Arousal Cues Arousal-Related Material in Memory

17 Pages·2004·1.5 MB·English

LANRUOJ FO LABREV GNINRAEL DNA LABREV ROIVAHEB 22, 633--649 (1983)
Arousal Cues Arousal-Related Material in Memory: Implications for
Understanding Effects of Mood on Memory
MARGARET S. CLARK, SANDRA MILBERG, AND JOHN Ross
nolleM-eigenraC ytisrevinU
Evidence is presented that )1( arousal acts as a cue for arousal related material in memory,
and )2( this effect may partially underlie mood cuing mood-related material from memory.
Specifically, two studies are presented showing that material people learn when in a high
arousal state and material they learn when in a normal arousal state is subsequently best
recalled when they are in a similar arousal state. Then a brief review of literature and a
third study are presented supporting the claim that arousal acting as a cue for arousal-
related material in memory may provide a partial explanation for why moods cue similarly
toned material from memory. The literature review reveals that positive mood inducers
elicit elevated autonomic arousal and the third study shows that elevated arousal in com-
bination with a positive mood enhances the effect the positive mood has on subsequent
judgments, whereas arousal by itself has no impact on judgments. Finally, it is suggested
that arousal acting as a cue for arousal-related material in memory may account for such
previously observed phenomena as state-dependent memory effects produced by alcohol.
and drugs, "misattribution" of arousal, "excitation transfer," and effects that arousal has
been shown to have on self-focus and humor.
Recently, the idea that moods may cue be part of how a mood is stored in memory
similarly toned material in memory, causing and that changes in arousal that reoccur with
that material to be more likely to come to subsequent moods may be part of what
mind and to influence judgments and be- primes affectively toned material stored
havior, has received much attention and earlier. Our case will be made using posi-
support (e.g., Bartlett & Santrock, 1979; tive moods as an example, but we believe
Bower, Monteiro, & Gilligan, 1978; Isen, our argument applies to negative moods as
1975; Isen, Shalker, Clark, & Karp, 1978; well. Before making our arguments, though,
Leight & Ellis, 1981; Teasdale & Fogarty, a brief review of some previous work on
1979). However, this research leaves some the effects of positive moods on memory
important questions unanswered. Among and judgments is necessary.
those is the following: Just what is it that is
Positive Moods Prime Positive Material
stored representing mood in memory and is
from Memory
subsequently cued when the mood reoc-
curs? There is now considerable evidence that
In this paper we will present a case that material stored in a person's memory linked
storage of information about changes in au- with a positive feeling will be more likely
tonomic arousal that accompany moods may to come to mind (e.g., Isen et al., 1978) and
will come to mind faster (Teasdale & Fo-
This research was supported by a grant from the
garty, 1979) when a person is in a positive
Ford Motor Company Research Fund to the first au-
feeling state than when he or she is not.
thor. We thank Gina Fleitman, Jeanne Jessup, and
Barbara Waddell for their help in conducting Study 3, Further, this evidence has been considered
and Gordon Bower, Sheldon Cohen, Eric Johnson, important by social psychologists because
Henry Ellis, Lynne Reder, and Michael Scheier for it provides an explanation for why positive
their helpful comments on this research. Correspon-
moods increase attraction (Gouaux, 1971),
dence regarding this paper should be addressed to
helping (Isen, 1970), expectations of future
Margaret S. Clark, Dept. of Psychology, Carnegie-
Mellon University, Pittsburgh, Pa, 15213. positive events (Masters & Furman, 1975),
336
38/1735-2200 $3.00
thgirypoC © 3891 yb cimedacA ,sserP .cnI
llA sthgir fo noitcudorper ni yna mrof ,devreser
634 CLARK, MILBERG, AND ROSS
and the perceived pleasantness of scenes the proposal that information about changes
(Forest, Clark, Isen, & Mills, 1979; Isen & in arousal that accompany positive moods
Shalker, 1982). Specifically, if being in a may be stored in memory, and that this in-
positive mood increases the likelihood of formation and material linked to it may be
thinking of positive aspects of other people, primed when later positive mood inducers
helping, future events, and everyday situ- once again produce similar states of arousal.
ations, these thoughts may increase helping, Before going on, it should also be noted
attraction, expectations of future success, that our argument need not imply that
and perceived pleasantness of situations arousal accompanying positive moods
(Isen, 1975; Clark & Waddell, 1983). should make negatively toned material in
Isen et al. (1978) and Clark and Isen (1982) memory (which may also be stored with
have discussed in some detail the question arousal) more likely to come to mind or vice
of through what processes increased acces- versa. While positive and negative mood in-
sibility of similarly toned material might ducers may both elicit changes in arousal,
come about. One possibility is that when they also elicit other reactions that are dis-
information is stored in memory, feelings tinct from one another. Positive and nega-
experienced at the same time are also tive mood inducers, for example, produce
stored, linked to that information. Given distinct facial expressions, physical pos-
this, the accessibility effect may be pro- tures, and evaluations of how fortunate one
duced by a process of "spreading activa- is or how "nice" the world is. Information
tion" (Collins & QuiUian, 1972; Collins & about such events may also be stored in
Loftus, 1975). This involves several as- memory providing links between past and
sumptions. First, current moods may acti- present affectively toned material. Thus,
vate nodes representing moods in memory when one is in a positive mood, positive
and activation may spread to other material facial expressions, positive postures, and
in memory linked to those nodes. Second, positive evaluations, as well as arousal may
the longer the mood nodes are stimulated contribute to priming positive material,
the longer activation spreads to connected while only arousal may prime negative ma-
material. Third and finally, this activation terial. The positive material should there-
may vary in quantity and nodes at intersec- fore receive the most priming and should
tions of a network will fire, bringing the ma- be most likely to come to mind (Clark,
terial to awareness when activation sur- 1982a).
passes a threshold. Consequently current At this point it is also important to note
moods may cause mood-related material in that although arousal is discussed
memory to be more likely to come to mind throughout this paper as if it were a general
than usual. state, we do not mean to imply that the
As noted above, however, this leaves un- arousal accompanying different emotional
answered the question of how moods are states is undifferentiated. There is consid-
represented in memory. We began to an- erable evidence to the contrary (e.g., Ax,
swer this question by asking what effects 1953; Funkenstein, 1956; Lacey, 1967;
the diverse events that induce positive Schwartz, Weinberger, & Singer, 1981). In-
moods have in common. We suspect it might stead our arguments are based on the idea
be information about these events that is that information stored in memory about
stored in memory, constitutes "storage of the arousal states that accompany different
a mood," and is reactivated later when a emotions and exercise may be quite similar.
new event induces a similar mood. Since a This information may be similar either be-
change in autonomic arousal seemed to be cause different patterns of arousal still have
one effect various positive mood inductions much in common and/or because people's
might have in common, this strategy led to phenomenal experience of different pat-
AROUSAL CUES AROUSAL-RELATED MATERIAL IN MEMORY 635
terns of arousal is very similar (Cannon, Testing the Idea that Arousal Can Serve
1927; Mandler, 1975). as a Retrieval Cue
To make a case that arousal is one ele-
Our first two studies examined the idea
ment linking present moods to similarly, af-
that information about arousal can be stored
fectively toned material in memory requires
in memory and that subsequent arousal can
a three-part presentation. First, evidence
serve as a retrieval cue for material presum-
that positive moods are accompanied by
ably linked to that information. We hypoth-
changes in arousal is needed. This evidence
esized that (1) material learned when one is
is presented in the remainder of the intro-
experiencing high arousal will be best re-
duction. Second, support must be found for
called when one is once again experiencing
the idea that information about a change in
high arousal (this is important because pos-
arousal can be stored in memory and can
itive moods are associated with enhanced
serve as a retrieval cue. This support is pro-
arousal and we want to argue that enhanced
vided by our first two studies. Finally, the
arousal when one reexperiences a positive
idea that arousal may contribute to the
mood may cue that earlier material), and (2)
priming of emotional material in memory
material learned when one is experiencing
must be supported and our third study does
a resting level of arousal will be best re-
so. We turn now to the first part of our
called when one is once again experiencing
argument.
that resting level of arousal.
Are Positive Moods Accompanied
by Arousal? STUDY 1
Method
Fortunately for the present argument,
several studies have established that "ev- Overview. Under the guise of a study on
eryday" positive experiences of just the sort learning under distracting conditions, sub-
psychologists have often used to induce jects learned a list of phrases while exper-
positive moods--i.e, success (Isen, 1970), iencing enhanced arousal and a second list
free gifts (e.g., Isen, Clark & Schwartz, while experiencing normal arousal. Later,
1976), funny movies (e.g., Averill, 1969; they were given a recall test for phrases on
Gouaux, 1971), and imagining happy or sad both lists when they were experiencing ei-
thoughts (e.g., Cialdini & Kenrick, 1976; ther enhanced arousal or normal arousal.
Masters & Furman, 1976)--are accompa- Subjects. Subjects were 37 college stu-
nied by elevations in autonomic arousal. dents enrolled in their first psychology class.
Specifically, using such diverse measures of They partially fulfilled a course require-
autonomic arousal as heart rate, systolic ment by participating.
blood pressure, adrenaline excretion in Stimulus materials. Two lists of phrases
urine and palmar conductance, Mueller and were prepared for use in this study. To con-
Donnerstein (1981) and Patkai (1971) have struct them, a large number of two-word
provided evidence that success and free phrases consisting of an adjective and a
gifts are accompanied by elevations in noun were generated (e.g., failed exam,
arousal, Averill (1969) and Levi (1972) have wide road). Then 8 judges rated how ab-
shown that humorous movies elicit ele- stract and how emotionally toned each
vated arousal, and Schwartz et al. (1981) phrase was. Finally, 20 phrases were se-
have shown that thinking happy thoughts lected for each of 2 lists, hereafter referred
produces elevations in arousal. Given this to as List A and List B, such that the av-
evidence, we may turn to the second part erage abstractness and emotionality ratings
of our argument. That is that arousal may of the phrases on each list did not differ.
serve as a retrieval cue for arousal-related Procedure. Upon each subject's arrival,
material in memory. the experimenter said she was testing the
636 CLARK, MILBERG, AND ROSS
effects of various types of distraction on the subjects had originally exercised or re-
learning. Subjects would try to learn phrases laxed (a time period that we had established
following exercise, a visual distraction, and to be sufficient for the exercise subjects'
a noise distraction, as well as following re- arousal to return to baseline). Next subjects
laxation (the supposed control condition). who had originally exercised performed the
Each subject expected to be run in all con- relaxation task (normal arousal) and sub-
ditions but, it was explained, the conditions jects who had originally relaxed performed
would not be run one right after the other. the exercise task (high arousal). Then each
Rather, there would be two sessions of two group learned the list of phrases to which
conditions each. First the relaxation and they had previously not been exposed in the
exercise conditions would be run. Then same manner they had learned the first list.
there would be a break during which sub- Thus arousal at learning was a within-sub-
jects would perform some analytical tasks, jects variable.
followed by the visual and noise distraction After the second learning session, all
conditions. subjects worked on some "spatial and an-
At this point half the subjects started the alytical tasks" for approximately 51 min-
study with the relaxation task (normal utes, after which the experimenter apolo-
arousal) and half with the exercise task (high getically, and with some embarrassment,
arousal). Those assigned to perform the re- said an error had been made. One of the
laxation task simply sat in a lounge chair previous conditions would have to be run
for 7 minutes and relaxed. Those assigned again. For a randomly chosen half of the
to the exercise condition stepped up and subjects the condition to be run again was
down on a cinder block for 7 minutes. the exercise condition. For the remaining
After this the first learning task took place half it was the relaxation condition. All sub-
under the same conditions (with the excep- jects repeated the required condition, and
tion of the subject's arousal level) regard- once again sat down expecting to learn a
less of the preceding task. The subject was new list of phrases. Instead the experi-
seated and listened to a list of 20 phrases menter unexpectedly asked the subject to
being read aloud by the experimenter. The recall as many phrases as possible from the
experimenter presented 1 phrase every 4 two lists previously learned. (Half the sub-
seconds. Then the experimenter repeated jects who had learned List A first and half
the list a second time, and asked the subject the subjects who had learned List B first
to recall the phrases aloud. Finally the ex- recalled the two lists in the order in which
perimenter repeated the list a third time, they had been learned. The remaining sub-
and as subjects recalled the phrases aloud jects recalled the lists in the reverse order.)
a second time the experimenter recorded Thus arousal at recall was a between-sub-
recall. They were given 3 minutes for recall. jects variable. Subjects were given 5 min-
Subjects' recall at this point will later be utes for recall and responses were re-
referred to as their original learning. (Half corded.
the subjects who had exercised immedi- Finally subjects were probed for suspi-
ately thereafter learned List A and half cion and debriefed. One subject in the High
learned List B. Half the subjects who had Arousal at Recall condition did not believe
relaxed immediately thereafter learned List an error caused the necessity for repeating
A and half learned List B.) the task. His data were excluded from all
At this point, a questionnaire asking about analyses. Four subjects who had taken a
aspects of the preceding task was given to course in which various aids to memory had
achieve some spacing between conditions. been taught reported using the "method of
After subjects filled out the questionnaire, loci" to store and to recall the phrases.
approximately 25 minutes had elapsed since Since it seemed likely that use of this tech-
AROUSAL CUES AROUSAL-RELATED MATERIAL IN MEMORY 736
nique would obscure the usefulness of - High Arousal at Recall
e---e Normal Arousal at Recoil
arousal as a retrieval cue, it was decided,
_j 95
prior to examination of the data, to exclude ~ c9
%-,. .I Study 2
these subjects' data from the analyses. Two # 58
"""-,~ ~ Study 1
were in the High Arousal at Recall condi- ~ 08
tion and two were in the Normal Arousal ~ 57
w
at Recall condition. :~ 7o "~'o Study t
w
a. 65 "~ Study 2
60
RESULTS
I I
Normal High
Manipulation Checks on Arousal
AROUSAL AT LEARNING
No direct manipulation check on arousal FIG. .1 Percentage of phrases recalled as a function
was taken during the course of the study. of arousal at time of learning and at time of recall.
However, six additional subjects were asked
to step up and down on the cinder block for F(1,31) = .39, NS, no effect of arousal at
7 minutes and to relax in a lounge chair for recall, F(1,31) = .78, NS, and a significant
7 minutes. (A 25-minute break was allowed interaction between arousal at learning and
between tasks.) Half performed the exer- arousal at recall, F(1,31) = 23.48, p < .0001.
cise task first, and half performed the re- A planned comparison indicated that when
laxation task first. Results supported the ef- subjects were highly aroused at time of re-
fectiveness of the arousal manipulation. call, recall of material heard when highly
Exercise produced significantly greater in- aroused was significantly greater than re-
creases in pulse from pre- to post-task (mean call of material heard when normally
change = + 45) than did relaxation (mean aroused, t(31) = 7.9, p < .01. A second
change in pulse = -0.15), t(4) = 8.1, p < planned comparison indicated that when
.)lOO. subjects were experiencing normal arousal
Dependent measures. The primary mea- at recall, recall of material heard when nor-
sure was the number of phrases recalled mally aroused was significantly greater
relative to the number originally learned, than recall of material heard when highly
expressed as a proportion. The means for aroused, t(31) = 15.6, p < .01.1
this measure clearly fell in the predicted Since the "emotional tone" of phrases
pattern. For subjects who were highly on both lists had been prerated, it was pos-
aroused at recall this proportion was greater sible to examine whether arousal at recall
for phrases originally heard when experi- enhanced recall of material prerated as
encing high arousal (85%) than for phrases "emotional," as some might expect, given
originally heard when experiencing normal our theorizing. To accomplish this, the 01
arousal (70%). For subjects who were ex- phrases in each list which received the
periencing normal arousal at recall this pro- highest emotionality ratings were classified
portion was greater for phrases originally as "emotional" phrases and the 01 phrases
heard when experiencing normal arousal receiving the lowest emotionality ratings
(91%) than for phrases originally heard when were classified as "nonemotional." A 2 x
they were experiencing enhanced arousal 2 mixed ANOVA was conducted with emo-
(69%). These results are depicted in Fig- tionality of phrases as one variable, arousal
ure .1 at recall as the second independent vari-
A 2 × 2 mixed ANOVA, with arousal at
learning (normal or high) as the within-sub- 1 It should be noted that when the four subjects who
used the "method of loci" as a memory aid are in-
jects variable and arousal at recall (normal
cluded in the analyses, neither the results of the overall
or high) as the between-subjects variable,
analysis nor of the two planned comparisons are
revealed no effect of arousal at learning, changed.
836 CLARK, MILBERG, AND ROSS
able, and percentage of phrases recalled as Before going on to show that arousal may
the dependent variable. There were no sig- enhance priming of positive material by
nificant effects. positive moods, we felt it necessary to rep-
licate the findings of the first study using
DISCUSSION manipulations of arousal at recall which
Our first study supports the hypothesis were distinct from those used at learning.
that level of arousal serves as an effective Although in the first study there were no
retrieval cue for material previously stored differences between conditions except in
with information about similar levels of arousal at the time subjects actually sat
arousal. Material learned when in a high or down to learn and to recall phrases, there
low arousal state was subsequently best re- is an alternative explanation for the results
called when subjects were again in a similar of that study. Cues associated with the
arousal state. In other words, level of au- arousal and relaxation tasks other than level
tonomic arousal appears to serve as a re- of arousal may have been associated with
trieval cue in much the same way as posi- the phrases learned following those tasks.
tive moods have been shown to serve as For example, posture and movement varied
effective retrieval cues in recent studies. In- between the two arousal tasks. Further,
deed, we suggest it is in part because arousal while the arousal tasks ended prior to ac-
serves as an effective retrieval cue that pos- tual learning and recall, Smith (1979) has
itive moods serve as effective retrieval cues. shown that recalling a prior context can
The plausibility of this argument is sup- serve as a retrieval cue. It is not a far step
ported by the present study in combination from that finding to assume our subjects may
with the research cited in the introduction have been thinking about the immediately
showing that events known to induce moods preceding situations while learning and re-
have also been shown to elicit arousal. calling word phrases and that aspects of
In this study arousal did not enhance those prior situations other than arousal
memory for the phrases prerated as being produced the state-dependent effects we
most emotional, relative to other phrases. observed. Therefore in our second study,
We do not find this to be surprising. The level of arousal at learning was manipulated
fact that our judges could, when required, in a manner distinct from the way it was
prerate some phrases as being more emo- manipulated at recall.
tional than others (e.g., "failed exam" was
YDUTS 2
rated as more emotional than "wide road"),
does not mean those phrases were actually Method
stored with enhanced arousal in our sub-
Subjects. Sixteen college students partic-
jects' memories. Most students do not, in
ipated. They were either paid or partially
fact, fail exams, for example. Conse-
fulfilled a departmental requirement by par-
quently, there is no reason why they should
ticipating.
have arousal information stored along with
Procedure. The word lists and procedure
the phrase "failed exam" although they may
were the same as in the first study with the
recognize that it would be an emotional ex-
exception of the manipulation of arousal at
perience if they did. Indeed, we suspect that
in the present study most of what was
viously reported by Isen et al. (1978, Study 2). In that
learned was not stored with any affective study positive moods were shown to be effective re-
information and that that accounts for the trieval cues for previously stored positive trait words.
failure of arousal to enhance memory for However, while the majority of subjects in psychology
experiments may not have encountered the specific
words prerated as being emotional. 2
experiences described by our phrases, they have, no
2 Upon reading this explanation, one reviewer ar- doubt, encountered many manifestations of positive
gued that it does not fit well with some findings pre- and neutral personality traits.
AROUSAL CUES AROUSAL=RELATED MATERIAL IN MEMORY 639
recall. That manipulation took place as fol- heard when experiencing high arousal (88%)
lows. The experimenter began the study by than phrases originally heard when experi-
telling the subject there were three rather encing normal arousal (60%). In addition,
than four distraction conditions--relaxa- for subjects experiencing normal arousal at
tion, exercise, and visual distraction. Fur- recall this proportion was greater for phrases
thermore, when it came time for the third orginally heard when experiencing normal
condition, the visual distraction condition, arousal (77%) than phrases originally heard
she did not tell the subject she had erred when experiencing high arousal (65%).
and one of the previous conditions would A 2 x 2 mixed analysis of variance was
have to be repeated. Instead, she stayed performed on these data with arousal at
with her original story that the third task learning (normal or high) as the within-sub-
would involve visual distraction. This task jects variable and arousal at recall (normal
turned out to be watching a film. Half of or high) as the between-subjects variable.
the subjects, those assigned to the High This analysis revealed a marginally signifi-
Arousal at Recall condition, watched a sex- cant effect of arousal at learning, F(1,14) =
ually explicit film depicting a heterosexual 4.31, p < .06, such that subjects tended to
couple engaging in various sexual acts. 3 The recall more phrases learned when aroused
remaining subjects, those assigned to the (76.5%) than phrases learned when not
Normal Arousal at Recall condition, aroused (68.5%), no effect of arousal at re-
watched a film about a chimpanzee learning call, F(1,14) = .39, NS, and a significant
sign language. Each film lasted approxi- interaction between arousal at learning and
mately 6 minutes. Here it should be noted arousal at recall, F(1,14) = 26.49, p < .0001.
that many studies provide evidence that A planned J comparison indicated that
films depicting sexual encounters increase when subjects were highly aroused at re-
both self-reports and physiological indices call, recall of material heard when aroused
of autonomic arousal (e.g., Baron & Bell, was significantly greater than recall of ma-
1977; Byrne, Fisher, Lamberth, & Mitchell, terial heard when not aroused, t(14) =
1974; Zillmann, Bryant, Comisky, & Me- 26.27, p < .01. A second planned compar-
doff, 1981; Zillmann, Mody, and Cantor, ison indicated that when subjects were not
1974). After viewing the erotic or nonerotic highly aroused at recall, recall of material
film, subjects in the second study were given previously heard when not aroused was sig-
a surprise free recall task for the two lists nificantly greater than recall of material
of phrases learned earlier. heard when aroused, t(14) = 4.72, p < .01.
As in the first study, a second 2 x 2 mixed
RESULTS analysis of variance was conducted with
As in the first study, the primary depen- emotionality of phrases and arousal at re-
dent measure in the second study was call as the independent variables, and
number of word phrases recalled relative to number of phrases of each type recalled rel-
the number originally learned, expressed as ative to the number learned as the depen-
a proportion. Once again, as can be seen in dent variable. As in Study ,1 this analysis
Figure ,1 the means for this measure clearly revealed no significant effects.
fell in the predicted pattern. For subjects
who were highly aroused at recall this pro- DISCUSSION
portion was greater for phrases originally The results of the second study closely
parallel the results of the first study and
3 All subjects were called and told that the study provide additional support for our hypoth-
might involve viewing some sexually explicit material
esis that level of arousal serves as an effec-
prior to agreeing to participate. They also were in-
tive retrieval cue for material stored in
formed they could leave the study at any time and still
receive full credit or pay. memory. Again, material learned when ex-
640 CLARK, MILBERG, AND ROSS
periencing a given arousal state was sub- material coming to mind than would oth-
sequently best recalled when subjects were erwise have been the case.
in a similar arousal state. Furthermore, since As noted above, enhanced arousal ac-
the two sets of arousal manipulations in this companying a positive mood should not
study had little in common except for the make material linked with the opposite (in
ability to elicit arousal, the results of this this case negative) feeling tone more likely
study seem to rule out the possibility present to come to mind. Since other aspects of the
in the first study that something about the positive mood (e.g., facial expressions,
arousal manipulations other than arousal posture, evaluations) are also priming pos-
produced the effects observed. itive material, the positive material will re-
We tend to believe that the marginally ceive the most priming and be most likely
significant effect for phrases learned when to come to mind. Furthermore, once posi-
highly aroused to be better recalled than tive thoughts do come tO mind they (in
phrases learned when experiencing normal combination with continuing arousal, facial
arousal that was observed in this second expressions, and global evaluations),
study was due to chance. This effect did not should prime additional positive thoughts,
approach significance in the first study. In- prolonging the positive mood.
deed, in that study there was a slight ten- On the basis of the above reasoning, the
dency in the opposite direction. following specific predictions were tested
in Study 3: (1) as Isen has suggested and
Testing the Idea that Arousal is One Link
found risen, 1975; Isen et al., 1978), a pos-
between Present Mood States and
itive mood should cause people's judg-
Similarly Toned Material in Memory
ments of something about which they have
Our next step in this program of research stored varied, affectively toned memories
was to conduct a study to support the idea (in this case their university) to become
that arousal priming similarly toned mate- more positive, and (2) this effect should be
rial from memory may account in part for enhanced by increased arousal which ac-
mood priming similarly toned material from companies, and presumably prolongs and/
memory. Specifically the third study was or intensifies, a positive feeling state. No
designed to demonstrate that additional prediction was made about the effect of
arousal at the time a person is experiencing arousal in the absence of positive feedback.
a positive mood will enhance priming of Without knowing what mood subjects are
positive material from memory. If arousal in when they begin the study, it is impos-
acting as a retrieval cue for arousal-related sible to make such a prediction. For sub-
material in memory underlies the effect of jects in positive moods to begin with, pos-
mood on memory, then additional arousal itive associations to the university should
at the time one experiences a positive mood already be "primed." Additional arousal
induction should enhance the effect of that may provide another cue to the same ma-
mood induction on judgments for one or terial, causing it to "come to mind" and
both of the following reasons: (1) arousal making judgments more positive. However,
persists over time and thus may prolong the subjects who are in negative moods to begin
priming of similarly-toned material from with should already have negative associ-
memory, resulting in more of this material ations to the university "primed," and fur-
coming to mind, and (2) the more extremely ther arousal may provide another cue to that
positive material is in memory, the more material, causing it to "come to mind" and
intense the arousal tied to it may be (Clark, making judgments more negative. Finally,
1982a). Thus, enhanced arousal along with for subjects in no particular feeling state,
a positive mood induction may result in a the position taken thus far makes no pre-
different, more extremely positive set of dictions regarding the effects of further
AROUSAL CUES AROUSAL-RELATED MATERIAL IN MEMORY 146
arousal. Although there is some evidence were exposed to a manipulation modeled
that extremely intense, unexpected arousal after one used by Zillmann and Bryant
may produce negative feelings (Marshall & (1974). They were told their distracting task
Zimbardo, 1979; Maslach, 1979), there is no was to slide disks with holes onto a string.
evidence that less intense arousal states The experimenter pointed to a large card-
have the same effects. Less intense eleva- board carton with two holes for the sub-
tions in arousal, such as those involved in ject's arms. The disks and string were in-
the studies in this report, may in the ab- side. The subject was to put his or her hands
sence of a mood inducer cause both posi- into the box and string the disks together.
tively and negatively toned material to be The High Arousal subjects were told that
more likely than usual to come to mind. their task involved stepping up and down
These thoughts may then cancel out each on a cinder block.
other's effects on judgments. Thus, as Next, all subjects were told that since
stated above, no prediction about the effect some conditions involved physical exer-
of arousal in the absence of a mood manip- cise, and (in the Normal Arousal condi-
ulation was made. tions) that because everyone had to be run
through the same procedure, all subjects
YDUTS 3
were to sign a consent form. Also, each
Method
subject's blood pressure and pulse had to
Overview. Under the guise of a study on be taken before and after the distracting
distraction and memory, subjects listened task. Subjects then signed the form and had
to a story while either stepping up and down their blood pressure and pulse taken. At this
on a block (High Arousal) or stringing card- point all subjects began their assigned task
board disks together (Normal Arousal). Af- while listening to a taped story about a stu-
terwards, all subjects took a memory test. dent's first day at law school. The tasks and
Half the subjects from each arousal condi- tapes lasted about 7 minutes.
tion were then told they had done well (Pos- Upon completion, all subjects immedi-
itive Feedback conditions); half were told ately wrote down as many things from the
that their test had not yet been scored (No story as they could remember. Ninety sec-
Feedback conditions). Finally, all subjects onds later, although the subjects were still
filled out a survey about their university. writing, the experimenter stopped them. She
Subjects. Forty-four college students said the study was over, that the only thing
participated. They partially fulfilled a class left to do was to take the final measure of
requirement by participating or were paid blood pressure and pulse to make sure they
$3.00. Each was randomly assigned to one were back to normal before leaving. Before
of the following conditions: (1) High that, though, she said she wanted to take
Arousal-Positive Feedback, (2) High their memory results into her supervisor's
Arousal-No Feedback, )3( Normal Arousal- office "because she likes to have a look at
Positive Feedback, or (4) Normal Arousal- them after each person is finished."
No Feedback. The first experimenter took the results
Procedure. Upon arrival the experi- across the hall to the second experimen-
menter greeted the subject saying that she ter's office. Upon returning, the first ex-
was a research assistant and reminded the perimenter took the subject's blood pres-
subject that the study concerned memory sure and pulse for a second time, assured
and distraction as had been indicated on the the subject he or she was "back to normal,"
sign up sheet. Each subject would perform when in fact subjects in the High Arousal
a distracting task while listening to a story condition were not, and sent the subject to
on tape. Afterwards they would be given a the second experimenter's office to be paid
memory test. The Normal Arousal subjects or to pick up a credit slip.
246 CLARK, MILBERG, AND ROSS
The second experimenter, who was un- The first experimenter, who was unaware
aware of the subject's arousal condition, in- of the feedback conditions, repeated the
troduced herself and asked for the subject's second experimenter's explanation of why
name. She picked up the subject's memory she was conducting the survey, and handed
results and in the No Feedback conditions subjects a survey on campus attitudes. She
simply said, "Barb gave your memory re- told them to give their first reactions to each
suits to me, but I really haven't had a chance question.
to look at them, so I can't tell you how you The survey asked subjects to rate (a) the
did." In the Positive Feedback conditions quality of teaching in general at the univer-
she had circled individual facts from what sity, (b) the quality of teaching in their major,
the subject had recalled and had put a (c) how worthwile required courses outside
"memory score" at the bottom of the sheet. their major had been, (d) how concerned
In these conditions she picked up the sheet, with students faculty in their department
showed it to the subject and said, "Barb were, (e) the quality of the campus li-
gave your memory results to me, and I've braries, (f) the friendliness of students at
had a chance to look them over. You really the university, and (g) how pleasant the
did very well! Your score is way above the campus was. The ratings were made on
norm. Most people have a very difficult time scales ranging from -4 (indicating an ex-
just remembering a couple of facts while tremely negative attitude) to + 4 (indicating
they are distracted, and you did much better an extremely positive attitude).
than that. You ought to be very pleased be- Finally, the subjects reported back to the
cause the ability to remember things when second experimenter, who started to fill out
you're distracted is related to general intel- their credit form or to pay them. Before she
ligence."4 finished she casually said, "By the way,
The second experimenter asked if the there's more to my study than I explained
subject was doing the study for credit or so far. I'm curious as to whether you have
pay and said she would give it to the subject any idea of what it is?" Four subjects ex-
shortly. First, though, she wanted the sub- pressed suspicion. Two thought the survey
ject to do one more thing. Since the study was a part of the actual study, one did not
took such a short time, she was asking ev- believe the positive feedback, and one ex-
eryone to help out her research assistant. pressed suspicion about both the survey and
The research assistant had been sick the whether arousal was being studied. One of
previous semester, had an "incomplete" in the first two subjects was in the Normal
a research methodology class, and was now Arousal-No Feedback condition, the other
conducting a survey to finish the course. was in the Normal Arousal-Positive Feed-
The survey was short, and the second ex- back condition. The third subject was in the
perimenter asked the subject to fill it out. High Arousal-Positive Feedback condition,
No subject objected, and all were directed and the fourth was in the High Arousal-No
back to the first experimenter. Feedback condition. Data from these sub-
jects were not included in any of the anal-
4 A very similar manipulation was used by Clark and yses. Finally, all subjects were thoroughly
Waddell (1983). In that study additional subjects were
debriefed, given pay or credit, and dis-
run solely to check on the effectiveness of the manip-
missed.
ulation. Those subjects received success or no feed-
back from one experimenter, then entered another ex-
RESULTS
perimenter's room who was blind to conditions. That
experimenter first judged their moods (unobtrusively) Manipulation Checks on Arousal
then, under the guise of a pretest for a separate study,
The change in subjects' blood pressure
asked them to rate their own moods. Both checks pro-
and pulse from just before the distraction
vided evidence for the effectiveness of our manipula-
tion. task to 2/12 minutes after the task was the