This article was downloaded by: [New York University]

On: 24 March 2015, At: 16:10

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Advertising
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/ujoa20

Affect Intensity and the Consumer's Attitude

toward High Impact Emotional Advertising Appeals
a a
David J. Moore & William D. Harris
a
Department Sport Management and Communication and The School of Business ,
University of Michigan , Ann Arbor , MI , USA
b
Department of Marketing, College of Business , Quinnipiac College , Hamden , CT ,
USA
Published online: 31 May 2013.

To cite this article: David J. Moore & William D. Harris (1996) Affect Intensity and the Consumer's
Attitude toward High Impact Emotional Advertising Appeals, Journal of Advertising, 25:2, 37-50, DOI:
10.1080/00913367.1996.10673498

To link to this article: http://dx.doi.org/10.1080/00913367.1996.10673498

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy, completeness, or suitability
for any purpose of the Content. Any opinions and views expressed in this publication are the opinions
and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of
the Content should not be relied upon and should be independently verified with primary sources of
information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands,
costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or
systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in
any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Affect Intensity and the Consumer's Attitude toward
High Impact Emotional Advertising Appeals
David J. Moore and William D. Harris

Subjects scoring high on the Affect Intensity Measurement (AIM) scale responded with greater emotional
intensity than low AI subjects to both positive and negative emotional appeals. These high AI individuals also
expressed more positive attitudes and higher levels of enjoyment of the positive emotional appeal. However, in
response to the negative emotional appeal, high AI and low AI subjects did not differ in ad enjoyment level or
attitude toward the ad. Emotional responses mediated the effects of affect intensity on attitude toward the ad
only when subjects were exposed to the positive emotional appeal. Theoretical and managerial implications of
the effect of affect intensity on the recipient's attitude toward high impact emotional advertising appeals are
discussed.
Downloaded by [New York University] at 16:10 24 March 2015

David J. Moore is Associate
Professor of Marketing, Department
Sport Management and Communica-
tion and The School of Business, The
University of Michigan, Ann Arbor,
MI.
William D. Harris is Associate
Professor of Marketing, Department
of Marketing, College of Business,
Quinnipiac College, Hamden, CT.
The authors wish to thank Meryl
Gardner, Curt Haugtvedt, the editor,
Les Carlson, and three anonymous
reviewers for their helpful comments
on earlier drafts of this manuscript.
Journal of Advertising,
Volume XXV; Number 2
Summer 1996
One of the most critical concerns advertisers express about television
advertising is the potential decline of the persuasive impact of TV spots
because of increasing clutter in the media environment. To make advertis-
ing appeals more distinctive and hence, perhaps more persuasive, advertis-
ers frequently use dramatic emotional ads-messages designed to "shock
the emotions and make the brain itch'" (Moore 1989). Such advertising
executions often feature high impact, sensually evocative appeals that stimu-
late strong positive emotions (e.g., the Hallmark greeting card ads drama-
tizing precious moments and nostalgic memories in life). However, advertis-
ers also employ graphic and sensational negative emotional advertising
messages such as those used by insurance companies, healthcare institu-
tions, drug and alcohol abuse clinics, food manufacturers exploiting con-
sumers' fears of cancer and cardiac diseases (Alsop 1988), and public service
agencies making appeals to prevent child abuse or to promote the use of
contraceptive devices (LaTour and Pitts 1989). For example, one recent TV
advertisement produced by the American Diabetes Association featured a
male victim of diabetes removing his right boot revealing a prosthesis in-
stead of a foot. The voice-over appeal warned, "Give diabetes an inch and it
will take a foot. '" The appeal generated such mixed emotional reaction from
the media audience that most TV stations refused to broadcast the ad.
The use of high impact negative emotional appeals is the subject of much
debate in the advertising community. One advocate of the "blunter-is-bet-
ter" approach, the Advertising Council, claims that its own research shows
the public to be more willing than ever before to confront serious issues that
affect life and health (King 1989). In contrast, critics have warned that the
use of fear-inducing "shock" ads may produce excessive levels of anxiety
that may pose a genuine threat to the psychological well-being of the mes-
sage recipient (Henthorne, LaTour, and Nataraajan 1993; Hyman and Tansey
1990). Previous advertising research on fear appeals has suggested that
when the intensity of the message exceeds normal thresholds of severity,
the message recipient typically develops an avoidance response that limits
the persuasive impact of the appeal (Ray and Wilkie 1970). The avoidance
response may be linked to a negative attitude toward the ad itself.
Attitude toward the ad can be influenced by the emotions (both positive
 38
and negative) elicited by exposure to an advertising
appeal (Bagozzi and Moore 1994; Stayman and Aaker
1988). However, individuals may differ significantly
in the level of emotional intensity with which they
respond to an advertising stimulus, and the intensity
level may have a parallel influence on attitude forma-
tion (Moore, Harris, and Chen 1995). In identifying
the structural antecedents of attitude toward the ad
CA.), Mackenzie and Lutz (1989) hypothesized that
individual differences among consumers should be
reflected in the message recipients' moods which in
turn should influence their A.d , Interestingly, what is
still not clear from prior marketing research is whether
the presumed individual differences in emotional re-
activity do influence the way consumers respond to
high impact emotional appeals, particularly the
Downloaded by [New York University] at 16:10 24 March 2015
consumer's attitude toward the ad.
Research has shown that individuals do differ widely
in the intensity of their emotional response to affect-
laden stimuli (Larsen 1984; Larsen and Diener 1987).
Hence, because of differences in the magnitude of
their affective response to emotionally provocative
appeals, some individuals may experience intense
emotional discomfort when exposed to negative emo-
tional appeals and others may be only mildly affected.
We conducted a study to examine the extent to which
individual differences in emotional reactivity influ-
ence consumers' response to emotionally provocative
advertising appeals and to observe how those emo-
tional reactions influence attitude toward the ad.
Literature Review
Affect Intensity
The affect intensity measurement (AIM) scale mea-
sures the strength with which individuals experience
their emotions in response to emotion-eliciting stimuli
(Larsen 1984). Larsen and Diener (1987) confirmed
that when people are exposed to equal levels of affect-
producing stimuli, some individuals consistently re-
spond with high levels of emotional intensity while
others respond with only moderate levels. Moreover,
the emotional reactivity seems to generalize across
both positive and negative emotional domains. For
example, Larsen, Diener, and Emmons (1986) found
that subjects classified as high on the AIM scale,
when faced with a positive emotion-eliciting event,
reported stronger positive affect than subjects classi-
fied as low on the AIM scale. Correspondingly, when
faced with a negative emotion-eliciting event, the same
high AI individuals reported stronger negative emo-
Journal of Advertising
tional responses than their low AI counterparts. When
subjects were exposed to a neutral or nonemotional
stimulus or event, the individual differences tended
to disappear. Similarly, Diener et al, (1985), using
longitudinal studies, tracked the daily moods of indi-
viduals over a period of eight weeks. They found that
the intensity of the respondents' positive emotions
correlated approximately .70 with the intensity of
their negative emotions. Hence, some people may be
inclined to manifest more intensity in their emotions
regardless of the valence of the emotions (Larsen,
Diener, and Emmons 1986).
Validity and Reliability of the AIM Scale. Assess-
ments of the validity of the AIM scale have been
conducted in a variety of studies. In one study, Larsen
and Diener (1985) found that the correlation between
respondents' self-reported AIM scores and the par-
ents' reports of those subjects' emotional response
intensity correlated .50. Peer reports of affect inten-
sity have been shown to be correlated at .41 with self-
reported AIM scores (Larsen and Diener 1987). Three
separate studies have compared AIM with an aver-
age daily assessment of emotional response intensity
(Larsen and Diener 1987). AIM correlated .61 with
emotional intensity in the first study, .52 in the sec-
ond study, and.49 in the third study. Assessments of
the reliability of the AIM scale have been quite im-
pressive. Larsen (1984) found that the test-retest re-
liability for AIM with one-month, two-month and
three-month time intervals was .80, .81 and .81, re-
spectively. In another study the test-retest reliability
of AIM was reported to be .75 over a two-year time
interval (Larsen and Diener 1987) .
Actiuity Correlates of Affect Intensity. Larsen,
Diener, and Emmons (1986) found affect intensity to
be related to four dimensions of temperament: socia-
bility, activity level, arousability, and emotionality.
They found that high AI individuals were more ac-
tive, sociable, physically arousable and emotionally
reactive than their low AI counterparts. Evidence
from daily activity reports (Larsen 1984) suggests
that individuals scoring high on the affect intensity
measure tend to seek emotional stimulation from day-
to-day activities that are highly emotionally provoca-
tive (e.g., sexual activity, attending a sports event,
going to a party, and singing/dancing). In contrast,
low affect intensity individuals tend to engage in daily
activities that are less emotionally provocative, such
as leisure reading, shopping, letter writing, and bi-
cycle riding. In spite of those findings, Larsen, Diener,
and Emmons (1986) reported that affect intensity
showed a zero correlation with Zuckerman's (1979)
 Summer 1996
sensation seeking scale. The reason is that, unlike
the AIM scale that measures affective reaction to
normally occurring day-to-day activities, the sensa-
tion-seeking construct incorporates behaviors associ-
ated with risky and thrilling activities that are un-
usual and infrequent and serve to provide a change
from the daily routine of life (Larsen and Diener 1987).
High AI individuals tend to maintain strong and con-
sistent emotional responses by engaging in day-to-
day activities that are most likely to stimulate emo-
tions. In other words, "individuals high on the affect
intensity dimension do not seek out-of-the-ordinary
experiences as much as they seek out an ordinary
daily life that is more emotionally stimulating" (Larsen
Downloaded by [New York University] at 16:10 24 March 2015
and Diener 1987, p. 24).
The Regulation of Sensory Stimulation. Research
has confirmed that people tend to differ in their
baseline level of arousal (Eysenck 1967). That is, some
individuals may be quite underaroused at baseline
and others may be very overaroused at baseline. Given
the assumption that people are normally driven to-
ward some optimal level of arousal (Geen 1984), the
individual who is underaroused at baseline is expected
to be motivated to seek stronger forms of stimulation
to compensate for the presumed low level of arousal,
whereas the already overaroused individual is ex-
pected to seek to minimize strong forms of stimula-
tion (Larsen and Diener 1987, p. 28). Affect intensity
theorists suggest that high AI individuals who may
be under-aroused at baseline may modulate the in-
tensity of the impact of emotional stimuli and there-
fore tend to manifest stronger or more intense emo-
tional reactions. In contrast, low AI individuals who
may be already over-aroused at baseline may be much
less emotionally reactive to equivalent levels of emo-
tion-provoking stimulation (Larsen and Diener 1987).
That rationale suggests that for high AI individuals,
positive emotional stimulation is likely be a very
strong form of enjoyment, but negative emotional
stimulation may be unbearably unpleasant.
Research also suggests that dimensions of tempera-
ment serve as mechanisms for the regulation of emo-
tional arousal (Strelau 1982). Larsen and Diener
(1987) found that high levels of affect intensity were
associated with individuals who were elevated on the
four fundamental dimensions of temperament (socia-
bility, activity, arousability, and emotionality). In
other words, highly emotionally reactive individuals
are likely to show preferences for frequent social in-
teraction, an activity-driven lifestyle, and physically
arousing sensory expefiences (Larsen and Diener
1987).
39
Affect Intensity, Emotional Response, and Attitude
Formation. Although previous research has estab-
lished that individuals' affect intensity levels have a
direct impact on their emotional responses (Larsen,
Diener, and Emmons 1986), very little research is
available on the relationship between affect intensity
and attitude. Advertising studies featuring self-moni-
toring (Snyder and DeBono 1985) and need for cogni-
tion (Cacioppo and Petty 1982) have provided encour-
aging support for a linkage between individual differ-
ences in personality traits and attitude formation.
For example, Haugtvedt, Petty, and Cacioppo (1992)
have shown that need for cognition can influence atti-
tude formation through the process of differentiating
strong from weak arguments underlying the mes-
sage. In a corresponding manner, Moore, Harris, and
Chen (1995) found that affect intensity influenced
attitude formation through the mediation of emotional
responses. Their study featured a public service ad-
vertisement about child abuse that elicited very strong
negative as well as empathic emotions. High AI indi-
viduals reported more favorable attitudes toward the
organization sponsoring the ad and showed more posi-
tive attitudes toward helping to support the organi-
zation. However, because high AI individuals suppos-
edly experience their emotions with great intensity,
such individuals may tend to avoid any stimuli (or
advertising appeals) that evoke uncomfortable nega-
tive emotions. If so, high AI individuals may be more
likely than their low intensity counterparts, to report
unfavorable attitudes toward a negative emotional
advertising appeal. To provide a theoretical founda-
tion for that proposition, we examined some of the
literature on approach/avoidance responses to emo-
tional stimuli.
Approach/Avoidance Responses to
Emotional Stimuli
Research assessing approach/avoidance behavior
toward emotional stimuli that differ in hedonic qual-
ity indicates that high AI individuals do have a dis-
taste for negative stimuli and seek positive sensory
stimulation (Gallagher, Diener, and Larsen 1989).
Research associated with optimal stimulation theory
suggests that individuals, in general, are likely to
select positive situations and avoid negative ones
(Gallagher, Diener, and Larsen 1989; Petrie 1967).
Moreover, some individuals supposedly experience
their emotions with greater intensity than others,
and it may be more difficult for such high AI indi-
viduals to tolerate the experience of intense negative
 40
emotional stimulation. Hence, they tend to dislike
exposure to such stimulation.
Research has shown that ad-induced emotions have
a direct impact on attitude formation (Batra and Ray
1986; Edell and Burke 1987; Holbrook and Batra
1987). In our study, we conceptualized attitude to-
ward the ad as the individual's favorable or unfavor-
able evaluation of the advertising appeal to which he
or she has been exposed (MacKenzie and Lutz 1989).
Because high AI individuals tend to express dislike
or avoidance of intense negative emotional stimuli
and a corresponding preference for positive emotional
stimuli (Gallagher, Diener, and Larsen 1989), indi-
viduals with high scores on the AIM scale should be
expected to have significantly stronger negative atti-
tudes toward the ad when exposed to a negative emo-
Downloaded by [New York University] at 16:10 24 March 2015
tional appeal and stronger positive attitudes toward
the ad when exposed to a positive emotional appeal.
Study Objectives and Hypotheses
Our primary objective was to examine the relation-
ship between affect intensity and attitude toward the
ad in a setting where subjects were exposed to both
negative and positive emotional advertising appeals.
That approach would facilitate the measurement of
both Aad and the extent to which subjects enjoyed the
experience of viewing the ad. We were also interested
in confirming the flndings of previous studies (Larsen,
Diener, and Cropanzano 1987; Moore, Harris, and
Chen 1995) on the influence of affect intensity on
emotional responses.
HI: High AI subjects manifest stronger emo-
tional reactions than their low AI coun-
terparts, when exposed to a positive as
well as a negative emotional appeal, but
the two groups do not differ in the inten-
sity of their emotional responses to a
nonemotional appeal.
H2: High AI subjects report more positive
(negative) attitudes toward an ad than
their low AI counterparts, when exposed
to a positive (negative) emotional appeal,
but the two groups do not differ signifi-
cantly in attitude scores in response to a
nonemotional appeal.
H3: High AI subjects express greater (less)
enjoyment of an ad than their low AI coun-
terparts, when exposed to a positive (nega-
tive) emotional appeal, but the two groups
do not differ significantly in ad enjoyment
levels in response to a nonemotional ap-
peal.
Journal of Advertising
Another objective our study was to examine the
interrelationship of affect intensity, emotional re-
sponse, and attitude formation. Because affect inten-
sity has a direct influence on the strength of emo-
tional stimulation (Larsen and Diener 1987) and be-
cause emotional responses have been shown to influ-
ence attitude toward the ad (Holbrook and Batra
1987), we expected the effect of affect intensity on Aad
to be mediated by emotional responses only when
subjects were exposed to emotional ads.
H4: The effect of affect intensity on Aad is
mediated by emotional responses when
subjects are exposed to emotional ads, but
not when they are exposed to a
nonemotional ad.
Method
Subjects and Experiment Design
Subjects were 131 marketing undergraduates (52010
female) who agreed to participate to obtain course
credit at a large midwestern university. Ages ranged
from 19 to 28 years. The experiment featured a 2(AI:
high vs. low) x 3(ad type) within-subjects repeated
measures factorial design whereby all subjects were
exposed to a negative emotional ad, a positive emo-
tional ad, and a nonemotional ad.
Stimuli
Subjects were exposed to three commercials care-
fully selected by a panel of judges consisting of two
marketing professors and two MBA students. The ad
selection process was extensive, requiring a review of
almost 100 taped TV commercials. The first type of
ad selected was classified by the panel of judges as
evoking primarily positive emotions such as joy, hap-
piness, and warmth. The ad was for special occasion
greeting cards by Hallmark Inc. and featured a joy-
ous but moving wedding scene. The second type of ad
had a strong negative emotional appeal described as
touching, moving, sad, and sympathetic. The ad was
sponsored by a internationally known disaster relief
organization (Red Cross) seeking financial contribu-
tions for needy victims of recent disasters around the
world. The third type of ad was nonemotional in con-
tent, featuring a presentation of factual information
about Colgate toothpaste. The ad was selected to serve
as a contrasting ad exposure condition. In other words,
high and low AI subjects were not expected to differ
significantly in emotional intensity or attitude for-
mation in response to the nonemotional ad. The three
 Summer 1996
target ads along with several other ads were embed-
ded within a 30 minute segment of a classic TV movie.
Procedure
Six weeks prior to the experiment, subjects were
administered the 40-item 6-point AIM scale (Larsen
1984). They were told that the purpose ofthe exercise
was to gain insight about their personal values and
TV viewing habits. Using upper and lower quartiles,
we classified 33 subjects as high on the AIM scale and
37 as low on the scale. Those 70 selected subjects
were later contacted and invited to make appoint-
ments to participate in an unrelated experiment.
In the experiment, subjects were assigned to small
Downloaded by [New York University] at 16:10 24 March 2015
groups ranging in size from 5 to 12 per session. Sub-
jects were told that the experimenter was interested
in their attitudes and opinions about TV movies. To
support the cover story, the questionnaire began with
several questions about TV viewing habits. Having
completed those questions, subjects were invited to
relax and view the movie as though they were seated
at home. At the end of the show, subjects were in-
formed that they would be given an opportunity to
view some of the ads that had appeared during the
showing of the movie. Those ads were in fact the
target ads. The order of presentation of the ads was
counterbalanced and rotated to minimize order ef-
fects and carryover biases. For example, some sub-
jects saw the positive ad first and others saw either
the negative ad or the nonemotional ad first. The
entire sequence of exposure to the ads was completely
randomized. After viewing each ad, subjects recorded
the extent to which they felt each of several emotions,
their enjoyment of the ad, and their attitude toward
the ad.
Measures
Affect Intensity and TV Program Preferences. As
part of the cover story procedure, subjects were asked
to indicate the extent to which they were likely to
watch each of several categories of movies. The cat-
egories were selected to represent emotionally stimu-
lating types of movies such as romance, mystery, com-
edy, and war, as well as informational programs such
as news and documentaries. A 4-point scale was used
= =
(1 very unlikely and 4 very likely to watch the
program). Although no formal hypotheses were de-
veloped, we expected high AI subjects to show a greater
preference for emotionally stimulating programs than
their low AI counterparts. In contrast, we expected
41
no significant difference in preference scores between
high and low AI subjects in response to informational
or documentary news programs.
Manipulation Checks. Our manipulation checks had
two objectives. First, we wanted to determine whether
the emotional ads did have a greater impact on the
emotions than the nonemotional appeal. Second, be-
cause only one ad from each of the three ad-type
categories was used in the experiment, we wanted to
determine the extent to which the target ads may
have differed on dimensions other than emotionality.
Major differences among the ads on irrelevant di-
mensions could potentially distort the results of the
study. Using 7-point scales, we measured (1) the ex-
tent to which subjects were familiar with the ads, (2)
the degree of liheability of the ads, (3) the extent to
which the ads were perceived as crude, and (4) the
extent to which they were offensive to the viewer.
Dependent Measures. The measures of emotions used
for each ad were selected from a list of positive and
negative emotional adjectives drawn from the litera-
ture (Batra and Holbrook 1990; Edell and Burke 1987;
Holbrook and Batra 1987; Izard 1977). Using 7-point
scales, subjects indicated the strength of their emo-
tional reaction to the ad. The list of emotions in-
cluded feeling emotional, happy, joyous, warm, moved,
touched, sympathetic, and sad. Attitude toward the
ad (A.) was measured on four 7-point scales anchored
by good/bad, like/dislike, effective/not effective, and
interesting/not interesting (Holbrook and Batra 1987).
Enjoyment of the ad (Adonjoymont) was measured on a 7-
point scale on which subjects indicated the extent to
which they enjoyed watching the ad. Having com-
pleted the questionnaire, subjects were debriefed and
dismissed.
Results
Factor Analysis of the AIM Scale
To confirm the reliability of the affect intensity con-
struct, we factored and obliquely rotated all 40 items
by the Harris-Kaiser method (SAS User's Guide 1985),
which is consistent with the rotation method used by
Larsen (1984). The data (Appendix A) supported five
factors similar to the factor pattern initially reported
by Larsen (1984). Factor 2 (general intensity), factor
4 (intra-personal positive affect), and Factor 5 (pref-
erence for arousal) are similar to those reported by
Larsen (1984). However, the factor loadings for factor
1 (negative and positive affective reactions to life situ-
 42 Journal of Advertising

Table 1
Preferences Scores for TV Programs by High and Low AI Respondents·

TV Program Preferences
TV Program Category High AI Low AI FValue

Romance 2.36 1.47 7.68 d

(.78) (1.39)
War 1.55 .77 2.90 b
(.91) (2.02)
Musical 1.73 .86 7.18 d
(.98) (1.33)
Mystery 2.13 .86 7.18 d
(.56) (1.33)
7.51 d
Downloaded by [New York University] at 16:10 24 March 2015

Comedy 2.86 1.71

(.35) (1.94)
Documentary/news (60 Minutes) 2.25 2.82 3.69 b
(1.29) (1.01)

·N - 71 sUbjects; AI - affect intensity; standard deviations in parentheses.

bSignilicant at the .05 level.
CSignllicant at the .01 level.
dSignilicant at the .001 level.

ations) and factor 3 (intra-personal affective reac-
tions) were different from Larsen's (1984) findings in
that there was a definite tendency for both positive
and negative emotional reactions to load on the same
factor. Those factor patterns are consistent with the
notion that some individuals respond with greater
emotional intensity regardless of whether the emo-
tions are positive or negative (Larsen and Diener
1987).
TV Program Preferences
Table 1 reports the cell means showing the differ-
ences between high AI and low AI subjects' prefer-
ences for various categories of TV programs. Consis-
tent with the theory that affect intensity is related to
the temperamental dimension of emotional
arousability, high AI individuals expressed a prefer-
ence for programs that are typically emotionally stimu-
lating (e.g., romance, war, musicals, mystery, and
comedies), and low AI respondents expressed a pref-
erence for less emotionally charged news programs
such as CBS 60 Minutes (F 1.66 =3.69, p < .05).
Manipulation Checks
In the first manipulation check, paired-comparison
t-testa showed that the emotional ads had a signifi-
cantly stronger appeal to the emotions than the
nonemotional ad, (tOO> =6.20, p < .0001). The second
set of manipulation checks (all with 7-point scales)
were designed to determine whether subjects per-
ceived the target ads as being significantly different
in areas other than the emotional dimension. For the
measure of ad familiarity, we found no significant
differences in the scores associated with the emo-
tional and the nonemotional advertising appeals (t 66
= 1.36, p < .17). Similarly, we found no significant
differences in likeability between the emotional ads
and the nonemotional ad (t 66 =1.39,p < .17). Also, the
emotional and nonemotional ads did not differ in the
extent to which subjects perceived them to be offen-
sive (t 66 = 1.70, p < .09). On the question of whether
the ad was too crude to be shown on TV, we found no
significant differences in the scores associated with
the emotional and nonemotional ads (t 66 = .13, p <
.89). The apparent success of the manipulation checks
suggests that other factors such as prior familiarity
with the ad, variances in ad likeability, and other
reactions to the ad did not confound the results.
Reliability Checks
We conducted two types of reliability checks on the
emotions measures associated with each of the three
target ads. First, Cronbach alpha correlation coeffi-
 Summer 1996
cients for the variables (emotional, happy, joyous,
warm, moved, touched, sympathetic, and sad) linked
to the positive emotional ad, the negative emotional
ad, and the nonemotional ad were .90, .88, and .79,
respectively. Second, a principal components analy-
sis on all emotions variables revealed two dominant
factors (see Appendix B). All measures of emotions
associated with the positive and the negative emo-
tional advertising appeals loaded on the first factor
(eigenvalue =8.26). The second factor (eigenvalue =
3.78) contained all measures associated with the
nonemotional ad with the exception of two variables,
HAPPY3 and JOYOUS3, which presumably were in-
compatible with the tone of the nonemotional infor-
mational ad for Colgate toothpaste. In sum, the fact
Downloaded by [New York University] at 16:10 24 March 2015
that the loading pattern clustered into two main fac-
tors linked to the emotional and nonemotional ads,
respectively, suggests that the ads were perceived by
respondents as either emotional or nonemotional.
Emotions, Attitudes, and Ad Enjoyment
For data analysis on the dependent measures, we
used a 2(AI) x 3(ad type) within-subjects ANOYA
factorial design in which exposure to the three types
of ads was the repeated measure. The results showed
significant main effects for affect intensity on the
emotion measures associated with each of the two
emotional ads, but no significant effects for the same
measures in response to the nonemotional advertis-
ing appeal. Interaction effects were observed for the
following dependent measures: felt emotional (F2.64 =
3.80, p < .02), felt happy, (F2.64 =3.33, P < .04), and ad
enjoyment (F2,64 = 3.93, P < .02. The attitude toward
the ad measure was not significant, (F2.64 = 2.28, p <
.10). Table 2 reports the means of the high AI and low
AI respondents for the positive emotional ad, the nega-
tive emotional ad, and the nonemotional ad.
H1:Emotional Responses (Table 2). In response to
the positive emotional ad, high AI subjects had sig-
nificantly stronger emotional responses on all posi-
tive measures (happy, joyous, warm, moved, and
touched). High AI respondents also reported signifi-
cantly higher levels of sympathy than low AI respon-
dents (X = 5.03 vs. X = .11, (F1.6 = 6.98, P < .01).
Evidently, high AI individuals experienced greater
empathy with the feelings expressed by actors in the
ad. On the sadness measure we found no significant
difference between groups (X =4.12 vs. X =3.86, (FI.69
=.36, p < .23). That result was expected because high
AI subjects were not expected to manifest significant
increases in negative emotions in response to a posi-
43
tive emotional appeal.
In response to the negative emotional appeal, high
and low AI subjects did not differ significantly in
their expressions of how happy they felt. However,
low AI subjects had significantly higher scores (X =
1.43 vs. X=.91) than high AI individuals (F1.69 =4.18,
P < .01) in reporting how joyous they felt. Because the
negative emotional ad was designed to elicit sympa-
thy, sadness, and other negative types of emotions,
we expected no significant differences between high
AI and low AI subjects. The rest of the emotional
responses to that ad followed the expected pattern.
For the adjective "warm," we found no significant
difference between high AI and low AI subjects, as
expected. When asked the extent to which they felt
moved, touched, sympathetic, and sad in response to
the negative appeal, high AI subjects reported sig-
nificantly stronger emotional responses than their
low AI counterparts. Interestingly, when asked to
report their emotions using positive adjectives such
as "happy" and "warm" that are incongruent with the
negative valence ofthe advertisement, the two groups
did not differ significantly in the intensity of their
emotional reactions. That finding is not surprising as
we did not expect a positive emotional ad to evoke a
negative emotional response and vice versa.
When exposed to the nonemotional ad, high and
low AI subjects did not differ significantly on any of
the seven emotional measures. That finding is consis-
tent with our expectation that the two groups would
differ only in the presence of an emotion eliciting
stimulus (Larsen and Diener 1985). Our findings
therefore support HI.
H2: Attitude Toward the Ad. As predicted, in re-
sponse to the positive emotional ad, high AI subjects
reported significantly stronger attitudes toward the
ad than low AI subjects, (X =6.22 vs. X =5.70, F 1.69 =
6.21, P < .01). In contrast, the two groups did not
have significantly different A. d scores when exposed
to the negative emotional ad. Those results do not
support H2, that high AI subjects would exhibit
stronger negative attitudes than low AI subjects in
response to a negative emotional ad. When exposed
to the nonemotional ad, high AI and low AI subjects
did not have significantly different A. d scores. That
finding is consistent with our expectation.
H3: Ad Enjoyment. As expected, in comparison with
low AI subjects (X =5.30), high AI subjects (X =6.30)
expressed significantly higher levels of enjoyment for
the positive emotional ad (F2.l3l = 7.37, P < .001).
However, we found no significant differences in ad
enjoyment when subjects were exposed to the nega-
 44 Journal of Advertising

Table 2
Comparison of Responses to Positive Emotional, Negative Emotional, and Nonemotional Ads·

Positive Emotional Ad Negative Emotional Ad Nonemotional Ad

Measures High Low High Low High Low

AI AI FValue AI AI FValue AI AI FValue

Emotional appeal 6.58 5.60 8.34d 5.64 4.76 4.98 b 1.97 2.45 1.28
Happy 4.33 3.46 4.66 b 1.24 1.68 2.42 1.42 1.22 .33
Joyous 4.39 3.16 9.73 d .91 1.43 4.18 c .78 .67 .13
Warm 5.24 4.32 9.08 d 2.68 2.88 .29 .73 .76 .01
Moved 5.39 4.03 13.61· 4.49 3.65 5.36 c .86 .73 .18
Touched 5.52 4.43 11.09 d 4.56 3.35 8.36 d .61 .65 .02
Sympathetic 5.03 4.11 6.98 c 4.91 4.16 4.19 c .58 .51 .06
Downloaded by [New York University] at 16:10 24 March 2015

Sad 4.12 3.86 .36 4.36 3.27 9.04d .97 .76 .46
A od 6.22 5.70 6.21 c 3.33 3.37 .03 4.76 4.09 .12
Ad IlI!gymonI 6.30 5.30 7.37 d 4.67 4.57 .06 2.70 2.62 .02

·N·71 sUbJects; AI • affect Intensity.

bSlgnificantat the .05 level.
CSlgnlflcant at the .01 level.
dSignificantat the .001 level.
·Significant at the .0001 level.

tive emotional ad. Although that finding is contrary
to H3, it is consistent with the lack of significance
associated with attitude toward the negative emo-
tional ad. To some extent, the finding is not surpris-
ing. We did not expect high AI respondents to express
higher levels of enjoyment for a negative emotional
ad. Perhaps high intensity individuals were uncom-
fortable with the higher levels of negative emotional
stimulation that they presumably experienced.
H4: Mediation. The regression analysis provided
support for the hypothesis that positive emotions
mediate the effect of affect intensity on Aod (Table 3).
We used Baron and Kenny's (1986) four-step proce-
dure to establish mediation. First, the predictor vari-
able affect intensity, taken alone, was related to the
mediator, positive emotions (~= 1.03, W = .14, F I •67 =
10.91, p < .(01). Affect intensity also significantly
influenced the dependent variable, Aod (P = 75, W
=.08, F l o67 = 6.21, P < .01). Second, the mediator vari-
able, positive emotions, had a significant effect on Aod
(~ = .61, W = .42, F I ,67 = 49.88, p < .0001). As a final
condition for proving mediation, Table 3 shows that
when the positive emotions variable was included
with affect intensity as a predictor variable in the
regression equation, the previously significant effect
of affect intensity on Aod was eliminated (P =.14, W =
.43, F I ,67 = 24.83, p < .35), but the effect of positive
emotions on A. d remained significant (p = .59, R2 =
.43, F 1,67 = 24.83, p < .(01). Thus, positive emotions
served as the mechanism through which affect inten-
sity influenced A.d' H4 is therefore supported.
For the negative emotional ad and the nonemotional
ad, it was impossible to show any mediation role for
emotions because affect intensity had no relationship
to Aod in either of those ad exposure conditions. How-
ever, in the negative emotional ad condition, emo-
tions had a negative and significant relationship with
Aod (p = -.29, W =.21, F I ,67 = 17.61, p < .0001), which
suggests that the stronger the expression of negative
emotions, the more unfavorable was the attitude to-
ward the ad.
Discussion
We examined the extent to which the affect inten-
sity construct (Larsen 1984) can be applied to the
understanding of message recipients' response to dra-
matic and provocative emotional advertising appeals.
In general, the results support Larsen and Diener's
(1987) contention that certain individuals are predis-
posed to respond with significantly greater emotional
intensity than others when exposed to emotion-elicit-
ing stimuli. As predicted, the results demonstrate
that when the ads were emotional, whether positive
 Summer 1996
Table 3
Regression Analysis Showing Positive Emotions Mediating the Effect of Affect Intensity on
Attitude Toward the Ad (A..Ja
Positive
Emotions
Affect intensity 1.03c
Positive emotions
R2 .14
Fvalue 10.91
aN • 71 subjects.
bSignificant at the .01 level.
cSignificant at the .001 level.
Downloaded by [New York University] at 16:10 24 March 2015
dSignificant at the .0001 level.
or negative, individuals who were classified as high
AI had stronger emotional reactions than their low
AI counterparts. In contrast, no significant differ-
ences in the intensity of emotional responses were
observed when subjects were exposed to a
nonemotional ad. We also found that high AI indi-
viduals had stronger positive attitudes toward the ad
and greater levels of enjoyment only when exposed to
a positive emotional ad and not in response to a nega-
tive emotional ad or a nonemotional ad. Interestingly,
emotions served as the mechanism through which
affect intensity influenced attitude only when sub-
jects were exposed to a positive emotional appeal.
Attitude toward High Impact
Emotional Ads
From the proposition of Gallagher, Diener, and
Larsen (1989), we expected high AI individuals to
express a greater distaste for the negative feelings
stimulated by exposure to the negative emotional ap-
peal and therefore to express more negative attitudes
toward the ad than their low AI counterparts. Our
results show no significant differences in ad attitudes.
A. d scores are 3.33 and 3.37, respectively, for high AI
and low AI respondents. On a 7-point scale, those
scores are somewhat near the midpoint range, per-
haps suggesting a certain degree of indifference rather
than a very strong dislike for the ad.
The neutrality in attitudes expressed by both high
AI and low AI subjects could be interpreted in a vari-
ety of ways. For example, subjects could have been
involved in a process of avoidance of an emotionally
upsetting stimulus-a type of psychological disen-
gagement that may have inhibited further elabora-
45
AdM Ad ad Ad ad
.75b .14
.61d .59d
.08 .42 .43
6.21 49.88 24.83
tion of the appeal. Interestingly, recent research has
shown that elaborative processing of the message is
important in the formation of strong attitudes (Petty
et al. 1993). In any case, the arousal of negative emo-
tions in response to the ad may have been an impor-
tant factor in attitude formation in our study. Indeed,
our results showed that when subjects were exposed
to a negative emotional appeal, negative emotions
had a significant but negative impact on attitudes
toward the ad. For high AI subjects, the experience of
negative emotional arousal did not produce a favor-
able attitude toward the advertisement. In contrast,
the data show that emotions mediated the influence
of affect intensity on attitude toward the ad only when
subjects were exposed to a positive emotional appeal.
Alternatively, the neutrality in ad attitudes could
be interpreted to mean that message recipients are
really indifferent in their attitudes toward "shock"
ads and may even enjoy some degree of thrill when
exposed to such appeals. Although the high AI and
low AI groups do not differ in level of enjoyment of
the ad, their mean scores are slightly above the mid-
point range on a 7-point scale (Table 2), perhaps sug-
gesting that both groups experienced some mild de-
gree of enjoyment of the negative emotional ad. One
possible explanation for that finding is that even
though the Red Cross appeal evoked very strong emo-
tions such as sadness and sympathy, that form of
emotional arousal was still tolerable, perhaps enjoy-
able, even for high AI individuals. Therefore, if such
ads can induce the intended behavioral change or
influence compliance among the target audience, ad-
vertisers should continue to use them. That conclu-
sion would suggest empirical support for the "blunter-
is-better approach" and undermine the warnings of
 46
critics who claim that the use of high powered, emo-
tionally shocking appeals may produce excessive lev-
els of anxiety and thereby cause psychological harm
(Henthorne, LaTour, and Nataraajan 1993; Hyman
and Tansey 1991).
Nevertheless, our study provides further evidence
of the importance of using individual difference vari-
ables in studies of advertising effects. The findings
also enhance our understanding of the reasons why
strong, emotionally charged ads vary so widely in
their effects on the message recipients' responses.
Future Research and Limitations of
the Study
Downloaded by [New York University] at 16:10 24 March 2015
Future research should examine more carefully the
extent to which emotionally provocative advertising
stimuli influence the attitudes of high AI individuals.
For example, it is still not clear whether the arousal
of strong negative emotions actually diminishes the
recipient's motivation to process the message and
whether the lack of cognitive elaboration reduces the
strength of the attitude toward the ad. When exposed
to an affect-laden advertising appeal, high intensity
subjects are likely to experience a significantly height-
ened level of emotional arousal that may reduce the
motivation for cognitive elaboration, thus rendering
such respondents less capable of reporting their true
attitudes (Sanbonmatsu and Kardes 1988).
Before any extensive generalizations can be drawn
from our findings, certain limitations of our study
must be identified. First, although affect intensity
performed as expected across the three different cat-
egories of advertising appeals, only one ad represented
each category. Pretest results confirmed convergent
validity-the assessment of the impact of the inde-
pendent variables on the dependent measures (Cook
and Campbell 1979}-but the design of our experi-
ment does not enable us to rule out divergent validity
concerns such as the possibility that factors other
than the emotional character of the target ads influ-
enced the results. We attempted to address that limi-
tation by conducting manipulation checks, which in-
dicated that the three ads did not differ on potentially
confounding factors such as perceived ad likeability
and ad familiarity. Nevertheless, other divergent va-
lidity concerns may remain. One way to address the
problem in future research is to expose subjects to
multiple ads representing each of the three catego-
ries and use the ad as a nested factor.
The crucial issue is the need for a more rigorous
Journal of Advertising
manipulation of the emotional versus nonemotional
advertising stimuli so that the validity of the affect
intensity effect can be established more convincingly.
One parsimonious method of reducing the potential
for confounding effects is to ensure that the emo-
tional and nonemotional versions of the target ad
have a high degree of similarity and then manipulate
only one specific feature of the target ad. Such ma-
nipulation should produce a difference between the
emotional and nonemotional ads (Stayman and Batra
1991). We used that method in a follow-up study to
provide support for the validity of the affect intensity
construct in a more rigorously controlled experiment.
The follow-up study had a simple 2 x 2 ANOVA
factorial design. Subjects previously classified as high
and low in affect intensity were invited to a lab to
listen to a new format for radio broadcasting. During
the last commercial break of a 20 minute broadcast
segment, they listened to either an emotional or a
nonemotional version of a radio advertisement for a
new burglar alarm from Honeywell. The essential
features (such as the ad copy and the spokesperson)
of both versions of the ad were the same. However, in
the emotional version, dramatic sound effects were
used to create an emotional impact. Immediately af-
ter hearing the commercial, subjects recorded the
strength of their emotional response to the ad, their
attitudes toward the ad, and the extent of their em-
pathic involvement. The results of the follow-up study
provided remarkable support for the findings of the
main study. For measures of emotional response, we
found main effects for the affect intensity and ad-type
predictor variables and a significant AI by ad-type
interaction: high AI subjects experienced stronger
emotional reactions than low AI subjects but the two
groups did not have significantly different reactions
when exposed to the nonemotional ad. The attitude
toward the ad measure revealed no significant differ-
ences between high AI and low AI subjects and no
significant interaction effects. The results also showed
that high AI subjects, in comparison with their low
AI counterparts, perceived the exposure to the ad as
a more painful experience only when exposed to the
emotional version of the ad. Altogether, the findings
of the follow-up study are consistent with those ofthe
main study and therefore bolster our confidence in
the validity of the affect intensity construct.
A second limitation of our study pertains to the
practical use of the AI construct for target marketing
and segmentation. For example, if individual differ-
ences in affect intensity prove to be related to
persuadability, the types of appeals that may be sue-
 Summer 1996
cessful with one audience segment may not be as
successful with another segment. Hence, to enhance
the advertiser's ability to identify relevant market
segments, some degree of linkage should be estab-
lished between individual differences in affect inten-
sity and readily observable and measurable patterns
of behavior such as a person's lifestyle and personal
preferences. Larsen, Diener, and Emmons (1986)
found affect intensity to be related significantly to
four dimensions of temperament. High AI individu-
als were more active, sociable, physically arousable,
and emotionally reactive than their low AI counter-
parts. Larsen (1984) suggested that high AI individu-
als should have a greater tendency to seek out emo-
tional stimulation from day-to-day activities and ex-
Downloaded by [New York University] at 16:10 24 March 2015
periences that are emotionally provocative. Indeed,
our preliminary findings (Table 1), indicating a stron-
ger preference by high AI respondents for the more
emotionally stimulating TV programs and a weaker
preference for the less emotionally stimulating pro-
grams (such as news documentaries), provide encour-
aging support for the theoretical rationale underly-
ing the affect intensity construct. A systematic pro-
gram of research for the future should include the
development of linkages between individual differ-
ences in affect intensity and consumer lifestyles, atti-
tudes, and personal preferences.
References
Alsop, Ronald (1988), "Drug and Alcohol Clinics Vie for
Patients,"Wall Street Journal, November 15, B1.
Bagozzi, Richard P. and David J. Moore (1994), "Public Service
Advertisements: Emotions and Empathy Guide Prosocial Be-
havior," Journal of Marketing, 58 (January), 56-70.
Baron, Ruben M. and David A. Kenny (1986), "The Moderator-
Mediator Distinction in Social Psychology Research: Concep-
tual, Strategic and Statistical Considerations," Journal ofPer-
sorwlity and Social Psychology, 51, (6), 1173-1182.
Batra, Rajeev and Michael L. Ray (1986), "Affective Responses
Mediating Acceptance of Advertising, " Journal of Consumer
Research, 13, (September), 234-249.
- - - - - and Morris B. Holbrook (1990), "Developing a Ty-
pology of Affective Responses to Advertising," Psychology &
Marketing, 7, (Spring), 65-81.
Cacioppo, John T. and Richard E. Petty (1982), "The Need for
Cognition," Journal of Personality and Social Psychology, 42,
(I), 116-131.
Cook Thomas D. and Donald T. Campbell (1979), Quasi-Experi-
~entation: Design & Arwlysis Issues for Field Settings, Bos-
ton: Houghton Mifflin.
Diener, Ed, Randy J. Larsen, Steven Levine, and Robert Emmons
(1985), "Frequency and Intensity: The Underlying Dimen-
sions of Affect," Journal ofPersonality and Social Psychology,
48, (5), 1253-1265.
Edell, Julie A. and Marian Chapman Burke (1987), "The Power of
Feelings in Understanding Advertising Effects," Journal of
Consumer Research, 14, (December), 421-433.
47
Eysenck, Hans J. (1967), The Biological Basis ofPersonality, Spring-
field, IL: Thomas.
Gallagher, Dennis, Ed Diener, and Randy J. Larsen (1989), "Indi-
vidual Differences in Affect Intensity: A Moderator of the
Relation Between Emotion and Behavior," working paper,
Department of Psychology, University of Illinois, Urbana-
Champaign.
Geen, Russell G. (1984), "Preferred Stimulation Levels in Intro-
verts and Extroverts: Effects on Arousal and Performance,"
Journal of Pereonality and Social Psychology, 46, (6), 1303-
1312.
Haugtvedt, Curtis P., Richard E. Petty, and John T. Cacioppo
(1992), "Need for Cognition and Advertising: Understanding
the Role of Personality Variables in Consumer Behavior,"
Journal of Consumer Psychology, 1 (3), 239-260.
Henthorne, Tony L., Michael S. LaTour and Rajan Nataraajan
(1993), "Fear Appeals In Print Advertising: An Analysis of
Arousal and Ad Response," Journal ofAdvertising, 22 (June),
59-68.
Holbrook, Morris B. and Rajeev Batra (1987), "Assessing the Role
of Emotions as Mediators of Consumer Responses to Advertis-
ing," Journal ofConsumer Research, 14, (December), 404-420.
Hyman, Michael R. and Richard Tansey (1990), "The Ethics of
Psychoactive Ads," Journal of Business Ethics, 9 (February),
105-114.
Izard, Carroll E. (1977), Human Emotions, New York: Plenum
Press.
King, Thomas R. (1989), "More Public Service Commercials Take
Blunter-Is-Better Approach," Wall Street Journal, June 5, B1.
LaTour, Michael S. and Robert E. Pitts (1989), "Using Fear Ap-
peals in Advertising for AIDS Prevention in the College-Age
Population," -Jaurnal of Health Care Marketing, 9, (Septem-
ber),5-14.
Larsen, Randy J. (1984), '"!'heory and Measurement of Affect In-
tensity as an Individual Difference Characteristic," Disserta-
tion Abstracts International, 85, 2297B (University Micro-
films No. 84-22112).
- - - - - a n d Ed Diener (1987), "Affect Intensity as an Indi-
vidual Difference Characteristic: A Review," Jaurnal of Re-
search in Personality, 21 (I), 1-39.
-----, , and Russell S. Cropanzano (1987),
"Cognitive Operations Associated with Individual Differences
in Affect Intensity," Journal of Personality and Social Psy-
chology, 33 (4),767-774.
-----, , and Robert A. Emmons (1986), "Af-
fect Intensity and Reactions to Daily Life Events," Journal of
Persorwlity and Social Psychology, 51 (4), 803-814.
MacKenzie, Scott B. and Richard J. Lutz (1989), "An Empirical
Examination of the Structural Antecedents of Attitude To-
ward the Ad in an Advertising Pretesting Context," dournol
of Marketing, 53 (April), 48-65.
Moore, David J. (1989), "Advertising That Makes the Brain Itch,"
MBA Update, Bureau of Business Practice, Old Lyme, CT:
Simon & Schuster Inc.
Moore, David J., William D. Harris, and Hong C. Chen (1995),
"Affect Intensity: An Individual Difference Response to Ad-
vertising Appeals," Journal of Consumer Research, 22 (Sep-
tember), 154 - 164.
Petrie, Asenath (1967), Individuality in Pain and Suffering, Chi-
cago: University of Chicago Press.
Petty, Richard E., David W. Schumann, Steven A. Richman, and
Alan J. Strathman (1993), "Positive Mood and Persuasion:
Different Roles for Affect Under High and Low Elaboration
Conditions," Journal ofPersonality and Social Psychology, 64
(1),5-20.
 48
Ray, Michael L. and William Wilkie (1970), "Fear: The Potential of
an Appeal Neglected by Marketing," Journal of Marketing
Research, 34 (January), 54·62.
Sanbonmatsu, David M. and Frank R. Kardes (1988), 'The Effects
of Physiological Arousal on Information Processing and Per-
suasion," Journal ofConsumer Research, 15 (December), 379·
385.
SAS User's Guide: Statistics Version & Edition (1985), Cary, NC:
SAS Institute, Inc. 27511.
Snyder, Mark and Kenneth G. DeBono (1985), "Appeals to Image
About Quality: Understanding the Psychology of Advertis·
ing," Journal of Personality and Social Psychology, 49 (3),
586-597.
Downloaded by [New York University] at 16:10 24 March 2015
Journal of Advertising
Stayman, Douglas M. and David A. Aaker (1988), "Are All the
Effects of Ad-induced Feelings Mediated by Awl?" Journal of
C07IBumer Research, 15, (December), 368-373.
- - - - - and Rajeev Batra (1991), "Encoding and Retrieval
of Ad Mfect in Memory," Journal of Marketing Research, 28
(May), 232-239.
Strelau, Jan (1982), "Biologically Determined Dimensions of Per-
sonality or Temperament?" Personality and Indiuidual Dif-
ferences, 3, (4), 355-360.
Zuckerman, Marvin (1979), Sensation. Seeking: Beyond Optimum
Leuel ofArousal, Hillsdale, NJ: Erlbaum.
 Summer 1996 49

Appendix A
Rotated Factor Pattern of 40·ltem AI Scale
(factor loadings in parenthesis)

Factor 1: Negative and Positive Affective Reactions to Life Situations (eigenvalue - 2.46)
13. (.69) When I talk in front of a group for the first time my voice gets shaky and my heart races
23. (.62) When I receive a reward I become overjoyed
25. (.54) When I do something wrong I have strong feelings of shame and guilt
30. (.48) When I do feel anxiety it is normally very strong
39. (.80) When I am nervous I get shaky all over
40. (.68) When I am happy the feeling is more like contentment and inner calm than one of exhilaration
and excitement (reversed)

Factor 2: General Intensity (eigenvalue - 2.85)

Downloaded by [New York University] at 16:10 24 March 2015

26. (.78) I can remain calm even on the most trying days (reversed)
28. (.68) When I get angry it's easy for me to still be rational and not overreact (reversed)
31. (.80) My negative moods are mild in intensity (reversed)

Factor 3: Intrapersonal Affective Reactions (eigenvalue - 3.44)

1. (.73) When I accomplish something difficult I feel delighted or elated
4. (.72) I feel pretty bad when I tell a lie
5. (.47) When I solve a small personal problem, I feel euphoric
9. (.48) If I complete a task I thought was impossible, I am ecstatic
11. (.38) Sad movies deeply touch me
12. (.79) When I'm happy it's a feeling of being untroubled and content rather than being zestful and
aroused (reversed)
16. (.85) The memories I like the most are of those times when I felt content and peaceful rather than
zestful and enthusiastic (reversed)
33. (.37) When I feel happiness, it is a quiet type of contentment (reversed)
36. (.44) When I feel gUilty, this emotion is quite strong

Factor 4: Intrapersonal Positive Affect (eigenvalue - 3.28)

2. (.47) When I feel happy it is a strong type of exuberance
7. (.59) My happy moods are so strong that' feel like I'm in heaven
8. (.49) I get overly enthusiastic
14. (.91) When something good happens, I'm usually much more jubilant than others
18. (.81) When I'm feeling well it's easy for me to go from being in a good mood to being really joyful
27. (.95) When things are going good I feel "on top of the world"

Factor 5: Preference for Arousal (eigenvalue - 4.39)

19. (.70) "Calm and cool" could easily describe me (reversed)
24. (1.07) When I succeed at something, my reaction is calm and contentment (Reversed)
29. (.99) When I know I have done something very well, I feel relaxed and content rather than excited and
elated (reversed)
37. (.95) I would characterize my happy moods as closer to contentment than joy (reversed)
 50 Journal of Advertising

Appendix B
Factor Loading Patterns of All Emotional Responses Associated with Positive Emotional Ad, Negative
Emotional Ad, and Nonemotional Ad

Emotional Variable Factor 1 Factor 2

Positive Emotional Ad
Emotional1 .49894
Happy1 .73116
Warm1 .68415
Moved1 .77741
Touched1 .79173
Sympathetic1 .66304
Sad1 .38866
Downloaded by [New York University] at 16:10 24 March 2015

Negative Emotional Ad
Emotional2 .74087
Happy2 .64380
Joyous2 .69299
Warm2 .72132
Moved2 .75260
Sympathetic2 .63566
Sad2 .54800

Nonemotional Ad
Emotional3 .71786
Happy3 .37916
Joyous3 .28369
Warm3 .61753
Moved3 .75478
Touched3 .74138
Sympathetic3 .67072
Sad3 .52376