key response with buttons

[bisserai]

Dear FF team,

I'm setting up a norming study for which i would like to have participants listen to an audio stimulus (e.g. a1, a2, a3...), then be presented with two text options on screen (a1_1 vs a1_2, a2_1 vs a2_2...) and use the keyboard only to select whether they choose the left option or the right option. I'd like to record that they chose a1_1 or a1_2 regardless of where on the screen they were presented. At first i had my audio and text options on different trials and tried using a choice type response with key_mapping "L" and "R" and a key_only response, but that gave me L and R buttons, as well as additional text instructions telling ppts to make a choice. Now i have the audio and the text options on the same trial, however i get a continue button and the fixation cross that's supposed to be in the middle is shifted left. I thought i could use "location" but that's only for responses, and my text options are stimuli - i don't want to make a different response for each audio stimulus. Ideally i'd like to have my fixation cross centered (location 5) and once the audio is done (this bit is fine, the text options appear after the audio) have the text options appear in locations 1 and 3.

Thanks a lot for your help!
bissera

[Rachel - Participant Affairs]

Hi bissera, 
Can you say more about what you are trying to achieve? It would be helpful if you could break your question into parts, as it is difficult for me to determine how to answer without more information. 
A few thoughts that might help: does using the choice response with key_mapping, key_only, and instruction set to "" achieve the desired result? This will suppress the instructions, but you mention that you don't want the "L and R buttons." Do you mean that you don't want any text displayed at all? As far as adjusting the location of the fixation cross, take a look at the custom option for specifying the location of stimuli within a trial template. 

Best,
Rachel

[bisserai]

Hi Rachel,

Thanks so much for your message! Yes, of course, here's some more detail.

I want to have a trial on which i have a fixation cross in location 5 while my audio a1 is playing and once the audio is finished i want to have two text options t1a and t1b appear in locations 1 and 3 or if the fixation cross is gone, which i have no preference for, locations 4 and 6. Indeed all i want on screen is the text options and no instructions or buttons (but could keep the cross, i don't mind). Then i want my participant to use the keys to choose one of these text options and to automatically proceed to the next audio once they've made their choice (no buttons). The location of the text options should be randomised, but not the location of the fixation obviously, which is why at the moment, since i have everything in one trial, i haven't successfully managed to use the "location" and "randomize_location" options and keep one stimulus stable and the others moving. The reason why i put them on one trial is that before they were on two separate trials, but then i wasn't able to coordinate the audios with the text options and that's vital. I have:

• 50 audio pairs: a1/a1x, a2/a2x, a3/a3x
  
• 50 text pairs which go with the audios: t1a/t1b for audios in pair 1, t2a/t2b for audios in pair 2 etc
  

I want to have the two appropriate text options appear after my randomised audio file is played and i want my response to record which text option was selected.

So the problem is that if i have separate trials i can't successfully coordinate the audios and the text and if i have them on a single trial i can't place them where i want them on screen. In terms of the response, the problem is that i want the response button to be linked to a stimulus and not a button i've created.



Thanks so much,
bissera

[Rachel - Participant Affairs]

Hi bissera,
Thanks for providing more info. Based on your description, I think there are two ways of achieving (most of) what you want:

1. Specify the audio, fixation cross, text options, and a keypress response within the same trial. Use the custom option for specifying the location of stimuli and set auto_advance to true. When you create your keypress response, make sure to set multiple to false. For the audio stimulus, set barrier to true. If you'd like the fixation cross to disappear once the audio is finished playing, you can set its duration to the length of the audio clip. See below for an example of how you can set up this trial template.

   

We discovered a bug in the process of testing this solution, so note that for now, defining the 'whitelist' property for keypress responses will result in the study getting stuck. But this issue will be fixed in the next week or so, now that we are aware of it. 

Randomizing the location of your text options will be an issue with this solution, however, as you've discovered.

2. Specify the audio, fixation cross, text options, and a choice response within the same trial. Again, make sure that barrier is set to true for the audio stimulus. For the choice response, make sure that key_only is true and key_mapping is defined. As I mentioned before, setting instruction to "" will suppress the instructions. You can randomize the location of the text options with the "locations": "random" property of the choice response.

   

I think this almost gets you exactly what you want, aside from the 'Press (defined key)' labels underneath the choice options. There shouldn't be any clickable buttons using this approach. 

Please report back if either of these solutions work for you. If not, we can try to troubleshoot any continued issues you are having. 


Best,

Rachel

[bisserai]

Hi Rachel,

Thanks so much for your message. The position of the text stimuli must be randomised, so option1 won't be a possibility i've tried option 2 though, but for some reason the choice doesn't come up at all at the moment. i did everything the way you list it and really don't see a reason why it shouldn't, but it just doesn't turn up at all, the end of the audio is the end of the trial. I've added instructions in order to be able to detect it if it's there, but i can't see them and i can't see a log of it appearing in the csv file with the data. I've tried defining the text options both as stimuli as well as inside the choice response, and they don't turn up, except when they are defined as stimuli they push my fixation cross over to the left, but don't actually appear... not sure what i'm doing wrong...

Another problem is that if i understand correctly, if i do option2, i'll have to make 100+ responses and 100+ trials for all the audio files i'd like people to respond to. It's not the end of the world of course, but if there's another possible implementation that involves a little less manual work and a little more automation that would be great. i didn't think this study would be hard to implement!

Best,
bissera

[Ting]

Hi Bissera, can you try removing "auto_advance"? or just setting it to false.

To catch a remote possibility, did you accidentally type "response" (singular) instead of "responses" (plural)? That's a common typo too.

[bisserai]

Hi both,

Sorry for a delay in response - i was on holidays and had a very hectic first week back to work!

Indeed, you were right Ting, i had missed the s in responses!! Something i hate about coding, the tiniest error can result in seemingly undecypherable havoc. So the study seems to be doing what i need it to: it presents a fixation cross, an audio file and follows with two text options, which one chooses from using a key. Thank you both very much! Now i just have to create an individual response for each stimulus, but with the csv response definition file that should be rather quick...

Thanks again!

Best,
bissera

[Ting]

Glad to hear that, Bissera! Happy to help!