(update 03/18/2005)
Is category learning affected differently by test vs. study?
Task:
Learn visuospatial category task by classifying novel exemplars from three categories à transfer to novel exemplars during final test.
Stimuli:
Categorize 10 x 10 grid patterns of red and black squares, each generated from one of three prototypes (A, B, or C) by randomly transposing each square with probability .09 (same as in Holyoak Space 4).
Design:
Conditions: Subjects randomly assigned to 2 between-subjects conditions:
1. Test Condition: Present grid à make category judgment (A, B, or C?) à make representativeness judgment (“How representative is the current grid of others in its category?”) à get auditory feedback.
2. Study Condition: Present grid + auditory presentation of category à select correct category (A, B, or C) à make representativeness judgment (“How representative is the current grid of others in its category?”).
Sessions:
Session 1 (Learning): One learning session of 240 trials (8 blocks of 30 trials), either test or study, with correct category presented (study condition) or feedback after retrieval (test condition). » 5 sec each for category and representativeness judgments + 1(?) sec feedback (11 sec x 240 trials = 44 min)
Session 2 (Test): One test session of 60 trials (2 blocks of 30 trials) without correct category presentation or feedback, and without representativeness judgment. » 5 sec for category judgment (5 sec x 60 trials = 5 min)
Total time: Assume 3 min for consent and instructions + Session 1 (44 min) + Session 2 (5 min) = 52 min.
Retention interval: 7 days
Subjects: 80 web-based paid subjects to pilot (40 per condition).