and cognitive processing (Linnenbrink & Pintrich, 2002b). For elaborative
strategy use, there was a significant positive relation for study 1 (r = .20,p<
.05); however, the correlation was not significant for study 2 (r = .16,p> .05).
Similarly, metacognitive strategy use was not significantly related to positive
affect in study 1 (r = .13,p> .05) but there was a significant positive relation
for study 2 (r = .33,p< .001). Thus, the findings linking positive affect to
strategy use were somewhat small and inconsistent between the two studies.
For negative affect, we found that students with negative affect consis-
tently scored worse on the post-test measure of physics understanding (study
1: r = –.21,p< .05; study 2: r = –.36,p< .001, Linnenbrink & Pintrich,
2002b). However, it was not clear whether negative affect was related to con-
ceptual change. In particular, additional regression analyses designed to ex-
amine the change in understanding of Newtonian physics (pre-test score in-
cluded as a control) showed no significant relation between negative affect
and the post-test measure of physics understanding for study 1 (b= –.05,p>
.05) but did show a significant negative relation for study 2 (b= –.34,p£
.001). Thus, while the findings suggest that negative affect may be detrimen-
tal for conceptual change in physics, further investigation is needed to clarify
these findings. In terms of strategy use, negative affect was unrelated to
metacognitive (study 1: r = –.03,p> .05; study 2: r = –.06,p> .05) and elabo-
rative strategy use (study 1: r = –.10,p> .05; study 2: r = –.10,p> .05). Thus,
while negative affect did not seem to change the types of strategies that stu-
dents used, it was associated with poorer performance on the post-test meas-
ure of understanding projectile motion and seemed to be associated with
lower levels of conceptual change.
We are unaware of other empirical research linking affect to conceptual
change in science understanding. Therefore, we focus on the interpretation of
the results from our laboratory. Overall, we found that positive affect was un-
related to conceptual change but was at least moderately associated with
adaptive strategy use and processing. Furthermore, negative affect was either
unrelated or hindered conceptual change but did not seem to alter students’
strategy use.
However, before we interpret our findings based on the theoretical mod-
els, it is important to keep a few limitations in mind. First, our measure of af-
fect was not clearly a measure of moods or emotions. Therefore, we need to
use some caution in evaluating the efficacy of the psychological models based
on mood to our findings. Second, our reliance on self-report measures to as-
sess affect does not parallel much of the experimental work used to develop
these models in which mood was manipulated. It is possible that students
may not accurately report on their own affect or that their reports are altered
by their performance on the post-test exam such that students who felt they
did poorly may report that they had higher levels of negative affect during the
task when in fact the high levels of negative affect emerged as they completed
- AFFECT AND COGNITIVE PROCESSING 69