medin87train {catlearn}

Input representation of Exp. 1 in Medin et al. (1987) for models input-compatible with slpALCOVE or slpSUSTAIN.

Description

Creates randomized training blocks for Experiment 1 in Medin et al. (1987), in a format that is suitable for slpALCOVE, slpSUSTAIN, and other models that use either of those input-representation formats.

Usage

medin87train(blocks = 2, subjs = 2, seed = 7649, missing = 'pad')

Arguments

subjs

Number of simulated participants to run.

blocks

Number of blocks to generate. The ten trial types are randomized within each block.

seed

Set random seed.

missing

If set to 'geo', output missing dimension flags (see below). If set to 'pad', use the padded stimulus representation format of slpSUSTAIN.

Details

A matrix is produced, with one row for each trial, and with the following columns:

ctrl - Set to 4 on the first trial for each participant - 4 resets the model to the initial state and does unsupervised learning afterwards. Set to 3 for unsupervised trials - normal unsupervised learning trial.

blk - Training block.

stim - Stimulus number, ranging from 1 to 10. The numbering scheme is the same as in Medin et al. (1987, Fig. 1).

x1, x2, ... - input representation. Where missing='geo', x1, x2, and x3 are returned, each set at 1 or 0. This is the binary dimensional representation required by models such as slpALCOVE, where e.g. x2 is the value on the second dimension. Where missing='pad', w1, w2, x1, x2, y1, y2, z1, z2, are returned. This is the padded represenation required by models such as slpSUSTAIN; e.g. y1 and y2 represent the two possible values on dimension 3, so if y1 is black, y2 is white, and the stimulus is white, then [y1, y2] = [0, 1].

Although the trial ordering is random, a random seed is used, so multiple calls of this function with the same parameters should produce the same output. This is usually desirable for reproducibility and stability of non-linear optimization. To get a different order, use the seed argument to set a different seed.

Value

R by C matrix, where each row is one trial, and the columns contain model input.

Author(s)

Lenard Dome, Andy Wills

References

Medin, D. L., Wattenmaker, W. D., & Hampson, S. E. (1987). Family resemblance, conceptual cohesiveness, and category construction. Cognitive Psychology, 19(2), 242–279.