Announcement

Collapse
No announcement yet.

Data Output in a Recognition Test

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Data Output in a Recognition Test

    I am currently working on a recognition test. I’m having an issue with my data output and I will outline the experiment in order to allow for a better understanding of my issue.

    [Block 1] Participants are presented with instructions to study the stimuli (simple sentences such as ‘the professor likes candy or water’) carefully,

    [Block 2] followed by a cue, and a list of ‘original statements’ (i.e., statements that they will be comparing with those that are presented in the recognition phase of the experiment). Note: The ‘original statements’ all belong to one stimulus list. Each item is presented after a cue and is visible for a fixed time during this portion of the experiment.

    These first two blocks have gone off without a hitch. The cue is presented between each stimulus item.

    [Block 3] Participants are given instructions to respond with a key stroke to signify whether the statement is ‘old’ (i.e., belongs to the list of ‘original statements’) or ‘new’ (i.e., does not belong to the list of ‘original statements).


    Note: The recognition phase is filled with several types of statements:

    -original statements

    -several types of foils, i.e. several sets of statements that are similar, but differ from the original statement in some significant way.

    -- these foils I have grouped into different sets, with the aim of having data analysis simplified. That is, with the aim of being able to examine whether there is a relationship between the types of foils that are selected and the type of original statement (there are three types of original statements, i.e., claims made using the connective, ‘or’, ‘and’, or ‘if,then’).

    In the process of making these different sets I created different stimuli lists for each type of foil. For example, here is a list of an original statement and four possible foils:

    Original statement (or): This professor runs marathons or eats pasta

    Different-form foil (and): This professor runs marathons and eats pasta

    Different-form foil (if,then): If this professor runs marathons, then he eats pasta

    Different-noun foil: This professor runs marathons or eats chocolate

    Different-connective foil: This professor runs marathons __(but/whenever/unless)_ he eats pasta


    Is it possible to have my data output signify not only whether the person correctly identifies an original statement as an original statement or a foil as a foil (simple right or wrong recognition test), but also the prevalence of a certain type of foil being selected as an original foil? For example, could I have my output include the variable of the frequently that an ‘and’ foil is selected when the original statement is an ‘or’ statement? Could I have my output tell me what the frequency of an ‘if,then’ foil is selected when the original statement is an ‘or’ statement?

    Thank you for your time.

  • #2
    It is possible to do these things, I think, by using a combination of tags, parameters, and rules. You can tag trials directly in the Trial Editor. But you can also modify that tag at run-time using a trial-level rule, e.g. “If <criteria is true> then Set trial’s tag to ___”. Every tag set that you create results in an additional column in the data file.

    For the output to indicate a frequency, you will need to create a parameter of type Counter and then use a rule to increment the parameter.

    Comment


    • #3
      Is it possible to tag at the event level?

      Thanks for the advice! It's greatly appreciated.

      You say, “If <criteria is true> then Set trial’s tag to ___”. This may seem trivial, but would it be possible to do this for events? That is, to create trial level rules for each stimulus list?

      For example, (as said before in this recognition test I have a stimulus list for 'original' items and a different list for each type of foil) would it be possible to create rules such that items in EACH list were given a unique tag depending on whether they were correctly (or incorrectly) recognized as 'new' or 'old' statements?

      Comment


      • #4
        Correction: "That is, to create trial level rules for each stimulus list?" changed to
        "That is, to create rules for each stimulus list?"

        Comment


        • #5
          I think this question is better answered on the phone. Call me at 1-800-233-7871.

          Comment


          • #6
            Wonderful Phone Support!

            Thank you, Hisham, for the phone support! You clearly guided me through this issue and I am extremely thankful.
            All the best!

            Comment

            Working...
            X