Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Combine text and key press responses
Hello world!
I'm creating a word association task. Participants must write a word in response to a cue word. I´m using the response option"keypress" because it is important to evaluate the first key press and the interval between each key press; however, it is laso important to me that participants can see what they write.
I tried to combine two resposes "keypress" and "text" as explainned in the tutorial Recording participant responses, but F5 ignore the "keypress" response and only record the "text".
I'm not shure that this is the best approach to solve my problem.
My current code is:
  "type": "basic",
  "stimulus_pattern": {"order": "random"},
    "responses": [["k_resp","t_resp"]]

Thanks to everyone
Hello! This combination of responses does not currently work. The text response grabs all input to itself and doesn't let the keypress response listen to the keypresses. This is something that we can work on! We have some questions about what you want to achieve in your design:

1. In text responses, participants can hit backspace to delete what they have written and can write over it. Do you want to be grabbing all of the deletions?
2. Participants can also potentially copy and paste text into the text box. This wouldn't tell you much about reaction time, so do you need copy and pasting disabled?

Any other information you can provide about what you are hoping to achieve is welcome!
Thank you for you response!
1) I think the best option in our design is to disable the backspace key; however, grabbing all deletions would be an interesting option for data analysis.
2) Yes, we need the option "copy/paste" disabled.
Hi - We are working on this as fast as we can. We expect to have a solution for you by the end of this week. Thanks for suggesting these tweaks!
Alright! We have something for you to test now. We host a testing version of FindingFive at On the testing version:

(1) the keypress response can now listen to the key presses in a text response. That is, as participants are typing their answer in the text response, the individual keystrokes are also recorded by a background "keypress" response.

(2) the "text" response can now optionally take a new property:

"prevent_paste": false

which will disable the pasting function in a text response

(3) the "keypress" response can now optionally take a new property:

"functional_keys": true

which will not only record the individual character keystrokes, but also functional keys like "Backspace", "Delete", etc.

Whenever you have a chance to test these new features, please let us know if they resolve the issues in your design? If they work as intended, we'll push them to the production server so that you can use them in studies! Thanks Big Grin
I am very grateful to finding five.

I already test those functions; however, I found two bugs marked in the attached document.
1) Messy responses: I suppose that after collecting each key with the "keypress function," the complete response should be collected when participants submit their responses; nevertheless, there is a delay when text response is saved. You can see in the attached file, in the tab "words" cells marked in gray that correspond to each keypress for the word "cup," then several keypresses from other words, and finally the text response "cup."
2) Repetition of keypress responses: In the attached document, you can find in the tab "words" some cells marked in blue that exemplify how keypress is repeated throughout the document. This word (wall) is only one example, but there are several keypress repetitions if you see it in detail. To test this problem, I performed a second test responding only with one keypress using the "a" character. Forty-eight stimuli should be 48 keypresses cells and 48 text submissions; however, there are 365 keypresses responses and 48 text responses (this last number is correct).

I appreciate your help.

Attached Files
.xlsx   test_1.xlsx (Size: 98.27 KB / Downloads: 1)

Unfortunately I'm somewhat completely at a loss on how to interpret your data without seeing your study. The color coding you were referring to did not seem to show up on my end.

In general, the keypress response records all keystrokes, including the entire history of deletion, changes, and whatnot. The result will be messy. It will not be the same as the value collected in a text response.

I'll PM you regarding what to do next.
Alrighty! I think we have resolved this issue - thanks so much by the way, it's actually a latent bug of the keypress response that had always existed, and was only exposed by the particular setup you are using. Smile We are glad to be able to identify the issue.

Can you try your study on the testing server again and see if the issues still exist? Thanks!
I already test this correction, and it works great. I cannot find any other issue.

I appreciate your help. Smile
Fantastic! We'll move these new features / bug fixes to the production server today. They will show up on the production server ( tomorrow.

Forum Jump:

Users browsing this thread: 1 Guest(s)