USA college students were tested pre-teaching with comparative pencil and paper tests.

Learners’ ideas

• Heavier objects fall faster than light objects.
• Objects with the same speed that approach an incline will reach a higher height if they have a greater mass.
• Acceleration and velocity are synonymous.
• The total volume of a liquid affects the pressure because of the increased weight of the liquid present.
• Some students had no instruction on topics but still showed a systematic way of making the predictions, whether correct or not.
• Students cling strongly to certain patterns of thinking e.g. heavier objects fall faster, and they also cling to a particular strategy.
• In questions involving the pendulum, students can find it difficult to decide how to use angular values (the angle at which the pendulum bob is released) but were more eager to use variables such as the 'mass of the bob' and the 'distance out to the side'.
• Even though most students included irrelevant variables in their answers, no students ignored the relevant variables.
• Students' strategies may start to attribute different levels of importance to the different variables. E.g. For the heavier objects fall faster pattern of thinking, the importance of mass and height of the object can start to differ if their values are varied in a question.
• Students made decisions about a strategy and then applied them consistently.

The study

Aims

To identify which strategies students use in an introductory physics course to predict the behaviour of common physics systems.

How the evidence was collected

Before the teaching commenced, all students underwent testing through paper-and-pencil tasks. Almost all of these tasks involved presenting students with two similar physical scenarios and asking them to predict which of the two would exhibit a specific behavior, such as moving faster or producing a greater rate of water flow. Students were required to provide answers and explain their reasoning for each question.

The data was analysed using pattern recognition. This involved identifying all the plausible strategies that individuals might employ to predict the behavior of a system. For each strategy, an ideal sequence of answers was generated. These ideal sequences were then compared to the actual sequences provided by students until matches were identified. If the ideal and actual sequences differed on no more than four items, they were considered as matching.

Details of the sample

An unknown number USA students (probably over 18) on a college general physics course.