A new sentence generator providing material for maximum reading speed measurement

Area of Science:

• Cognitive Psychology
• Psycholinguistics
• Human-Computer Interaction

Background:

• Assessing maximum reading speed is crucial for understanding reading abilities.
• Existing methods may have limitations in generating sufficient equivalent stimuli for repeated testing.
• Controlling for lexical and psycholinguistic variables is essential for accurate reading speed measurements.

Purpose of the Study:

• To introduce a new method for generating a large number of equivalent short sentences for assessing maximum reading speed in adults.
• To validate the proposed method against traditional reading speed assessment tools.
• To enable reliable repeated measurements by minimizing sentence learning effects.

Main Methods:

• Development of a sentence generator utilizing an ontology for concept distribution.
• Generation of true/false sentences based on common knowledge.
• Implementation of a psychophysical threshold algorithm for determining maximum reading speed.
• Comparison of reading speed scores with traditional MNREAD sentences.

Main Results:

• The proposed generator produces a vast number of equivalent sentences.
• Maximum reading speed scores from the new generator strongly correlated with traditional MNREAD scores (r = .836).
• The method facilitates repeated measurements, mitigating sentence familiarity issues.

Conclusions:

• The new sentence generation method is a valid and reliable tool for assessing maximum reading speed.
• Its capacity for generating numerous equivalent sentences is advantageous for research on reading performance variability.
• This method enhances the ability to conduct precise and repeatable reading speed assessments.