COMPUTATIONAL ESTIMATION STRATEGIES OF PROFESSIONAL MATHEMATICIANS

Forty-four academic pure mathematicians were given Levine's (1982) computational estimation task, which involved estimating the answers to 20 multiplication and division problems and describing the strategies used. The mathematicians were accurate estimators, and they used a great variety of strategies, as many as 23 for a single problem. All but 2 of the mathematicians used some strategy that was not used by the other 43 for at least one problem. When 18 mathematicians were retested after an interval of several months, they used different strategies on the second test for 9 to 17 of the 20 problems. Some theoretical implications of these findings are discussed.