your help.

• Practice_type is the name of the textbox component. I have two routines with loops, one is PracticeTrial with the trial_prac loop.
  

  

• The other loop:
  

  

• The code runs relative to each frame.

• The data output ‘correct’ still shows up as all 0 when I insert
  print(practice_type) print(type(practice_type)) print(corrAns)

• Glad you mentioned the capitalization, makes sense. The code string doesn’t work though.
  

  

[Michael]

@carlyleannah

Practice_type is the name of the textbox component.

OK, then comparing practice_type to your corrAns variable will never evaluate to True, as they are quite different things. practice_type is a complicated object (a TextBox2 stimulus) that has lots of attributes (like a position, font style and colour, whether it is currently visible, and so on). The text it contains is just one of those attributes. An analogy would be that you can’t directly compare the identity of a person to a car, just because that car happens to contain a driver. You need to directly compare like-with-like, in this case:

if practice_type.text  == corrAns: # refer directly to the text attribute

@carlyleannah

The code runs relative to each frame.

This code should only run at the end of the routine, not on every frame. That way the comparison is only made when the response is complete, not while it is still being typed (and possibly getting edited).

@carlyleannah

The data output ‘correct’ still shows up as all 0 when I insert
print(practice_type) print(type(practice_type)) print(corrAns)

That wasn’t designed to change what happened in the data but to give you debugging information in real-time: it should have revealed that you were comparing incompatible object types (a TextBox2 stimulus begin compared to a string of characters).

@carlyleannah

Glad you mentioned the capitalization, makes sense. The code string doesn’t work though.

That is because you have an if: line with nothing following it. Also, as above, a TextBox stimulus won’t have a .lower() method, only strings do, so you need to apply it there:

if practice_type.text.lower() == corrAns.lower():
    # having something here is important

EDIT: PS there does seem to be a bug at the moment about the content of the text box not being stored correctly at the end of a routine, so for the timbering at least, you might indeed need to do the comparison on each frame, and just trust that the last one will supersede any previous ones:

Use of textbox answer to code correct answer

Hi There, Aha OK I’ve looked into this more and think I have it. Currently there is a bug whereby text.text is replaced with “default text” at the end of the routine (the developers @TParsons are aware of this and it is fixed for the next release). For your case though this is causing problems, because the correct answer is not “default text” of course… For now, you need to check if the correct answer has been typed in the textbox on each frame. I have added the code component “checkCorrect” t…