Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Area of Science:

• Cognitive psychology
• Computational geometry
• Experimental psychology

Background:

• Controlled stimuli are crucial for figure recognition research.
• Previous studies indicate axisymmetric pairs are harder to discriminate than non-identical pairs.
• Isomorphic figures share structural properties but may differ in shape.

Purpose of the Study:

• To determine if identical line lengths in figure pairs increase discrimination difficulty to the level of axisymmetric pairs.
• To explore the impact of superficial features on shape discrimination within isomorphic figures.
• To systematically generate and categorize stimuli for figure recognition experiments.

Main Methods:

• Developed a method to generate controlled (6, n) figures using points on a regular hexagon.
• Calculated and stored structural (graph invariants) and superficial (non-graph invariants) features in a database.
• Generated three types of isomorphic (6, 4) figure pairs: rotated-to-be-identical (Idr), axisymmetric (Ax), and non-identical, non-axisymmetric (Nd).

Main Results:

• Identical line lengths between figures increased discrimination difficulty, comparable to axisymmetric pairs.
• Non-identical, non-axisymmetric (Nd) pairs were further classified based on line length differences.
• The study provides a framework for generating and analyzing visual stimuli for cognitive experiments.

Conclusions:

• Line length similarity is a significant factor influencing the difficulty of discriminating between visual figures.
• The findings contribute to understanding the perceptual basis of shape recognition and feature importance.
• The developed protocol offers a robust method for creating precise experimental stimuli in vision research.