Outcomes: Comparing Stimulus and Response

Was I right or wrong?

On each trial, when we make a response, we might wonder about the relationship between our response and the actual state of affairs — in other words, we are interested in the outcome. If the stimulus and response match, we are correct. If the stimulus and response don’t match, we are in error. And sometimes that may be all that really matters - was my answer correct or was it an error?

Try a few trials of the task:

As you have no doubt noticed, if you have given it a try, you receive feedback.

The feedback is presented to the right of the response buttons. At the top, you are shown the current trial and the total trials. Below that, for the moment, the feedback is either Correct, Error, or if you don’t respond in time, No Response.

How was I right or wrong?

If we think about it a bit more, we realize that there are actually two ways we can be correct, and two ways we can be in error, leading to four possible outcomes: hit, miss, correct rejection, or false alarm.

  • A hit, also called a true positive, is when the signal is actually present, and we correctly respond ‘present’
  • A miss, also called a false negative or type II error, is when the signal is actually present, and we erroneously respond ‘absent’
  • A false alarm, also called a false positive or type I error, is when the signal is actually absent, and we erroneously respond ‘present’
  • A correct rejection, also called a true negative, is when the signal is actually absent, and we correctly respond ‘absent’

We have two possible states of the world (signal present or absent) and two possible responses (‘present’ or ‘absent’), leading to a nice two by two table of possible outcomes:

Try a few more trials of the task, and this time the feedback will inform you of which of the four possible outcomes occurred:

Now our feedback is Hit, Miss, False Alarm, Correct Rejection or if you don’t respond in time, No Response.

Now that we know how to classify the outcome of individual trials, we might wonder how we can aggregate and summarize these results…