Affective Cues and Processing Strategy: Color-Coded Examination Forms Influence Performance

Robert C. Sinclair
Alexander S. Soldat
University of Alberta

Melvin M. Mark
Pennsylvania State University

We argue that external cues provide affective information that
influences processing strategy and, therefore, examination perform-
ance. Participants completed 2 supposedly different forms of a
midterm examination; in fact, randomly assigned participants com-
pleted identical midterm examinations printed on either red or blue
paper. Blue paper led to superior performance, especially for
difficult questions. We note a method for appropriately adjusting
scores to control for form effects while maintaining percentile
rankings within and across form. We discuss implications for test
administration, examination form effects, and adjustment of scores.

Considerable research has demonstrated that affective
states influence processing strategy and judgmental accuracy
(e.g., Sinclair, 1988; Sinclair & Mark, 1992, 1995 ). Generally,
happy moods lead to nonsystematic, less detail-oriented, and
more heuristic processing; whereas sad moods lead to more
systematic, more detail-oriented, and less heuristic processing.
Consistent with this position, Sinclair ( 1988 ) and Sinclair and Mark ( 1995 ) demonstrated that happy people were less accu-
rate on performance appraisal and statistical judgments than
were sad people. Cognitive response analyses demonstrated
that these effects occurred because of the differential process-
ing strategies associated with happy versus sad moods.

One explanation for these differential processing strategy
effects is the cognitive tuning extension of the affect-as-infor-
mation hypothesis ( Clore, Schwarz, & Conway, 1994; Sinclair, Mark, & Clore, 1994). According to this hypothesis,
affective states provide informational signals about situations
and the conduct of tasks. Happy moods signal that situations
are benign and that no extra cognitive resources are necessary
for processing and judgment, whereas sad moods signal that
situations are threatening or important, and that detailed
processing is necessary (for empirical evidence supporting this
position, see Sinclair et al., 1994).

Recently, Soldat, Sinclair, and Mark ( 1997 ) extended the
cognitive tuning branch of the affect-as-information hypothe-
sis to environmental affective cues (e.g., color) that do not
directly influence perceivers' moods. In particular, they ar-
gued that colors serve as affective cues, thereby signaling the
degree of processing required in a particular situation. For
example, although red conveyed positive affect and blue
conveyed negative affect, color had no actual impact on
mood. This pattern of effects emerged in a study in which
participants rated their current affective state or the affect
conveyed by the paper. Paper color did not influence reported
affective state (cf. Jacobs & Blandino, 1992; Jacobs & Suess, 1975 ) but did influence ratings of the affect conveyed by the
paper. An additional study replicated this effect and demon-
strated that white fell between red and blue on the measure
of conveyed affect (again, no effects emerged on affective
states; see Soldat et al., 1997). In Soldat et al., another group
of participants completed both simple and complex Graduate
Record Exam-like questions, involving analytic problem solv-
ing, on either red or blue paper and evaluated their current
affective states and the difficulty of reading the materials on
Likert scales. Blue paper led to greater accuracy, especially for
complex questions. However, paper color did not influence
mood or reading difficulty. Furthermore, analyses of covari-
ance controlling for arousal (as reported on the arousal com-
ponent of the measure of current affective state)
demonstrated no effects of arousal.

The Soldat et al. ( 1997 ) findings, replicated in a second
study that included a white-paper control group that felt
between the performance levels of blue and red, appear to
have implications for testing situations. For example, instruc-
tors may create different forms of multiple-choice examina-
tions to prevent cheating. Generally, instructors present
questions in different orders on the different forms. Further-
more, instructors may print different forms on differently
colored paper to make it clear that there are different forms,
and to provide a straightforward visual check that copies of
the same form are not adjacent to one another. In this
situation, students whose examination forms are on colors
that convey more positive affect may process information less
systematically, leading to differential performance on the
examination. Of course, in these naturally occurring situ-
ations, exam form is confounded with paper color, so whether
form effects are due to color or question order is unclear.
Soldat et al. conducted their research demonstrating differen-
tial accuracy as a function of paper color in laboratory settings

-130-