Counter model for word identification: reply to Bowers (1999)

Area of Science:

• Cognitive Psychology
• Psycholinguistics
• Computational Neuroscience

Background:

• The counter model (Ratcliff & McKoon, 1997) models word identification processes.
• It explains how prior word encounters influence current identification.
• Prior encounters typically create response biases, favoring previously seen words.

Purpose of the Study:

• To demonstrate how the counter model explains improved performance with low-frequency words after prior encounters.
• To refute claims that data supporting the counter model in forced-choice experiments are spurious.
• To advocate for a comprehensive theoretical framework for word identification.

Main Methods:

• Analysis of word identification experiments, particularly those involving brief word presentations.
• Examination of data concerning low-frequency words and prior encounter effects.
• Re-evaluation of findings from forced-choice experiments with dissimilar alternatives.

Main Results:

• The counter model successfully accounts for improved performance on low-frequency words following prior exposure.
• The authors demonstrate that criticisms regarding data supporting the counter model are unfounded.
• The model's explanatory power extends across various tasks and variables.

Conclusions:

• The counter model provides a robust explanation for word identification phenomena, including the effects of prior encounters.
• The study reaffirms the validity of data supporting the counter model.
• A detailed cognitive account of word identification is essential for understanding performance variations.