[solved] Correcting Response IAT with OpenSesame

fdelange

I would like to ask a question about creating the IAT with OpenSesame. When participants are pressing the wrong key I want to insert a screen were they can correct themselves. So after the Feedback (which they will see a X) what do I do ?

lvanderlinden

Hi Florence,

Welcome to the forum!

I think the best way to give participants a second chance after an incorrect answer is by executing the following steps:

1. Append some 'second chance' items (e.g. a second sketchpad, for (almost) the same stimulus presentation, and a second keyboard_response item, for the new response) to your trial_sequence. Place them after your feedback item, but before the logger item.
2. Set their "Run if" statements, in the trial_sequence item, to [second_chance] = yes.
3. Append an inline_script item immediately after the feedback item, but before the 'second chance' items.
4. Place something like the following code to its Run phase tab (see comments for a more detailed explanation):

# First, check whether the response was correct by using the 
# built-in OpenSesame variable 'correct', It has the
# value 1 if the most recent response was correct, and the 
# value 0 if it was incorrect.

# We retrieve this variable by using the experiment function 
# 'self.get()'.

# If the response was correct, we DON'T need to run the 'second
# chance' items:
if self.get("correct") == 1:
    exp.set("second_chance", 'no')

# If the response was incorrect, we DO need to run the 'second
# chance items:
elif self.get("correct") == 0:
    exp.set("second_chance", 'yes')

After applying those suggestions, your overview area and trial_sequence item should look something like this:

An example experiment can be found here (simply download the script, save with the extension .opensesame, and open as normally):

• http://pastebin.com/FUVQSpZh

Does this help?

Best,

Lotje

amorphia

Thanks! This was helpful for me at least: I also just implemented an IAT in OpenSesame. Although, I eventually realised it can be even simpler than this. You don't have to use an inline script - you can just use the inbuilt variable "correct" directly in the Run if fields. First I use it directly after the first Response item to see if I need to display the stimuli again but with a red cross (indicating a mistake was made), then directly after that I use it to see whether I need to collect a second response. So I skip the feedback item and the inline item, and just have a conditional second sketchpad item and a conditional second response item.

LenaFrescamente

Dear open Sesame Team & Users,

First of all, thank you very much for opensesame, it is really great!

Yet, I currently have a problem with adjusting the logger.
My experiment contains an implicit association test, which is basically 7blocks (loops) with each running a sequence containing a list of single words (sketchpad Items). Participants are asked to respond to each word with pressing either the “e” or the “i” key (checked via keyboard response “KEY”).

Depending on the valence of the word either the one or the other is correct. If they make a mistake, another sketchpad item appears (“Error”) which also comes along with a keyboard response waiting allowing only for “e” and “I” keys. This so far works fine (see loop structure below).

However, I would need the response time (aka time since the sketchpad item has been displayed till the respective key was pressed) for each of the single words (within each block) plus the time they need to correct their answers (this can be in two variables, I can summarize later in SPSS).

As far as I understood it, the logger does not store the response time not for the single items within. (Currently I have 1 logger as last step in the experiment with the default "check everything").
As I have used the same key for each word within the sequence as well as within all blocks, the “new response_time_KEY” overwrites the former.
I guess I need some kind of inline scripting that creates a new variable called response_time_1 for each row of the cyrcle. Moreover, some of the items are repeatedly shown (cycle repeated 2 times) so it would be great if this variable would also contain the cycle_nr. For instance “response_time_1_1”. I hope my question makes any sense to you and I thank you very much for your help!

Here is the script for the first block:

set repeat "1"
set description "Repeatedly runs another item"
set item "Pos_Neg_Sequence"
set column_order "Valenz;correct_response"
set cycles "16"
set order "random"
setcycle 0 correct_response "e"
setcycle 0 Valenz "einfühlsam "
setcycle 1 correct_response "e"
setcycle 1 Valenz "engagiert"
setcycle 2 correct_response "e"
setcycle 2 Valenz "geistreich "
setcycle 3 correct_response "e"
setcycle 3 Valenz "hilfsbereit"
setcycle 4 correct_response "e"
setcycle 4 Valenz "interessiert"
setcycle 5 correct_response "e"
setcycle 5 Valenz "sympathisch "
setcycle 6 correct_response "e"
setcycle 6 Valenz "verantwortungsbewusst"
setcycle 7 correct_response "e"
setcycle 7 Valenz "zuverlässig"
setcycle 8 correct_response "i"
setcycle 8 Valenz "aggressiv "
setcycle 9 correct_response "i"
setcycle 9 Valenz "dumm "
setcycle 10 correct_response "i"
setcycle 10 Valenz "gewissenlos"
setcycle 11 correct_response "i"
setcycle 11 Valenz "habgierig "
setcycle 12 correct_response "i"
setcycle 12 Valenz "intolerant "
setcycle 13 correct_response "i"
setcycle 13 Valenz "kaltherzig "
setcycle 14 correct_response "i"
setcycle 14 Valenz "mitleidslos "
setcycle 15 correct_response "i"
setcycle 15 Valenz "zänkisch "

The sequence than looks like this:

• Pos_Neg_Sequence
  • Flush pending key presses at sequence start = checked
  • Sketchpad: "Pos_Neg" displays [Valenz]
  • Keboard_response "KEY" checks for the correct_response
  • Sketchpad: Error run if [correct] = 0
  • KEY_Error run if [correct] = 0 waits for the correct response

LenaFrescamente

BTW Sorry for the terribly formated post! I

sebastiaan

Hi Lena,

Currently I have 1 logger as last step in the experiment with the default "check everything"

The logger item should be placed at the end of the trial sequence, which in your case is called Pos_Neg_Sequence. If you do this, everything should be logged as a normal multirow spreadsheet, which I'm guessing is what you need, right?

For examples on how to use the logger, see the tutorials:

• http://osdoc.cogsci.nl/tutorials/step-by-step-tutorial/
• http://osdoc.cogsci.nl/tutorials/aps-tutorial/#step3additemstoyourtrialsequence

BTW Sorry for the terribly formated post!

For formatting tips, see

• http://www.cogsci.nl/forum/index.php?p=/discussion/2/tips-and-tricks-e-mail-alerts-formatting-account-settings#formatting

I hope this clears things up!

Cheers!
Sebastiaan

LenaFrescamente

Dear Sebastian,

thank you very much for your help and fast answer. Now the recodring of the single items works very good! Thank you!

loics

Hi everyone,

I try to find out an easier way to get a feedback INTO the display on the IAT (when the answer is wrong). In this way, we only have to configure the logger to collect the correct answer, and we don't need to input another display "second chance".

Fort that, I made :

• a list of 3 variables: [Stimulus] (which is printed into the display) ; [Corr] (for the correct answer associated ; [Wron] (for the wrong answer associated)
• the duration's display on "0"
• the keyboard_response on:
  correct answer: [Corr]
  allowed answer: [Corr]

Something like that :

My idea is to print the red cross only when the participant press the key associated in the variable [Wron]. Is it possible ?

Thanks for your help !

sebastiaan

Hi @loics,

Sure, you can use define a correct-response variable, and use run-if statements to do something only when a preceding response was incorrect. I would recommend following the beginner tutorial, which explains this in detail:

• http://osdoc.cogsci.nl/tutorials/step-by-step-tutorial/

Good luck!
Sebastiaan

loics

Hi @sebastiaan,

Thank you for your answer,
I will propose my version of IAT when I finish it !