Arithmetic word problem solving: evidence for the construction of a mental model.

Area of Science:

• Cognitive Psychology
• Psycholinguistics
• Artificial Intelligence

Background:

• Problem-solving research often focuses on successful resolution.
• Understanding how individuals represent and recognize problems is crucial for cognitive modeling.
• Previous studies have explored propositional and non-propositional representations in reasoning.

Purpose of the Study:

• To investigate the nature of mental representations formed during arithmetic word problem-solving.
• To determine if individuals rely on abstract relational structures or literal wording for problem recognition.
• To provide evidence for the construction of mental models in problem resolution.

Main Methods:

• Two experiments involving adult participants solving and then recognizing arithmetic word problems.
• Presentation of original, inconsistent, and paraphrased problems during recognition tasks.
• Analysis of recognition rates to differentiate between representation types.

Main Results:

• Paraphrased problems, despite linguistic differences, were recognized at higher rates than inconsistent problems.
• Recognition was not solely based on remembering exact numerical values from the problem text.
• Results indicate a reliance on the underlying relational structure of problems.

Conclusions:

• Individuals construct non-propositional representations when solving arithmetic word problems.
• These findings support the mental model theory, suggesting abstract representations are key.
• Understanding problem structure, not just surface features, is vital for cognitive processing.