About the task
RNC comment: This was referred to by Sidse as "paired associates learning". I can see no reason to call it that. It doesn't seem to resemble the human CANTAB PAL test (though I haven't got a good description of that). I've called it "delayed matching to location (DMTL)".
I'm also not sure why we need to have a whole library of stimuli for this task. [Excerpt adapted from Sidse's wish-list: In Monkey-CANTAB, a unique object is used for each trial, so the task requires a large library of images. Pigs are poorer at visual discrimination; therefore, Sidse wants to ensure that the test only uses objects known to be discriminable by pigs. She suggests (1) that the objects are combinations of one of 6 colours and one of 6 patterns; (2) that the target object is selected pseudorandomly, such that it is never an object that has been presented in the last 10 trials.] Object identity has no bearing on the task, so why not dedicate one object to this task (perhaps even as a discriminative stimulus indicating that "now we're doing DMTL") and just use that one? I can't think of an obvious reason why changing the task stimulus every task is important for DMTL. There are only two tasks in which the subject sees a whole array of identical stimuli - this and Spatial Working Memory, which is a one-stage task and different in several other ways. Probably not wise to use the same object for SpatialWM and DMTL, and other than that I can't see any problem.
So we'll use one object for the moment. Simpler.
I seem to have a lot of queries about this task, so maybe I've misinterpreted the whole thing... in which case we'll have to do PAL, whatever that might be - and consider this a bonus task! :-)
Further information on human paired-associates learning (PAL) from Shibley Rahman: at the highest level, eight boxes are displayed around the edge of the screen. A number of stimuli (up to eight) are shown sequentially in different locations. In the test phase, a stimulus is shown in the centre and the subject must indicate which location it was in previously. It's therefore a delayed matching-sample-to-location test (the association under test is between stimulus and location).
This would fit with Sidse's comments re having lots of stimuli, but she says "... Showing an object in one OBL [meaning location] and then after a delay presenting this object again in all locations and the correct location has to be touched." That's not the same as human PAL as described by Shibley, who should know.
Mike agrees that DMTL only becomes PAL when you are presented >1 stimulus/location pair in the first phase, and then have to match them in a subsequent phase.
So we need to ask the Danes about this.
| • | Marker 1 signals the beginning of a trial. |
| • | One object is shown, somewhere. (Optionally, the subject must touch this, in which case it is punished and the trial ends if it fails to respond. As a further option, it may be rewarded for touching it.) |
| • | Then there's a "memory" delay. (Configurable.) |
| • | Marker 2 signals end of delay. |
| • | Several (1-4) copies of the original object are presented in random locations. Only one is in the same location as the original object. This one must be touched (rewarded). |
| • | Timeouts/incorrect responses punished. |
Levels: same as number of objects shown in second phase (1-4).
Option to repeat the second stage of the trial (once) if the subject gets it wrong. (We''ll present this repeat of phase 2 straight after it's been punished for getting it wrong in phase 2.) (Sidse requested an unspecified "correction procedure" in addition to this, but this is a correction procedure, so we'll leave it at this unless she requests something else.)
Configuring the task
| • | Maximum number of trials. When the subject has performed this number of trials, the task ends. (You may specify 0 for no limit, though you must specify a limit on the number of trials, the time, or both.) |
| • | Maximum time. When this time elapses, the task is terminated as soon as the current trial has finished. (You may specify 0 for no limit, though you must specify a limit on the number of trials, the time, or both.) |
| • | Must touch phase 1 stimulus. If this is selected, then the subject must respond to the Phase 1 stimulus in order to proceed to phase 2. If you choose this option, you may also choose whether or not the subject should be rewarded for touching the Phase 1 stimulus. If you do not want your subject to have to touch the stimulus, you must specify the Phase 1 stimulus duration instead. |
| • | Maximum time to wait for a response. If the subject fails make a response within this time, the subject fails the trial. (This time limit applies to Phase 1, if you require your subjects to touch the Phase 1 stimulus, and Phase 2, and the correction procedure if one is used.) |
| • | Correction phase if subject fails Phase 2. Optionally, phase 2 can be repeated immediately (once) if the subject fails it the first time. |
| • | Time between trials. Specify a minimum and a maximum intertrial time (they may be the same). The actual time is chosen with a rectangular probability distribution within these values. |
| • | Main object. This shows the object that will be used as the stimulus. Click Set to choose the stimulus. |
| • | Starting number of stimuli. Obvious! |
| • | Increase number of stimuli during task. If enabled, the number of stimuli will increase as the task goes on. You can either increase the number of stimuli after a fixed number of trials, or when the subject reaches a criterion of a certain number of "correct" trials within the last 20 trials. Fill in this number (either the number of trials, or the number to get right out of 20) in the box labelled X. |
Options for reward and punishment are set in the General Parameters section; visual objects are defined in the Visual Object Library.
Screenshots from the task
Phase 1
Phase 2
