Multiplicative computational estimation with decimal numbers

Abstract

Computational estimation is a basic skill, useful in solving arithmetic problems. In this paper, we analyze the difficulties, strategies and errors in computational estimation of multiplications and divisions with natural and decimal numbers. We use a mixed methodology, quantitative and qualitative, collecting data by means of an estimation test, followed by a phase of interviews. Participants in the study were 131 pre-service teachers. Among the results, tasks with decimals less than 0.1 are more difficult than others; tasks with decimals less than 1, more difficult than those with natural or decimals greater than 1; and B divisions (with dividend less than divisor) are more difficult than A divisions (with dividend larger than divisor). Qualitative study shows 3 basic estimation strategies (First digits, fractions, and alternative algorithm) and 15 types of errors (4 in interpretation phase, 10 in computation phase, and 1 in evaluation).