Effects of Attentional Cueing on Novice
Problem Solvers
Tanner J. Stevens
,
HOME PROJECT OTHER REU EXPERIENCES ABOUT ME
My project investigated how students react to visual cueing when
solving introductory physics problems.
Previous studies have suggested that students who answer visual
problems correctly with the right
conceptual understanding tend to look at certain areas more
frequently, while students with incorrect
conceptual understanding look at other areas of the
problem. Our goal is to cue students to
look in areas
where most “expert” problem solvers look while tracking the
students’ eye movements to see whether we
can help students activate prior knowledge, solve the problem
and gain a better understanding of the
concepts
used.
Students were asked to solve conceptual physics problems with
highly visual components while we tracked
their eye movements.
There were four groups of problems, each being on a problem from a
previous study
which showed
large differences in eye movements between correct and incorrect students. After solving each
problem, the students were asked to explain their
reasoning. If the student answered
incorrectly or gave
incorrect reasoning, they were shown a similar problem. Students in the cued condition saw colored
shapes
flash across the screen in the similar problems, while students
in the non-cued condition saw the problem as
normal. Once the student
answered correctly with correct reasoning, or if they didn’t answer correctly
after
three similar problems, students were shown a transfer problem,
involving the same concept but in a
different context, to determine whether the student could still
apply the reasoning to other situations.
Answers for all problems were recorded, and each interview was
video recorded and audio recorded for
qualitative analysis on students’ explanations.
I first looked at students’ accuracy on the transfer
problems. Over all problems, students in
the cueing
condition correctly solved the transfer problem with correct
reasoning 34.4% of the time, while students in
the non-cueing condition gave correct answers and reasoning
25.9% of the time. Students verbal
responses
were also analyzed and categorized into different mental models
used to solve each problem. It was found
in
the cued group of students that the average number of changes in
mental model per problem group was
higher than the non-cued group in three of the four problem
groups. More in-depth data analysis will
be
available soon in my final presentation below.
I am currently finalizing a proposal to present my project at
the National Association of Research in Science
Teaching (NARST) Annual Conference. If my proposal is accepted, I will complete
and present a full paper
at the conference in Orlando in April, 2011. Wish me luck!
Unfortunately, since the review process is
double-blind, I will not be posting my paper here. I have posted my final presentation slides
below.