# Preservice elementary teachers' performance on tasks involving building, interpreting, and using linear mathematical models based on scientific data as a function of data collection activities

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Anita Hill Bowman (Creator)
Institution
The University of North Carolina at Greensboro (UNCG )
Web Site: http://library.uncg.edu/
George Bright

Abstract: A modification of Janvier's "star" model of understanding mathematical function is proposed as a theoretical basis for framing this experimental study of the relationship between preservice elementary teachers' performance on tasks involving building, interpreting, and using linear mathematical models based on physical science data and whether or not the subject participated in data collection tasks. Fifty-two elementary education majors enrolled at a small university in the southeastern region of the United States participated in this experiment by completing two 2-hour workshops and a 50-minute, 36-item posttest. Twenty-seven subjects were randomly assigned to the "data collection" group and 25 to the "no data collection" group. All participants used TI-81 graphing calculators to analyze the relationships between four pairs of variables: (a) total mass of a liquid and its container (Y) versus the volume of liquid used (X), (b) total height from the table top to the water level in a beaker (Y) versus the volume of water in the beaker (X), (c) total mass of coins and the cup containing the coins (Y) versus the number of coins in the cup (X), and (d) the length of a spring (Y) versus the total mass of objects attached to the spring (X). Data analysis via TI-81 calculators included entering data from tables, constructing scatter plots, and determining the least squares linear regression model. For each mathematical model constructed, subjects identified the slope and y-intercept, including units of measure; constructed a contextual interpretation of the slope and y-intercept; and solved verbal problems using the model to predict outcomes.