Response time measure

Comments

• eduard

  July 2019

  Hi Nasia,

  You can either do it offline (after testing), in your data analysis software (Excel, SPSS, R,...), or in the experiment itself, by subtracting the time stamp of when the first sketchpad was shown from the time the response was given. If you do it in the experiment, you have to do it in an inline_script (var.time_keyboard_response - var.time_sketchpad_1). Per default Opensesame stores the times at which all items were called, relative to the beginning of the experiment. Those variables all start with "time_XXX". So when you check your logfile, you can find the variables that code the onset of the sketchpad and the response item, and compute your response times.

  
  

  Good luck,

  Eduard

  
• hulya

  August 2020 edited August 2020

  Hi @eduard ,

  I have a very similar question, but I have a self-paced reading task.

  Basically, I have a "response_time_Word1" and I have "response_time_Word2". As far as I know, substracting the former from the latter should give me my response time, no? To read the sentences, participants press the space bar, and they see one word at a time. So a sketchpad = a word, and a keypress to keep reading.

  Is this "response_time_Word1" or etc. meaning that, for example if it is 654 ms., the participant spent time looking at the sketchpad from the point it showed up on the screen until their space bar press? I mean, that's how I interpret it but I don't want to analyze the wrong variable...

  But then when I see the "time_Word1" and "time_Word2" and "response_time" and "avg_rt", I got all confused.

  And also here you say "time_variablename" means the timestamp of when that sketchpad is shown since the beginning of the experiment. In my case, do I need to use that?

  I am reaaaaaaaally confused, I would be glad for help.

  Best,

  H
• eduard

  August 2020

  Hi,

  And also here you say "time_variablename" means the timestamp of when that sketchpad is shown since the beginning of the experiment. In my case, do I need to use that?

  That is up to you, most (if not all) of the response time measure are based on these internal time_item variables. But I think the response_time variables are only measured from when the keyboard item was presented until the key was pressed. If you want to measure the time between two sketchpads you probably have to use the variables time_sketchpad1 and time_sketchpad2, to compute the time in between them.

  If you do several different of such computations I would recommend to explicitly compute the RTs (based on the time_xx variables), because it can happen that intermediate items confuse opensesame (e.g. having multiple response items with different settings). When you compute these variables explicitly you are in control of what is happening and don't need to rely on measures computed under the hood.

  "avg_rt": This is the running average of the variable response_time in your log file. In most cases, this variable is not relevant for you.

  Hope this clears things up.

  Eduard

  
• hulya

  August 2020

  Hi Eduard,

  Thank you for your answer! I think I might be a little bit more confused, though...

  But, so, should I disregard all the response_time_items? I mean, you say when the keyboard item was presented, but, what do you mean by keyboard item? Because, I just have sketchpads with "keypress duration set to 0" and a keyboard_response item following that. So, would I be wrong to say, "response_time_item is calculating when the sketchpad (keyboard item?) is presented until the key was pressed" ? And if so, won't I need that?

  I am sorry for the confusion...and what do you mean by explicitly computing them? One by one substraction for every two consecutive item?

  Best,

  H
• eduard

  August 2020 edited August 2020

  I am sorry for the confusion...and what do you mean by explicitly computing them? One by one substraction for every two consecutive item?

  Yes. If you want to know the time difference between sketchpad1 and sketchpad2, you compute it by presentation_time = time_sketchpad2-time_sketchpad1

  But, so, should I disregard all the response_time_items?

  Not, necessarily. That depends on your experimental sequence. Normally, they are quite accurate in straightforward stimulus-response sequences. I guess my main point is, that you can easily compute them yourself if you have doubts that they are accurate (see above the way to compute).

  So, would I be wrong to say, "response_time_item is calculating when the sketchpad (keyboard item?)

  If the sketchpads have duration of 0, then I don't think they have a response_time variable associated with them. I am not 100% sure on the internals of all those variables, but I think all the response_time variables require a user response (keyboard, mouse, etc), and are computed by subtracting the time, when the item was executed and the time the user response was registered. Therefore, two sketchpads with 0 duration should not have a default response time variable associated with them.

  And if so, won't I need that?

  That I don't know. It is a little awkward to explain these things and give concrete advice without a concrete example. So, if you really want to know these things, please give an example.

  Eduard