### Abstract:

The purpose of the study was to investigate students' learning of the function concept and the role of the graphing calculator in a College Algebra course. Differences between students with high symbolic manipulation skills. and students with low symbolic manipulation skills were also examined. On the basis of an algebraic skills test administered by the instructor (high/low) and students' academic majors (math & science, business, and liberal arts), 25 students from one College Algebra class were placed into six categories.
To gather data on students' understanding of functions, a pretest and posttest were administered. The Function Test consisted of four identification questions given in each of the representations, three questions asking for the definition, an example, and a nonexample of functions, and 15 questions consisting of three problem situations given in the numerical, graphical, and symbolic representations. To gather data on the role of the graphing calculator, daily classroom observations were conducted. To verify students' responses and classroom observations, formal interviews with students and informal interviews with the instructor were conducted.
Students' personal definition progressed towards the formal definition of functions. Yet, students had difficulties with the univalence requirement in three areas:
(a) order of domain and range, (b) preference for simple algorithms, and (c) the restriction that functions were one-to-one. Compared to students with low symbolic manipulation skills, students with high symbolic manipulation skills were more flexible working between representations of functions. Half of the interviewed students with low symbolic manipulation skills perceived a single function given in numerical, graphical, and symbolic representations as separate entities.
The graphing calculator played a role in all phases of the solution process. During the initial phases, students used calculators to develop a symbolic approach. The prime motivation for using graphing calculators during the solution-execution phase was to avoid careless errors. The most common use of graphing calculators was to check answers during the solution-monitoring phase.
However, graphing calculators created difficulties for students who accepted graphs at face value. Interpreting the truncated graph shown by the calculator, students determined that exponential functions possessed a bounded domain because they did not explore the graph.