THE MEDIATING EFFECT OF EMOTIONS ON THE COMPARISON BETWEEN
DESCRIPTIVE NORM INFORMATION AND BEHAVIORS: THE CASE OF RECYCLING

By
Hye Seung Koh

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Communication – Master of Arts
2013

ABSTRACT
THE MEDIATING EFFECT OF EMOTIONS ON THE COMPARISON BETWEEN
DESCRIPTIVE NORM INFORMATION AND BEHAVIORS: THE CASE OF RECYCLING

By
Hye Seung Koh
Based on the focus theory of normative conduct, discrete emotion approaches and the
heuristic-systematic model, this study examined the effects of descriptive norms message and
prior recycling behavior on affective responses, cognitive responses, attitudes, and intentions
toward recycling. It was predicted that the effect of descriptive norm messages on young
people’s feelings of satisfaction and guilt, number and valence of thoughts, attitudes, and
behavioral intentions toward recycling would be moderated by prior recycling behavior. It was
also predicted that the descriptive norm message – attitude/intention relationship is mediated by
emotional responses which result from a discrepancy between prior recycling activity and the
descriptive norm message. In the main experiment, one hundred-eighty-eight college students
were exposed to either a high prevalence or a low prevalence message. Prior recycling behavior,
aroused emotions, message processing, attitude, and intention to recycling were measured to see
if any differences existed as a function of the individual’s prior recycling behavior and
descriptive norm message exposure. Overall, results indicated that prior recycling moderated
only the relationship between descriptive norms messages and intention to recycling. Descriptive
norms message was a significant predictor of feelings of satisfaction while prior recycling
behavior was a significant predictor of feelings of guilt. There was no significant difference in
the number of thoughts across experimental condition. There was a significant difference
between recyclers in the high prevalence condition and those in the low prevalence condition in

valence of thoughts such that recyclers in the high prevalence condition had greater number of
positive thoughts while recyclers in the low prevalence condition had fewer number of positive
thoughts.

Key-word: descriptive norm messages, prior behavior, emotion, cognitive processing,
intention to recycling.

TABLE OF CONTENTS

LIST OF TABLES ...................................................................................................................... vi
LIST OF FIGURES ................................................................................................................... vii
INTRODUCTION ........................................................................................................................1
LITERATURE REVIEW .............................................................................................................3
The Focus Theory of Normative Conduct (FTNC) ............................................................3
Previous Literature: Persuading People to Recycle ............................................................6
Emotions and Social Norms..............................................................................................10
HYPOTHESES AND RESEARCH QUESTIONS ....................................................................15
METHOD ...................................................................................................................................25
Design Overview ..............................................................................................................25
Stimulus Material ..............................................................................................................25
Pilot Study.........................................................................................................................26
Main Experiment ..............................................................................................................28
Participants..............................................................................................................28
Procedure ................................................................................................................28
Measures .................................................................................................................29
Independent Variables ..................................................................................29
Prior Recycling Behavior ....................................................................29
Dependent Variables .....................................................................................30
Aroused Emotions...............................................................................30
Happiness ..................................................................................30
Pride ..........................................................................................30
Satisfaction................................................................................30
Guilt ..........................................................................................31
Other Negative Emotions ..........................................................31
Thoughts about the Video Message ......................................................31
Attitudes ................................................................................................33
Behavioral Intent...................................................................................33
Manipulation Check ......................................................................................34
Perceived Descriptive Normative Content in the Message.................34
Perceived Prevalence of Recycling among MSU Students ................34
Other Message Factors........................................................................34
Control Variables ..........................................................................................34
Issue Involvement ...............................................................................34
Social Desirability...............................................................................34
Demographic Information ...................................................................35
iv

RESULTS ...................................................................................................................................36
Pilot Study.........................................................................................................................36
Main Experiment ..............................................................................................................37
Preliminary Analyses ..............................................................................................37
Main Analyses ........................................................................................................40
DISCUSSION .............................................................................................................................48
THEORETICAL IMPLICATIONS ............................................................................................55
LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH......................................57
FOOTNOTES .............................................................................................................................60
APPENDICES ............................................................................................................................62
Appendix A: Tables and Figures ......................................................................................63
Appendix B: Stimuli .........................................................................................................79
Appendix C: Questionnaire Items for Pilot Study ............................................................80
Appendix D: Informed Consent Form ..............................................................................83
Appendix E: Questionnaire for Main Experiment - Pre-test ............................................84
Appendix F: Questionnaire for Main Experiment - Post-test ...........................................88
REFERENCES .........................................................................................................................100

v

LIST OF TABLES

Table 1. Conceptual frameworks of discrete positive emotions .................................................64
Table 2. Coding variables, definitions and sample messages from the participants’ answers ...65
Table 3. Number of items retained, means, standard deviations, and alphas for all scales across
conditions ....................................................................................................................................66
Table 4. Correlation among variables .........................................................................................68
Table 5. Summary of hypotheses and research questions...........................................................69
Table 6. Mean and standard deviation, t-value, sig. and η2 for satisfaction and other positive
emotions among recyclers...........................................................................................................71
Table 7. Mean, standard deviation, t-value, and sig. for guilt and other negative or neutral
emotions among non- recyclers ..................................................................................................72
Table 8. The association between the valence of thoughts and experimental conditions...........73
Table 9. Regression results for attitude toward recycling...........................................................74
Table 10. Mean and standard deviation for attitude toward recycling across different
experimental conditions ..............................................................................................................75
Table 11. Regression results for intention to recycle..................................................................76

vi

LIST OF FIGURES

Figure 1. The association between the valence of thoughts and the experimental conditions ...77
Figure 2. The effects of the descriptive norm messages on intention to recycle depending on prior
recycling behavior.......................................................................................................................78

vii

INTRODUCTION
Socially desirable behaviors are those behaviors that the majority of people in society
think are the right things to do for the collective, although an individual might not directly
benefit from performing them (Cialdini, Kallgren, & Reno, 1991; Kallgren, Reno, & Cialdini,
2000). Recycling is a socially desirable behavior since most people acknowledge the necessity of
recycling although it is personally costly (Smith, Haugtvedt, & Petty, 1994). Some people still do
not recycle despite growing public concern about recycling and the recognition of the importance
of recycling (Harrison Poll, 2007). Efforts have been made to improve recycling rates using
social norms campaigns, however, some of them have cautioned that informing non-recyclers of
the prevalence of recycling or the descriptive norm, may lead to arousal of negative emotions
such as guilt (Dannenberg, Hausman, Lawrence, & Powell, 2012).
Studies have shown that pro-social behaviors including recycling can be predicted by the
perception of social prevalence or approval of the desirable behavior (Cialdini, 2003; Kaiser &
Shimoda, 1999). Such perceived public support is essential to social movements (Stern, Dietz,
Abel, & Kalof, 1999). Many studies have examined the cognitive mechanisms of how perceived
normative information influences attitudes, intention, and behavior (Jacobson, Mortensen, &
Cialdini, 2011; Rimal, 2008). However, there is little literature exploring the role of emotions in
the persuasive power of normative information. This study attempts to fill this gap by exploring
the emotions that emerge as a result of the discrepancy between one’s previous behavior and the
prevalence of a behavior described in norms messages; it also looks at the effects of emotioneliciting normative messages on attitudes and behaviors.
Focus theory of normative conduct (FTNC) (Cialdini et al., 1991; Cialdini, Reno, &
Kallgren, 1990) provides a theoretical framework to examine the effects of perceived social

1

norms to improve recycling among young people. According to the theory, when social norms
are salient, individuals behave based on their observation of what the majority does. This is
because the popularity of a particular behavior is an indicator of an effective or helpful behavior
in a given context (Cialdini, 2011; Goldstein & Cialdini, 2011).
Based on focus theory and the research on the role of discrete emotions in persuasion,
this study predicts the effects of descriptive normative information on young people’s attitudes
and behavioral intentions toward recycling. It is also predicted that the normative information –
attitude/intention relationship is mediated by emotional responses (satisfaction or guilt) which
result from a discrepancy between prior recycling activity and normative information. Thus, this
study attempts to explore the underlying cognitive and affective mechanisms of how normative
information influences people’s behaviors through arousing emotions. This paper begins with a
review of relevant literature and then hypotheses and research questions are proposed. To test the
hypotheses and to answer the research questions, the method for an experiment is detailed. Next,
in the results section, preliminary data analyses are described and the findings from the
experiment are addressed. Finally, the implications of the results are discussed along with the
limitations of the study and suggestions for future research.

2

LITERATURE REVIEW
The Focus Theory of Normative Conduct (FTNC)
Social science researchers have discussed the role of social norms in influencing
behavior and put effort into refining the definition of norms (Pepitone, 1976; Schaffer, 1983).
Cialdini and Trost (1998) define norms as standards which describe and guide behavior. Rimal
and Real (2003) conceptualize norms as the codes of conduct corresponding to one’s group,
which are communicated through social interaction. Social norms can be defined as social
standards which inform and guide behavior via social interaction (Rimal & Real, 2003). These
behavioral standards are understood, created, and maintained through people’s cognitive and
affective processing of nonverbal and verbal messages through observation or communication.
According to focus theory of normative conduct (FTNC), social norms can be
categorized into two distinct types, descriptive and injunctive norms, both of which influence
behaviors (Cialdini et al., 1991; Cialdini et al., 1990; Kallgren et al., 2000). Descriptive norms
refer to what is commonly done in a society while injunctive norms refer to that which is
commonly approved and disapproved.

A primary difference between descriptive norms and

injunctive norms is social sanctions; descriptive norms do not entail social sanctions for noncompliance with the norm (Cialdini et al., 1990). That is to say, descriptive norms motivate
individuals to behave by informing them of what is likely to be an effective or helpful behavior
in the society while injunctive norms motivate via real or anticipated social punishment. Cialdini
(2011) argued that the components of descriptive norms and injunctive norms in a message
should work in different ways because of the distinct motivational mechanism of each norm.
Descriptive norms are related to intrapersonal goals (e.g., making accurate/efficient
decisions) which can be achieved by applying the heuristic rule ‘most others do it so I will too’,

3

through observing others’ behaviors (Cialdini & Trost, 1998). On the other hand, injunctive
norms are more closely associated with interpersonal goals (e.g., gaining/maintaining social
approval) which demands an understanding of the culture’s moral rules, that is, what others tend
to approve or disapprove (Goldstein & Cialdini, 2011). For example, Goldstein, Cialdini, and
Griskevicius’s study (2008, Experiment 1) showed that using normative influence (i.e.,
informing targets about what the majority has been doing to protect the environment), it was
more effective to persuade people to reuse a hotel towel compared to just providing arguments
for the importance of environmental protection in a standard message. Specifically, the results
indicated that 44% of people who received the message including descriptive normative
information reused the hotel towel while 35% of people who received the standard message
without descriptive normative information reused it. This shows that it is persuasive to inform
people that the majority is following a particular desirable behavior. Also, a study conducted by
Cialdini et al., (1990) indicated that participants littered in an already littered environment (i.e.,
showing the prevalence of littering) rather than in a clean environment since the descriptive
norms regarding pro-littering was more prominent than the anti-littering one. From these results
it can be inferred that people are more likely to perform a certain behavior when they perceive
the behavior as socially prevalent regardless of whether the behavior is socially desirable or not.
In addition to the importance of distinguishing between the two different types of social
norms, FTNC addresses the issue of the salience of normative effects; explaining why normative
influence occurs. When descriptive norms or injunctive norms are salient, normative influence is
more likely to occur and people will engage in norms-congruent behavior (Cialdini, 2011;
Cialdini et al., 1991; Goldstein & Cialdini, 2011). Salient norms are those which are focal to
people. Enhancing salience occurs through highlighting or making people focused on the norms

4

and can occur when an individual views a message including normative information or where an
individual has distinctive dispositional factors activating a particular norm (Goldstein & Cialdini,
2011; Jacobson et al., 2011; Kallgren et al., 2000).

According to FTNC, salient norms can

activate behavior, whereas non-salient or less salient norms cannot. People should be more likely
to perform norm-congruent behaviors when the norm within a message is prominently
emphasized (Cialdini & Goldstein, 2004).
Several previous studies support the FTNC effects for descriptive norms. For example,
Cialdini et al., (1990) manipulated descriptive norms and salience to determine their effects on
littering. Specifically, these researchers manipulated the descriptive norms into either a dirty
environment (i.e., indicating the high prevalence of littering) or a clean environment (i.e.,
indicating the low prevalence of littering). Normative salience was manipulated as high or low
by including a confederate. In the high-norm salience condition, the confederate dropped garbage
into either a dirty or clean environment while in the low salience condition, (s)he just walked by
a garbage can in either a dirty or clean environment without littering. The results showed that
participants in the littered environment with high-norms salience were most likely to litter than
others following the salient high descriptive norms. On the other hand, participants who were in
the clean environment with high-norms salience condition littered the least because it was salient
to them that littering was as an unpopular behavior. That is to say, the stark contrast between the
clean environment and the confederate’s behavior (littering), led to norms-salience effects. These
findings could be interpreted as indicating that people tended to perform certain behaviors when
they were exposed to descriptive normative information showing that a particular behavior was
socially prevalent (Goldstein et al., 2008) because it enhances the salience of norms. Importantly,

5

a number of studies have been designed to understand the motivators of recycling; several of
which address the role of social norms in this process.
Previous Literature: Persuading People to Recycle
There have been a number of studies designed to improve recycling rates, some of which
have addressed the role of social norms in recycling. For example, Park, Levine, and Sharkey
(1998) applied theory of reasoned action (TRA; Fishbein & Ajzen, 1975) and self-construal to
predict one’s motivation toward recycling. The two predictors of intention toward recycling,
attitude and subjective norms, had a significant effect on behavioral intention toward recycling.
However, independent self-construal and interdependent self-construal had no direct effect on
behavioral intention toward recycling. Theory of planned behavior (TPB; Ajzen, 1991) has also
been used to predict paper recycling in an inter-cultural context: Thailand and the U.S.
(Chaisamrej & Zimmerman, 2007). In their study, personal values, altruism and cultural value
(self-construal) were added to three major components of TPB, attitude, subjective norms, and
perceive behavioral control (PBC). They found that subjective norms and PBC were the
predictors of one’s intention to engage in paper recycling for the two different cultures
depending on the level of altruism. For both countries, participants who had high levels of
altruism were more likely to intend to recycle than those who had low levels of altruism.
Furthermore, one’s altruistic value predicted attitude and PBC toward recycling. On the other
hand, the attitude component had an effect on only the U.S. participants’ intention toward paper
recycling.
Although previous findings have demonstrated the effectiveness of the TRA and TPB
concepts, including subjective norms, to predict behavioral intention, many studies indicate the
need to improve the predictive power of these theories by including factors such as affective

6

states (i.e., anticipated regret) and moral norms (Conner & Armitage, 1998; Parker, Manstead, &
Stradling, 1995; Rivis, Sheeran, & Armitage, 2009), descriptive norms and group identification
(Norman, Clark, & Walker, 2005), and personal- and societal-levels of social norms (Park &
Smith, 2007). Some of this research (Norman et al., 2005; Parker et al., 1995; Rivis et al., 2009)
indicates that the inclusion of additional normative (i.e., moral norm and descriptive norms) and
affective (i.e., anticipated regret) components in the TPB substantially improved prediction of
behavioral intentions.
In addition to normative influence, previous literature also has explored various ways to
persuade people to actively engage in recycling by varying the nature of the messages used to
influence people. For example, recent research by Kronrod, Grinstein, and Wathieu (2012)
compared the relative persuasive effects between assertive message appeals and suggestive
message appeals. Previous studies indicated assertive requests would violate one’s drive for
freedom (Dillard & Shen, 2005; Dillard, Steven, James, & Terry, 1997). Consequently, Kronrod
et al., (2012) predicted assertive appeals would be less effective in persuading people to engage
in pro-social behavior. However, they found that the effect of assertive appeals depended on
one’s level of perceived importance of the issue of recycling. In other words, the assertive appeal
was effective among people with higher levels of perceived importance of recycling while the
suggestive phrasing had persuasive power for people who attributed lower levels of importance
to the behavior. Although this study did not measure participants’ affective responses to different
message appeals, it might be reasonable to assume that if people strongly believe that recycling
is the right thing to do, they might rarely perceive the violation of their freedom when they are
exposed to the assertive appeal message because the assertive phrasing corresponds with their
beliefs.

7

Loroz (2007) also examined the interaction between message frames (positive vs.
negative) and reference points (self vs. self-other) on attitudes and intentions toward prosocial
behaviors including recycling and prevention of HPV. The results illustrated that participants
who received a positively framed message paired with self-other reference point or a negatively
framed message paired with self- reference point had more positive attitudes and intentions
toward recycling compared to others who received unmatched messages. White, MacDonnell,
and Dahl (2011) showed similar message effects between gain/loss-framed messages and one’s
construal level. Specifically, when the loss-framed message (i.e., one that describes recycling as
a way to avoid losing/wasting resources) was matched with concrete construal, and the gainframed message (i.e., one that describes recycling as a way to save/gain resources) was matched
with abstract construal, their persuasive effects were greater than when the messages were
unmatched. Further, they explored the role of mediators of 1) processing fluency and 2)
increased efficacy in explaining matching effect on intention toward recycling. According to
their results, individuals who received matched messages were more likely to process messages
easily relative to those who received unmatched messages; this led to enhanced self-efficacy.
Consequently, this process generated greater intention to be active recyclers. Findings from these
studies (Kronrod et al., 2012; Loroz, 2007; White et al., 2011) provide guidelines for how to
create effective persuasive messages to persuade people to recycle considering the influence of
individual differences.
There is some research to suggest that beyond the social-psychological drivers of
behaviors described above, emotional response to messages may play a role in motivating
recycling behaviors. In a study designed to describe young people’s recycling behaviors,
Dannenberg et al., (2012) conducted in-depth interviews with college students. These interviews

8

indicated that students recognized the importance of environmental issues and showed intentions
to engage in sustainable behaviors including recycling. At the same time, students expressed the
lack of time and resources, which resulted from conflicting demands. The authors argued that
social norms might have little effect on behavioral change because young people already know
that recycling is necessary, but they rarely recycled because of the lack of time. They concluded
that young people might feel guilty when they are informed that recycling is the norm because of
their own inaction. Accordingly, the present study takes into account this possibility by exploring
the effect of discrepancy between one’s prior recycling activity and the normative information
indicating the prevalence of recycling on young people’s attitude and intention toward recycling,
which is possibly mediated by emotional mechanisms, specifically: satisfaction and guilt.
In the current study, it is expected that the discrepancy between an individual’s previous
recycling behavior and the prevalence of recycling portrayed in a message would differently
influence one’s attitude and intention toward recycling through emotional responses. Little is
known about the role of emotion generated by appraisal of descriptive normative information in
changing people’s attitudes and behavioral intentions despite the fact that emotional response to
messages is an important motivator of individual behavior (Nabi, 2010). Accordingly, this study
attempts to explore the role of emotions as mediators of the discrepancy between one’s prior
behavior and descriptive normative information, and these effects on attitude and intention. The
following section reviews in detail two approaches to studying the effects of emotions on
persuasion with a particular focus on two emotions thought to result from social norms messages:
guilt and satisfaction.

9

Emotions and Social Norms
Emotion, a psychological construct, can be defined as “internal mental states
representing evaluative valenced reactions to events, agents, or objects that vary in intensity”
(Nabi, 2010, p. 153). Approaches to emotion can be broadly categorized into two perspectives:
dimensional and discrete (Bolls, 2010; Lazarus, 1991; Nabi, 2010). Both approaches commonly
posit that emotion plays a significant role in persuasion. In the dimensional perspective, emotion
emerges from basic motivational processes, either appetitive or aversive motivational subsystems.
Through the process, the direction of emotion can range from a pleasant response to an
unpleasant response with different levels of arousal (Bolls, 2010). The effect of emotions on
attitude, intention and/or behavior change mainly depends on the direction and intensity of the
aroused emotion. However, several studies contended that a discrete emotional perspective
allows researchers to better understand the persuasion process compared to the dimensional,
which is simply based on valence (Dillard, Plotnick, Godbold, Freimuth, & Edgar, 1996; Dillard
& Nabi, 2006; Dillard & Peck, 2000; 2001; Nabi, 1999; 2002; 2010; O’Keefe, 2002). This is
because the discrete emotion perspective specifies differences between different emotions and
allows prediction of how people respond to each distinct emotion (Cornner & Armitage, 1998;
Griskevicius, Goldstein, Mortensen, Sundie, Cialdini, & Kenrick, 2009; Nabi, 2010).
According to the discrete perspective, emotions can be divided into discrete categories.
Each discrete emotion emerges from one’s evaluative appraisal of the relationship between one’s
goal and the environment (Nabi, 1999). Each emotion has a distinctive goal and its
corresponding action tendency to serve adaptive functions, which can influence attitudes,
intention, and/or behavior. The perceived discrepancy between the goal and environment
generates negative emotional states while the perception of compatibility of goal-environment

10

yields positive emotions (Dillard & Nabi, 2006). For example, guilt is aroused when an
individual violates a moral or ethical code against oneself or society, and thus another person
may be harmed by the individual’s action or inaction (Lazarus, 1991). Bierbrauer (1992)
contended that people feel guilty when they violate norms; when people feel guilty, they tend to
blame themselves through self-reflection. From the evolutionary perspective, this negative affect
needs to be alleviated through the corresponding action tendency such as reparation for the harm
or seeking punishment (Baumann, Cialdini, & Kendrick, 1981; Dillard & Peck, 2000; Kugler &
Jones, 1992; Nabi, 1999; 2010). The feeling of guilt is a form of self-sanction, which motivates
appropriate deeds and instills a sense of social responsibility in a constructive way. However, too
strong a feeling of guilt may arouse psychological resistance eliciting resentment, annoyance,
and irritation, which may lead to a decrease in the effect of persuasive message on attitude or
intention

(O’Keefe, 2002).

Previous studies indicate that guilt is motivational in that it encourages various prosocial behaviors including helping others (Boster, Mitchell, Lapinski, Cooper, Orrego, & Reinke,
1999; Carlsmith & Gross, 1969), giving donations (Basil, Ridgway, & Basil, 2006; 2008;
Hibbert, Smith, Davies, & Ireland, 2007; Massi, 2005), and getting health screening for others
(Hullett, 2004). In these studies, one’s perception of responsibility elicited feelings of guilt,
which, in turn, motivated people to perform an action that increased positive emotion thereby
reducing feelings of guilt.
Research has also explored the association between persuasion and positive discrete
emotions such as satisfaction, happiness and pride (De Young, 1986; Dillard et al., 1996; Dillard
& Peck, 2000; 2001, Lazarus 1991). Theories of positive emotion indicate that positive emotions
are attained when one’s goal and environment are compatible (Dillard & Nabi, 2006; Dillard &

11

Peck, 2001; Lazarus, 1991; Nabi, 1999; 2010; Shaver, Schwartz, Kirson, & O’Connor, 1987) and
the concept of discrete emotional response has dominated this paradigm. Yet, the identification
of discrete positive emotions, like negative emotions, is complex (see Table 1). For example,
Shaver et al., (1987) categorized human emotions into six basic categories (i.e., love, joy, anger,
sadness, fear, and surprise) and explored hierarchical structures of these basic human emotions.
Their analysis indicated that joy was a higher-order concept embracing sub-categories such as
pride, or satisfaction. According to their findings, people who succeeded in their tasks and had a
sense of achievement felt joy. They were also joyful when they were accepted by the
surrounding environment and had a feeling of belonging (Shaver et al., 1987). According to
Lazarus (1991), happiness as feeling people have when they make reasonable progress toward
their goals and when they expect positive outcomes, which is similar to satisfaction or
contentment. Pride was conceptualized as a feeling people have when they enhance their egoidentity (i.e., self- and social- esteem) by taking credit for an achievement for themselves or for
someone or groups with whom they identify (Lazarus, 1991). That is to say, the critical others
between happiness and pride is whether credit is given to oneself or to others.
Dillard and Peck (2000; 2001) conceptualized happiness and contentment/satisfaction
distinctively, and measured the two discrete positive emotions. They defined happiness as a
feeling people have when they achieve or make progress toward their goals and as a function of
self-rewarding behavior. Contentment or satisfaction was defined as a feeling people have when
they are absent from threats and as a function of conserving resources.
Among the discrete positive emotions, De Young (1985-1986; 1986; 1996) found that
intrinsic personal satisfaction was the most important motivational emotion for recycling.
Recyclers who actively participated in recycling were satisfied with their behaviors and

12

continued to perform them on their own since they wanted to maintain their positive emotions.
According to Oliver (1980), people tend to be satisfied with themselves when they have an initial
standard/goal and experience some positive discrepancy from the standard. Along with the role
of descriptive normative information as an indicator of an efficient or helpful behavior, it is
reasonable to assume that an individual would have his/her own standard in accord with
descriptive normative information and experience a discrepancy between his/her prior behavior
and the standard. Satisfaction, as a sub-category of happiness/joy, reflects those antecedents and
responses (Lazarus 1991; Shaver et al., 1987). Accordingly, because experiencing positive affect
is rewarding, when people experience positive affect such as satisfaction, they tend to maintain
or elevate their emotional states by performing behaviors that result in positive emotion.
In a study of recycling and emotional appeals, Lord (1994) explored the effect of
affective appeals (fear vs. satisfaction) and source types (advertising, publicity, and personalinfluence) on motivating people to engage in their community’s recycling program. In the study,
fear was manipulated by showing risks from the failure to recycle while satisfaction was
manipulated by indicating environmental benefits and personal and social satisfaction arising
from recycling. Also, each source type was manipulated by presenting the source directly. For
example, the advertising type was designed to appear as an ad for a fake company name. The
publicity was a type of news article, and the personal type was in the form of a letter. The results
showed that the positively framed satisfaction appeal message led to more favorable beliefs and
attitudes toward recycling. On the other hand, the negatively framed fear appeal message was
more effective to increase recycling behavior.
In their work examining the weak relationship between attitudes and behaviors for prosocial behaviors, Smith et al., (1994) argued for the importance of the consideration of the

13

affective aspects of such behaviors. In their survey, participants were asked to indicate their
feelings, both satisfaction and guilt when they do or do not recycle. The results showed that the
affect associated with recycling had greater power to predict one’s recycling behavior among
people who had weak attitudes toward recycling than among people with strong attitudes. Other
research on recycling provides evidence that one’s attitude, intention and/or behavior are closely
associated with affect. Meneses’ (2010) study indicated that one’s level of involvement with
recycling was more dependent on cognitive factors such as knowledge of recycling and
ecological conscience than emotions, while actual recycling behavior was more dependent on
emotion rather than cognition since people have directly or indirectly experienced what recycling
is and how to recycle (making cognitions less important). The finding implies that if people have
knowledge of how to recycle and self-efficacy about recycling behaviors, they may rely on
aroused emotion when they decide to recycle. This section has addressed the affective predictors
of recycling. Given the persuasive effects of descriptive normative information and emotions
discussed above, the current study makes predictions about their effects on attitude and intention
toward recycling in the following section.

14

HYPOTHESES AND RESEARCH QUESTIONS
The literature reviewed above suggests that recycling behaviors might be susceptible to
normative influence, but that emotions could be elicited from norms messages. Thøgersen (1996)
contended that ecologically-oriented behaviors such as recycling are categorized within the area
of morality. Although external incentives such as monetary benefits also have an effect on
recycling behavior, people are more likely to regard the conduct of recycling as beliefs about
right and wrong rather than comparing costs and benefits (Viscusi, Huber, & Bell, 2011). In this
way, recycling is a behavior guided by the norm that recycling is the morally correct behavior
(Cialdini et al., 1991; Kallgren et al., 2000). Previous studies examined the effect of moral norms
on recycling and found the predictive power of moral norms on intention to recycle (Kaiser,
Ranney, Hartig, & Bowler, 1999; Kaiser & Shimoda, 1999). These studies indicate that it may be
reasonable to regard recycling as a socially desirable activity. Thus, people are encouraged to
participate in recycling, yet not everyone does participate. To promote recycling, it might be not
enough to only argue the social desirability of the behavior.
According to FTNC, each type of social norm has a distinctive motivational mechanism;
descriptive norms motivate people to follow the normative behavior by showing them the
prevalence of the behavior while injunctive norms stimulate people to perform the normative
behavior by showing them the approved behavior (Goldstein & Cialdini, 2011). Descriptive
normative information indicating what the majority is doing plays the role of a guide for popular
and effective behavior. That is to say, descriptive normative information can be a standard or
goal when a person makes a decision about what s/he will do later. Given the role of descriptive
normative information as a standard, recyclers should experience satisfaction when they are
exposed to a descriptive norm message indicating low prevalence of recycling relative to when

15

they are exposed to the message indicating high prevalence of recycling. That is, when viewing
messages about low prevalence, recyclers will think that they have been performing better than
other people and are likely to subsequently feel satisfied.
According to Oliver (1993), satisfaction occurs when one’s positive expectations are
fulfilled or exceeded. Lazarus (1991) also contended that individuals will be satisfied with
themselves or their behaviors when they perform well to achieve their goals. That is, satisfaction
will be derived from a sense of achievement. It seems reasonable to expect that recyclers who
have actively engaged in recycling would be more satisfied with themselves when they are
informed that this behavior exceeds the social norm. Thus, recyclers who are exposed to a
message indicating low prevalence of recycling would be more satisfied with themselves than
recyclers who are exposed to a message indicating high prevalence of recycling.
On the contrary, it is predicted here that people who rarely recycle will experience guilt
when they are exposed to a message indicating high prevalence of recycling relative to when
they are exposed to a message indicating low prevalence of recycling. This is because the
message highlights the fact that they have failed to perform a behavior that the majority is
actively performing. Baumeister, Stillwell, and Heatherton (1994) argued that guilt is aroused by
the perception that one has performed non-normatively or that one has failed to behave
normatively. Considering the nature of recycling behavior in a society, and the causes and
processes which result in emotional consequences, the current study predicts:
H1: The effect of descriptive norm messages on satisfaction will be moderated by prior
recycling behavior such that recyclers, relative to non-recyclers, will experience the most
satisfaction in the low versus high prevalence message condition.

16

H2: The effect of descriptive norm messages on guilt will be moderated by prior
recycling behavior such that non-recyclers, relative to recyclers, will experience the most guilt in
the high versus low prevalence message condition.
Research on discrete emotions indicates a number of different potential categories of
positive emotion that might be driven by norms messages; additional research is needed to
clarify the structures and relationships among discrete positive emotions (Bagozzi, 1991). In
order to address this issue, the present study includes measures of multiple discrete positive and
negative emotions and explores whether satisfaction is the positive emotion most closely related
to recycling among recyclers relative to other positive emotions including happiness and pride.
Similarly, it seeks to determine whether guilt is the primary emotion elicited by the high
descriptive norms message relative to other negative emotions including fear and anger. To help
better understand the characteristics of recycling as well as of the relationships among the
discrete positive emotions, this study poses the following research questions:
RQ1: Do recyclers experience greater satisfaction than other positive emotions when
they are exposed to a message indicating low prevalence?
RQ2: Do non-recyclers experience greater guilt than other negative or neutral emotions
when they are exposed to a message indicating high prevalence?
Conner and Armitage (1998) contend that past behavior (i.e., the consumption of sweet
and fried foods, smoking, and exercise) is a powerful predictor of a future behavior. Research
has found that prior behavior predicts future behavior, but it remains relatively unclear the
process by which one’s past behavior impacts one’s future behavior. For example, Aarts,
Verplanken, and van Knippenberg (1998), in a study of transportation preferences, showed that
habitual past behaviors (i.e., frequent performance) were more closely related to simplified

17

decision making processes rather than rational information processing. This means that people
were likely to perform the same behavior they conducted in the past, such as using the same
mode of transportation, rather than thinking carefully about new information about it. Past
behavior can be an influential source of information in some cases; however, its influence
depends on the nature of the behavior (Aarts et al., 1998). Accordingly, in an effort to extend the
explanatory power of prior behavior as a source of information, this study examines the extent to
which

prior behavior plays a role in triggering information processing in response to normative

messages about recycling behavior. It is predicted here that when people are exposed to
normative messages indicating the prevalence of recycling, they compare their past performance
with other people’s behaviors.
According to the heuristic-systematic model (HSM) of persuasion, there are two distinct
routes of information processing: systematic processing and heuristic processing (Chaiken, 1987;
Chaiken, Liberman, & Eagly, 1989; Zuckerman & Chaiken, 1998). HSM posits that two routes
of information processing can occur simultaneously with one predominating in some cases.
Systematic processing involves careful and effortful consideration of information and results in
more thinking about the message; motivation and/or ability drive people to carefully scrutinize
the message content. For example, when people perceive a message as relevant or important to
them, they are more likely to embark on systematic processing. Subsequently, people who
primarily process a message systematically tend to generate more thoughts about issue-relevant
arguments than those who heuristically process it because they analyze the given message more
carefully. On the other hand, heuristic processing occurs when people depend on simple decision
rules without analytical and effortful scrutiny of information.

18

Dillard and Nabi (2006) contend that emotions can have a direct effect on information
processing through their distinct action tendency of each discrete emotion but also that emotions
may work with cognitions during the processing. That is, affect may serve as an additional
source of information. Previous studies have provided mixed explanations regarding the role of
affect in information processing. Some studies have shown that people in a positive mood
engage in increased systematic processing in order to maintain the positive mood while other
studies have shown that those in a positive mood engage in reduced systematic processing
because the cognitive capacity was reduced while activating memories that produce positive
feelings (Zuckerman & Chaiken, 1998). Accordingly, Wegener, Petty, and Smith (1995)
specified the relationship between positive mood and information processing under certain
conditions; people in a positive mood tend to engage in systematic processing when they process
uplifting messages but tend to engage in less systematic processing when processing depressing
messages. Along with these findings, it is reasonable to predict that recyclers who have actively
engaged in recycling will be more likely to engage in systematic processing when they are
exposed to a descriptive normative message indicating low prevalence of recycling which
arouses higher levels of satisfaction, compared to that indicating high prevalence of recycling
which arouses lower levels of satisfaction.
Some research has shown that negative feelings such as fear lead to reduced systematic
processing because defensive avoidance resulting from strong negative feelings causes people to
feel high levels of stress, and thus discourages them from processing a message carefully
(Gleicher & Petty, 1992). On the other hand, other research has argued that negative feelings
may increase systematic processing in some cases (Schwarz, 1990; Zuckerman & Chaiken,
1998). For example, people who have negative feelings such as guilt will be motivated to process

19

a message systematically if they want to be more confident in their judgement or decision
because the uncomfortable feelings such as guilt need to be resolved.
In the present study, it is expected that non-recyclers either in the high-prevalence
condition or the low-prevalence condition will not experience strong feelings of guilt or fear
because the descriptive normative information does not imply sanction or punishment. In regard
to this expectation, the greater feeling of guilt people experience, the greater needs to resolve the
uncomfortable feelings they will have. Naturally, one’s motivation to engage in systematic
processing will depend on the extent to which emotions, either satisfaction or guilt, are aroused.
It is reasonable to predict that non-recyclers will be more likely to process information
systematically when they are exposed to a descriptive normative message indicating the high
prevalence of recycling which arouses higher feeling of guilt, compared to that indicating low
prevalence of recycling which arouses lower feelings of guilt. With reference to the cognitive
mechanisms between normative information and behavior in FTNC, this study predicted that as
discrepancy between one’s prior recycling behavior and the prevalence of recycling portrayed in
a descriptive norm message is greater, one will show greater affective response. Accordingly,
both the greater discrepancy and the greater degree of aroused emotions will activate systematic
processing. Considering the nature of recycling behavior in a society, and the causes and
processes which result emotional consequences, the current study predicts:
The effects of the descriptive norm messages on message-relevant thoughts (both positive and
negative) will be moderated by prior recycling behavior such that:
H3a: Recyclers will produce a higher number of total message-relevant thoughts, both
positive and negative, in the low versus high prevalence message condition.

20

H3b: Non-recyclers will produce a higher number of total message-relevant thoughts,
both positive and negative, in the high versus low prevalence message condition.
The ways in which the normative messages and prior behavior will be linked with the
valence of the thoughts are less clear. Mathur and Chattopadhyay (1991) explored that the effects
of television program-induced moods on cognitive processing of embedded advertisements. In
their experiment, participants were asked to write down their thoughts about the advertisements
after being exposed to either a happy or sad program. Their findings showed a significant
positive association between the valence of moods and the valence of thoughts; participants who
were exposed to a happy program generated a greater number of positive thoughts about
advertisements than did those who are exposed to a sad program. In line with this finding, it may
be assumed that recyclers, relative to non-recyclers, will produce a higher number of positive
message-relevant thoughts in the low versus high prevalence message condition since they will
experience the most positive affect, satisfaction. On the contrary, non-recyclers, relative to
recyclers, will produce a higher number of negative message-relevant thoughts in the high versus
low prevalence message condition because they will experience the most negative emotion, guilt.
At least one study, however, would lead to a different hypothesis. Dillard et al. (1996)
predicted a positive association between the valence of affective response and the valence of the
dominant cognitive response while examining affective responses to HIV/AIDS PSAs. Their
findings indicated that there was no statistically significant association between six affective
responses (i.e. surprise, puzzlement, happiness, fear, anger and sadness) and the valence of the
dominant cognitive response. In reference to the mixed results, this study posed a research
question regarding the association between the affective responses and the valence of thoughts
about the messages indicating prevalence of recycling.

21

RQ3: Does the pattern of the valence of thoughts vary by experimental conditions?
Previous studies have found that satisfaction is one of the most critical motivations to
encourage recycling (De Young, 1985-1986; 1986; 1996; Lord, 1994; Smith et al., 1994). For
example, positively framed messages appealing to satisfaction led to more positive attitudes
toward recycling (Lord, 1994). Attitude, for the purpose of this study, is defined as a
combination of an individual’s evaluations about a given issue, object, person, or behavior
including both affective (i.e. evaluations) and cognitive (i.e. beliefs) properties (Eagly & Chaiken,
1993; Petty & Cacioppo, 1986). Affective properties refer to one’s feelings toward the behavior
(i.e. excited or bored), and cognitive properties refer to traits and characteristics of the behavior
(i.e. beneficial/harmful) (Crites, Fabrigar, & Petty, 1994). In this study, affective properties in
attitude may be closely associated with one’s evaluation of recycling behavior itself, while the
emotions (i.e. satisfaction or guilt) which are induced by the discrepancy between one’s prior
recycling behavior and the prevalence of recycling in a message may be closely related to one’s
assessment of one’s past performance. Based on the findings from previous studies (De Young,
1985-1986; 1986; 1996; Lord, 1994), it seems reasonable to predict that a behavior that brings
about positive emotions about oneself such as satisfaction will yield a positive attitude.
Furthermore, recyclers who experience satisfaction are likely to continue to recycle because of
the need to maintain positive emotions (De Young, 1985-1986; 1986; 1996; Meneses, 2010).
Based on these findings, this study predicts:
H4: Recyclers in the low prevalence message condition will have more positive attitude
toward recycling than those who in the high prevalence message condition.
H5: Recyclers in the low prevalence message condition will be more likely to intend to
recycle than those in the high prevalence message condition.

22

H6: For recyclers, the effects of descriptive norm messages on behavioral intent will be
mediated by satisfaction.
Once guilt generates an unpleasant affective state, people are motivated to engage in
recommended or desirable behaviors which produce positive affect in order to alleviate such
negative feelings (Baumann et al., 1981). Accordingly, people who have higher levels of feelings
of guilt, as a result of perceived discrepancy via self-reflection between one’s previous inaction
and the high prevalence of the action, will actively attempt to alleviate their strong negative
states indicating positive attitudes and strong intentions toward recycling in order to reduce their
feelings of guilt. Taking into account aversive aspect of feelings of guilt and an individual’s need
to alleviate it (Baumann et al., 1981; Nabi, 1999; O’Keefe, 2002), the current study predicts:
H7: Non-recyclers in the high prevalence message condition will have more positive
attitude toward recycling than those in the low prevalence message condition.
H8: Non-recyclers in the high prevalence message condition will be more likely to
intend to recycle than those in the low prevalence message condition.
H9: For non-recyclers, the effects of descriptive norm messages on behavioral intent will
be mediated by guilt.
To sum up, the discrepancy between one’s previous recycling activity and the
descriptive norm messages will lead to emotional consequences and cognitive appraisal, which,
in turn, influence attitude and/or intention toward recycling.

Nabi (2010) contended that it

is important to research the specific emotional state as well as the onset of emotional experiences
since it may improve the predictive power of human’s behavioral outcomes. This study attempts
to provide an explanation of the effects of normative influence by exploring the emotional

23

process as a mediator in the relationship between descriptive normative information and young
people’s attitudes and intentions toward recycling.

24

METHOD
Design Overview
To test the study hypotheses and answer the research questions, this study consisted of a
pilot study and a main experiment. The purpose of the pilot study was to check the manipulation
of the independent variable (i.e., messages about the prevalence of recycling) as well as the
quality of messages (i.e., believability and persuasiveness). In the main experiment, a 2
(prevalence of recycling: High versus Low) x 2 (prior recycling behavior: Recycler versus Nonrecycler) pre-post test mixed design was employed. Participants’ past recycling behavior was a
measured within-subject factor and the prevalence of recycling was manipulated in the messages;
emotions, attitudes, and behavioral intent were the measured dependent variables. In the pre-test,
all participants were asked to report their participation in environmental activities including prior
recycling behavior. After completing these questions, all participants were randomly assigned to
watch a video message indicating the prevalence of recycling (i.e., high or low). After watching
a video message, they were asked to report their emotional responses, thoughts about recycling,
attitudes, and behavioral intentions toward recycling as outcome variables.
Stimulus Material
This study used a video as the stimuli to persuade participants to recycle (Appendix A).
1

The documentary film, Making it Work, was created by students at Michigan State University.
The short film has a running time of about 11 minutes and includes interviews with students,
faculty, and staff around from MSU. The main story of the film is what people at MSU can
recycle and what they think about recycling. For the interest of current study, the video was

shortened to a running time of around three minutes to reduce participants’ fatigue. The edited
video included content about the type of materials students recycle, how they recycle, and why

25

2

they do it. The descriptive normative content was added in the first and last scenes of the video
to show the prevalence of recycling among MSU students. The messages for the high and low
prevalence conditions were exactly the same except the prevalence manipulation. The descriptive
normative information (i.e., the prevalence) was manipulated using either a high or low
percentage of students who recycle. The high prevalence message stated: In 2012, 80% of MSU
students recycled the paper they used. The low prevalence message stated: In 2012, 20% of MSU
3

students recycled the paper they used.
Pilot Study

The purpose of the pilot study was to test whether the independent variable was
manipulated as intended. In this study, for a message containing high descriptive normative
information, participants should perceive the message as showing the high prevalence of
recycling as well as perceive that recycling is a prevalent behavior among MSU students; for the
message containing low descriptive normative information, participants should perceive the
message as showing low prevalence of recycling as well as perceive that recycling is a less
prevalent behavior among MSU students than those receiving the high prevalence message. A
small sample (N = 68) was used in the pilot study. Participants were recruited from a participant
pool maintained in the Department of Communication at a Midwestern university in the U.S.
They were randomly assigned to watch one of two video messages.
After viewing the messages on the website, an online survey was completed by
participants to measure the manipulation of the independent variable (all pilot test items were
presented in Appendix B). Perceived descriptive normative content in the message was measured
with four items with a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly
agree). Evidence for the reliability and validity of the scale using a 5-point response format was

26

provided in a previous study (Lapinski, Maloney, Braz, & Shulman, 2013, α = .89). The
examples included such as: “The information presented above indicates that most students at
MSU engage in recycling the paper they use.” and “Clearly, this message shows that not many
students at MSU are recycling the paper they use.” The perceived prevalence of recycling
among MSU students was measured with four items with a 7-point Likert-type scale ranging
from 1 (strongly disagree) to 7 (strongly agree) adopted from a study conducted by Lapinski et
al., (2013; α = .91). The examples included such as: “Most students at MSU actively recycle the
paper they use.” and “The majority of the students at MSU actively recycle the paper they use.”
To determine the possible effects of confounds, the extent to which people perceived the
approval of recycling among MSU students and perceived the injunctive normative content in
the message were measured. Each measure was adopted from a previous study (Lapinski, Rimal,
DeVries, & Lee, 2007, α = .80) and consisted of four items with a 7-point Likert-type scale
ranging from 1 (strongly disagree) to 7 (strongly agree). The measure for the perceived approval
of recycling included examples such as: “It is clear that many students believe paper recycling is
important.” and “People at MSU approve of paper recycling.” The measure for the perceived
injunctive normative content in the message included examples such as: “Based on the video, I
feel like students at MSU would think less of me if I didn’t recycle the paper I use.” and “This
video shows that people around me would respect me more if I recycled paper I use.”
In addition, the quality of messages (i.e. believability and persuasiveness) was checked.
To check the believability of a message, three items with a 7-point Likert scale ranging from 1
(strongly disagree) to 7 (strongly agree) were adapted from Chen’s study (2011, α = .72) and
one item was created for the current study. For example, the believability measurement included
items such as: “I understand what the video is talking about.” and “I think the statistic in the

27

video seems believable.” The persuasiveness of a message was assessed with three 7-point
semantic differential items. Two items were adopted from Dillard and Peck’s study (2000, α
= .95), which anchored with “not at all convincing/very convincing” and “not at all
persuasive/very persuasive”. Additionally, one item was created for the current study anchored
with “not at all influential/very influential”.
Main Experiment
Participants. Participants for this study were undergraduate college students at a large
public university in the Midwest. One hundred eighty eight college students were recruited from
a participant pool maintained in the Department of Communication or from courses opened in
the College of Arts and Science at the university. They were offered extra course credits for their
participation. The age of participants ranged from 18 to 33 years old. The mean age of the
participants was 20.90 years (SD = 1.86 years). Most participants were Caucasian (61.7%), but
the sample included people who reported as Asian (20.7%), African-American (12.2%), AsianAmerican (2.1%), Hispanic (1.6%), and Hispanic-American (.5%). Around two percent of the
participants reported their race as “other”. The gender split was approximately even (female54.8%). The sample comprised 43.9% seniors followed by juniors (29.4%), sophomores (17.6%),
and freshmen (8%). Two participants (1.1%) did not report their year. Most participants lived
off-campus (83.5%), and lived with roommate(s) (85.6%). Eleven participants (5.9%) affiliated
to at least one pro-environmental group. Just over half of participants identified themselves as
recyclers (58.5%).
Procedure. Participants were recruited from a participant pool in the Department of
Communication or received an invitation email from their instructor. Once voluntarily choosing
to participate in the study, they were provided a consent form (Appendix C). After agreeing to

28

the terms on the consent form, they were asked to complete an “Environmental Behaviors Scale”
adopted from Lapinski et al. (2007). To reduce the possibility of demand characteristics, the
scale included not only questions about recycling behavior but also filler questions about other
environmentally friendly behaviors. First, participants answered screening questions asking
about their use of recycling materials. When participants answered “Yes” to the screening
question, they were asked to report how many times they engaged in recycling including paper,
glass, metal cans, and plastic at home or while in the class (all pre-test items were presented in
Appendix D).
After completing the pre-test, participants were told that each individual would view a
video related to environmental behaviors in order to evaluate the quality of the video message.
They were then randomly assigned to one of the two experimental conditions: the high or low
prevalence message condition. Subsequently, all participants completed measures of main
outcome variables, aroused emotions, cognitive processing, attitude, and intention. Additionally,
they completed the measures for manipulation check and several demographic items in the posttest (all post-test items were presented in Appendix E). The experiment took approximately 2530 minutes for participants to complete.
Measures. Each measure was checked for inter-item correlations, item contribution to
scale reliability, and internal consistency.
Independent Variables
Prior Recycling Behavior. Two independent scales were employed to measure the prior
recycling behavior. First, the extent to which participants engaged in recycling several materials
(i.e. paper, plastic, metal can, and glass) during the last 6 months was assessed with 8 items
ranging from 1 Never to 5 Always. The measure included items such as: “In the last 6 months,

29

how often did you recycle the paper you use at home or in the dorm? and “In the last 6 months,
how often did you recycle the paper you use while in the class?” Additionally, participants were
asked to identify themselves as recyclers or non-recyclers with a single item.
Dependent Variables
Aroused Emotions. Participants were asked to answer questions about their feelings of
happiness, pride, satisfaction, and guilt for the main outcome variables. Measures of other
negative or neutral emotions were also completed.
Happiness. The extent to which participants experienced happiness when watching the
video messages was measured with five items which were created for the current study. Five
items used a 7-point Likert-type scale anchored by 1 (strongly disagree) and 7 (strongly agree).
The measure included items such as: “While watching a video I felt happy”, “I felt joyful”, “I felt
glad”, “I felt delighted”, and “I felt cheerful.”
Pride. The extent to which participants felt proud while viewing the message was
measured with three items which were created for this study. The items used a 7-point Likerttype scale anchored by 1 (strongly disagree) and 7 (strongly agree). The measure included items
such as: “While watching a video I felt proud”, “I felt accomplished”, and “I felt credited for
myself because of my own behavior.”
Satisfaction. The extent to which participants experienced satisfaction when watching
the video was measured with four items. One item was derived from previous literature in which
satisfaction was measured with six items (Oliver, 1980, α = .82) and three items were created for
this study. The items used a 7-point Likert-type scale anchored by 1 (strongly disagree) and 7
(strongly agree). The measure included items such as: “While watching a video I felt satisfied”,
“I felt good”, “I felt pleased”, and “I felt contented.”

30

Guilt.

The extent to which participants felt guilty while watching the video was

measured with four items modified from Harris’s study (2007, α = .92). The guilt measure used a
7-point Likert-type items ranging from 1 (strongly disagree) to 7 (strongly agree). The measure
included items such as: “While watching a video I felt guilty”, “I felt regretful”, “I felt bad about
myself”, and “I felt remorseful.”
Other Negative Emotions. To check possible influence of other negative or other
emotions (neither positive nor negative), a measurement was adopted from previous research
(Dillard et al., 1996). The scale consisted of 15 single items to measure 5 negative or neutral
emotions anchored by 0 (None of this feeling) and 6 (A great deal of this feeling). Each emotion
comprised of three items: puzzlement (puzzled, confused, and bewildered, α = .74), surprise
(surprised, astonished, and amazed, α = .74), fear (fearful, afraid, and scared, α = .94), anger
(angry, irritated, and annoyed, α = .84), and sadness (sad, dreary, and dismal, α = .78).
Thoughts about the Video Message. The procedure for capturing and analyzing the
number and valence of message-relevant thoughts was adapted from previous studies (Dillard et
al., 1996). Participants were asked to write the thoughts they had while watching the video
message with a single open-ended question. Participants were allowed two minutes to type these
thoughts. Each participant’s thoughts were unitized by independent coders based on a single
stated idea regardless of grammatical errors. The criteria to unitize the thoughts were derived
from Cacioppo and Petty (1981), which included 1) surface feature markers (i.e. periods,
commas, and semicolons) and 2) conjunctions – coordinate conjunction (i.e., and, or, but, so),
correlative conjunction (i.e., both A and B, not only A (but) also B, B as well as A, either A or
B), and subordinate conjunction (i.e., since, because, if, unless, though). The unit of analysis was
a single thought unit. For example, a participant who answered like “I feel surprised that we only

31

have 20% of MSU student to recycle things (1). I would recycle plastic (1), papers (2) and metal
(3) in future to make contribution to our environment.” had four thoughts while watching the
video. After unitizing the thoughts, two coders counted the total number of thoughts. Then, the
unit was coded as being relevant or irrelevant to the content of the video. A message relevant
thought was thoughts about the message contents such as one that was related to the content of
the message (i.e., “I realized how much waste I produce every single day”), or thoughts about
emotions experienced (i.e., “I felt guilty that I do not recycle the paper I use”). If irrelevant, the
unit was not coded. Irrelevant thoughts were those that did not relate to the video in any way (i.e.,
“Trying to figure out what the question is asking.”). Subsequently, the valence of thought units
was coded as one of three categories: positive (i.e., “[I was] Proud to be a part of a university
that is making a difference.”), negative (i.e., “I was embarrassed at how little I recycle in
comparison with the statistics mentioned in the video.”) or neutral (i.e., “I had no idea that MSU
students were so active in recycling.”). Overall, the dependent variables assessed here were:
number of message-relevant thoughts, types of thoughts (relevant or irrelevant), and the valence
of thoughts (positive, negative, or neutral). Thought coding variables and examples can be found
in Table 2. Inter-coder reliability was assessed for all steps of the process.
Two trained, independent coders unitized the thoughts for 40 participants (about 20% of
the participants) and coded the types of and the valence of thoughts. Guetzkow’s U was
4

5

calculated to check inter-coder reliability for unitization. Guetzknow’s U was .003, which
could be regarded as being reliable since a small figure indicates successful unitization
(Guetzkow, 1950). Then, to check inter-coder reliability for the types and valence of thought
units, Cohen’s Kappa (Cohen, 1960) was calculated but the values were under .7. To improve the
quality of coding process, disagreements in coding were resolved by discussion and the two

32

coders were re-trained. Guetzknow’s U and Cohen’s Kappa were calculated for the thoughts
6

from 20 participants. Guetzknow’s U was .018. Cohen’s Kappa for the types of thoughts
was .78 and for the valence of thoughts was .72. Coders then coded the remaining thoughts.
Attitudes. Participants’ attitudes toward recycling were assessed with semantic
differential items including both measures of general attitudes and cognitive items (Crites et al.,
1994). Five cognitive word pairs were used to assess participants’ beliefs about the traits of
recycling (i.e. useful/useless, wise/foolish, beneficial/harmful, valuable/worthless, and
wholesome/unhealthy). The attitude items which asked participants’ general evaluations about
recycling constituted four pairs of words (i.e. positive/negative, like/dislike, good/bad, and
desirable/undesirable). All items were assessed on 7-point bipolar semantic differential scales
ranging from -3 to 3 with negative numbers indicating the negative evaluations and the positive
numbers indicating the positive evaluations and 0 indicating a neutral point. Each participant’s
answer for each attitude scale was summed and then the average was calculated to test
hypotheses.
Behavioral Intent. Two scales were used to measure behavioral intention to engage in
recycling two weeks later. The first measurement consisted of three items adopted from a study
conducted by Anderson (2012, α = .96). Example items are such as “In the next two weeks, how
likely it is that I will continue/start to sort trash in order to recycle the paper I use”,
“continue/start recycling paper I use”, and “continue/start to put paper into the corresponding
recycle bin.” Four items in the second measurement were created for the current study.
Example items were: “In the next 2 weeks, I plan to recycle the materials I use” and “I will
recycle any paper I use over the next 2 weeks.” The two scales consisted of 7-point Likert-type
scale anchored by 1 (not likely at all) and 7 (extremely likely). To test hypotheses, the second

33

measurement was employed because all participants answered it while some participants missed
the first measurement.
Manipulation Checks
Perceived Descriptive Normative Content in the Message. The extent to which
participants perceived that the messages indicate the high or low prevalence of recycling was
assessed on the same scale used in the pilot test (Lapinski et al., 2013, α = .89).
Perceived Prevalence of Recycling among MSU Students. The extent to which
participants perceived that the recycling is prevalent among MSU students was assessed on the
same scale used in the pilot test (Lapinski et al., 2013, α = .91).
Other Message Factors. To ensure the possible effect of confounds in the message
manipulation and check the quality of message, the perceived injunctive normative content in the
message, perceived approval of recycling, and the believability and persuasiveness of the video
message were measured with the same scale applied in the pilot test.
Control Variables
Issue Involvement. Four items were used to measure the extent to which participants
regard recycling as an important issue from a previous study (Lapinski et al., 2013; α = .72). The
original measure was developed and validated by Cho and Boster (2005). The items were
measured on a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree).
Examples included: “Recycling is an important issue to me” and “I really don’t care about the
issue of recycling.”
Social Desirability. The measurement for social desirability bias in participants’
responses was adapted from a study conducted by Schuessler, Hittle, and Cardascia (1978). The
original scale consists of 16 items with a 9-point bipolar rating scale ranging from 1 - low social

34

desirability, to 5 - neutral social desirability, to 9 - high social desirability (K- R 21 α = .64). For
this study, the scale was adapted to a 6-item scale which used 7-point Likert-type responses
ranging from 1 (strongly disagree), 7 (strongly agree).
Demographic Information. Questions regarding demographic information included age,
gender, ethnicity, major, and on vs. off-living situation. Also, the amount of exposure to the
video was measured with a single item, “How much of the video about recycling did you watch?”
with 7-point scale. The response range was from 0 = “None of the video message” to 6 = “All of
the video message.”

35

RESULTS
Pilot study
To determine whether the descriptive normative information was manipulated as
intended and check for the possible effects of confounds a pilot study was conducted. After
watching the stimulus message, participants (N = 68) completed an online questionnaire asking
about their perceptions of the video message. Participants rated the video message on several
items with 7-point Likert-Type scales – including perceived descriptive normative content of the
message (α = .89), perceived prevalence of recycling (α = .94), perceived injunctive normative
contents in the message (α = .65), perceived social approval of recycling (α = .57), message
believability (α = .77), and message persuasiveness (α = .93).
An independent sample t-test was performed to determine whether participants who
watched the video message with high descriptive norm information (N = 33) perceived the video
message as showing higher prevalence of recycling as well as perceived that recycling is a more
prevalent behavior among MSU students than participants watching the video message with low
descriptive norm information (N = 35). Participants viewing the high descriptive norm message
(M = 5.02, SD = 1.20) perceived the video message as showing higher prevalence of recycling
than participants watching the low descriptive norm message (M = 3.62, SD = 1.06, t (68) = 5.11,
2

p = .001, η = .28). Likewise, participants who were in the high descriptive norm condition (M =
4.53, SD = 1.38) perceived that recycling is a more prevalent behavior among MSU students
relative to participants in the low descriptive norm condition (M = 3.41, SD = 1.15, t (68) = 3.66,
2

p = .001, η = .16).
Other differences between the groups were evidenced.

Participants viewing the high

descriptive norm message (M = 4.84, SD = 0.93) perceived that the video message content

36

indicated greater approval of recycling than participants who viewed the low descriptive norm
2

message (M = 4.31, SD = 0.65, t (68) = 2.71, p = .009, η = .10). Participants in the high
descriptive norm condition (M = 4.93, SD = .76) perceived that recycling is a behavior more
strongly approved among MSU students than did those in the low descriptive norm condition (M
2

= 4.48, SD = .73, t (68) = 2.49 p = .02, η = .08).
Regarding the quality of the video message, there was no significant difference between
the two groups in perceived believability and persuasiveness of the message. Participants who
were exposed to the message containing high descriptive norm information did not differ from
participants watching the video message containing low descriptive norm information in
perceived believability (High M = 5.64, SD = 0.83; Low M = 5.58, SD = 0.80, t (68) = .33, p
2

= .74, η = .002) or persuasiveness (High M = 4.85, SD = 1.04; Low M = 4.52, SD = 1.16, t (68)
2

= 1.22, p = .23, η = .021). From the results above, it was reasonable to assume that the
manipulation of the current study was successful (Smith, Atkin, Martell, Allen, & Hembroff,
2007). However, the perceived injunctive normative information in the message and perceived
social approval of recycling among MSU students were significantly different between two
experimental conditions. It is not unusual for descriptive normative information to influence
people’s perception of injunctive norms (Rimal, 2008). These findings resulted in inclusion of
measures of the variable described above in the main experiment.
Main Experiment
Preliminary Analyses. Before testing hypotheses and answering the research questions,
preliminary analyses were conducted. All data were cleaned. In total, 188 participants completed
the survey for this study. One participant reported s/he did not watch the video and her/his data

37

were removed from additional analyses. The mean exposure of participants was 5.12 (SD = 1.32;
mid-point of the scale: 3). Thus, 187 participants were used for further analyses. Of 187, one
participant did not identify him/herself as either a recycler or a non-recycler. His/her data were
automatically excluded in the SPSS when analyses were conducted.
All scaled items were examined for positive contribution of items to scale reliability,
item-total correlations, overall scale reliability, and the extent to which the distributions
approximated normality. The scale means, standard deviations, and alphas for all scales across
conditions were presented in Table 3. This analysis revealed scale alphas were largely within
acceptable ranges (α > .75) with several exceptions. The involvement (α = .64), social
desirability (α = .64), perceived injunctive normative content in the message (α = .71), and
perceived approval of recycling (α = .64) scales indicated relatively low reliability. To improve
the reliability, items that did not contribute to scale reliability were removed from each scale:
involvement (α = .75, removed item 3), social desirability (α = .840 removed item 4, 5, and 6),
and perceived approval of recycling (α = .75, removed items 1 and 4).
To check the stimulus induction, an independent t-test was employed. The results
showed that Levene's Test for Equality of Variances was violated so that the corrected values
were reported. The descriptive normative information was manipulated as intended; participants
in the high prevalence condition (N= 98, M = 5.38, SD = 1.08) perceived that the message
content indicated higher prevalence of recycling than those in the low prevalence condition (N=
2

88, M = 3.49, SD = 1.30, t (169.57) = 10.71, p = .001, η = .40). Also, participants in the high
prevalence condition (N= 99, M = 5.31, SD = 1.13) perceived that recycling was a more
prevalent behavior among MSU students than those in the low prevalence condition (N=88, M =
2

3.50, SD = 1.60, t (154.57) = 8.84, p = .001, η = .34).

38

Participants in high prevalence condition (N= 99, M = 4.89, SD = 0.88) perceived the
video message as showing greater approval of recycling at MSU than those in low prevalence
2

condition (N= 88, M = 4.25, SD = 1.07, t (185) = 4.53, p = .001, η = .10). Also, participants
exposed to the high descriptive norm information (N= 99, M = 5.58, SD = 1.08) perceived that
recycling is a behavior more strongly approved among MSU students than participants exposed
to the low descriptive norm information (N= 88, M = 4.89, SD = 1.04, t (185) = 4.47, p = .001, η

2

= .10).
Regarding the quality of the video message, there was no significant difference in
perceived persuasiveness of the video message between the two groups (High: N= 99, M = 5.11,
2

SD = 1.37; Low: N= 88, M = 5.05, SD = 1.28, t (185) = 0.32, p = .75, η = .001). However, there
was a significant difference in participants’ perceived believability between the high prevalence
2

condition and the low prevalence condition [t (185) = - 2.20, p = .029, η = .03]. Participants in
the low prevalence condition (N= 88, M = 5.68, SD = 1.00) perceived that the message was more
believable than those in the high prevalence condition (N= 99, M = 5.34, SD = 1.12). Importantly,
the overall mean of perceived persuasiveness was 5.09 (SD = 1.33) and the perceived
believability was 5.50 (SD = 1.08). The mid-point of the scales for the persuasiveness and the
believability was four (i.e., like 7-point Likert type scale); one-sample t-tests indicated that the
2

mean of persuasiveness (t (186) = 11.21, p = .001, η = .40) and believability (t (186) = 19.09, p
2

= .001, η = .66) significantly differed from the mid-point. It was reasonable to assume that the
stimulus was of good quality (Smith et al., 2007).

39

Main Analyses. To test the study hypotheses and answer the research questions,
hierarchical regression analyses, single sample t-tests, and chi-square tests were employed. Prior
to conducting hierarchical regression analyses experimental conditions were dummy-coded, with
the high descriptive norm condition being 0 and the low descriptive norm condition coded as 1.
The participants’ prior recycling behavior was similarly dummy-coded, with a recycler coded as
0 and a non-recycler coded as 1. Other continuous variables, except the dependent variables were
mean-centered to avoid potential multicollinearity. For decisions about the inclusion of
covariates, correlations between all study-variables were calculated; only potential covariates
exhibiting significant correlations with the study dependent variables were included as covariates
in the analyses. For a full overview of these results, see Table 4. For each analysis performed, the
first block of the regression analyses contained study covariates including demographic variables.
The second block of the regression analyses contained the main predictors followed by an
interaction term between the main predictors (i.e., experimental condition and prior recycling
behavior) in the third block. A simplified summary of each hypothesis and research question and
its result were provided in Table 5.
Hypotheses 1 and 2 predicted that the effect of the descriptive normative information on arousal
of emotions depended on prior recycling behavior. Specifically, Hypothesis 1 predicted that
recyclers would experience the most satisfaction in the low versus high prevalence condition,
relative to non-recyclers. Hypothesis 2 predicted that non-recyclers would experience the most
7

guilt in the high versus low prevalence condition.

For the satisfaction dependent variable, a total of 187 participants were included in the
analyses. To test Hypothesis 1, two variables which were highly correlated with the dependent
variable, involvement and persuasiveness were included as covariates in the analysis. The overall

40

2

model was significant F (5, 180) = 8.15, p < .001, adjusted R = .162. The analysis showed that
the descriptive norm information had a significant direct effect on satisfaction, unstandardized B
2

= -0.40, t (185) = -2.75, p = .006, sr = .03 such that participants in the high prevalence condition
experienced greater satisfaction than those in the low prevalence condition. However,
participants’ prior recycling behavior was not a significant predictor of feelings of satisfaction,
2

unstandardized B = -0.30, t (185) = -1.91, p = .058, sr = .02. Furthermore, there was no
significant interaction effect between descriptive normative information and prior recycling
2

behavior on satisfaction, unstandardized B = 0.12, t (185) = 0.42, p = .68, sr = .001. Thus, it
was concluded that the data were not consistent with Hypothesis 1.
For the guilt dependent variable, two variables which were highly correlated with the
dependent variable, involvement, and social desirability, were included as covariates in the
2

analysis. The overall model was not significant F (5, 180) = 6.34, p < .005, adjusted R

= .06.The results indicated there was no significant direct effect of the descriptive normative
2

information on feelings of guilt, unstandardized B = 0.36, t (181) = 1.83, p = .07, sr = .02.
However, prior recycling behavior was a significant predictor of guilt (unstandardized B = 0.44, t
2

(181) = 2.08, p = .04, sr = .02) such that non-recyclers experienced greater feelings of guilt than
recyclers did. Furthermore, there was no significant interaction effect between descriptive
normative information and prior recycling behavior on feeling of guilt, unstandardized B = -0.27,
2

t (180) = -0.67, p = .51, sr = .002. Consequently, it was concluded that the data were not
consistent with Hypothesis 2.

41

Research question 1 and Research question 2 asked whether there were different patterns
of positive emotions (RQ1) and negative emotions (RQ2) across different levels of descriptive
norm message and prior recycling behavior. Single-sample t-tests were used to answer these
questions. RQ1 asked whether recyclers would experience more satisfaction than other positive
emotions including happiness and pride when they were exposed to the low descriptive norm
message relative to the high descriptive norm message. A total of 109 participants who identified
themselves as a recycler were included in the analyses. Among the recyclers, those in the high
prevalence condition (N = 59) experienced greater positive emotions overall including
satisfaction, happiness, and pride than those in low prevalence condition (N = 50, see Table 6).
The mean score of satisfaction in the low prevalence condition (M = 4.73, SD = 0.98) was used
as a test value in a series of single-sample t-tests. Among recyclers in the low prevalence
condition, there was no significant differences among the positive emotions [Happiness, M =
2

4.72, SD = 1.08, t (49) = -0.07, p = .95, η = .001; Pride, M = 4.91, SD = 1.01, t (49) = 1.23, p
2

= .22, η = .03]. That is to say, the recyclers in the low prevalence condition did not experience
different amounts of positive emotions nor did those in the high prevalence condition.
RQ2 asked whether non-recyclers would experience more guilt than other negative or
neutral emotions including puzzlement, surprise, fear, anger, and sadness when they were
exposed to the high descriptive norm message relative to the low descriptive norm message. A
total of 77 participants who identified themselves as non-recyclers were included in the analyses.
Non-recyclers in the low prevalence condition (N = 40) experienced greater feeling of guilt,
puzzlement, fear, and sadness than those in high prevalence condition (N = 37; see Table 7). On
the other hand, non-recyclers in high prevalence condition experienced greater feelings of
surprise and anger than those in low prevalence condition. None of the negative or neutral

42

emotions was significantly different between the high prevalence condition and the low
prevalence condition. Then, the mean score of guilt in the high prevalence condition (M = 3.5,
SD = 1.41) was used as a test value in the one-sample t-tests. Among non-recyclers in the high
prevalence condition, there were significant differences such that non-recyclers in the high
prevalence condition experienced more guilt than puzzlement (M = 1.88, SD = 1.28, t (39) = 2

2

8.00, p = .001, η = .62), fear (M = 1.93, SD = 1.39, t (39) = -7.18, p = .001, η = .57), anger (M
2

= 2.06, SD = 1.19, t (39) = -7.67, p = .001, η = .60), and sadness (M = 1.83, SD = 1.22, t (39) = 2

8.69, p = .001, η = .66). However, there was no significant difference between guilt and surprise
2

(M = 3.53, SD = 1.72, t (39) = 0.12, p = .90, η = .001). Thus, the non-recyclers in the high
prevalence condition experienced more guilt and surprise than other emotions.
Hypotheses 3a and 3b predicted that the effect of descriptive norm information on the
number of total message-relevant thoughts would depend on prior recycling behavior.
Hypothesis 3a predicted that recyclers in the low prevalence condition would have greater
number of message-relevant thoughts than those in high prevalence condition while Hypothesis
3b predicted that non-recyclers in the high prevalence condition would have greater number of
message-relevant thoughts than those in the low prevalence condition. To test these hypotheses,
participants’ answers for the thought listing question were coded by two independent coders.
Of all 187 participants, 3 participants did not answer the thought listing question. Thus,
the total number of participants was 184 for testing these hypotheses. The total number of
thought units was 682. The average number of thought units per each participant was 3.71 (SD =
2.42). The total number of message-relevant thoughts was 641 (94%). Specifically, 454 units
(70.8%) were coded as thoughts about the message content and 187 units (29.2%) were coded as

43

thoughts about emotions experienced. Forty one thought units (6%) were irrelevant or
incomplete and thus excluded from further analyses.
A hierarchical regression analysis was conducted to test whether there was difference in
the number of message-relevant thoughts across different experimental conditions. In the
regression model, the number of message-relevant thought units per each participant was
regressed onto descriptive normative information and prior recycling behavior in the first block
and then the interaction term in the second block. The overall model was not significant, F (3,
173) = 0.48, p = .69, adjusted R2 = -.01. The results indicated that no significant main effects
2

(descriptive normative information, unstandardized B = -0.29, t (174) = -0.85, p = .40, sr = .004;
2

prior recycling behavior, unstandardized B = -0.03, t (174) = -0.10, p = .93, sr = .0001) on the
number of message relevant thoughts. The interaction effect on the number of message-relevant
2

thoughts was also not significant, unstandardized B = -0.59, t (173) = -0.86, p = .39, sr = .004.
Research question 3 asked whether the patterns of the valence of message-relevant
thoughts (positive, negative or neutral) varied by experimental conditions. To answer the
research question, four categories were created and a chi-square test of independence was
conducted: 1 - recyclers in the high prevalence condition, 2 - recyclers in the low prevalence
condition, 3 - non-recyclers in the high prevalence condition, and 4 - non-recyclers in the low

χ2 (6, N=637) = 21.97, p = .001, α

prevalence condition. A chi-square analysis showed a significant association between the
experimental conditions and the valence of thought units,

= .05, Cramer’s V = .13 (see Table 8). The results indicated that there was a significant

difference in the positive message-relevant thoughts across different experimental conditions
such as recyclers in the high prevalence condition had a significantly greater number of positive

44

thoughts than those in the other conditions (standardized residual = 2.3). Similarly, recyclers in
the low prevalence condition had a significantly fewer number of positive thoughts than people
in the other conditions (standardized residual = -2.7, see Figure 1).
Hypotheses 4 and 7 dealt with the relationship between descriptive norm information
and attitudes toward recycling based on prior recycling behavior. Hypothesis 4 predicted that
recyclers in the low prevalence condition would have more positive attitudes toward recycling
than those in high prevalence condition while Hypothesis 7 predicted that non-recyclers in the
high prevalence condition would have more positive attitudes toward recycling than those in low
prevalence condition. To test these hypotheses, a hierarchical regression analysis was conducted
after the two scales for attitude were summed and the mean was calculated. In the regression
model, attitude toward recycling was regressed onto five covariates, namely, believability,
persuasiveness, involvement, social desirability, and perceived approval, in the first block and
then the descriptive normative information and prior recycling behavior in the second block, and
the interaction term in the third block.
2

The overall model was significant, F (8, 177) = 3.08, p = .003, adjusted R = .08.
However, the results indicated that no significant main effects or interaction effects on attitudes
toward recycling (see Table 9). Thus, it was concluded that the data were not consistent with
hypotheses 4 and 7. However, it is worth mentioning that the overall mean of participants’
attitudes toward recycling was 6.25 (SD = 1.19) on a 7-point scale. There appears to be a ceiling
effect for attitude; participants had positive attitudes toward recycling across all experimental
conditions (see Table 10).
Hypotheses 5 and 8 predicted that the relationship between descriptive norm information
and behavioral intent to recycle would depend on prior recycling behavior. Hypothesis 5

45

predicted that recyclers in the low prevalence condition would have greater intention toward
recycling than those in the high prevalence condition whereas Hypothesis 8 predicted that nonrecyclers in the high prevalence condition would have greater intention toward recycling than
those in the low prevalence condition. Of the two intention scales included in the questionnaire,
the second intention scale was used for the analysis because all participants completed it (N =
187) compared to the first intention scale (N = 176). Hierarchical regression analysis was
conducted to test these hypotheses. In the regression model, intention to recycle was regressed
onto five covariates, believability, persuasiveness, involvement, social desirability, and perceived
approval, in the first block, and then the descriptive normative information and prior recycling
behavior in the second block followed by the interaction term in the third block. The overall
2

model was significant, F (8, 177) = 22.32, p < .001, adjusted R = .48 (see Table 11).
The results indicated that involvement, social desirability, and persuasiveness were
significant predictors in the first block. There was no significant direct effect of descriptive norm
message on intention toward recycling. Prior recycling behavior was a significant predictor for
intention toward recycling such that recyclers were more likely to intend to recycle in the future
2

than non-recyclers, unstandardized B = -0.67, t (178) = -3.96, p = .001, sr = .05. Furthermore,
there was a significant interaction effect between descriptive normative information and prior
recycling behavior on intention toward recycling, unstandardized B = -0.68, t (178) = -2.12, p
2

= .035, sr = .01. Recyclers were more likely to intend to recycle in the low prevalence condition
(M = 6.10, SD = 0.98) than in the high prevalence condition (M = 5.92, SD = 1.27) while nonrecyclers were more likely to intend to recycle in the high prevalence condition (M = 4.98, SD =

46

1.63) than in the low prevalence condition (M = 4.57, SD = 1.56). Thus, it was concluded that the
data were consistent with the hypotheses 5 and 8 (see Figure 2).
To test the mediated model predicted in Hypotheses 6 and 9, steps for testing mediation
outlined in Baron and Kenny (1986) were followed. This consists of four conditions necessary to
establish mediation; the first of these steps is to establish an association between the causal
variable and the mediator variable (Kenny, 2013). Because the data were not consistent with the
predicted interaction between normative message and prior recycling behavior on either guilt or
satisfaction, the other steps were not necessary and it was determined that the data were not
consistent with the predicted mediated model.

47

DISCUSSION
Previous literature has focused on examining the effect of social norms messages on the
cognitive mechanism of normative influence. The aim of this study was to fill the gap in the
literature by exploring the affective mechanisms; particularly the role of emotions in the
relationship between descriptive norm information, prior recycling behavior, and
attitude/intention toward recycling. The study tested a number of predictions and research
questions about these relationships. Although the data were not consistent with many of the
study predictions, the data were consistent with the predicted interaction effect of descriptive
norm message and prior recycling behavior on intention to recycle. Along with this finding, there
are several interesting findings in this study.
First, the descriptive norm messages designed to manipulate prevalence perceptions as
either high or low prevalence were successful. Participants in both the pilot study and the main
experiment perceived the high descriptive norm message content as showing high prevalence of
recycling and the low descriptive norm message content as indicating low prevalence of
recycling at MSU. In terms of influencing normative perceptions, participants exposed to the
high descriptive norm message perceived that recycling was more prevalent behavior among
MSU students than those exposed to the low descriptive norm message. Although there were
differences in believability, perceived injunctive norm content in the messages, perceived social
approval of recycling among MSU students between the experimental conditions, it is not
unusual for descriptive norm information to influence people’s perceptions of injunctive norms
(Rimal, 2008). Given these data, it is reasonable to conclude that the stimulus messages were of
good quality.

48

In the main experiment, this study examined whether recyclers and non-recyclers
experienced different affective responses when they were exposed to different levels of
descriptive norm messages. Specifically, Hypotheses 1 and 2 predicted that recyclers would
experience the greatest satisfaction in the low prevalence condition relative to the high
prevalence condition whereas non-recyclers would experience the greatest guilt in the high
prevalence condition relative to the low prevalence condition. The data were inconsistent with
this prediction. There was no interaction effect for descriptive norm information and prior
recycling behavior on satisfaction or guilt. This indicates prior recycling behavior does not
moderate the relationship between descriptive norms and emotion in the current research.
Contrary to the prediction, descriptive normative information was a significant predictor of
feeling of satisfaction such that participants in the high prevalence condition experienced greater
satisfaction than those in the low prevalence condition. Prior recycling behavior was a significant
predictor of feelings of guilt such that non-recyclers experienced greater feelings of guilt than
recyclers.
According to previous literature on discrete emotions, satisfaction or happiness can be
aroused by experiencing enjoyable situations or events such as observing a desirable
phenomenon regardless of their contribution to it (Ekman & Cordaro, 2011) while guilt or pride
can be regarded as self-conscious emotions which require one to hold responsibility for or take
credit for the situation (Baumeister et al., 1994; Sauter, 2010; Weiner, 1985). In order to better
understand these findings, post hoc analyses were conducted to examine the effects of the
independent variables on the other emotions. In the post-hoc tests, similar directions were
detected in terms of the effect of the descriptive norm messages on happiness and the effect of
prior recycling behavior on pride. Like satisfaction, the results indicated that the prevalence of

49

recycling in the descriptive norm messages was a significant predictor of feelings of happiness
such that participants in the high prevalence condition experienced greater happiness than those
in the low prevalence condition. Like guilt, pride was predicted only by prior recycling behavior
such that recyclers experienced greater pride than non-recyclers.
These findings may be explained by different referent points of the emotions; that is, the
messages may cause people to reflect on their own or others behaviors. For example, the people
presented in the descriptive norm message or the descriptive norm message itself may be the
referent point of the feelings of satisfaction and happiness whereas oneself like one’s prior
performance may be the referent point of the feelings of guilt and pride. It will be meaningful to
examine the referent point of each discrete emotion in the future research to help researchers
better understand the mechanism of discrete emotions.
In order to untangle the effects of messages on different emotions, research questions 1
and 2 asked whether recyclers experienced greater satisfaction than either happiness or pride in
the low prevalence condition while non-recyclers experienced greater feelings of guilt than other
negative or neutral emotions such as puzzlement, surprise, fear, anger, and sadness in the high
prevalence condition. Regarding research question 1, recyclers in the low and high prevalence
conditions experienced similar amounts of the three positive emotions. Recyclers also indicated
greater positive emotions in the high prevalence condition relative to the low prevalence
condition. These findings, along with the findings for Hypothesis 1, call into question whether
satisfaction and happiness are distinct from each other in the current study. In this study, many of
the positive emotions were highly correlated each other ranging from r = .66 (happiness and
pride), to r = .69 (happiness and satisfaction), to r = .74 (satisfaction and pride). However, there

50

was a significant difference between experimental conditions only in recyclers’ satisfaction but
not happiness or pride, indicating different antecedent conditions for different positive emotions.
Some previous research has conceptualized happiness and satisfaction in the same
category (Ekman & Cordaro, 2011; Lazarus, 1991) while other research has conceptualized them
differently (Dillard & Peck, 2000; 2001; Oliver, 1980). Sauter (2010) argued that pride is a selfconscious emotion and is distinct from basic emotions like happiness. Previous literature has
argued that future studies should clarify the structures and associations among discrete positive
emotions (Bagozzi, 1991; Lazarus, 1991), but the results of the current study cannot provide
solid evidence regarding how these emotions should be categorized. Future research may
examine whether these emotions are distinct by considering other moderators such as group
identity in the relationship between descriptive norms and emotions. For example, Rimal (2008)
found that descriptive norms have a stronger effect on college students’ intentions to drink as the
extent which their group identification (i.e., aspiration and similarity) increased. It is possible
that recyclers in the high prevalence condition might find greater similarity while watching the
message and thus experience greater satisfaction than those in the low prevalence condition.
For research question 2, non-recyclers experienced significantly greater guilt than other
negative emotions in both the high and low prevalence conditions. That is, across message
conditions, the non-recyclers experienced guilt; indicating that even low prevalence messages
may motivate behavioral comparison. Previous literature pointed out that people feel guilty when
they violate norms (Baumeister et al., 1994). This study provides evidence for this. When nonrecyclers were exposed to a descriptive norm message, they may become conscious of their
violation of the norms both in the high and low prevalence condition because they already know
the importance of recycling.

51

This study also explored whether prior recycling behavior moderated the relationship
between the descriptive norm messages and information processing. Hypothesis 3a predicted that
recyclers in the low prevalence condition would have greater number of message-relevant
thoughts than those in high prevalence condition whereas hypothesis 3b predicted that nonrecyclers in the high prevalence condition would have greater number of message-relevant
thoughts than those in the low prevalence condition. The data were not consistent with this
prediction. Neither the prevalence of recycling nor prior recycling behavior was a significant
predictor of the number of message-relevant thoughts. It was found that participants perceived
injunctive norm content in the message as well as the social approval of recycling among MSU
students in the manipulation check and it may be the case that social approval of recycling
predicts the number of thoughts. In the thought listing, participants expressed statements like, “I
thought that it was surprising that so many students recycle and that I should be better about it”,
“Like I should start recycling more often”, and “I felt hopeful for the future, happy, good. I felt
like I should recycle more.” Jacobson et al. (2011) indicated that participants needed more
cognitive effort to process injunctive normative information than descriptive normative
information in their experiment. Future research can examine whether descriptive and injunctive
norm messages are processed in different ways by comparing the number of thoughts from one
who receives a descriptive norm message with the number of thoughts from one who receives an
injunctive norm message.
Research question 3 explored the patterns of valence of message-relevant thoughts
across experimental conditions. The results showed that there were different patterns of valence
across experimental conditions. Specifically, recyclers had more positive message-relevant
thoughts in the high prevalence condition than other conditions and had fewer positive message-

52

relevant thoughts in the low prevalence condition. This is consistent with other research on
emotional response to messages. Mathur and Chattopadhyay (1991) found a positive relationship
between mood and the number of thoughts such as that participants exposed to a happy program
generated a greater number of positive thoughts about advertisements than did those who were
exposed to a sad program. In the current study, there was a significant difference in recyclers’
satisfaction between the high and low prevalence condition among all emotion measures. It may
be the case that satisfaction facilitates message processing as an additional source of information
in a positive direction.
Next, this study sought to examine whether the effect of descriptive norm messages on
attitudes toward recycling depended on prior recycling behavior. Hypothesis 4 predicted that
recyclers in the low prevalence condition would have greater positive attitudes toward recycling
than those in the high prevalence condition whereas hypothesis 7 predicted that non-recyclers in
the high prevalence condition would have greater positive attitudes toward recycling than those
in the low prevalence condition. The data were not consistent with this prediction. Participants,
regardless of the different levels of descriptive norm messages and prior performance, had
positive attitudes toward recycling. That is, across conditions, participants had very positive
attitudes toward recycling.
Finally, this study explored whether the effect of descriptive norm information on
intention to recycle depended on prior recycling behavior. Hypothesis 5 predicted that recyclers
in the low prevalence condition would be more likely to intend to recycle than those in the high
prevalence condition whereas Hypothesis 8 predicted that non-recyclers in the high prevalence
condition would be more likely to intend to recycle than those in the low prevalence condition.
The data were consistent with these predictions. Recyclers were more likely to intend to recycle

53

in the low prevalence condition than in the high prevalence condition whereas non-recyclers
were more likely to intend to recycle in the high prevalence condition than in the low prevalence
condition.
Although the data were in line with the predictions, it is unclear why recyclers tended to intend
to recycle more in the low prevalence condition than in the high prevalence condition because
they experienced greater satisfaction in the high prevalence condition. In addition, there was no
significant difference in non-recyclers’ guilt between the two experimental conditions. One
possible explanation is that prior behavior drives outcome expectations; a moderator of the
descriptive norm-behavior relationship identified in earlier research (Rimal, 2008). Additional
research might examine this issue in more depth. In the following section, theoretical
implications of the current findings are discussed.

54

THEORETICAL IMPLICATIONS
The current findings have implications for extending the explanatory power of prior
behavior. Previous research suggested that past behavior was a powerful predictor of a future
behavior (Conner & Armitage, 1998), however, that its influence would depend on the nature of
the behavior (Aarts et al., 1998). In the current study, prior recycling behavior was a significant
predictor of future recycling behavior; that is, recyclers were more likely to intend to recycle
than non-recyclers. Prior recycling behavior also predicted feelings of guilt such that nonrecyclers experienced greater feelings of guilt than recyclers.
Furthermore, the current findings have important implications for theories of social
norms to explain behaviors. By exploring the affective mechanism of social norms, the current
study extends the previous literature that mainly focused on explaining the cognitive mechanism
of social norms. In the current study, descriptive norms messages predicted feelings of
satisfaction such that participants in the high prevalence condition experienced greater
satisfaction than those in the low prevalence condition. This finding may be related to the distinct
level of descriptive norms.
Park and Smith (2007) examined the effects of five distinctive norms on the intention to
enroll on a state organ-donor registry and the intention to talk about organ donation with family
members based on TPB and SNA (Social Norm Approach). In their study, they separated the
levels of perceived social norms into either a personal- or a societal-level. Regarding the
personal-level of social norms, the degrees of popularity and approval of a behavior among an
individual’s important people would be a determinant of the behavior while in the societal-level,
the degrees of popularity and approval of a behavior in an individual’s society would be a critical
referent point. They found that personal level-descriptive norms were a major predictor of intent

55

to register in a state organ-donor program while subjective norms were a critical predictor of
intent to discuss about organ donation with family. These findings imply the close connection
between each norm and each distinctive behavior.
When a person individually performs a behavior alone, such as signing an organ-donor
registry, personal level-descriptive/injunctive norms may have greater power in predicting future
behavior than other norms. On the other hand, when a person performs a behavior that involves a
social aspect, such as talking with family about organ donation, subjective norms or societal
level-descriptive/injunctive norms may be more critical determinant of future behavior. In the
current study, participants may regard recycling behavior as a social behavior so that they simple
thought about what other people did in the descriptive norms message but did not think about
themselves. That is, it is the effect of societal level-descriptive norms on feelings of satisfaction
that participants in the high prevalence condition experienced greater satisfaction than those in
the low prevalence condition. In the following section, several limitations of this study and some
additional suggestions for future research are discussed.

56

LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH
First, this study only considered attitude and intention toward recycling as dependent
variables without addressing actual recycling behavior. This study could be extended to assess
the influence of descriptive normative information on the actual recycling behavior. Recycling is
a behavior for which attitudes are overwhelmingly positive, yet many people do not actually do it.
Furthermore, recycling involves in several materials, not only paper, but also metal cans, plastic,
glass, electronics, etc. The current experiment only focuses on paper recycling as an outcome
variable. Future research could deal with other recyclable materials.
Another limitation is related to the use of a convenience sample. Although the target
population of this study is young people, using an undergraduate sample in one university in the
Midwest may weaken the generalizability of the current findings to other populations such as
young people at other universities. The findings may not be generalized to other groups of people
including older adults and less educated young people. Also, there was a difference in
believability between high prevalence condition and low prevalence condition when checking
the quality of descriptive norm messages. This was measured so it could be controlled as a
covariate when it was associated with the dependent variables, so may be only a minimal
concern. An additional limitation is that this study used an on-line survey design. Because of this,
participants could fast forward through the video if they wanted. However, judging by
participants’ answers to the exposure item, most participants did watch the videos.
Lastly, this study only considered the persuasive effect of descriptive norm information
on attitude and intention toward recycling. In the manipulation check, not only descriptive
normative content/perceived prevalence of recycling but also injunctive normative
content/perceived approval of recycling were significantly different between the high and low

57

prevalence conditions. This was an unintended effect, and therefore a limitation of this study; to
deal with this issue, perceived approval of recycling was included as a covariate in the analyses
when appropriate. Future research should also assess the independent effect of injunctive norm
information on emotions and subsequent intentions to recycle. This is particularly important
because the theoretical framework of this study, FTNC, posits that descriptive and injunctive
social norms affect behavior in distinct ways (Cialdini et al., 1990). It may be more desirable to
examine the impact of both descriptive norms and injunctive norms than considering only one of
them.
Although participants showed relatively strong intention to recycle, the study finding
that the interaction effect of the descriptive norm information and prior recycling behavior on
intention to recycle could be interpreted as a case of the destructive power of descriptive norms.
A previous study found that descriptive norm information could decrease desirable behaviors
among people who have already engaged in the desirable behaviors. For example, a field study
conducted by Schultz, Nolan, Cialdini, Goldstein, and Griskevicius (2007) showed that the
descriptive normative information actually led to increased energy consumption among the
households consuming less energy relative to the neighborhood average. Their findings also
illustrated the effect of injunctive norms on reducing the negative influence of descriptive norms.
Again, combining both descriptive and injunctive norms in a message could be explored in the
future research to determine whether injunctive norms messages can temper the effects of
descriptive norms messages. Cialdini (2003) also contended if two norms in a message are
consistent with each other rather than in opposition to each other, it may have constructive
implications for pro-social behaviors like recycling.

58

Despite the several limitations of the current study, the research has a key strength in
that it attempted to fill a gap in the literature by exploring the roles of discrete emotions in the
relationship between descriptive norms and attitude/intention. As Nabi (2010) argued,
understanding the antecedents and states of discrete emotions will allow researcher to better
predict and explain human behaviors. In addition to the directions for future research described
above, future research should continue to explore the mediating effects of emotions on the
relationship between social norms and attitude, intention, and behavior.

59

FOOTNOTES
1.

The original film was created by David Cooper, Carol Baldwin, and Andy Childers. It

was the third prize winning video of the 2009 AHECTA Student Production Awards; our
gratitude to the creators for allowing us to use the film for this study. Thanks to Bob Albers in
the Department of Telecommunication, Information Studies, and Media for his help identifying
the creators.
2.

The original film is not publically available. The contents in the edited video are not

the same with those in the publically available video.
3.

The current study focused on paper recycling because it has been established for a

long time in MSU.
4.

Guetzkow’s U
=

5.

6.

7.

the number of units from coder1−the number of units from coder 2

the number of units from coder 1+ the number of units from coder 2

Guetzkow’s U =

Guetzkow’s U =

139−132

139+132
57−55
57+55

For better understanding of the effect of descriptive norm messages and prior behavior

on the emotional responses, hierarchical regression analyses were conducted with happiness and
pride as outcome variables.
In the regression model, feeling of happiness was regressed onto four covariates,
believability, persuasiveness, involvement, and perceived approval, in the first block and then the
descriptive norm condition and prior recycling behavior in the second block. The interaction
60

term was added in the third block. The overall model was significant F (7, 178) = 10.07, p < .001,
2

adjusted R = .26. The results were similar to those of the satisfaction as a dependent variable.
The descriptive norm information had a significant direct effect on happiness, unstandardized B
2

= -0.37, t (179) = -2.35, p = .02, sr = .02 such that participants in the high prevalence condition
experienced greater happiness than those in the low prevalence condition. However, the prior
recycling behavior was not a significant predictor of the feeling of happiness, unstandardized B =
2

-0.08, t (179) = -0.55, p = .58, sr = .001. Furthermore, there was no significant interaction effect
between the descriptive norm messages and prior recycling behavior on happiness,
2

unstandardized B = -0.06, t (178) = -0.19, p = .85, sr = .0001.
In addition, ratings of pride was regressed onto two covariates, persuasiveness and
involvement, in the first block and then the descriptive normative information and prior recycling
behavior in the second block followed by the interaction term in the third block. The overall
2

model was significant F (5, 179) = 10.81, p < .001, adjusted R = .21. Descriptive norm
information was not a significant predictor of feeling of pride, unstandardized B = -0.24, t (179)
2

= -1.52, p = .13, sr = .01. However, participants’ prior recycling behavior had a direct effect on
2

feeling of pride, unstandardized B = -0.67, t (179) = -4.05, p = .001, sr = .07 such that recyclers
experienced greater feeling of pride than non-recyclers. There was no significant interaction
between the descriptive norm messages and prior recycling behavior on pride, unstandardized B
2

= -0.06, t (178) = -0.19, p = .85, sr = .0001.

61

APPENDICIES

62

Appendix A: Tables and Figures

63

Table 1. Conceptual frameworks of discrete positive emotions

Happiness

Lazarus
(1991)

Ekman &
Cordaro
(2011)
Dillard & Peck
(2000; 2001)
Weiner
(1985)

A feeling people have when
making reasonable progress toward
their goals and when they expect
positive outcomes

Feelings people have when
something is enjoyable

Pride
A feeling people have when
they enhance their ego-identity (i.e.
self- and social- esteem) by taking
credit for an achievement for
themselves or for someone or
groups with who they identify
A feeling people have when
they faced with or challenged
difficulties, or achieved goals

A feeling people have when
they achieve or make a progress
toward their goals and as a function
of self-rewarding

Satisfaction/Contentment
A feeling same with
happiness

A feeling people have when
they are absent from threats and as a
function of conserving resources
A feeling people have when
they can attribute a positive
outcome to themselves
A feeling people experience
when they have an initial
standard/goal and experience some
positive discrepancy from the
standard

Oliver
(1980)

64

Table 2. Coding variables, definitions and sample messages from the participants’ answers
Coding variable

Thoughts
about the
message
content

Definition

Sample messages

Thoughts that were related to

“I realized how much

the content of the message or

waste I produce every

that referenced any features of

single day.”

the message such as credibility

or “I feel like this

and message quality

percentage is not
accurate.”

Types
Thoughts
about
emotions
experienced

Thoughts that referenced any
emotional responses regarding
the message content

“I felt guilty that I do
not recycle the paper I
use.”

MessageRelevant

Thoughts that expressed

Thoughts

positive emotions including
Positive

happiness, satisfaction, pride,
and joy

“[I was] Proud to be a
part of a university that
is making a difference.”

“I was embarrassed at
Valence

Thoughts that expressed
Negative

negative emotions including
guilt, anger, fear, etc.

Neutral

Thoughts that were neither
positive or negative

65

how little I recycle in
comparison with the
statistics mentioned in
the video.”

“I had no idea that
MSU students were so
active in recycling.”

Table 3. Number of items retained, means, standard deviations, and alphas for all scales across
conditions
Scale

No. of Items

Mean

SD

Α

Happiness

5

4.76

5.71

.97

Pride

3

4.69

3.53

.86

Satisfaction

4

4.78

4.30

.94

Guilt

4

3.26

5.56

.95

Attitude 1

5

6.24

6.47

.94

Attitude 2

4

6.26

5.24

.96

Intention 1

3

5.44

4.29

.92

Intention 2

4

5.51

5.90

.97

Involvement

4

4.78

4.01

.64

3

4.98

3.35

.75

6

4.89

4.98

.64

Improved
Involvement
(removed item3)
Social Desirability
Improved Social
Desirability
(removed item 4,5,6)
Believability

3

5.53

2.76

.84

4

5.48

4.31

.87

Persuasiveness

3

5.09

3.97

.93

Descriptive
Normative Content
Perceived
Prevalence
Injunctive
Normative Content
Perceived Approval

4

4.48

6.11

.87

4

4.47

6.55

.97

4

4.59

4.10

.71

4

4.68

3.70

.64

5

2.23

.75

Improved Perceived
Approval
(removed item 1,4)

2

66

Table 3. (cont’d)
Scale

No. of Items

Mean

SD

Α

Puzzlement

3

1.98

1.24

.93

Surprise

3

3.27

3.53

.84

Fear

3

2.08

4.30

.97

Anger

3

2.14

5.56

.94

Sadness

3

1.88

6.47

.94

67

Table 4. Correlation among variables
Satisfaction

Happiness

Pride

Guilt

Attitude

Intention

Involvement

Social
Desirability

Happiness

.694**

Pride

.741**

.659**

Guilt

-.405**

-.295**

-.451**

Attitude

.081

.107

.086

-.156*

Intention

.111

.211**

.240**

.027

.318**

Involvement

.267**

.358**

.343**

-.172*

.268**

.609**

Social
Desirability
Believability

.126

.139

.124

-.167*

.192**

.378**

.303**

.148*

.323**

.164*

-.113

.236**

.387**

.414**

.308**

Persuasiveness

.307**

.423**

.282**

-.069

.201**

.391**

.359**

.212**

Social
Approval
Puzzlement

.210**

.313**

.164*

-.143

.236**

.193**

.221**

.339**

-.192**

-.213**

-.112

.236**

-.079

-.018

-.027

-.090

Surprise

.005

.081

.012

.121

.114

.106

.182*

.130

Fear

-.216**

-.088

-.121

.310**

.019

-.005

.083

-.089

* p < .05, ** p < .01, α < .05 (2-tailed).

68

Table 4. (cont’d)
Satisfaction

Happiness

Pride

Guilt

Attitude

Intention

Involvement

Anger

-.276**

-.338**

-.163*

.245**

-.037

.039

-.027

Social
Desirability
.034

Sadness

-.292**

-.231**

-.173*

.349**

-.024

.053

-.024

-.061

Experimental
-.194**
-.177*
Condition
Prior
-.239**
-.177*
Recycling
* p < .05, ** p < .01, α < .05 (2-tailed).

-.098

.138

-.041

-.022

.057

-.047

-.377**

.204**

-.146*

-.408**

-.314**

-.107

69

Table 4. (cont’d)
Believability Persuasiveness

Social
Puzzlement
Approval

Surprise

Fear

Anger

Sadness

Persuasiveness

.489**

Social
Approval
Puzzlement

.250**

.314**

-.185*

-.034

-.119

Surprise

-.018

.197**

.138

.466**

Fear

-.054

.026

-.106

.655**

.542**

Anger

-.199**

-.081

-.253**

.646**

.414**

.683**

Sadness

-.150*

.053

-.192**

.670**

.456**

.810**

.793**

-.312**

.079

-.108

.155*

.122

.150*

-.085

.017

.039

.008

-.070

-.008

Experimental
Condition

Experimental
.160*
-.023
Condition
Prior
-.138
-.198**
Recycling
* p < .05, ** p < .01, α < .05 (2-tailed).

70

.022

Table 5. Summary of hypotheses and research questions
Hypothesis

Result

H1: The effect of descriptive norm messages on satisfaction

Not consistent with data

will be moderated by prior recycling behavior such that
recyclers, relative to non-recyclers, will experience the
most satisfaction in the low versus high prevalence message
condition.
H2: The effect of descriptive norm messages on guilt will

Not consistent with data

be moderated by prior recycling behavior such that nonrecyclers, relative to recyclers, will experience the most
guilt in the high versus low prevalence message condition.
RQ1: Do recyclers experience greater satisfaction than

Non-significant

other positive emotions when they are exposed to a

difference

message indicating low prevalence?
RQ2: Do non-recyclers experience greater guilt than other
negative or neutral emotions when they are exposed to a
message indicating high prevalence?

Experienced greater guilt
and surprise than other
emotions

H3a: Recyclers will produce a higher number of total

Not consistent with data

message-relevant thoughts, both positive and negative, in
the low versus high prevalence message condition.
H3b: Non-recyclers will produce a higher number of total

Not consistent with data

message-relevant thoughts, both positive and negative, in
the high versus low prevalence message condition.
RQ3: Does the pattern of the valence of thoughts vary by
experimental conditions?

Non-significant
difference

H4: Recyclers in the low prevalence message condition will
have more positive attitude toward recycling than those
who in the high prevalence message condition.

71

Not consistent with data

Table 5. (cont’d)
H5: Recyclers in the low prevalence message condition will

Consistent with data

be more likely to intend to recycle than those in the high
prevalence message condition.
H6: For recyclers, the effects of descriptive norm messages

Not consistent with data

on behavioral intent will be mediated by satisfaction.
H7: Non-recyclers in the high prevalence message

Not consistent with data

condition will have more positive attitude toward recycling
than those in the low prevalence message condition.
H8: Non-recyclers in the high prevalence message

Consistent with data

condition will be more likely to intend to recycle than those
in the low prevalence message condition.
H9: For non-recyclers, the effects of descriptive norm
messages on behavioral intent will be mediated by guilt.

72

Not consistent with data

Table 6. Mean and standard deviation, t-value, sig. and η2 for satisfaction and other positive
emotions among recyclers

Satisfaction

Recyclers
(N = 109)

Happiness

Pride

High
(N = 59)
Low
(N = 50)
High
(N = 59)
Low
(N = 50)
High
(N = 59)
Low
(N = 50)

.03

.14

.02

1.08

5.2

.06

1.18

4.72a

.05

0.98

5.13

.02

1.09

4.73a

η2

1.49

5.2

Sig.

1.86

SD

t

2.36*

Mean

1.05

4.91a

1.01

Note. Independent t-test was conducted with the descriptive norm messages as an independent
variable and each discrete positive emotion as dependent variables. *p < .05.
Means with the same subscript (a) are not significantly different from one another in the low
prevalence condition at p < .05.

73

Table 7. Mean, standard deviation, t-value, and sig. for guilt and other negative or neutral
emotions among non- recyclers
Mean
High (40)
Guilt
Puzzlement
Nonrecyclers
(N = 77)

Low (37)
High (40)
Low (37)
High (40)

Surprise

Low (37)
High (40)

Fear
Anger
Sadness

Low (37)
High (40)
Low (37)
High (40)
Low (37)

1.83 b

SD
1.41
1.32
1.28
1.27
1.72
1.41
1.39
1.62
1.19
1.32
1.22

1.91

1.49

3.5 a
3.71
1.88 b
2.13
3.53a
3.1
1.93 b
2.26
2.06 b
1.95

t

Sig.

-0.67

.50

-0.84

.41

1.21

.23

-0.98

.33

0.39

.70

-0.28

.78

Note. Means with different subscript (a and b) are significantly different from one another in the
high prevalence condition at p < .05.

74

Table 8. The association between the valence of thoughts and experimental conditions
The Valence of
Thoughts
Positive
Negative
Neutral

Experimental Condition
High Prevalence
Low Prevalence
NonNonRecyclers
Recyclers
Recyclers
Recyclers
58
34
18
29
(2.3)*
(-0.5)
(-2.7)*
(0.7)
17
27
17
11
(-1.1)
(1.8)
(-0.1)
(-0.6)
121
108
120
77
(-0.9)
(-0.5)
(1.6)
(-0.1)

χ2

Cramer’s
V

21.97

.13

Note. Adjusted standardized residuals appear in parentheses below group frequencies.
* p = <.05.

75

Table 9. Regression results for attitude toward recycling
B

SE

β

t

Involvement

0.17

0.09

.16

1.98*

.14

Social Desirability

0.07

0.10

.06

0.74

.05

Believability

0.11

0.10

.10

1.11

.08

Persuasiveness

0.04

0.08

.04

0.49

.03

Perceived approval

0.16

0.08

.15

1.87

.13

Sr

First Block

2

F (5, 180) = 4.83, p = .001, adj R = .094
Second Block
Descriptive normative

-0.05

0.18

-.02

-0.25

-.02

-0.15

0.18

-.06

-0.84

-.06

0.11

.01

information
Prior recycling behavior

2

Fchange (2, 178) = 0.39, p = .68, R change = .004
Third Block
Descriptive normative

0.04

0.34

.01

information x
Prior recycling behavior
2

Fchange (1, 177) = 0.01, p = .91, R change < .001

* p = <.05,
Note. Descriptive normative information: high prevalence was coded as 0 and low prevalence
was coded as 1.
Prior recycling behavior: recyclers were coded as 0 and non-recyclers were coded as 1.
sr: semipartial correlation

76

Table 10. Mean and standard deviation for attitude toward recycling across different
experimental conditions

Recycler

Non-recycler

Sum

N
M
SD
N
M
SD
N
M
SD

High Prevalence
59
6.45
0.95
40
6.07
1.12
99
6.30
1.04

Low Prevalence
50
6.33
1.36
37
6.01
1.35
87
6.19
1.36

77

Sum
109
6.40
1.15
77
6.04
1.23
186
6.25
1.19

Table 11. Regression results for intention to recycle
B

SE

β

t

Sr

Involvement

0.62

0.09

.47

7.32**

.41

Social Desirability

0.32

0.10

.20

3.18*

.18

Believability

0.08

0.09

.06

0.88

.05

Persuasiveness

0.19

0.08

.17

2.48*

.14

Perceived approval

-0.06

0.08

-.05

-0.73

-.04

First Block

2

F (5, 180) = 28.29, p = .001, adj R = .424
Second Block
Descriptive Normative
Information
Prior Recycling Behavior

-0.19

0.17

-.07

-1.11

-.06

-0.67

0.17

-.23

-3.96*

-.21

-2.12**

-.11

2

Fchange (2, 178) = 8.63, p = .001, R change = .049
Third Block
Descriptive Normative
Information Χ
Prior Recycling Behavior

-0.68

0.32

-.18

2

Fchange (1, 177) = 4.49, p = .035, R change = .013*

* p = <.05, ** p < .001
Note. Descriptive normative information: high prevalence was coded as 0 and low prevalence
was coded as 1.
Prior recycling behavior: recyclers were coded as 0 and non-recyclers were coded as 1.
sr: semipartial correlation

78

Figure 1. The association between the valence of thoughts and the experimental conditions

The number of message-relevant thoughts

140

Positive
Negative
Neutral

120

121

108

120
100

77
80

58

60
34
40
17

18

27

29 11

17

20
0
Recyclers in the high
prevalence condition

Recyclers in the low
prevalence condition

Non-recyclers in the
high prevalence
condition
Experimental conditions

Non-recyclers in the
low prevalence
condition

Note. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this
thesis.

79

Figure 2. The effects of the descriptive norm messages on intention to recycle depending on prior recycling behavior.

6.5
Low Prevalence
Condition
6.1

High Prevalence Condition
Low Prevalence Condition

6
High Prevalence
Condition
5.92
Intention
to
recycler

5.5
High Prevalence
Condition
4.98

5

4.5

Low Prevalence
Condition
4.57

4
Recyclers

Non-recyclers

Note. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this
thesis.

80

Appendix B: Stimuli

A video clip indicating other students are actively recycling at MSU will be presented. In
the beginning and the end of video clip the descriptive normative information (either high or low)
will be shown.

High prevalence - http://www.youtube.com/watch?v=gV6PmhUqK04

“In 2012, 80% of MSU students recycled the paper they used.”

Low prevalence - http://www.youtube.com/watch?v=jG692jN3c70

“In 2012, 20% of MSU students recycled the paper they used.’

81

Appendix C: Questionnaire Items for Pilot Study

Direction: Please choose the answer that best represents your understanding of what the video
message shows.
Believability
1. I understand what the video is talking about.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

6

7

Strongly agree

6

7

Strongly agree

7

Strongly agree

2. I think the statistics in the video seems believable.
Strong disagree

1

2

3

4

5

3. The statistics in the video seem plausible to me.
Strong disagree

1

2

3

4

5

4. I think the information provided in the video is believable.
Strong disagree

1

2

3

4

5

6

Persuasiveness
To me, the message is
1. not at all convincing

1

2

3

4

5

6

7

very convincing

2. not at all persuasive

1

2

3

4

5

6

7

very persuasive

3. not at all influential

1

2

3

4

5

6

7

very influential

82

Direction. Please choose the answer that best represents your understanding of what the video
message shows.

Perceived Descriptive Normative Content in the Message
1. The information presented above indicates that most students at MSU engage in recycling the
paper they use.
Strong disagree
2.

1

2

3

4

5

6

7

Strongly agree

Clearly, this message shows that not many students at MSU are recycling the paper they use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

3. It is clear that many students have recycled the paper they use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

4. From the information above, it seems that the majority of students at MSU recycle the paper
they use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

6

7

Strongly agree

7

Strongly agree

Perceived Prevalence of Recycling
1. Most students at MSU actively recycle the paper they use.
Strong disagree

1

2

3

4

5

2. Most of the students at MSU actively engage in paper recycling.
Strong disagree

1

2

3

4

5

6

3. Most of the students at MSU recycle the paper they use as actively as they are supposed to.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

4. The majority of the students at MSU actively recycle the paper they use.
Strong disagree

1

2

3

4

83

5

6

7

Strongly agree

Perceived Injunctive Normative Content in the Message
1. Based on the video, I feel like students at MSU would think less of me if I didn’t recycle the
paper I use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

2. It is clear from this information that many students believe paper recycling is important.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

3. The video shows that people at MSU approve of paper recycling.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

4. This video shows that people around me would respect me more if I recycled paper I use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

Perceived Social Approval of Recycling
1. I feel like people would think less of me if I didn’t recycle the paper I use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

2. It is clear that many students believe paper recycling is important.
Strong disagree

1

2

3

4

5

6

7

4

5

6

7

Strongly agree

3. People at MSU approve of paper recycling.
Strong disagree

1

2

3

Strongly agree

4. I think that people around me would respect me more if I recycled the paper I use.
Strong disagree

1

2

3

4

84

5

6

7

Strongly agree

Appendix D: Informed Consent Form - “Environmental Behaviors Study”

Welcome to the “Environmental Behaviors Study” research. We would like to provide
you with a consent form to inform you about the research and convey that participation is
voluntary. If you have any question while you read this consent form, feel free to ask the
researchers.
In the current study, you are being asked to participate in research of Environmental
Behaviors. This research study is being conducted by Michigan State University. This research
project attempts to evaluate a message regarding environmental behaviors. To be an eligible
participant in this study, you must be a MSU student. If you choose to participate in this study,
you will view a video message and be asked to complete a questionnaire that assesses the quality
of the message and your ideas and attitude toward it.
It will take around 30 minutes or less to complete the questionnaire. If you may feel
uncomfortable while answering questions in the current study, do not have to answer any
question you do not want to. Participation in this study is voluntary and you may withdraw your
consent to participate at any time without penalty. The risks of participating are not greater than
those usually met in mundane life. You may not directly benefit from participation in this study,
however, you will contribute to the understanding of young people’s behaviors regarding
environment issues. You will receive an extra credit for the participation after completing the
questionnaire but will not be given money. You may speak with the instructor regarding
completing an alternative assignment.
Your privacy will be protected to the maximum extent allowable by law. No personally
identifiable information will be reported in any research product. Moreover, only trained
research staff will have access to your responses. None of the information that you report will be
linked to you individually. Any individually identifying information that is incidentally obtained
will be deleted from the research record.
This is a study led by researchers in the Department of Communication at Michigan
State University. If you have any question about this study, such as scientific issues, how to do
any part of it, or to report an injury, please contact, Maria Lapinski, Associate Professor, 297
Communication Arts & Sciences, 517-355-9569, lapinsk3@msu.edu.
If you have questions or concerns about your role and rights as a research participant,
would like to obtain information or offer input, or would like to register a complaint about this
study, you may contact, anonymously if you wish, the Michigan State University’s Human
Research Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or
regular mail at 408 W. Circle Dr. Rm 207 Olds Hall, MSU, East Lansing, MI 48824.
By going forward in the study, you indicate your agreement to participate in this research and
have your answers included in the data set.
If you agree please click the button.

85

Appendix E: Questionnaire for Main Experiment
Pre-test

We are interested in learning about how people perceive typical behaviors of MSU
students. We are asking you to estimate the percentage of MSU students who engage in certain
behaviors as well as how often you engage in these behaviors.
For example, recycling is a behavior including sorting recyclable materials such as used
paper, glass, metal cans, plastic, and electronics, and throwing the used materials into each
appropriate recycle bin.
It is not a test, so there are no right or wrong answers. Please answer each item as
carefully and accurately as you can by moving the bar below.
For example, if one thought about half the teenagers in the U.S. drink alcohol once
each week, one would answer 50% by adjusting the bar below.
Instruction. The following items ask you to estimate the percentage of MSU students who
engage in certain behaviors. For each question, your answers can range from 0 (none of MSU
students) to 100% (all of MSU students).
1. In the last 6 months, what percentage of students at MSU do you think recycled the paper they
used (i.e. office paper, magazine, and newspaper) at home or in the dorm? ______
2. In the last 6 months, what percentage of students at MSU do you think recycled the paper they
used (i.e. office paper, magazine, and newspaper) while in class? ______
3. What percentage of students at MSU do you think smoke at least a pack of cigarettes each
week? ______
4. What percentage of students at MSU do you think are at risk for eating disorders (anorexia,
bulimia, or binge eating)? ______
5. In the last 6 months, what percentage of students at MSU do you think recycled the metal cans
they used at home or in the dorm? ______
6. In the last 6 months, what percentage of students at MSU do you think recycled the metal cans
they used while in class? ______
7. In the last 6 months, what percentage of students at MSU do you think recycled the plastic
they used at home or in the dorm? ______
8. In the last 6 months, what percentage of students at MSU do you think recycled the plastic
while in class? ______

86

9. In the last 6 months, what percentage of students at MSU do you think recycled the glass they
used at home or in the dorm? ______
10. In the last 6 months, what percentage of students at MSU do you think recycled the glass
they used while in class? ______
11. What percentage of students at MSU do you think are at risk for obesity? ______
12. What percentage of students at MSU exercises at least 4 times each week? ______
13. In the last 6 months, what percentage of students at MSU do you think recycled the
electronics (a cell phone or a laptop) they used? ______
14. What percentage of students at MSU walks to school or class in order to conserve energy?
______

Instruction. The following items ask you to estimate how often you engage in these behaviors.
For example, if you think that you drink alcohol once each week, you would answer
"Always".

1. In the last 6 months, have you used paper such as office paper, magazine, or newspaper?


1) Yes

2) No  I don't know
3)

If the question 1’s answer is 1) Yes following two questions were asked
1-2. In the last 6 months, how often did you recycle the paper you use at home or in the
dorm?


1) Never

2) Rarely

3) Sometimes

4) Often

5) Always


1-3. In the last 6 months, how often did you recycle the paper you use while in class?


1) Never

2) Rarely

3) Sometime

4) Often

2. In the last 6 months, have you used plastic (e.g. packaging or containers)?


1) Yes

2) No  I don't know
3)

If the question 2’s answer is 1) Yes following two questions were asked

87

5) Always

2-1. In the last 6 months, how often did you recycle the plastic you use at home or in the
dorm?


1) Never

2) Rarely

3) Sometimes

4) Often

5) Always

2-2. In the last 6 months, how often did you recycle the plastic you use while in class?


1) Never

2) Rarely

3) Sometimes

4) Often

5) Always

3. In the last 6 months, how often did you exercise 4 times a week?


1) Never

2) Rarely

3) Sometimes

4) Often

5) Always

4. In the last 6 months, have you used metal cans?


1) Yes

2) No  I don't know
3)

If the question 4’s answer is 1) Yes following two questions were asked
4-1. In the last 6 months, how often did you recycle the metal cans you use at home or in
the dorm?


1) Never

2) Rarely

3) Sometimes

4) Often

5) Always

4-2. In the last 6 months, how often did you recycle the metal cans you use while in
class?


1) Never

2) Rarely

3) Sometimes

4) Often

5) Always

5. In the last 6 months, have you used glass?


1) Yes

2) No

 I don't know
3)

If the question 5’s answer is 1) Yes following two questions were asked
5-1. In the last 6 months, how often did you recycle the glass you use at home or in the
dorm?


1) Never

2) Rarely

3) Sometimes

4) Often

5) Always

5-2. In the last 6 months, how often did you recycle the glass you use while in class?


1) Never

2) Rarely

3) Sometimes

88

4) Often

5) Always

6. In the last 6 months, how often did you walk to school or class in order to conserve energy?


1) Never

2) Rarely

3) Sometimes

89

4) Often

5) Always

Appendix F: Questionnaire for Main Experiment
Post-test

The following questionnaire asks you to evaluate some information about recycling.
Paper recycling is sorting paper which can be recycled and putting the paper into the
right corresponding trash bin before disposal. The recyclable paper includes office paper,
magazine, and newspaper.
The following items are designed to measure feelings you have while watching the
video.
It is not a test, so there are no right or wrong answers. Answer each item as carefully and
accurately as you can.
Aroused Emotions
Direction: The following items are designed to measure feelings you have while watching the
video. It is not a test, so there are no right or wrong answers. Answer each item as carefully and
accurately as you can (1 indicates strongly disagree, and 7 indicates strongly agree).

Happiness
1. While watching the video, I felt happy.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

4

5

6

7

Strongly agree

4

5

6

7

Strongly agree

5

6

7

Strongly agree

5

6

7

Strongly agree

2. While watching the video, I felt joyful.
Strong disagree

1

2

3

3. While watching the video, I felt glad.
Strong disagree

1

2

3

4. While watching the video, I felt delighted.
Strong disagree

1

2

3

4

5. While watching the video, I felt cheerful.
Strong disagree

1

2

3

4

90

Pride
1. While watching the video, I felt proud.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

6

7

Strongly agree

2. While watching the video, I felt accomplished.
Strong disagree

1

2

3

4

5

3. While watching the video, I felt credited for myself because of my own behavior.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

4

5

6

7

Strongly agree

4

5

6

7

Strongly agree

5

6

7

Strongly agree

4

5

6

7

Strongly agree

4

5

6

7

Strongly agree

5

6

7

Strongly agree

6

7

Strongly agree

Satisfaction
1. While watching the video, I felt satisfied.
Strong disagree

1

2

3

2. While watching the video, I felt good.
Strong disagree

1

2

3

3. While watching the video, I felt pleased.
Strong disagree

1

2

3

4

4. While watching the video, I felt contented
Strong disagree

1

2

3

Guilt
1. While watching the video, I felt guilty.
Strong disagree

1

2

3

2. While watching the video, I felt regretful.
Strong disagree

1

2

3

4

3. While watching the video, I felt bad about myself.
Strong disagree

1

2

3

4

91

5

4. While watching the video, I felt remorseful.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

Other Emotions
None of this Feeling(0) __1__:__2__:__3__:__4__:__5__ (6)A Great Deal of this Feeling
While watching the video, I was/felt
1. Puzzled
2. Confused
3. Bewildered
4. Surprised
5. Astonished
6. Amazed
7. Fearful
8. Afraid
9. Scared
10. Angry
11. Irritated
12. Annoyed
13. Sad
14. Dreary
15. Dismal

Thoughts about Paper Recycling
What kind of thoughts do you have while watching the video message? Please list all of the
thoughts you had while watching the video. There is no right or wrong answers.
You will be given 2 minutes to answer it.

92

Attitudes
Direction: The following statements describe your feelings and thoughts on paper recycling. On
the scale below, please choose one answer that best represents your feelings or thoughts about
the paper recycling. There is no right or wrong answer.

Remember,

paper recycling is sorting

paper which can be recycled and putting the paper into the right corresponding trash bin before
disposal. The recyclable paper includes office paper, magazine and newspaper.
To me, paper recycling is
1. Useless

-3

-2

-1

0

1

2

3

Useful

2. Foolish

-3

-2

-1

0

1

2

3

Wise

3. Harmful

-3

-2

-1

0

1

2

3

Beneficial

4. Worthless

-3

-2

-1

0

1

2

3

Valuable

5. Unhealthy

-3

-2

-1

0

1

2

3

Wholesome

To me, paper recycling is
1. Negative
2. Dislikable

-3
-3

3. Bad
4. Undesirable

-2

-1

-2

0

-1

1

0

2

1

3

2

3

-3
-3

-2

-1

0

1

2

-2

-1

0

1

2

3

Positive
Likable
3

Good

Desirable

Behavioral Intent
Direction: For the following items, please indicate your behavior and how much you agree with
the behavior described in each of the following statement.

93

1. Have you recycled paper you use?


1) Yes

2) No

If the question 1’s answer is 1) Yes, the following question will be presented.
In the next two weeks, how likely is it that I will…
1. Continue to sort trash in order to recycle the paper I use.
Not likely at all

1

2

3

4

5

6

7

Very much likely

3

4

5

6

7

Very much likely

2. Continue recycling paper I use.
Not likely at all

1

2

3. Continue to put paper into the corresponding recycle bin.
Not likely at all

1

2

3

4

5

6

7

Very much likely

If the question 1’s answer is 2) No, the following question will be presented.
In the next two weeks, how likely is it that I will…
4. Start to sort trash in order to recycle the paper I use.
Not likely at all

1

2

3

4

5

6

7

Very much likely

3

4

5

6

7

Very much likely

7

Very much likely

5. Start recycling paper I use.
Not likely at all

1

2

6. Start to put paper into the corresponding recycle bin.
Not likely at all

1

2

3

4

94

5

6

Direction: Please indicate how much you agree with the behavior described in each of the
following statement. Your answer can range from not likely at all to very much likely.
1. In the next two weeks, I plan to recycle the paper I use.
Not likely at all

1

2

3

4

5

6

7

Very much likely

2. In the next two weeks, I intend to recycle the paper I use.
Not likely at all

1

2

3

4

5

6

7

Very much likely

7

Very much likely

3. I will recycle any paper I use over the next two weeks.
Not likely at all

1

2

3

4

5

6

4. I plan to engage in paper recycling over the next two weeks.
Not likely at all

1

2

3

4

5

6

7

Very much likely

Issue Involvement
Direction: Please indicate your agreement on each of the following statements (1 indicates
strongly disagree, and 7 indicates strongly agree).
1. Paper recycling is an important issue to me.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

5

6

7

Strongly agree

7

Strongly agree

2. Paper recycling is an issue I care about.
Strong disagree

1

2

3

4

3. To me, paper recycling is a trivial issue. (reverse coded)
Strong disagree

1

2

3

4

5

6

4. I really don't care about the issue of paper recycling. (reverse coded)
Strong disagree

1

2

3

4

95

5

6

7

Strongly agree

Social Desirability
Direction: Please indicate your agreement on each of the following items (1 indicates strongly
disagree, and 7 indicates strongly agree).

1. I find that I can help others in many ways.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

2. In spite of many changes, there are still definite rules to live by.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

4

5

6

7

Strongly agree

3. One can always find friends if he tries.
Strong disagree

1

2

3

4. It is difficult to think clearly about right and wrong these days.
Strongly disagree

1

2

3

4

5

6

7

Strongly agree

5. Many people are friendly only because they want something from you.
Strongly disagree

1

2

3

4

5

6

7

5

6

Strongly agree

7

6. At times I feel that I am a stranger to myself.
Strongly disagree

1

2

3

4

Strongly agree

<The same measures used in pilot study>
Direction: Please choose the answer that best represents your understanding of what the video
message shows.
Believability
1. I understand what the video is talking about.
Strong disagree

1

2

3

4

96

5

6

7

Strongly agree

2. I think the statistics in the video seems believable.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

6

7

Strongly agree

7

Strongly agree

3. The statistics in the video seem plausible to me.
Strong disagree

1

2

3

4

5

4. I think the information provided in the video is believable.
Strong disagree

1

2

3

4

5

6

Persuasiveness
To me, the video message is
1. not at all convincing

1

2

3

4

5

6

7

very convincing

2. not at all persuasive

1

2

3

4

5

6

7

very persuasive

3. not at all influential

1

2

3

4

5

6

7

very influential

Direction. Please choose the answer that best represents your understanding of what the video
message shows.

Perceived Descriptive Normative Content in the Message
1. The information presented above indicates that most students at MSU engage in recycling the
paper they use.
Strong disagree

2.

1

2

3

4

5

6

7

Strongly agree

Clearly, this message shows that not many students at MSU are recycling the paper they use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

3. It is clear that many students have recycled the paper they use.
Strong disagree

1

2

3

4

97

5

6

7

Strongly agree

4. From the information above, it seems that the majority of students at MSU recycle the paper
they use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

Perceived Prevalence of Recycling
1. Most students at MSU actively recycle the paper they use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

7

Strongly agree

2. Most of the students at MSU actively engage in paper recycling.
Strong disagree

1

2

3

4

5

6

3. Most of the students at MSU recycle the paper they use as actively as they are supposed to.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

4. The majority of the students at MSU actively recycle the paper they use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

Perceived Injunctive Normative Content in the Message
1. Based on the video, I feel like students at MSU would think less of me if I didn’t recycle the
paper I use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

2. It is clear from this information that many students believe paper recycling is important.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

3. The video shows that people at MSU approve of paper recycling.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

4. This video shows that people around me would respect me more if I recycled paper I use.
Strong disagree

1

2

3

4

98

5

6

7

Strongly agree

Perceived Social Approval of Recycling
1. I feel like people would think less of me if I didn’t recycle the paper I use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

2. It is clear that many students believe paper recycling is important.
Strong disagree

1

2

3

4

5

6

7

4

5

6

7

Strongly agree

3. People at MSU approve of paper recycling.
Strong disagree

1

2

3

Strongly agree

4. I think that people around me would respect me more if I recycled the paper I use.
Strong disagree

1

2

3

4

5

6

7

Strongly agree

Exposure
1. How much of the video about recycling did you watch?
I watched…..
None of the video message (0)

__1__:__2__:__3__:__4__:__5__ (6) All of the video message

Demographic
Directions: You are about to answer the last set of questions.

The following questions ask

you to describe YOURSELF. (Note. Please choose an answer that best represents you.)
1. I see myself as a….
1) Recycler

2) Non-recycler

3) Neither

2. I know where I can recycle the paper I use around my house.
1) Yes

2) No

3) No place to recycle

99

3. I know where I can recycle the paper I use on campus.


1) Yes

2) No

4. I know how to recycle the paper I use.


1) Yes

2) No

5. Please answer questions below in number. What is your age? For example, if you are 21 yearsold, put the number 21 below.
Your age is

_________ years-old. (e.g. 21)

6. When were you born? For example, if you were born in 1992, put the number 1992 below.
I was born in _________ (e.g. 1992)
7. Gender:

1) Male

2) Female

8. Your ethnicity is:
1) Caucasian
2) African American
3) Asian
4) Asian-American
5) Hispanic-American
6) Native American
7) Hispanic
8) Pacific Islander
9) Other (_________)
9. What are you studying?
My major is _________ (e.g. communication, or advertising etc.)

100

10. What year are you in?
1) Freshman

2) Sophomore

3) Junior

4) Senior

5) Other

(_________)
11. I live
1) on-campus

2) off-campus

3) Other (_________)

12. I live
1) Alone

2)

With roommate(s)

3) With parent(s)

4) Other

(_________)
13. Are you affiliated to any pro-environmental group?
1)

Yes (Please Specify the name of the group_________________)
and how long you have participated in the group _______months)

2)

No

Thank you for your participation!

Please click the button to leave your name, PID and

email address for credit record.

101

REFERENCES

102

REFERENCES

Aarts, H., Verplanken, B., & van Knippenberg, A. (1998). Predicting behavior from actions in
the past: Repeated decision making or a matter of habit? Journal of Applied Social
Psychology, 28, 1355–1374. doi: 10.1111/j.1559-1816.1998.tb01681.x
Anderson, J. (2012). Value-relevant involvement, value-expressive communication, and health
behaviors Retrieved from ProQuest Digital Dissertations. (AAT 3518428)
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human
Decision Processes, 50, 179-211.
doi: http://dx.doi.org.proxy2.cl.msu.edu/10.1016/0749-5978(91)90020-T
Bagozzi, R. P. (1991). Further thoughts on the validity of measures of elation, gladness, and joy.
Journal of Personality and Social Psychology, 61, 98-104.
doi: 10.1037/0022-3514.61.1.98
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182.
doi:http://dx.doi.org/10.1037/0022-3514.51.6.1173
Basil, D. Z., Ridgway, N. M., & Basil, M. D. (2006). Guilt appeals: The mediating effect of
responsibility. Psychology & Marketing, 23, 1035-1054.
doi: http://dx.doi.org/10.1002/mar.20145
Basil, D. Z., Ridgway, N. M., & Basil, M. D. (2008). Guilt and giving: A process model of
empathy and efficacy. Psychology & Marketing, 25, 1–23.
doi: 10.1002/mar.20200
Baumann, D. J., Cialdini, R. B., & Kendrick, D. T. (1981). Altruism as hedonism: Helping
and self-gratification as equivalent responses. Journal of Personality and Social
Psychology, 40, 1039-1046. doi: http://dx.doi.org/10.1037/0022-3514.40.6.1039
Baumeister, R. F., Stillwell, A. M., & Heatherton, T. F. (1994). Guilt: An interpersonal
approach. Psychological Bulletin, 115, 243-267.
doi: http://dx.doi.org/10.1037/0033-2909.115.2.243
Bierbrauer, G. (1992). Reactions to violation of normative standards: A cross-cultural
analysis of shame and guilt. International Journal of Psychology, 27, 181-193.
doi: 10.1080/00207599208246874

103

Bolls, P. D. (2010). Understanding emotion from a superordinate dimensional perspective: A
productive way forward for communication processes and effects studies.
Communication Monographs, 77, 146-152.
doi: http://dx.doi.org/10.1080/03637751003790477
Boster, F. J., Mitchell, M. M., Lapinski, M. K., Cooper, H., Orrego, V. O., & Reinke, R. (1999).
The impact of guilt and type of compliance-gaining message on compliance.
Communication Monographs, 66, 168-177.
doi: http://dx.doi.org/10.1080/03637759909376470
Cacioppo, J.T., & Petty, R.E. (1981). Social psychological procedures for cognitive response
assessment: The thought listing technique. In T. Merluzzi, C. Glass, & M. Genest (Eds.),
Cognitive Assessment (pp. 309-342). New York: Guilford.
Carlsmith, J. M., & Gross, A. E. (1969). Some effects of guilt on compliance. Journal of
Personality and Social Psychology, 11, 232-239.
doi: http://dx.doi.org/10.1037/h0027039
Chaiken, S. (1987). The heuristic model of persuasion. In Zanna, M. P., Olson, J. M., & Herman,
C. P. (Eds.), Social influence: The Ontario symposium, 5, (pp. 3 – 39). Hillsdale, NJ,
England: Lawrence Erlbaum Associates.
Chaiken, S., Liberman, A., & Eagly, A. H. (1989). Heuristic and systematic information
processing within and beyond the persuasion context. In J. S. Uleman & J. A. Bargh
(Eds.), Unintended thought (Vol. 16, pp. 212-252). Guilford Press. Retrieved from
http://books.google.com/books?hl=nl&lr=&id=HT6ddclz6EAC&pgis=1
Chaisamrej, R., & Zimmerman, R. (2007, May). Integrating the theory of planned behavior,
altruism, and self-construal to predict paper recycling behavior in individualistic and
collectivistic societies: Implications for communication campaign. Paper presented at
the conference on International Communication Association, San Francisco, CA.
Retrieved from
http://web.ebscohost.com.proxy1.cl.msu.edu/ehost/pdfviewer/pdfviewer?sid=b8b710
c4-b306-4182-9f32-379395877c47%40sessionmgr104&vid=6&hid=118
Chen, M. (2011). Examining the effectiveness of a fear appeal message regarding calling while
driving: The role of perceived behavioral control, social norms, and the third-person
perception. Unpublished master’s thesis. University of Houston, Houston, TX.
Cho, H., & Boster, F. J. (2005). Development and validation of value-, outcome-, and
impression-relevant involvement scales. Communication Research, 32, 235-264.
doi: 10.1177/0093650204273764
Cialdini, R. B. (2003). Crafting normative messages to protect the environment. Current
Directions in Psychological Science, 12, 105-109. doi: 10.1111/1467-8721.01242

104

Cialdini, R. B. (2011). The focus theory of normative conduct. In P. A. M. Van Lange, A. W.
Kruglanski, & E. T. Higgins (Eds.), Handbook of Theories of Social Psychology: Vol. 2
(pp. 295-313). Thousand Oaks, CA: Sage. doi: 10.4135/978-1-4462-4922-2
Cialdini, R. B., & Goldstein, N. J. (2004). Social influence: Compliance and conformity. Annual
Review of Psychology, 55, 591-621. doi: 10.1146/annurev.psych.55.090902.142015
Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative conduct:
A theoretical refinement and re-evaluation. Advances in Experimental Social
Psychology, 24, 201-234. doi: 10.1016/S0065-2601(08)0330-5
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct:
Recycling the concept of norms to reduce littering in public places. Journal of
Personality and Social Psychology, 58, 1015-1026.
doi: 10.1037/0022-3514.58.6.1015
Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity, and
compliance. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The Handbook of
Social Psychology (pp. 151–192). New York, NY: McGraw-Hill.
Cohen, J. A. (1960). A coefficient of agreement for nominal scales. Education and
Psychological Measurement, 20, 37-46. doi: 10.1177/001316446002000104
Conner, M., & Armitage, C. J. (1998). Extending the theory of planned behavior: A review
and avenues for further research. Journal of Applied Social Psychology, 28, 1429- 1464.
doi: 10.1111/j.1559-1816.1998.tb01685.x
Crites Jr, S. L., Fabrigar, L. R., & Petty, R. E. (1994). Measuring the affective and cognitive
properties of attitudes: Conceptual and methodological issues. Personality and Social
Psychology Bulletin, 20, 619-634. doi: 10.1177/0146167294206001
Dannenberg, C. J., Hausman, B. L., Lawrence, H. Y., & Powell, K. M. (2012). The moral
appeal of environmental discourses: The implication of ethical rhetorics. Environmental
Communication, 6, 212-232. doi:10.1080/17524032.2012.668856
De Young, R. (1985-1986). Encouraging environmentally appropriate behavior: The role of
intrinsic motivation. Journal of Environmental Systems, 15, 281-292.
doi:10.2190/3FWV-4WM0-R6MC-2URB
De Young, R. (1986). Some psychological aspects of recycling: The structure of
satisfactions. Environment and Behavior, 18, 435-449.
doi:10.1177/0013916586184001

conservation-

De Young, R. (1996). Some psychological aspects of reduced consumption behavior: The role
of intrinsic satisfaction and competence motivation. Environment and Behavior, 28,
358-409. doi: 10.1177/0013916596283005

105

Dillard, J. P., & Nabi, R. L. (2006). The persuasive influence of emotion in cancer prevention
and detection messages. Journal of Communication, 56, S123-S139.
doi: 10.1111/j.1460-2466.2006.00286.x
Dillard, J. P., & Peck, E. (2000). Affect and persuasion: Emotional responses to public service
announcements. Communication Research, 27, 461–495.
doi: 10.1177/009365000027004003
Dillard, J. P., & Peck, E. (2001). Persuasion and the structure of affect: Dual-systems and
discrete emotions as complementary models. Human Communication Research, 27,
38–68. doi: 10.1111/j.1468-2958.2001.tb00775.x
Dillard, J. P., Plotnick, C. A., Godbold, L. C., Freimuth, V. S., & Edgar, T. (1996). The multiple
affective outcomes of AIDS PSAs: Fear appeals do more than scare people.
Communication Research, 23, 44–72. doi: 10.1177/009365096023001002
Dillard, J. P., & Shen, L. (2005). On the nature of reactance and its role in persuasive health
communication. Communication Monographs, 72, 144-168.
doi: 10.1080/03637750500111815
Dillard, J. P., Steven, R. W., James, K. T., & Terry A. K. (1997). Politeness judgments in
personal relationships. Journal of Language and Social Psychology, 16, 297–325.
doi: 10.1177/0261927X970163003
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth,
TX: Harcourt Brace Jovanovich.
Ekman, P., & Cordaro, D. (2011). What is meant by calling emotions basic. Emotion Review, 3,
364-370. doi: http://dx.doi.org/10.1177/1754073911410740
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to
theory and research. Reading, MA: Addison-Wesley.
Gleicher, F., & Petty, R. E. (1992). Expectations of reassurance influence the nature of fearstimulated attitude change. Journal of Experimental Social Psychology, 28, 86-100.
doi: http://dx.doi.org/10.1016/0022-1031(92)90033-G
Goldstein, N. J., & Cialdini, R. B. (2011). Managing normative influences in organizations.
In D. D. Cremer, R.V. Dick, & J. K. Murnighan (Eds.), Social psychology and
organizations (pp.67-86). Routledge, NY: Taylor and Francis.
Goldstein, N. J., Cialdini, R. B., & Griskevicius, V. (2008). A room with a viewpoint: Using
social norms to motivate environmental conservation in hotels. Journal of Consumer
Research,35, 472-482. doi: 10.1086/586910

106

Griskevicius, V., Goldstein, N. J., Mortensen, C. R., Sundie, J. M., Cialdini, R. B., & Kenrick,
D. T. (2009). Fear and loving in Las Vegas: Evolution, emotion, and persuasion.
Journal of Marketing Research, 46, 385-395. doi:10.1509/jmkr.46.3.384.
Guetzkow, H. (1950). Unitizing and categorizing problems in coding qualitative data. Journal of
Clinical Psychology, 6, 47-58.
doi: 10.1002/1097-4679(195001)6:1<47::AID-JCLP2270060111>3.0.CO;2-I
Harris, M. (2007, November). The role of emotion in anti-drug PSAs: Investigating the
impact of guilt arousal on perceived message effectiveness and behavioral
intentions to use drugs. Conference Paper presented at the National Communication
Association, Chicago, IL.
Harrison Poll (2007, July 11). One-quarter of Americans do not recycle in their own home:
Majorities, however, do recycle aluminum, paper, plastic and glass.
Retrieved from http://www.harrisinteractive.com/vault/Harris-Interactive-Poll-ResearchFMI-recycling-column-2007-07.pdf
Hibbert, S., Smith, A., Davies, A., & Ireland, F. (2007). Guilt appeals: Persuasion
knowledge and charitable giving. Psychology & Marketing, 24, 723–742.
doi: 10.1002/mar.20181
Hullett, C. R. (2004). Using functional theory to promote sexually transmitted disease (STD)
testing the impact of value-expressive messages and guilt. Communication Research, 31,
363-396. doi: 10.1177/0093650204266103
Jacobson, R. P., Mortensen, C. R., & Cialdini, R. B. (2011). Bodies obliged and unbound:
Differentiated response tendencies for injunctive and descriptive social norms.
Journal of Personality and Social Psychology, 100, 433-448. doi: 10.1037/a0021470
Kaiser, F. G., Ranney, M., Hartig, T., & Bowler, P. A. (1999). Ecological behavior,
environmental attitude, and feelings of responsibility for the environment. European
Psychologist, 4, 59-74. doi: http://dx.doi.org/10.1027//1016-9040.4.2.59
Kaiser, F. G., & Shimoda, T. A. (1999). Responsibility as a predictor of ecological behavior.
Journal of Environmental Psychology, 19, 243-253.
doi: http://dx.doi.org/10.1006/jevp.1998.9123
Kallgren, C. A., Reno, R. R., & Cialdini, R. B. (2000). A focus theory of normative conduct:
When norms do and do not affect behavior. Personality and Social Psychology
Bulletin, 26, 1002-1012. doi: 10.1177/01461672002610009
Kenny, A. D. (2013). Mediation Retrieved from http://davidakenny.net/cm/mediate.htm
Kronrod, A., Grinstein, A., & Wathieu, L. (2012). Go Green! Should environmental messages
be so assertive? Journal of Marketing, 76, 95-102. doi: 10.1509/jm.10.0416

107

Kugler, K., & Jones, W. H. (1992). On conceptualizing and assessing guilt. Journal of
Personality and Social Psychology, 62, 318-327.
doi: http://dx.doi.org/10.1037/0022-3514.62.2.318
Lapinski, M. K., Maloney, E. K., Braz, M., & Shulman, H. C. (2013). Testing the effects of
social norms and behavioral privacy on hand washing: A field experiment. Human
Communication Research, 39, 21-46. doi: 10.1111/j.1468-2958.2012.01441.x
Lapinski, M. K., Rimal, R. N., DeVries, R., & Lee, E. L. (2007). The role of group orientation
and descriptive norms on water conservation attitudes and behaviors. Health
Communication, 22, 133-142. doi: 10.1080/10410230701454049
Lazarus, R. S. (1991). Emotion and adaptation. New York: Oxford University Press.
Lord, K. R. (1994). Motivating recycling behavior: A quasi-experimental investigation of
message and source strategies. Psychology & Marketing, 11, 341–358.
doi:10.1002/mar.4220110404
Loroz, P. S. (2007). The interaction of message frames and reference points in prosocial
persuasive appeals. Psychology & Marketing, 24, 1001-1023.
doi: http://dx.doi.org/10.1002/mar.20193
Massi, L. L. (2005). Anticipated guilt as behavioral motivation. Human Communication
Research, 31, 453–481. doi: 10.1111/j.1468-2958.2005.tb00879.x
Mathur, M., & Chattopadhyay, A. (1991). The impact of moods generated by television
programs on responses to advertising. Psychology & Marketing, 8, 59–77.
doi: 10.1002/mar.4220080106
Meneses, G. D. (2010). Refuting fear in heuristics and in recycling promotion. Journal of
Business Research, 63, 104-110. doi: http://dx.doi.org/10.1016/j.jbusres.2009.02.002
Nabi, R. L. (1999). A Cognitive-Functional Model for the effects of discrete negative
emotions on information processing, attitude change, and recall. Communication Theory,
9, 292–320. doi: 10.1111/j.1468-2885.1999.tb00172.x
Nabi, R. L. (2002).Discrete emotions and persuasion. In J. P. Dillard & M. Pfau (Eds.), The
persuasion handbook: Developments in theory and practice (pp. 289-308).
Thousand Oaks, CA: Sage.
Nabi, R. L. (2010). The case for emphasizing discrete emotions in communication research.
Communication Monographs, 77, 153-159.
doi: http://dx.doi.org/10.1080/03637751003790444

108

Norman, P., Clark, T., & Walker, G. (2005). The theory of planned behavior, descriptive
norms, and the moderating role of group identification. Journal of Applied Social
Psychology, 35, 1008-1029. doi: 10.1111/j.1559-1816.2005.tb02157.x
O’Keefe, D. J. (2002). Guilt as a mechanism of persuasion. In J. P. Dillard, & M. Pfau (Eds.),
The persuasion handbook: Developments in theory and practice (pp. 329–344).
Thousand Oaks, CA: Sage.
Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction
decisions. Journal of marketing research, 17, 460-469.
doi: http://www.jstor.org/stable/3150499
Oliver, R. L. (1993). Cognitive, affective, and attribute bases of the satisfaction response.
Journal of consumer research, 20, 418-430. doi: http://www.jstor.org/stable/2489356
Park, H. S., & Smith, W. S. (2007). Distinctiveness and influence of subjective norms,
personal descriptive and injunctive norms, and societal descriptive norms on
behavioral intent: A case of two behaviors critical to organ donation. Human
Communication Research, 33, 194-218. doi: 10.1111/j.1468-2958.2007.00296.x
Park, H. S., Timothy, R. L., & William, F. S. (1998). The theory of reasoned action and selfconstruals: Understanding recycling in Hawai’i. Communication Studies, 49, 196-208.
doi: 10.1080/10510979809368531
Parker, D., Manstead, A. S. R., & Stradling, S. G. (1995). Extending the theory of planned
behaviour: The role of personal norm. British Journal of Social Psychology, 34, 127137. doi: http://dx.doi.org/10.1111/j.2044-8309.1995.tb01053.x
Pepitone, A. (1976). Toward a normative and comparative biocultural social psychology.
Journal of Personality and Social Psychology, 24, 641-653.
doi: 10.1037/0022-3514.34.4.641
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion.
Advances in Experimental Social Psychology, 19, 123-205. Retrieved from
http://psychology.uchicago.edu/people/faculty/cacioppo/pubs.shtml#1986
Rimal, R. N. (2008). Modeling the relationship between descriptive norms and behaviors: A
test and extension of the Theory of Normative Social Behavior (TNSB). Health
Communication, 23, 103-116. doi: 10.1080/10410230801967791
Rimal, R. N., & Real, K. (2003). Understanding the influence of perceived norms on
behaviors. Communication Theory, 13, 184–203.
doi: 10.1111/j.1468- 2885.2003.tb00288.x

109

Rivis, A., Sheeran, P., & Armitage, C. J. (2009). Expanding the affective and normative
components of the theory of planned behavior: A meta-analysis of anticipated affect
and moral norms. Journal of Applied Social Psychology, 39, 2985-3019.
doi: 10.1111/j.1559-1816.2009.00558.x
Sauter, D. (2010). More than happy the need for disentangling positive emotions. Current
Directions in Psychological Science, 19, 36-40. doi: 10.1177/0963721409359290
Schaffer, L. S. (1983). Toward Pepitone’s vision of a normative social psychology: What is a
social norm? Journal of Mind and Behavior, 4, 275-294.
doi: 10.1037/0022-3514.34.4.641
Schuessler, K., Hittle, D., & Cardascia, J. (1978). Measuring responding desirably with attitudeopinion items. Social Psychology, 41, 224-235. doi: http://www.jstor.org/stable/3033559
Schultz, P. W., Nolan, J. M., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V. (2007). The
constructive, destructive, and reconstructive power of social norms. Psychological
Science, 18, 429-434. doi: 10.1111/j.1467-9280.2007.01917.x
Schwarz, N. (1990). Feelings as information: Informational and motivational functions of
affective states. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook of motivation and
cognition: Foundations of social behavior (Vol. 2, pp. 527-561). New York: Guilford.
Shaver, P., Schwartz, J., Kirson, D., & O'connor, C. (1987). Emotion knowledge: Further
exploration of a prototype approach. Journal of personality and social psychology, 52,
1061-1086. doi: 10.1037/0022-3514.52.6.1061
Smith, S. W., Atkin, C. K., Martell, D., Allen, R. and Hembroff, L. (2006). A social judgment
theory approach to conducting formative research in a social norms campaign.
Communication Theory, 16, 141–152. doi: 10.1111/j.1468-2885.2006.00009.x
Smith, S. M., Haugtvedt, C. P., & Petty, R. E. (1994). Attitudes and recycling: Does the
measurement of affect enhance behavioral prediction? Psychology & Marketing, 11,
359-374. doi: 10.1002/mar.4220110405
Stern, P. C., Dietz, T., Abel, T., Guagnano, G. A., & Kalof, L. (1999). A value-belief-norm
theory of support for social movements: The case of environmental concern. Human
Ecology Review, 6, 81-97. Retrieved from
http://ezproxy.msu.edu/login?url=http://search.proquest.com/docview/61453435?acc
ountid=12598
Thøgersen, J. (1996). Recycling and morality: A critical review of the literature. Environment
and Behavior, 28, 536-558. doi: 10.1177/0013916596284006

110

Viscusi, W. K., Huber, J., & Bell, J. (2011). Promoting recycling: Private values, social norms,
and economic incentives. American Economic Review, 101, 10-36.
doi: http://ssrn.com/abstract=1724924
Wegener, D. T., Petty, R. E., & Smith, S. M. (1995). Positive mood can increase or decrease
message scrutiny: The hedonic contingency view of mood and message processing.
Journal of Personality and Social Psychology, 69, 5-15.
doi:http://dx.doi.org/10.1037/0022-3514.69.1.5
Weiner, B. (1985). An attributional theory of achievement motivation and emotion.
Psychological Review, 92, 548-573. doi: http://dx.doi.org/10.1037/0033-295X.92.4.548
White, K., MacDonnell, R., & Dahl, D. W. (2011). It's the mind-set that matters: The role of
construal level and message framing in influencing consumer efficacy and
conservation
behaviors. Journal of Marketing Research, 48, 472-485.
doi: 10.1509/jmkr.48.3.472
Zuckerman, A., & Chaiken, S. (1998). A heuristic‐systematic processing analysis of the
effectiveness of product warning labels. Psychology and Marketing, 15, 621-642.
doi: 10.1002/(SICI)1520-6793(199810)15:7<621::AID-MAR2>3.0.CO;2-H

111