Let’s not be Indifferent about Neutrality: Neutral Ratings in the IAPS Mask Mixed Affective Responses

Abstract

The International Affective Picture System (IAPS) is a picture set used by researchers to select pictures that have been pre-rated on valence. Researchers rely on the ratings in the IAPS to accurately reflect the degree to which the pictures elicit affective responses. Here we show that this may not always be a safe assumption. More specifically, the scale used to measure valence in the IAPS ranges from positive to negative, implying that positive and negative feelings are end-points of the same construct. This makes interpretation of midpoint, or neutral ratings, especially problematic because it is impossible to tell whether these ratings are the result of neutral, or of mixed feelings. In other words, neutral ratings may not be as neutral as researchers assume them to be. Investigating this, in this work we show that pictures that seem neutral according to the valence ratings in the IAPS indeed vary in levels of ambivalence they elicit. Furthermore, the experience of ambivalence in response to these pictures is predictive of the arousal that people report feeling when viewing these pictures. These findings are of particular importance because neutrality differs from ambivalence in its specific psychological consequences, and by relying on seemingly neutral valence ratings, researchers may unwillingly introduce these consequences into their research design, undermining their level of experimental control.

Full text

BRIEF REPORT
Let’s Not Be Indifferent About Neutrality: Neutral Ratings in the
International Affective Picture System (IAPS) Mask Mixed Affective
Responses
Iris K. Schneider
University of Southern California and VU
University Amsterdam
Lotte Veenstra
VU University Amsterdam
Frenk van Harreveld
University of Amsterdam
Norbert Schwarz
University of Southern California
Sander L. Koole
VU University Amsterdam
The International Affective Picture System (IAPS) is a picture set used by researchers to select pictures
that have been prerated on valence. Researchers rely on the ratings in the IAPS to accurately reflect the
degree to which the pictures elicit affective responses. Here we show that this may not always be a safe
assumption. More specifically, the scale used to measure valence in the IAPS ranges from positive to
negative, implying that positive and negative feelings are end-points of the same construct. This makes
interpretation of midpoint, or neutral ratings, especially problematic because it is impossible to tell
whether these ratings are the result of neutral, or of mixed feelings. In other words, neutral ratings may
not be as neutral as researchers assume them to be. Investigating this, in this work we show that pictures
that seem neutral according to the valence ratings in the IAPS indeed vary in levels of ambivalence they
elicit. Furthermore, the experience of ambivalence in response to these pictures is predictive of the
arousal that people report feeling when viewing these pictures. These findings are of particular impor-
tance because neutrality differs from ambivalence in its specific psychological consequences, and by
relying on seemingly neutral valance ratings, researchers may unwillingly introduce these consequences
into their research design, undermining their level of experimental control.
Keywords: ambivalence, IAPS, neutrality, mixed feelings
Cited over 2,500 times (Google Scholar, April 2015), the Inter-
national Affective Picture System (IAPS; Lang, Bradley, & Cuth-
bert, 2008) is a widely used set of emotion-eliciting stimuli. Each
picture has a prerated valence score that shows how positive or
negative people feel, on average when viewing it. Researchers
select experimental stimuli based on these scores. Unfortunately,
the scores may be systematically misleading when researchers
select pictures with an apparently “neutral” score. Because the
IAPS stimuli are rated on bipolar scales, neutral ratings may
actually mask ambivalent responses.
Specifically, the IAPS norming procedure asked for ratings on a
scale that ranges from positive (happy, pleased, satisfied, contented,
hopeful) to negative (unhappy, annoyed, unsatisfied, melancholic,
despaired, bored) (Lang et al., 2008). This conceptualization implies
(i) that positive and negative end terms are two extremes of the same
underlying emotion construct and (ii) that positive and negative feel-
ings constitute a zero sum—The more positive respondents feel, the
less negative they feel, and vice versa. However, in daily life, people
may also experience instances in which they feel both positive and
negative at the same time, also referred to as ambivalence (Kaplan,
1972;Larsen, McGraw, & Cacioppo, 2001;Rees, Rothman, Lehavy,
&Sanchez-Burks,2013;Thompson, Zanna, & Griffin, 1995;Sch-
Iris K. Schneider, University of Southern California and VU University
Amsterdam; Lotte Veenstra, VU University Amsterdam; Frenk van Har-
reveld, University of Amsterdam; Norbert Schwarz, University of Southern
California; Sander L. Koole, VU University Amsterdam.
The authors gratefully acknowledge Evert-Jan van Doorn, Mark Rotte-
veel, and two anonymous reviewers for their useful suggestions. This work
was in part made possible by the Netherlands Organization for Scientific
Research Rubicon Grant awarded to Iris K. Schneider (No. 446-13-015),
and in part by a Consolidator Grant from the European Research Council
(ERC-2011-StG_20101124) awarded to Sander L. Koole.
Correspondence concerning this article should be addressed to Iris K.
Schneider, VU University, Department of Behavioral and Movement Sci-
ences, Room TR2E-29, van der Boechorstraat 1, 1081BT, Amsterdam, the
Netherlands. E-mail: schneider.ik@gmail.com
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Emotion © 2016 American Psychological Association
2016, Vol. 16, No. 2, 000 1528-3542/16/$12.00 http://dx.doi.org/10.1037/emo0000164
1
AQ: au
AQ: 1
AQ: 2
AQ: 9
tapraid5/emo-emo/emo-emo/emo00416/emo3124d16z
xppws S!1 2/24/16 1:36 Art: 2015-0884
APA NLM

neider et al., 2015;van Harreveld, van der Pligt, & de Liver, 2009).
The difference between the assumption underlying the bipolar scale
and people’s real-life affective experiences poses a problem for the
interpretation of the neutral ratings in the IAPS.
Imagine for instance, a chair. Most people may feel relatively
indifferent with regard to a chair, reflecting that they have neither
positive nor negative responses to it. On a bipolar scale, the chair
would have to be rated around the midpoint of the scale. However,
something that evokes both strong positive and strong negative
feelings (e.g., a tasty but fattening piece of chocolate cake) will
also be rated around the midpoint of the scale when people are
trying to do justice to both their negative and positive responses
(cf. Kaplan, 1972;Thompson et al., 1995). Accordingly, a neutral
rating along a bipolar scale may reflect either (correctly) the
absence of positivity and negativity, or the complete opposite, the
presence of relatively strong positivity and negativity (i.e., ambiv-
alence). This renders ‘neutral’ scores for IAPS pictures ambigu-
ous.
One rating already present in the IAPS that bears on this
possibility is the arousal that the pictures elicit. As previous
research has shown, the experience of ambivalence is associated
with increased arousal (Has, Katz, Rizzo, Bailey, & Moore, 1992;
Nordgren, van Harreveld, & van der Pligt, 2006;van Harreveld,
Rutjens, Rotteveel, Nordgren, & van der Pligt, 2009). If apparently
neutral IAPS pictures include pictures that elicit ambivalent re-
sponses, as well as pictures that elicit neutral responses, they
should show a wide range of arousal scores. Empirically, this is the
case. The arousal ratings of neutral IAPS pictures (i.e., within a
".05 range around the 5.0 midpoint of the IAPS scale) range from
M!2.00 to M!6.03 (on a 9-point scale) with an average arousal
score of 3.61 (SD !1.05) (Lang et al., 2008), a first indication that
people may actually may feel relatively ambivalent, about at least
some of them, rather than neutral.
This issue is of substantive importance for researchers, because
ambivalence is associated with distinct psychological conse-
quences, including uncertainty, biased information processing, and
increased illusionary pattern perception, in addition to increased
arousal (for overviews, see, e.g., Sparks, Conner, James, Shepherd,
& Povey, 2001;van Harreveld, Nohlen, & Schneider, 2015;van
Harreveld, Rutjens, Schneider, Nohlen, & Keskinis, 2014). Re-
searchers who intend to induce neutral feelings but select pictures
that elicit ambivalent responses run the risk of inadvertently intro-
ducing these consequences in their neutral condition. It is therefore
important to investigate whether neutral, midpoint-rated pictures in
the IAPS mask ambivalence.
The confusion between neutrality and ambivalence has been
addressed extensively in research on ambivalence and yielded at
least two ways of dealing with this ambiguity. First, because
ambivalence is made up of both positive and negative thoughts and
feelings, one way is to separately assess both these opposing
components of the attitude and submitting them to a formula that
takes into account both the strength and the similarity of the
ratings. Because ambivalence and neutrality differ in the extremity
of the evaluations (ambivalence is made up of strong opposing
feelings, whereas neutrality constitutes a lack of strong feelings
either way), this method allows distinction between the two (Ka-
plan, 1972;Refling, Calnan, Fabrigar, MacDonald, Johnson, &
Smith, 2013;Thompson et al., 1995). Given the focus of this
measure on the evaluations underlying the attitude, or its structure,
it is often referred to as a measure of objective ambivalence. A
second way to distinguish ambivalence from neutrality is to ad-
dress the experiential component of being ambivalent, or the
feelings of conflict people have, often referred to as subjective
ambivalence (Priester & Petty, 1996). Given its experiential na-
ture, arousal is more closely related to subjective rather than
objective ambivalence (Newby-Clark, McGregor, & Zanna, 2002).
We hypothesize that because neutral ratings are likely to repre-
sent both relatively neutral and relatively ambivalent ratings, levels
of ambivalence will differ between IAPS pictures with neutral
ratings. Furthermore, we hypothesize that these differences, spe-
cifically those in subjective ambivalence, are related to the amount
of arousal people experience in response to the pictures, reflecting
a crucial difference between neutrality and ambivalence. Previous
work on the separability of positive and negative valence process-
ing has examined independent assessments of valence, arousal,
positivity, negativity, and ambivalence (Ito, Cacioppo, & Lang,
1998). The current work goes beyond these findings by focusing
exclusively on neutral pictures and relating variation in ambiva-
lence directly to arousal.
Method
Participants and Design
Forty-one University of Amsterdam undergraduates (mean age
21.6 years, 11 men) participated in exchange for monetary reward
or course credit. Participants rated selected pictures (see below)
from the IAPS (Lang et al., 2008) on both original IAPS measures,
as well as measures of ambivalence.
Stimuli
We selected 31 pictures from the IAPS (Lang et al., 2008) for
which the valence ratings ranged from 4.95 to 5.05 (on a 9-point
scale, taken from the IAPS) resulting in an average rating of 4.99
(SD !.03).
1
Given that the midpoint of the valence scale is 5,
these pictures represent the most neutral pictures in the set. Pic-
tures depicted a range of subjects, such as animals, objects, and
people. As noted before, IAPS arousal scores for these pictures
ranged from M!2.00 to 6.03 (on a 9-point scale) with an average
arousal score of 3.61 (SD !1.05).
Measures
We measured valence and arousal with the Self-Assessment
Manikin (SAM) scales as used in the IAPS. SAM scales are a
visual depiction of a 9-point scale. For valence, 1 represents
positive feelings, and 9 represents negative feelings. For arousal, 1
represents no arousal, and 9 represents high arousal (see Lang et
al., 2008).
To measure objective ambivalence, we used two items assessing
participants’ negative and positive evaluations of each picture
1
Pictures used were as follows: 1645, 2002, 2214, 2220, 2397, 2484,
2749, 2890, 4302, 4550, 4561,7000, 7002, 7003, 7004, 7012, 7020, 7034,
7035, 7041, 7045, 7160, 7161, 7185, 7235, 7185, 7476, 7484, 7640, 9070,
and 9422. The range of 4.95–5.05 was chosen in order to keep as close as
possible to the true midpoint.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2SCHNEIDER ET AL.
Fn1
tapraid5/emo-emo/emo-emo/emo00416/emo3124d16z
xppws S!1 2/24/16 1:36 Art: 2015-0884
APA NLM

independently (Kaplan, 1972). Specifically, we asked participants
to ignore their negative (vs. positive) evaluations and rate how
positive (vs. negative) they thought the picture was. The items read
as follows: “Think about this picture. When you think about the
positive (negative) aspects of this picture, while ignoring the
negative (positive) aspects, how positive is your evaluation of this
picture?” Participants gave their ratings on a 9-point scale ranging
from “not at all positive” (negative) to “very positive” (negative).
To obtain an index of objective ambivalence, we submitted aver-
age scores on these items per picture to the following formula:
(P#N)/2 $|P—N|, where Pis the positive rating and Nis the
negative rating (cf. Thompson et al., 1995; also see Schneider et
al., 2015;van Harreveld et al., 2009;van Harreveld et al., 2014),
with lower scores indicating more neutrality and higher scores
indicating more ambivalence. Using this formula, we are able to
capture both the strength as well as the similarity of the ratings, the
two core components of ambivalence (for excellent reviews of
different measures of objective ambivalence, see Priester & Petty,
1996;Thompson et al., 1995).
Theoretically, this scale can range from $3 to 9. However,
consider that an objective ambivalence score of 1 is the result of a
rating of 1 on both the positive, as well as the negative, item,
reflecting true neutrality. Conversely, an objective ambivalence
score of 9 is the result of a rating of 9 on both items, reflecting
extreme ambivalence. Objective ambivalence scores below 1 can
only be the result of relatively univalent stimuli (i.e., stimuli that
are considered clearly more positive than negative or vice versa),
but univalent stimuli would not have yielded midpoint ratings in
the IAPS norming procedure to begin with, and therefore are not
part of our set of stimuli. Thus, because we selected only stimuli
around the midpoint score in the IAPS, here the objective ambiv-
alence scale ranges only from 1 to 9, with lower scores indicating
more neutrality and higher scores indicating more ambivalence.
To measure subjective ambivalence, we asked participants to
indicate for each picture the extent to which they had conflicting
thoughts and feelings about it, were indecisive about it, and expe-
rienced mixed feelings, on 9-point scales ranging from not at all
(1) to very much (9) (Priester & Petty, 1996;%!.97). Finally,
because we wanted to control for processing difficulty influencing
arousal levels, we included four items that asked participants to
rate the degree to which they thought the picture was hard to
understand, had multiple meanings, was interpretable in more than
one way, and was complicated, on a 9-point scale ranging from not
at all (1) to very much (9) to measure complexity of the pictures
(%!.96).
Procedure
Each picture was presented for 6 s (following the IAPS manual;
Lang et al., 2008), after which participants rated the picture on
valence and arousal using the SAM scales. After this, all pictures
were presented again and rated for objective ambivalence, subjec-
tive ambivalence, and complexity. Finally, participants were re-
warded and debriefed.
Results
Because of a technical error, Picture 7004 was omitted and
picture 7003 displayed twice. Excluding these pictures leaves 29
stimuli in our dataset.
2
For each picture, we computed mean scores
of valence, arousal, objective and subjective ambivalence, and
complexity. For the regression analyses below, we computed stan-
dardized Zscores to facilitate interpretation.
Recall that we hypothesized that apparently neutral scores may
mask underlying ambivalence. Confirming this expectation, the
neutral IAPS pictures varied in objective as well as subjective
ambivalence (see Table 1). The objective ambivalence scores had
a mean of 3.76, a standard deviation of 1.09, and a range from 1.80
to 5.78. The subjective ambivalence scores had a mean of 2.99, a
standard deviation of 0.68, and a range from 1.88 to 4.52. Scores
on objective and subjective ambivalence were in accordance with
previous studies in which ambivalence was experimentally manip-
ulated (e.g., Gillebaart, Schneider, & De Ridder, 2015;Nohlen,
van Harreveld, Rotteveel, Lelieveld, & Crone, 2014;Schneider et
al., 2013,2015;van Harreveld, Rutjens, et al., 2009;van Harreveld
et al., 2014), supporting the notion that these pictures elicit sub-
stantive levels of conflict. The standard deviation and range indi-
cate that the selected pictures, which were "0.5 scale points from
the neutral value of the IAPS scales, vary markedly in the extent
to which they elicit mixed feelings. These differing levels of
ambivalence are mirrored in the arousal ratings, which ranged
2
Coding error caused picture 7424 to be displayed instead of 7185.
However, due to the relatively neutral rating of 7424 (mean valence !
5.28) excluding this item did not alter the pattern of results, and was thus
included in all analyses.
Table 1
Means (SDs) for Objective and Subjective Ambivalence, for
Each of the Pictures
Description Number Objective ambivalence Subjective ambivalence
Skyscraper 7640 5.78 (3.00) 3.50 (2.50)
Boy 9070 5.73 (2.81) 4.52 (2.60)
Erotic female 4302 5.49 (2.78) 3.77 (2.60)
Hammer 7034 5.30 (2.75) 2.80 (2.14)
Male face 2220 5.09 (2.34) 3.69 (2.25)
Twins 2890 5.04 (2.55) 3.70 (2.15)
Erotic male 4550 4.68 (3.11) 3.03 (2.20)
Wolf 1645 4.49 (2.84) 4.06 (2.37)
Ramen 7476 4.21 (3.10) 2.49 (1.99)
Fan 7020 4.19 (2.57) 2.71 (2.08)
Smoking 2749 4.14 (2.29) 3.61 (2.48)
Man 2002 3.83 (2.36) 2.77 (1.96)
Abstract art 7185 3.80 (2.29) 3.36 (2.38)
Pole 7161 3.78 (2.33) 3.74 (2.47)
Rubberbands 7012 3.58 (2.62) 2.26 (1.84)
Rolling pin 7000 3.41 (3.08) 2.46 (2.03)
Mug 7035 3.38 (2.74) 2.13 (1.81)
Erotic male 4561 3.29 (2.66) 2.71 (1.98)
Fish 7484 3.15 (3.31) 2.41 (2.05)
Men 2397 3.15 (2.70) 2.88 (2.05)
NeutMan 2214 3.10 (2.36) 3.02 (2.13)
Baskets 7041 3.01 (2.82) 2.47 (2.09)
Zipper 7045 2.91 (2.52) 2.08 (1.50)
Amerindian 2484 2.86 (2.72) 3.96 (2.32)
Towel 7002 2.64 (2.81) 1.88 (1.49)
Chair 7235 2.78 (2.67) 2.08 (1.59)
Battleship 9422 2.43 (2.81) 3.15 (2.27)
Fabric 7160 2.09 (2.72) 2.80 (2.10)
Building 7242 1.80 (3.08) 2.64 (1.98)
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
3
IAPS NEUTRALS MASK MIXED FEELINGS
Fn2
AQ:3, T1
AQ: 4
AQ: 5
tapraid5/emo-emo/emo-emo/emo00416/emo3124d16z
xppws S!1 2/24/16 1:36 Art: 2015-0884
APA NLM

from 1.54 to 6.15, with a mean of 3.19 and a standard deviation of
1.12.
As predicted, these levels of arousal were positively correlated
with objective ambivalence, r(29) !.43, p!.019, and subjective
ambivalence, r(29) !.49, p!.007; the more ambivalence the
pictures induced, the more arousal participants experienced. Ad-
ditionally, objective ambivalence was positively correlated with
subjective ambivalence, r(29) !.58, p!.001, which replicates
previous findings (e.g., Newby-Clark et al., 2002) and negatively
related to valence, r(29) !$.52, p!.004, showing that the more
ambivalent people were, the less positive they were. Finally,
complexity was positively correlated with subjective ambivalence,
r(29) !.88, p&.001. No other correlations were statistically
significant.
We further predicted that subjective ambivalence, because of its
experiential nature, would be the primary predictor of arousal. To
test this, we performed a multiple regression analyses with arousal
as the dependent variable and valence and objective and subjective
ambivalence as predictors. We also added complexity ratings to
control for influence of processing difficulty on arousal. As pre-
dicted, only subjective ambivalence showed a unique statistical
significant relationship with arousal (see Table 2), such that the
stronger the experience of ambivalence was, the more arousal
people felt in viewing the pictures. Complexity had a marginally
significant relationship with arousal, albeit negatively. The model
was statistically significant and explains 40% of the variance. All
predictors were added simultaneously and entering the predictors
in different orders did not change the outcome of these regression
analyses.
As an exploratory analysis we assessed the correlation between
the standard deviation of valence and subjective and objective
ambivalence and complexity, building on the assumption that
ambivalent pictures are rated more positively by some participants
and more negatively by others, resulting in increased standard
deviations. Results indicate that standard deviation scores of va-
lence ratings were positively correlated with objective, r(29) !
.40, p!.030 and subjective ambivalence, r(29) !.88, p!.055,
but not with complexity, r(29) !.02, p!.090, revealing that the
standard deviation valence ratings of IAPS pictures are moderately
informative about the amount of ambivalence the picture elicits.
Discussion
Despite selecting the most neutrally rated IAPS pictures, within
a".05 range around the 5.0 midpoint of the IAPS scale, our
findings show marked variation in the degree to which people
experience ambivalence when viewing these pictures. Moreover,
the differing levels of ambivalence were predictive of the arousal
viewers reported. This highlights that some of the supposedly
neutral IAPS pictures do not elicit neutral responses. Instead, the
apparently neutral rating sometimes masks mixed feelings. This is
unavoidable on bipolar rating scales that conceptualize neutral as
the midpoint between polar opposites—Participants who want to
express “a bit of both” cannot be distinguished from participants
who want to express “neither.” This is important for researchers
who select neutral IAPS pictures for use as a neutral comparison
condition in their research. Ambivalence has well-documented
specific consequences for affect and cognition, including feelings
of uncertainty, biased information processing, and increased illu-
sionary pattern perception (e.g., Sparks et al., 2001;van Harreveld
et al., 2014,2015). Unknowingly introducing these factors into a
research design may lead to many surprises and unwarranted
conclusions.
Our results showed a strong positive correlation between sub-
jective ambivalence and complexity, reflecting that the more com-
plex an attitude is (i.e., has more attributes), the more likely it is
also experienced as ambivalent (cf. van Harreveld, van der Pligt,
de Vries, Wenneker, & Verhue, 2004). Notably, in the regression
analyses, complexity seemed negatively related to arousal. Poten-
tially, this reflects that when variance related to complexity as a
result of ambivalence is removed, the remaining variance reflects
complexity due to having to express an attitude toward something
completely neutral. This may in itself be a complex endeavor, but
not one that is related to strong arousal.
With regards to our measure of objective ambivalence, one
might argue that participants interpreted the low ends of our
unipolar scales valence (not at all negative,not at all positive,
respectively) as the opposite of the high ends (i.e., not at all
positive implies negative), making the scale essentially bipolar,
which could have potentially inflated participants’ responses (Rus-
sell & Carroll, 1999), and consequently, levels of objective am-
bivalence. However, in relation to our unipolar scales, we in-
structed participants to ignore evaluations of the opposite valence
while making their judgments, which has been shown to lead to an
underestimation of objective ambivalence (Refling et al., 2013).
As such, we do not believe that our levels of objective ambivalence
are inflated per se, but future work may examine differences in
ambivalence depending on methodological approach. More to the
point, the goal of the current work is to show (i) that levels of both
objective and subjective ambivalence vary in neutral IAPS pictures
and (ii) that subjective ambivalence is positively related to expe-
rienced arousal.
Related to this, in our studies participants first rated all pictures
on the original IAPS scales, and only then rated ambivalence.
Although this makes conceptual sense, future work could rule out
any potential effects (e.g., fatigue) of this specific sequence by
counterbalancing the order in which these measures are adminis-
tered.
A straightforward way to identify neutral pictures is to use
multiple unipolar scales, thus allowing respondents to express their
mixed feelings (as used in our objective ambivalence measure, cf.
Kaplan, 1972;Refling et al., 2013;Thompson et al., 1995). For the
pictures used in the present study, Table 1 presents the scores for
objective and subjective ambivalence. Because subjective ambiv-
Table 2
Predicting Arousal From Valence, Complexity, Objective, and
Subjective Ambivalence Using Multiple Regression Analyses, All
Predictors Entered Simultaneously
Predicting arousal beta t p
Valence .18 .97 .344
Objective ambivalence .1 .38 .706
Subjective ambivalence 1.14 2.59 .016
Complexity $.75 $1.99 .058
Model F(4, 24) !4.00, p!.013
Explained variance (R
2
)40%
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
4SCHNEIDER ET AL.
AQ:6,
T2
AQ: 7
tapraid5/emo-emo/emo-emo/emo00416/emo3124d16z
xppws S!1 2/24/16 1:36 Art: 2015-0884
APA NLM

alence is dependent on both the presence of negative and positive
evaluations as well as on situational factors (e.g., Nordgren et al.,
2006;Schneider et al., 2013;van Harreveld, Rutjens, et al., 2009)
and individual differences (e.g., Newby-Clark et al., 2002), objec-
tive ambivalence scores represent a relatively stable representation
of the structure of the evaluations. A selection based on these
scores reduces the risk that effects observed under supposedly
neutral control conditions are contaminated by the cognitive and
affective consequences of ambivalence.
References
Gillebaart, M., Schneider, I. K., & De Ridder, D. T. D. (2015). Effects of
trait self-control on response conflict about healthy and unhealthy food.
Journal of Personality. Advance online publication. http://dx.doi.org/10
.1111/jopy.12219
Has, R. G., Katz, I., Rizzo, N., Bailey, J., & Moore, L. (1992). When racial
ambivalence evokes negative affect, using a disguised measure of mood.
Personality and Social Psychology Bulletin, 18, 786 –797. http://dx.doi
.org/10.1177/0146167292186015
Ito, T. A., Cacioppo, J. T., & Lang, P. J. (1998). Eliciting affect using the
International Affective Picture System: Trajectories through evaluative
space. Personality and Social Psychology Bulletin, 24, 855– 879. http://
dx.doi.org/10.1177/0146167298248006
Kaplan, R. (1972). Augmented transition networks as psychological mod-
els of sentence comprehension. Artificial Intelligence, 3, 77–100. http://
dx.doi.org/10.1016/0004-3702(72)90043-4
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2008). International
Affective Picture System (IAPS): Affective ratings of pictures and in-
struction manual (Tech. Rep. No. A-8). Gainesville, FL: University of
Florida.
Larsen, J. T., McGraw, A. P., & Cacioppo, J. T. (2001). Can people feel
happy and sad at the same time? Journal of Personality and Social
Psychology, 81, 684 – 696. http://dx.doi.org/10.1037/0022-3514.81.4
.684
Newby-Clark, I. R., McGregor, I., & Zanna, M. P. (2002). Thinking and
caring about cognitive inconsistency: When and for whom does attitu-
dinal ambivalence feel uncomfortable? Journal of Personality and So-
cial Psychology, 82, 157–166. http://dx.doi.org/10.1037/0022-3514.82.2
.157
Nohlen, H. U., van Harreveld, F., Rotteveel, M., Lelieveld, G. J., & Crone,
E. A. (2014). Evaluating ambivalence: Social-cognitive and affective
brain regions associated with ambivalent decision-making. Social Cog-
nitive and Affective Neuroscience, 9, 924 –931.
Nordgren, L. F., van Harreveld, F., & van der Pligt, J. (2006). Ambiva-
lence, discomfort, and motivated information processing. Journal of
Experimental Social Psychology, 42, 252–258. http://dx.doi.org/10
.1016/j.jesp.2005.04.004
Priester, J. R., & Petty, R. E. (1996). The gradual threshold model of
ambivalence: Relating the positive and negative bases of attitudes to
subjective ambivalence. Journal of Personality and Social Psychology,
71, 431– 449. http://dx.doi.org/10.1037/0022-3514.71.3.431
Rees, L., Rothman, N. B., Lehavy, R., & Sanchez-Burks, J. (2013). The
ambivalent mind can be a wise mind: Emotional ambivalence increases
judgment accuracy. Journal of Experimental Social Psychology, 49,
360 –367. http://dx.doi.org/10.1016/j.jesp.2012.12.017
Refling, E. J., Calnan, C. M., Fabrigar, L. R., MacDonald, T. K., Johnson,
V. C., & Smith, S. M. (2013). To partition or not to partition evaluative
judgments comparing measures of structural ambivalence. Social Psy-
chological and Personality Science, 4, 387–394. http://dx.doi.org/10
.1177/1948550612460060
Russell, J. A., & Carroll, J. M. (1999). On the bipolarity of positive and
negative affect. Psychological Bulletin, 125, 3–30. http://dx.doi.org/10
.1037/0033-2909.125.1.3
Schneider, I. K., Eerland, A., van Harreveld, F., Rotteveel, M., van der
Pligt, J., van der Stoep, N., & Zwaan, R. A. (2013). One way and the
other: The bidirectional relationship between ambivalence and body
movement. Psychological Science, 24, 319 –325. http://dx.doi.org/10
.1177/0956797612457393
Schneider, I. K., van Harreveld, F., Rotteveel, M., Topolinski, S., van der
Pligt, J., Schwarz, N., & Koole, S. L. (2015). The path of ambivalence:
Tracing the pull of opposing evaluations using mouse trajectories. Fron-
tiers in Psychology, 6, 996. http://dx.doi.org/10.3389/fpsyg.2015.00996
Sparks, P., Conner, M., James, R., Shepherd, R., & Povey, R. (2001).
Ambivalence about health-related behaviours: An exploration in the
domain of food choice. British Journal of Health Psychology, 6(Pt. 1),
53– 68. http://dx.doi.org/10.1348/135910701169052
Thompson, M. M., Zanna, M. P., & Griffin, D. W. (1995). Let’s not be
indifferent about (attitudinal) ambivalence. Attitude strength: Anteced-
ents and consequences, 4, 361–386.
van Harreveld, F., Nohlen, H. U., & Schneider, I. K. (2015). The ABC of
ambivalence: Affective, behavioral, and cognitive consequences of atti-
tudinal conflict. In M. P. Zanna & J. Olson (Eds.), Advances in exper-
imental social psychology (Vol. 52, pp. 285–324). New York, NY:
Academic Press. http://dx.doi.org/10.1016/bs.aesp.2015.01.002
van Harreveld, F., Rutjens, B. T., Rotteveel, M., Nordgren, L. F., & van der
Pligt, J. (2009). Ambivalence and decisional conflict as a cause of
psychological discomfort: Feeling tense before jumping off the fence.
Journal of Experimental Social Psychology, 45, 167–173. http://dx.doi
.org/10.1016/j.jesp.2008.08.015
van Harreveld, F., Rutjens, B. T., Schneider, I. K., Nohlen, H. U., &
Keskinis, K. (2014). In doubt and disorderly: Ambivalence promotes
compensatory perceptions of order. Journal of Experimental Psychol-
ogy: General, 143, 1666 –1676. http://dx.doi.org/10.1037/a0036099
van Harreveld, F., van der Pligt, J., & de Liver, Y. N. (2009). The agony
of ambivalence and ways to resolve it: Introducing the MAID model.
Personality and Social Psychology Review, 13, 45– 61. http://dx.doi.org/
10.1177/1088868308324518
van Harreveld, F., van der Pligt, J., de Vries, N. K., Wenneker, C., &
Verhue, D. (2004). Ambivalence and information integration in attitu-
dinal judgment. British Journal of Social Psychology, 43(Pt. 3), 431–
447. http://dx.doi.org/10.1348/0144666042037971
Received April 16, 2015
Revision received December 22, 2015
Accepted December 30, 2015 !
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
5
IAPS NEUTRALS MASK MIXED FEELINGS
AQ: 8
tapraid5/emo-emo/emo-emo/emo00416/emo3124d16z
xppws S!1 2/24/16 1:36 Art: 2015-0884
APA NLM