Date of Graduation


Document Type



This study posed the following questions: (a) Does error-detection instruction improve students' confidence levels and/or reduce anxiety? (b) Do students find error-detection instruction helpful? (c) Do students' methods and awareness of error detection change after instruction? and (d) Does age, sex, math background or personality influence attitudes toward errors and error detection? College developmental algebra students received error-detection instruction during a semester-long beginning algebra course. They discussed errors and error detection mainly through the medium of "Pat Perplexed" problems. Pat was a mythical student who rarely got a problem correct. Pat's errors were corrected and explained collaboratively. Research on transfer, language, and metacognition was instrumental in the design of these problems. At the beginning and end of the semester students completed the Fennema/Sherman Confidence in Learning Mathematics and Anxiety Scales, a survey on error-detection methods, and personal interviews. An evaluation was completed at the end. Quantitative research used the entire sample. Qualitative research used 14 nontraditional students having the lowest combined initial score for the Fennema/Sherman Confidence and Anxiety Scales. Regression analysis indicated gender, which was highly correlated with entering attitude toward mathematics of interviewed students, and Pat involvement were factors related to change in Confidence scores. Instruction did not have a significant effect on anxiety. Small group involvement was important for perceived benefit, and those benefits differed for each student. Error-detection instruction increased ability, variety of methods and efficiency in finding errors. Students who viewed mistakes as a reflection on their intelligence did less well in the course or had less improvement on the Fennema/Sherman score than those who viewed mistakes as a challenge and part of the learning process. A strong affinity existed with "Pat" allowing for increased comfort and improved metacognition. Traditional and nontraditional students were more similar than dissimilar when it came to error detection. In conclusion, the opportunity for collaboration, metacognition, writing, verbalization, and identification with "Pat" were important inclusions in error-detection instruction.