FIRST THOUGHTS AND SOLVING PROBLEMS:

WHAT DO WE LEARN FROM PH.D. MATHEMATICIANS

 

Jean McGivney-Burelle

University of Connecticut

mcgivney@uconnvm.uconn.edu

 

Thomas C. DeFranco

University of Connecticut

defranco@uconnvm.uconn.edu

 

Although the nature of expert mathematical problem-solving performance has been documented, several issues need further investigation. As a result, 15 Ph.D. mathematicians from research I universities were asked to think aloud while solving four complex mathematics problems and respond to questions regarding their solutions. Information regarding participants’ “first thoughts” about a problem and control behavior exhibited on the problem was organized in a contingency table and a chi-square test was employed to analyze the data. Results indicated that: (1) in the majority of problem-solving protocols, participants were unable to identify the deep structure of the problems prior to solving them, and (2) there was no significant relationship between identifying the deep structure of a problem and exhibiting efficient control behavior. In addition, the 60 problem-solving protocols were coded and a cross-case analysis of the protocols was employed. General themes regarding efficient and inefficient control behavior with respect to domain knowledge, problem-solving skills, and beliefs were identified. Further analysis revealed that the use of appropriate heuristics and efficient control behavior enabled participants to solve the problems in spite of their inability to recognize the deep structure of a problem or identify an approach to the problem prior to solving it. This study was, in part, exploratory and more work needs to be done to understand control behavior exhibited by experts in the midst of solving complex mathematics problems.