Habituation example - word-object pairings
About this protocol
This protocol is based on Experiment 3 of the classic Werker et al. (1998) study cited below. Infants are habituated to a single word-object pair. Later, they are presented with four test items: familiar object with familiar word, familiar object with novel word, novel object with familiar word, and novel object with novel word. Pre-test and post-test trials, consisting of a novel object and word that do not appear in any other trials, are also included.
Werker, J. F., Cohen, L. B., Lloyd, V. L., Casasola, M., & Stager, C. L. (1998). Acquisition of word–object associations by 14-month-old infants. Developmental psychology, 34(6), 1289.
Starting the protocol
Starting definitions
This protocol will use one central display and a light positioned directly below the display. These would be our minimal starting definitions:
However, if you have more lights connected in your system (e.g. for headturn preference procedure) and will be leaving them all connected during testing, you may need to define more lights so that you can identify which one should be CENTER
.
Tags
Now we'll define our tags that reference files. We have three audio files and three video files - two of each for use in habituation/test trials, and one of each reserved for pre-test and post-test.
We could define these with LET
statements. But because these will always be paired, it is convenient to use TYPEDLET. This will allow us to define all of our possible word-object pairings and randomly select a pair for presentation, rather than independently selecting audio and video, and having less control over which appear together.
Our videos, which we'll call round
, green
, and blue
, show toys being rotated by a hand on a black background. Our audio files consist of infant-directed repetitions of a nonword: deeb, geff, or mip.
Having defined these tags with TYPEDLET
, we can use LINKED to define tags that pair them up appropriately. The tags mip
and round
will be reserved for pre-test and post-test, but all other pairings will occur in the test phase, and one of them will be featured in the habituation phase.
Groups
We'll define two groups consisting of our LINKED
tags. Both will contain all of the possible pairings of deeb
, geff
, blue
, and green
. At the start of each test session, we will randomly select one pairing from the first group for the infant to see during habituation. The test phase for all infants will consist of all four pairings, so our test_trials
group will also contain all four.
Experiment settings
Now we define our experiment settings. Because this protocol has a habituation phase, we must define all of our habituation criteria here.
We could also define key assignments for live-coding - without them, experimenters will use C for looks CENTER
to the light/screen, and W for AWAY
.
STEPS for execution
Pre-test
Now for the STEPs that will start once we run the protocol. First, we'll define our pre-trial phase. Before each trial, we'll have a center light flash until the infant is paying attention. We'll have the experimenter press C to begin a CENTER
look.
Once the infant is looking, we can start a trial. We'll turn off the light and display the pre-test stimuli. Note that because we defined the LINKED
tag prepost
with both an audio tag component and a video tag component, we can reference the LINKED
tag in the action statements for both AUDIO
and VIDEO
.
Our trials throughout the experiment will last a maximum of 14 seconds. Trials will end when we reach this limit, or when the infant is recorded as looking away for at least 1 second. There is only one pre-test trial, so once it is over, we end the phase and move on to habituation.
Habituation
As we start the habituation phase, the first thing we need to do is assign the participant randomly to a word-object pair that they will see during habituation. There were four possibilities in the habit_pairs
group.
Having selected in advance which word-object pair
the participant will see, we can define the inter-trial period (with the blinking light) and a habituation trial. Note that the pair
tag we selected from habit_pairs
is again a LINKED
tag that can be referenced in both of the action statements to play the audio and video.
We'll play the exact same stimuli for the rest of our habituation trials, so now we'll define a loop. We want to either keep playing trials until the infant meets our habituation criteria or reaches the maximum of 20 habituation trials (19 of them via the loop).
Note that our loop goes back to STEP 8
, where we started the light blinking for the inter-trial period, but excludes the assignment of the dynamic tag pair
back in STEP 7
. This is why we chose which word-object pair would be presented before we needed to use it to display trial stimuli in STEP 10
: we didn't want this LET
statement to be included in the loop. We want this dynamic tag to be assigned as one pair and stay that way so that we are repeatedly presenting the same word-object pair. If the LET
statement were inside the loop steps, we would repeat the choose statement on every loop iteration, and we would show different word-object pairings on different trials. In general, when you want to select from a group and be able to refer to the result of that selection throughout a phase, it's a good practice to make that selection in the same STEP
where you define your phase's start. "Getting it out of the way" like this makes it easier to not accidentally loop over and reassign a dynamic tag that you would prefer to stay static.
Test phase
Once the infant has habituated or met the maximum number of trials, we move on to the test phase. We'll begin again with the inter-trial light flashing before beginning a trial.
Recall that our four test items are all in the test_trials
group, and are LINKED
tags with all the possible pairings of the audio deeb
and geff
, and the videos of the objects blue
and green
. We want to display these four word-object pairs in a random order, without replacement. We'll define one trial, then use a loop to run the remaining trials.
The test phase is where we see the advantage of defining our pairs of tags via LINKED tags rather than selecting video and audio tags separately. If we had defined a test audio group and test video group, they would look like this:
LET test_audio = {deeb, deeb, geff, geff}
LET test_video = {blue, green, blue, green}
With random selection from each in turn across trials, there would be nothing to stop us from repeating a pairing, and thus failing to show all the combinations of words and objects. For example, on our first test trial we could randomly select deeb
and blue
- but there is no way to specify that if we choose deeb
again from the audio group, green
must be selected rather than blue
from the video group. We could define groups that would be chosen from in a fixed order, arranging each of the audio and video tags so that all the pairings are present when they are selected using FIRST (and creating multiple copies of this protocol to counterbalance test trial order.) But without LINKED
tags, we could not use RANDOM selection in this protocol.
We'll use another loop to play the remaining three test trials, after our first one is done, and this concludes our test phase.
Post-test
Lastly, we have a post-test trial, which is identical to our pre-test phase.
Now our experiment is done!
See the resources page for a copy of this protocol.
Last updated