LIBRARY
Michigan State

University

 

 

This is to certify that the
dissertation entitled

PERCEPTIONS OF THE AVERSIVENESS TEASING AS A
FUNCTION OF PERCENED IDENTITY CONFRONTATION,
AGGRESSION. AMBIGUITY & HUMOR, AND THE PERCEPTION
OF INTENT TO CAUSE HARM

presented by

Carrie Marie Oliveira

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Communication

744044454 blame/um, ._..

Major Professor’s Signature
5’"; ‘1‘ '67 007

Date

MSU is an affirmative-action, equal—opportunity employer

-.- - -.-.-.-.-.-.-,-.-.-4-
---‘-:---«-v-.-- ----------------.-‘--<- _.-.-.--I-.-.-.- -.-.-.-.-.-.-.

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:ICIRC/DaIeDue.indd-pr1

 

PERCEPTIONS OF THE; AVERSIVENESS TEASING AS A I-‘UNC’I‘ION OF
PERCEIVED IDENTITY CONFRONTA'I‘ION. AGGRESSION. AMBIGt.zI':"~.-' It»,
HUMOR, AND THE PERCEPTION OF [NTE‘NT TO CAUSE HARM
By

Carrie Marie Oliveira

- A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Communication

2007

ABSTRACT
PERCEPT IONS OF THE AVERSIVENESS TEASING AS A FUNCTION OF
PERCEIVED IDENTITY CONFRONTATION, AGGRESSION, AMBIGUITY &
HUMOR, AND THE PERCEPTION OF INTENT TO CAUSE HARM
By
Carrie Marie Oliveira

The goal of this paper was to examine factors contributing to the perceptions of the
aversiveness of teases. Kowalski (2004) asserts that numerous factors result in the
perception of message aversiveness. Based on these factors, Kowalski (2004) asserts
that identity confrontation, aggression, ambiguity and humor influence the extent to
which a tease is perceived as aversive. This contention is somewhat at odds with the
prevailing belief in the literature that the aversiveness of a tease is a fiJnction of the extent
to which the target of a tease perceives that the agent teased with malicious intent.
Tedeschi and Nesler (1993) provide a theoretical framework through which to test the
relationship between perceived intent to harm and perceptions of aversiveness. Both
Kowalski's definitional components model and a model of intent based on a theory of
grievances were tested in order to identify those variables that best predict the perception
of a tease as aversive. To test these models, participants responded to an online survey
asking for the description and evaluation of a recent interpersonal interaction in which
they had been the targets of a tease. Results indIcate that identity confrontation,
aggression. ambiguity and humor are all substantially and statistically significantly
related to perceptions of tease aversiveness, however only aggression and humor were

shown to be significant predictors of aversiveness. Both intent and foreseeability ofharm

were significant predictors of aversiveness. Implications of these findings suggest that a

comprehensive model of teasing should include aggression. humor, perceived intent to
harm and perceived foreseeable harm. Pending the retesting and replication of these
findings, this model provides greater understanding of teasing as a social phenomenon
and can inform policy and law-makers interested in implementing anti-teasing safe-
school initiatives. Future research should focus on the extent to which a revised model of

teasing impacts positive relational outcomes such as intimacy, closeness or solidarity.

To Gramps.

iv

ACKNOWLEDGMENTS
For their patience, guidance, open-mindedness, flexibility, and insightful feedback, I wish
to thank my committee members Dr. Nicole Ellison and Dr. Kami Silk. For asking the
hardest questions I have ever been asked to answer, and for his service on my committee,
I would like to thank Dr. John Sherry. For allowing me the freedom I needed and wanted
to make this paper my own, I wish to thank my advisor Dr. Bill Donohue. For their
endless support and encouragement, even from across half a continent and an ocean, I
wish to thank Ma, Daddy, and Nerbie. Finally, for being far closer to a family in
Michigan than I could have ever dreamed possible, and for being the most stellar of

cheering sections, I wish to thank Spicy, Magic, Rie, Mikey, and Mama Erin.

TABLE OF CONTENTS

LIST OF TABLES .................................................................... . ....................................... vii
LIST OF FIGURES ............................................................................... viii
INTRODUCTION
Purpose .............................................. . ........................................................... - ........ I
Rationale ............................................ . .................................................................... 2
Teasing Defined ...................................................................................................... 4
Aversiveness ........................................................................................................... 7
The Relationship between the Components of Teasing and Perceptions of
Aversiveness ......................................................................................................... I 3
lntentionality ......................................................................................................... 22
Summary and Hypotheses ..................................................................................... 25
METHOD
Participants ............................................................................................................ 28
Procedure .............................................................................................................. 28
Rationale for survey study .................................................................................... 3O
Instrumentation ..................................................................................................... 32
RESULTS
Participant Characteristics .................................................................................... 37
Tests of Linear Associations between Teasing Components and Aversiveness... 37
Test of a Component Model of Teasing ............................................................... 38
Test of Pattern of Different Message Types along Measures of Teasing
Components .......................................................................................................... 39
Test of a Model of Perceived lntentionality ......................................................... 41
Research Question: Which Model Do the Data Fit Better? .................................. 42
DISCUSSION
Interpretation of Findings ..................................................................................... 43
Limitations of the Study ........................................................................................ 47
Future Research .................................................................................................... 48
Summary ............................................................................................................... 50
APPENDICES
STATISTICAL TABLES ..................................................................................... 54
FIGURES .............................................................................................................. 58
SURVEY ITEMS .................................................................................................. 61
REFERENCES ................................................................................................................. 65

vi

LIST OF TABLES

Table 1. Scale Reliabilities and Descriptive Statistics across Four Teasing Contexts ..... 54

Table 2. Bivariate Correlation Matrix for Subject-Provided Teases ................................ 54
Table 3. Beta Coefficients for Teasing Components as Predictors of Aversiveness ........ 54
Table 4. Frequency of Message Classifications for Researcher-Provided Remarks ........ 55

Table 5. Means and Contrast Values for Researcher-Provided Remarks for Teasing

Components ...................................................................................................................... 55
Table 6. Beta Coefficients for Intent Variables as Predictors of Aversiveness ................ 55
Table 7. Beta Coefficients for Revised Model of Teasing ................................................ 56
Table 8. Scale Items and Factor Loadings ........................................................................ 61

vii

LIST OF FIGURES
Figure 1. Model of the aversiveness as a function of the components of a tease ............. 58

Figure 2. Model of aversiveness as a function of perceptions of intentional or
foreseeable harm ............................................................................................................... 58

Figure 3. Revised model of teasing ................................................................................... 59

viii

INTRODUCTION
Purpose

The purpose of this study is to address a currently divided body of literature
regarding teasing. Extant literature suggests that teases fall into one of two categories —
prosocial or antisocial. The dichotomization of teases results in two erroneous
impressions regarding teasing. First, that teases are prosocial or antisocial rather than
ranging on a continuum from prosocial to antisocial, and second that prosocial and
antisocial teases are two different types of messages. The goal of the current
investigation, then, is to clarify the conceptual definition of teasing so that it captures the
fiIIl range of teases and specifies a tease as a single message type.

As a subsequent goal to clarifying the conceptualization of teasing is the
identification of the features of teases that cause variability in the extent to which they are
perceived as aversive. Specific characteristics of a tease have been argued to have an
influence on the extent to which teases are perceived as negative. More commonly,
however, the intent of the agent enacting a tease is thought to have substantial influence
on the aversiveness of a tease. Empirical tests of the relationships between tease message
features and aversiveness will be compared to a test of the relationship between agent
intent and aversiveness in order to identify the variables that best explain the degree to
which teases are regarded as negative.

The overarching goal of the current paper, then, is to propose a comprehensive
and testable model of teasing. In doing so, a more theoretical approach to the study of

teasing and its effects on individuals and relationships can be taken. Improved theory

about teasing will enable a more unified body of literature to emerge which will
subsequently improve our understanding of teasing as a social phenomenon.
Rationale

The importance of understanding teasing as a social act goes beyond theory
building. Teasing has the ability to leave its victims with a variety of serious and
undesirable consequences. Targets of teasing have reported a variety of negative
outcomes including feelings of depression (Keery, Boutelle, van den Berg, & Thompson,
2005), anxiety (McCabe, Antony, Summerfeldt, Liss, & Swinson, 2003), neuroticism
(McCabe et al., 2003), diminished body image, eating disturbance (Keery et al., 2005;
Lunner et al., 2000) and decrease in self-esteem as it relates to body image (Gleason,
Alexander, & Somers, 2000). For a review of this literature, see McCabe et al. (2003).

The aforementioned consequences provide reasonable justification for the need of
a thorough understanding of teasing as a social phenomenon. This need increases
significantly, however, when one considers that teasing has been implicated as a
contributing factor in a number of recent violent and fatal incidents on campuses across
the country (Leary, Kowalski, Smith, & Phillips, 2003). These violent incidents have led
to the demand for safe-school initiatives designed to increase awareness of the effects of
teasing (Simmons, 2003) and to aid in the identification and assistance of students who
might be particularly susceptible to teasing and its adverse effects (Horowitz et al., 2004).

A problem arises, however, when an initiative is designed without a complete
understanding of the problem it is designed to resolve. Smith and Shu (2000) evaluated
the effectiveness of a government-mandated anti-bullying campaign in England schools.

Smith and Shu (2000) indicate that the most fi'equent type of bullying experienced by

IQ

subjects were single instances of name-calling. Single instances of name-calling however,
are not the type of teasing behavior that Leary et al. (2003) argue was responsible for a
number of school shootings. Instead, persistent and relentless victimization is associated
with enacting retaliatory violence (Leary et al., 2003). Persistent victimization was
reported in over 50 percent of a sample of over 2,300 students, with approximately 39
percent of the sample reporting having been victimized from anywhere between the entire
length of a school term to the course of several years (Smith & Shu, 2000). Regardless of
the number of students reporting persistent victimization in studies prior to 2000, for
nearly 40 percent of a sample to report being repeatedly teased over substantial lengths of
time suggests that anti-bullying initiatives may not be as effective as expected.

The question arises, then, as to whether anti-bullying initiatives are being
efficiently aimed. In a cursory review of titles of scholarly work articulating the need for
anti-bullying interventions, the words teasing and bullying are often used
interchangeably, suggesting that the experience of being teased is the same as the
experience of being bullied. The assumption that bullying and teasing are synonymous
becomes implausible given that bullying seems to be unambiguously negative while a
number of studies suggest that teasing is often positive and beneficial. Among the
positive outcomes of teasing are increased intimacy and a reinforcement of idiosyncratic
relational culture in close relationships (Baxter, 1992). Keltner, Oemig, & Monarch
(I998) review literature that suggests that teasing in close relationships serves as a face-
saving means of correcting nonn-violating behavior.

Given that bullying seems to be unambiguously negative, safe-school initiatives

are well-advised to focus their energy on the elimination of bullying from campuses.

Teasing, on the other hand, has been associated with both negative and positive outcomes
for individuals and relationships. The question then becomes whether teasing should be
eradicated from schools along with bullying, or whether it should be Iefi alone as a
normal and acceptable form of communication. The pragmatic aim of this study is to
address this question.

In order to gain a better, more thorough understanding of teasing and its place in a
safe-school initiative, the development of a clear conceptual definition of teasing and a
move toward a theoretical model of teasing and its outcomes is critical. Currently, the
literature betrays the discipline’s confusion about what constitute teasing. As such, it is
necessary to begin by defining what teasing is.

Teasing Defined

As with any abstraction in the social sciences, multiple conceptualizations of
teasing have been proffered in the literature (see Keltner, Capps, Kring, Young, &
Heerey, 2001 for an extensive review of different conceptual definitions of teasing).
Keltner et al. (2001) make a compelling argument as to why the various definitions of
teasing used in the upwards of fifty studies they examined were inadequate. Specifically,
Keltner et al. (2001) argue that the definitions used in the studies they reviewed suffered
from being overly context- or behavior-specific, components used in the definitions
suffer from ambiguity (such as in the case of the use of the word playful to describe a
tease), teases are often not differentiated from bullying (which may or may not involve
the teasing of a target), and that these definitions classify teasing as either prosocial or
antisocial and do little to resolve what Keltner et al. “a central tension in the literature on

teasing” (p. 232), further suggesting that the definition of a tease should account for both

antisocial and prosocial outcomes. In an attempt to correct their own concerns about
inadequate conceptualization, Keltner et al. (2001) propose a new definition, “a tease is
an intentional provocation accompanied by playful off-record markers that together
comment on something relevant to the target” (p. 234).

While Keltner et al.’s efforts at composing a definition that is not context-specific
and can account for teases that have both prosocial and antisocial outcomes, the
definition they propose is still plagued by the ambiguity of the term “playfulness” and
does not include any characteristics that might explain why a tease would be responded to
adversely. In an attempt to correct a definitional problem, Keltner et aI. bring important
conceptual issues to bear while simultaneously failing to account for all of their own
criticisms of existing definitions.

A definition of teasing proposed by Kowalski (2004) perhaps provides a more
suitable means of remedying Keltner et al.’s concerns. Kowalski defines teasing as “a
personal communication from an agent to a target that includes four components:
aggression, humor, ambiguity and identity confrontation”(p. 332). For the purposes of the
current study, this definition will be adopted, but adapted to include Keltner et al.’s
(2001) criterion that the message be about something personally relevant to the target. A
failure to adopt that criterion in the definition of teasing leaves room for a comment to be
made by an agent to a target that meets all four criteria but is about, for example,
someone close to the target. Such a comment might be likely to arouse the target, but is
not of the same character as a remark made to the target about herself. Thus, a tease for

the purposes of the current investigation is defined as a personal communication from an

agent to a target that comments on something personally relevant to the target, and that
includes four components: aggression, humor, ambiguity and identity confrontation.

The reason this definition is preferred is that it resolves all of the concerns raised
by Keltner et al. in their criticism of teasing definitions in existence at the time of their
2001 writing. The definition being used here is in no way context or situation specific. It
does include additional terms which need to be fiIrther explicated in order to fully
understand the nature of the definition of teasing, but the successful explication of the
four components of the definition of teasing resolves this problem.

Keltner et al. were also concerned about the failure of most definitions of teasing
to differentiate the construct from bullying. There seems to be a lack of a strong
conceptual definition of bullying in the literature. For example, Keltner et al. (200]) who
dismay over the fact that teasing and bullying are often conflated do not provide a
definition of bullying from which to differentiate teasing except to say that teasing is
likely a tactic used by bullies. Solber, Olweus, and Endresen (2007) define bullying more
concretely as a subset of aggressive behavior (that is, behavior that is designed to injure
or inflict pain) that is generally characterized by repetitive direction of hurtful actions to a
victim, who is often less-able or unable to defend against attack. Based on this definition,
bullying is characterized by repetitive or persistent attack. The means by which persistent
attacks occur might be teasing, but are also likely to include some types of physical
violence or intimidation. Teasing, while potentially employed by a bully, is not
synonymous with bullying. The definition of a tease being used here specifies a tease as a

particular message type rather than a pattern of behavior.

Finally, Keltner et al.’s primary concern, and the one they failed to adequately
address, is that a definition of teasing which relies on message components allows for the
explanation of both prosocial and antisocial teases. While Keltner et al.’s (2001)
definition did not specifically differentiate between prosocial and antisocial teases,
nothing in the definition explained the potential for adverse reactions to a tease. The
definition borrowed from Kowalski (2004), on the other hand, is accompanied by
Kowalski’s explanation that different responses to a tease are marked by different relative
levels of each of the four components of the tease. She provides the example that cruel
teasing and bullying are high in aggression and low in humor, where prosocial (using her
terminology) teasing is characterized by lower levels of aggression and high levels of
humor. Thus, all teases are defined by the same message features, and varying levels of
these message features result in teases that are variable in terms of their negative impact
on the target and the relationship.

Understanding the impact of the components of teasing on the perceptions of
aversiveness of the tease necessitates an understanding of the origin of perceptions of
aversiveness and an evaluation of the sources of aversiveness in social interactions. The
following section identifies the origin of aversiveness, and specifies the particular ways
that the components of the definition of teasing as argued above contribute to the
evaluation of a tease as aversive.

Aversiveness

The origin of perceptions of aversiveness. The evaluation of a stimulus as

aversive not only connotes that the stimulus was unpleasant but that there is a strong

desire for the stimulus to be subsequently avoided. Pain or other discomfort is often a

characteristic of aversive stimuli, often for good reason — if a stimulus hurts it is likely to
do damage, and is therefore wisely avoided. In social interactions, we very often do not
experience the sort of pain that keeps us from harmful stimuli in the physical world, but
we do very often experience hurt. Specifically, we experience hurt feelings as the result
of aversive occurrences in our social interactions. Hurt feelings generally come about as
the result of the perception that one has been relationally devalued (Leary & Springer,
2001). An individual feels relationally devalued when she feels as though her partner
values a relationship with her to a lesser degree than a) she desires, or b) than was
exhibited previously. At the center of feeling relationally devalued, is the fact that human
beings have a very basic and fundamental need to belong (Baumeister & Leary, 1995).
We desire close contacts with others, we want to know that we fit into a social group and
that our company is desired and valued by others. Feeling relationally devalued, then, is
in direct opposition to our need to for inclusion, and subsequently hurts. Leary and
Springer (2001) specifically note that teasing is likely to be hurtful to the extent that the
tease is perceived as being relationally devaluing.

Unfortunately, Leary and Springer (2001) do not specifically suggest what it is
about the nature and character of teasing or the teasing situation that may make it more or
less likely to result in the perception of relational devaluation. To answer this question,
we must understand the potential sources of aversiveness in social interactions as well as
how these sources give rise to the potential for teasing to hurt.

Sources of aversiveness. Perhaps the most commonly researched source of
aversiveness in interpersonal interaction is the intent to cause harm. The size of this

debate in the teasing literature makes it seem as though intentional ill-will is the sole

source of perceptions of aversiveness in teasing interactions. This impression, however, is
an artifact of a negative bias in the literature and is untrue for two reasons. First the
longer a relationship lasts, the more likely relational partners are to become careless with
one another and enact behaviors with aversive outcomes with no ill-intent whatsoever
(Kowalski, 2001; Miller, 2001). Miller (2001) argues that the character of close
relationships changes over time such that relational partners become careless with one
another, we take more liberties with one another, we become selfish, we become lax in
controlling our impulses to behave badly, and we engage in less effort to treat our
partners well. These lazy relationship behaviors increase the incidence of aversive
communication in even the closest of relationships. This suggests that relational partners
who inflict negative outcomes on a relational partner have no positive or negative
intention in mind. Secondly, Kowalski (Kowalski, 2001) contends that there numerous
factors that contribute to perceptions of the aversiveness of an action on the part of a
partner, only one of which is the perceived intent of a partner to do harm. Violations of
basic psychological needs, relationship characteristics, message characteristics and
individual differences all have the potential to influence the degree to which an
interpersonal exchange is regarded as aversive. The following section will discuss the
some of the proposed sources of aversiveness, specifying which ones are likely to be
usefiIl in understanding the aversiveness of teases.

Kowalski claims that all aversive interactions interfere with an individual’s basic
human needs for inclusion, affiliation and control (Kowalski, 2001, 2003). Teasing is
most certainly capable of making a person feel excluded, disliked, and out of control of

his or her social standing or situation, but these are not characteristics of a tease so much

as outcomes of teases perceived as aversive. To include them in the definition of a tease
as message characteristics makes the definition vague and conflates message features
with message outcomes.

lnappropriateness can lead to the perception of a message as aversive. While this
is a perfectly sound contention, it can be argued that a tease which is highly aggressive or
identity confronting, for instance, can be regarded as inappropriate. In short, any number
of factors combine to fit under the umbrella of propriety, thus assessing the
appropriateness of a tease is likely to provide no more information about perceptions of
aversiveness than would the assessment of specific violations of propriety.

Cultural or relational norms can also influence the perception of a tease as
aversive. For example, what is normative behavior between a man and his friends may be
entirely non-normative and ill-received if enacted between him and his girlfriend. The
trouble with including norrnativity in the definition of teasing makes the definition
context-bound, and violates Keltner et al.’s (200]) concern about the lack of
generalizability across contexts. Considering the extent to which a tease is normative is
also troublesome in light of another of Kowalski’s aversiveness-inducing message
feature: frequency. Kowalski contends that the more often a tease occurs the more likely
it is to be perceived as aversive. Kowalski argues that a repeated tease becomes
especially distressing when repeated requests for the behavior to stop have been made.
Such teases are classified as social allergens (Kowalski, 2003; Kowalski, Howerton, &
McKenzie, 2001) and have the same irritating effect as a persistent runny nose. It can be
argued, however, that a runny nose is fundamentally irritating, and the longer it persists,

the more irritating it becomes. This is likely also to be the case for a tease that has

10

become an allergen — that is the tease was likely regarded as aversive the first time it was
enacted and its repetition has only made it more aversive. Frequency does not at all
account for the initial perception of the tease as aversive, and largely uninformative in
understanding the basic perceptual process that results in an aversive evaluation of a
tease. Kowalski also discusses severity of a message in conjunction with frequency as
contributing to the allergic potential of a message. The extent to which a tease is severe
can be assessed by rating the extent to which it bears the characteristics of aggression,
ambiguity and identity confrontation. To argue that severity in a vague sense contributes
to perceptions of aversiveness is also uninformative.

Relational intimacy may also increase the aversive potential of an action because
the people with whom we keep particularly close ties are also more likely to impact our
emotional states (Kowalski, 2001 ). Moreover when a close other does impact our
emotions, the effect is felt more strongly than when our emotions are affected by
strangers. While this argument is compelling, intimacy is a feature of a relationship and
not a message. So while the intimacy of a relationship may influence the extent to which
a tease is regarded as aversive, it is outside the scope of the current investigation as it is
not a message feature.

Somewhat at odds with the contention that intimacy increases perceptions of
aversiveness is belief that relational type also influences aversive perceptions.
Specifically, Kowalski (2001) argues that we are more likely to overlook the potentially
aversive actions of a friend or family member when compared to the same action being
committed by an acquaintance or stranger. The argument here is that we are more likely

to make attributions of our intimates that they were out of control of the situation that

ll

caused the aversive behavior and forgive them, but we will make the attribution of a
stranger that they behaved badly on purpose. The different attributions we make about the
ability to control aversive actions for intimates and strangers are beyond the scope of the
current investigation. It should not, however, be disregarded as a potential contributing
factor to perceptions of tease aversiveness, especially when tested in conjunction with
intimacy.

Kowalski (2001) also includes individual differences as pctentially moderating
factors influencing the extent to which a behavior is likely to be construed as aversive.
For example, hurt proneness (Leary & Springer, 2001) or perceived risk in intimacy
(Kowalski, 2001) have been argued to increase the extent to which an individual
experiences negative outcomes as the result of aversive stimuli. Certainly, personality
attributes can influence a person’s experience of an event. However. there is very likely a
perceptual process that operates in the same manner for all people and individual
differences do nothing more than moderate that process. Understanding the moderating
role of individual differences is only meaningful and useful when the process these
differences are argued to moderate is well understood. Given the lack of a strongly
theoretical model of teasing, to specify the role of individual differences in the experience
of teasing is premature.

Kowalski’s specifies five additional message and communicator features that are
likely to influence perceptions of aversiveness. These include identity confrontation,
aggression, ambiguity, and intentionality. The first four of these factors are components
of the teasing definition borrowed from Kowalski (2004). lntentionality has been

purposefully left out of the definition of teasing, but is an important contributor to

12

perceptions of aversiveness and is the focus of most of the teasing research currently
written to date. In the following section, each of these four factors will be described and
an argument will be presented for their hypothesized influence on the aversiveness of a
tease. In addition to identity confiontation, aggression, ambiguity and intentionality,
humor will also be discussed as a factor contributing to aversiveness. Unlike the other
four concepts, however, humor is expected to diminish the aversive effects of a tease.
This argument will be further expanded upon in the following section.

The Relationship between the Components of Teasing and Perceptions of A versiveness

Identity and identity confiontation. The origin of an identity confrontation is the
failure of an interaction partner to behave in a manner that supports the image of
ourselves that we have chosen to portray. Successful support of our chosen identity has
been referred to as facework (Goffinan, I967), identity management (lmahori & Cupach,
2005), and communication competence (Spitzberg & Cupach, 1984). Despite varying
terminology, the crux of the argument is that in an interaction, an individual makes a
conscientious decision to portray herself in a particular manner (Goffman, 1959, 1967),
and in order for the interaction to be regarded as successful and satisfying, an individual
has the expectation that a conversational partner will behave in a way that allows her to
be seen in the manner in which she chooses.

An identity confrontation, on the other hand, directly challenges the image of
herself that a person has chosen to convey. It draws attention to incomplete or
inconsistent presentation of an identity, and arguably damages the target’s perception of
her ability to effectively engage in the interaction (Goffinan, I959). The more a tease is

in opposition to the aspect of identity being enacted at the time, the more likely the target

13

of the tease is likely to feel invalidated by a communicative partner (Spitzberg & Cupach,
1984). Further, a failure to have a preferred or enacted identity supported by a
communicative partner often leads to the perception of disrespect (Ting-Toomey, 2005).
While not specifically referenced as a source of relational devaluation (Leary & Springer,
2001), it stands to reason that feeling as though we are not validated or respected by a
communicative partner could lead to feelings of relational devaluation. Thus,

H I .' the greater the extent to which a tease is perceived as identity

confronting, the greater the extent to which that tease will be perceived as

aversive.

Aggression. The second message component of a tease, according to Kowalski
(2004), is aggression. Aggression has been used by scholars to describe a wide variety of
behaviors. In some cases aggression refers to acts of physical violence such as fighting
with or shooting another person (Larsen & Dehle, 2007). In other cases, aggression refers
to a coercive attempt such as in the case of rape (Warkentin & Gidycz, 2007). In yet other
cases, aggression refers to a verbal act designed to attack the self-concept of the target
(Infante & Wigley, 1986). Different uses of the word aggression (as is often the case in
the social sciences) lead to confiIsion and lack of clarity with regard to what the construct
is. The one aspect that all three of these usages have in common is that they are used to
refer to some kind of an attack on a target by an agent, whether by physical or
psychological means. Solberg et al. (2007) very articulately specify aggression as
behavior enacted with the goal of inflicting pain or injury.

There is very little discussion in any work that studies aggression on why it hurts

— presumably because the outcomes of being physically attacked, including physical pain,

14

injury, and death are invariably negative. Where teasing is concerned, however, the
discussion of why aggression hurts is a bit more important. The old children’s rhyme
about sticks and stones seems to suggest that words — even ones that attack — do not hurt.
Many a victim of a highly aggressive tease would beg to differ. So the question is: why
does a verbal attack hurt?

A simple explanation is that all aggression is construed the same way — as an
attack. Arguably, we develop the use of verbal aggression in childhood as an alternative
to physical aggression when physical displays of aggression are no longer tolerated by
parents and teachers (Forrest, Eatough, & Shelvin, 2005). Children move from slapping
and hitting to yelling and swearing. This often evolves into more indirect forms of
aggression such as manipulation and other sorts of interpersonal sabotage (Forrest et al.,
2005; Simmons, 2003). The adoption of newer, more subtle aggressive behaviors in place
of old, overt behaviors is a pattern of development that occurs over the lifespan. More
simply stated, verbally aggressive attacks become substitutes for physical attacks. The
implication of this in terms of the impact of aggression on perceptions of aversiveness is
that we come to translate different behaviors as attacking over the course of our
development. Regardless of the form an attack takes, any apparent attack causes
defensiveness and the feeling that we must protect ourselves from a partner. The result is,
once again, a feeling of relational devaluation. A partner whom we have trusted and to
whom we have made ourselves vulnerable has betrayed trust, taken advantage of
vulnerability, and reduced us to the object of attack rather than something valuable. Thus:

H2: the greater the extent to which a tease is perceived as aggressive, the

greater the extent to which the tease will be perceived as aversive.

15

Ambiguity. Ambiguity refers to the extent to which a message has multiple
meanings or possible interpretations. The more ambiguous a message is, the more
difficult it is to decide on the meaning of a message. Ambiguity is generally encoded into
a message by way of the use of off-record markers. Off-record markers are typically
nonverbal cues including changes in tone of voice, facial expressions, and gestures
Goffrnan (I967). The presence of these cues signal to the receiver that something other
than the verbal content of the message is what was actually meant by the utterance, and
that the utterance should not be interpreted literally.

In many social interactions, ambiguity has positive, face-saving functions. Take,
for example, an instance of flirting. Flirting is designed to provide a subtle signal of
interest in another person and is highly ambiguous in order to reduce the negative effects
of rejection (Baxter, 1992; Kowalski, 2001). Flirting is characterized by a set of
nonverbal cues that can either be interpreted as signs of attraction or that could be
attributed to environmental artifacts such as crowded space or loud music which
necessitate being close to another person. If a man is flirting with a woman in a bar, for
example, he is likely to lean closely toward her, possibly brush up against her or
otherwise touch her, make very direct eye-contact, and smile (Coker & Burgoon, 1987).
She has the choice to interpret these cues either as signals of attraction and reciprocate
them if she is equally attracted to him, or ignore them and not reciprocate. In the case that
she chooses not to reciprocate, she has rejected his signals of attraction, but he does not
have to feel as though his face was confronted because she did not reject him overtly.

Further, he can walk away fi'om the interaction and explain away his attraction behaviors

l6

simply as good listening skills, or as having happened as a result of environmental
constraints. In this regard, ambiguity has positive social outcomes.

Where teasing is concerned, however, ambiguity in a message is more likely to
cause distress or negative emotion because of the possibility that the ambiguity in a tease
may disguise a legitimate insult or attack. When an agent makes a remark that challenges
a target’s self-concept and attacks the target but is accompanied by off-record markers,
the target has the choice of two interpretations. The first of these interpretations is that the
tease was playful and lighthearted and enacted in the spirit of fun. The other
interpretation is that the tease was a thinly veiled and sincere criticism of the target. In
this case, the presence of off-record markers may be construed as having been attached to
an attack so that the attack could be made without seeming blatantly offensive. In this
way, ambiguity, rather than being face-saving, actually leaves room for a tease to be
construed as particularly offensive. The possibility that the agent is suggesting that there
is something wrong with the target and that off-record markers were attached to the
disapproval raises questions in the mind of the target as to whether the agent values him
at all. Again, we see that ambiguity can cause a feeling of relational devaluation, and lead
to the perception that the tease is aversive. Thus,

H3: the greater the extent to which a tease is perceived as ambiguous, the

greater the extent to which the tease will be perceived as aversive.

Humor. Humor is the final component of a tease. Unlike the other components of
identity confrontation, aggression, and ambiguity, humor is not expected to increase
perceptions of the aversiveness of a tease, but rather diminish or temper aversive

perceptions. Humor is arguably the defining characteristic between a tease and an insult.

l7

Humor moderates the effect of the negative components of a tease: all negative attributes
of a tease considered, a funny tease will be regarded as less aversive than will a tease that
is not funny.

Research has provided some evidence for the beneficial impact of humor. For
example, having a sense of humor appears to be positively associated with empathic
ability and the ability to initiate new relationships (Yip & Martin, 2006). Further, having
a sense of humor is negatively related to the experience of worry. Worry is positively
related to the experience of anxiety and pessimism (Kelly, 2002), suggesting that being
inclined toward laughing and the ability to perceive others as funny may have positive
psychological outcomes. Further, the use of humor by supervisors is positively associated
with reports of job satisfaction by employees (Hurren, 2006). This finding suggests that
humor might make an otherwise unpleasant or stressful task easier to complete.

Following this line of reasoning, it makes sense that the incorporation of humor in
a message with negative characteristics such as aggression and identity confrontation
may reduce the impact of those negative characteristics. For this reason, the inclusion of
humor in a tease functions to a) differentiate teases from insults — a message type that are
designed to attack and are used for their negative consequences, and b) reduces the extent
to which a tease is likely to be considered aversive. Because humor diminishes the
negative impact of an otherwise negative message type,

H4: the greater the extent to which a tease is perceived as humorous, the

lesser the extent to which the tease will be perceived as aversive.

l8

Given the expected relationships between each of the components of the
definition of teasing, a model combining the four components as predictors of
aversiveness can be tested. Thus,

H5: The combination of identity confiontation, aggression, ambiguity and

humor will, together, predict perceptions of the aversiveness of a tease.

To define a tease in terms of message components that are likely to lead to
aversive consequences makes substantial progress in terms of understanding teasing as a
type of message. Breaking a tease down into its four constituent parts and measuring
those parts makes observable the relationship between characteristics of a tease and
subsequent perceptions of aversiveness. An additional benefit of dissecting a tease into its
component parts is the ability to test the contention that different types of messages are
characterized by different levels of each of these components (Kowalski, 2004; Kowalski
etaL,2001)

Kowalski et al. (2001) suggest that different relative degrees of each of the
components of teasing ultimately combine to form different subcategories of teasing. For
instance, flirting is a combination of low aggression, low identity confrontation, and
moderate humor, where insults on the other hand are characterized by high aggression,
high identity confrontation and low humor. The crux of this hypothesis is that different
types of messages are objectively and systematically different from one another on each
of the dimensions that define a tease.

While such a contention is intuitively sound, Kowalski et al. (Kowalski, 2004;
Kowalski et al., 2001) provide no theoretical or empirical basis for such a claim. As such,

the belief that different message types systematically differ along each of the teasing

l9

dimensions should be tested. Empirical evidence for systematic differences between
different types of messages would help to define teasing as a message type, ideally
lending clarity to the conceptual definition of teasing, as well as demarcating differences
between message types based on a specific set of criteria.

In the current case. the message types of insult, tease, and joke are compared to
each other because they can be easily differentiated in terms of the degree to which they
vary along identity confrontation, aggression and humor. T cases, as defined earlier, are
characterized by some degree of identity confrontation, aggression and humor. Arguably,
if the tease is characterized by too much identity confrontation or aggression, it is likely
to be regarded as offensive and be decoded as an insult rather than a tease. As such,
insults should be regarded as more identity confronting and aggressive than teases.
Additionally, insults are unlikely to be regarded as funny because they are designed to
attack. Teases on the other hand are characterized by the presence of humor. Thus, insults
should be regarded as less humorous than teases.

The inclusion of jokes in this comparison was of particular importance in that a
joke is quite a different type of message than insults and teases. Jokes, unlike teases and
insults, do not contain components of aggression or identity confrontation. Jokes are
generally characterized as a type of humorous message that contains a punchline
designed to surprise the listener which then has to be refrarned in order to make sense
(Coulson & Kutas, 200i). Jokes are not specifically designed to challenge or attack
another person, but rather their aim is specifically to be funny. This means thatjokes
should rank lowest on aggression and identity confrontation among the three message

types of insults, teases and jokes. Further, jokes are told with the express purpose of

20

being funny, where teases feature humor as a means of tempering negativity, and insults
are not intended to be funny. As a result, jokes should rank highest on humor among the
three message types, followed by teases and insults. Thus,

H6a: 0n identity confrontation, insults will score highest, jokes will score

lowest, and teases will score between the insults and jokes.

65b: 0n aggression, insults will score highest, jokes will score lowest, and

leases will score between insults and jokes.

H66: 0n humor, insults will score the lowest, jokes will score the highest,

and teases will score between insults and jokes.

No hypotheses have been made regarding the role of ambiguity in determining
different message types, as its relation to aversiveness seems to change as a function of
the type of message of which it is a part. For instance, a remark is more likely to be
construed as either insulting or joking if it is unambiguously so. The clarity in the intent
of the message makes an insulting message insulting and a joking message amusing. On
the other hand, a tease is made aversive partly by the fact that it is ambiguous and that it
may be a sincere criticism being passed off as something funny. As a result of the
different outcomes ambiguity has depending on the type of message of which it is apart,
to make specific claims about the relationship between ambiguity and aversiveness
independent of the type of message is theoretically unsound.

Clearly, then this raises questions as to whether ambiguity is a reliable predictor
of aversiveness at all, or whether it is uniquely tied to perceptions of aversiveness of

teases. Understanding the essential nature of ambiguity in message processing is beyond

2l

the scope of the current investigation, though its place in a model of teasing seems to be
warranted and will be tested accordingly.
lntentionality

The issue of intentionality, while misplaced in a definition of a tease, is one of
particular importance in understanding the extent to which a tease is perceived as
aversive. As discussed earlier, the intent to do harm versus the intent to be lighthearted
and playful is at the center of the interesting, albeit, misguided decision for teasing
scholars to differentiate between antisocial and prosocial teases. A tease, after all, is a
tease regardless of the intent of the agent. That being said, however. it does not make
sense to disregard the role of intent in influencing aversiveness perceptions. Rather, the
intent of the agent is quite likely to influence perceptions of the agent but a) intent is not
hypothesized here to be a component of a message, and b) the true intent of the agent is
often unknown, and so questions of intent must be examined from the perspective of the
target.

Although the existing literature likely accurately reflects agent intent as a factor
influencing the aversiveness of a tease, previous investigations of the role of intent are
not driven by theory. While the belief that an agent intends to do a target harm might
seem intuitive enough to render theory unnecessary, the lack of a conceptual model
linking agent intent to aversiveness precludes obtaining empirical evidence in support of
this relationship. In order to address this current lack of theory and correct this oversight
in the literature, intentionality will be examined through a theory of grievances (Tedeschi

& Nesler, I993).

22

A theory of grievance s. In an attempt to understand the manner in which people
respond to perceived social injustices Tedeschi and Nesler (l 993) developed a theory of
grievances. The basic premise of the theory is that when an individual experiences a
social injustice, he feels compelled to assign responsibility for the injustice to an agent
who is then held responsible for reparations in response to the injustice. In the event that
a blameworthy agent is identified, adverse emotions, namely anger and resentment are
harbored by the victim until such a time that the agent corrects the wrong-doing. In the
absence of a blameworthy agent, on the other hand, there is no target at which a victim
can aim anger and resentment, and often feels less negative emotion in response to the
harmful even in the absence of an agent to hold responsible for the action. Tedeschi and
Nesler classify the kinds of social acts that are likely to cause a victim to seek out a
blameworthy agent in hopes of being compensated for harm done.

Origins of grievances. Tedeschi and Nesler (I993) define a grievance as “the
judgment that another social agent has performed an unjust or unfair action” (p. 14). This
naturally begs the question as to what kinds of events constitute injustices or unfair
actions. Tedeschi and Nesler describe four types of behaviors that constitute injustices: a)
the perception of the intent to inflict physical harm or pain, b) loss or damage to existing
or expected resources, c) perceived damage to social identities, including a failure to
demonstrate respect, and finally d) violations of rights or freedoms. Teasing can be
classified as falling into the damage to social identities category, due to the fact that
teases are inherently identity confronting. Identity confrontation has been previously
argued to demonstrate a lack of respect, and therefore makes the theory of grievances

applicable to understanding perceptions of teasing.

23

The appraisal process. As discussed above, the commission of an injustice leads
to the desire to identify a blameworthy agent responsible for the injustice as a means of
obtaining reparations for the harm done as a result of the wrong-doing. An appraisal
process is initiated in response to an injustice whereby the victim of an injustice attempts
to ascertain whether the agent who committed the injustice intended to do some harm to
the victim. In the event that the appraisal is made that the agent acted with intent to do
harm, the agent is considered blameworthy. The result of finding an agent responsible for
an injustice leads to the aversive experiences of resentment and anger.

On the other hand, if a victim fails to make the attribution that the agent acted
with the intent to do harm, a secondary appraisal process is initiated in which the victim
assesses the likelihood that the agent acted carelessly - that is whether the resulting harm
could have been foreseen and avoided. In the event that the harm was foreseeable, the
agent is again considered blameworthy. Neither in the event that the agent is appraised as
having intent to do harm, nor could any harm been foreseen, the agent is absolved of
responsibility. In the absence of an agent to hold responsible for an injustice, the victim’s
anger often dissipates, thereby reducing the aversive outcomes associated with the
injustice.

The result of this appraisal process is a set of hypotheses specifying the
relationship between perceptions of intent to do harm on the perception of the
aversiveness of a tease. Based on the contention that all teases are potentially grievable
offenses due to their identity confronting nature, the extent to which a tease is perceived
as aversive is a function of the extent to which a target perceives the agent as having the

intention to hurt such that:

24

H7: the greater the extent to which a target perceives that an agent

intended to do harm, the greater the extent to which the tease will be

perceived as aversive, and

H8: the greater the extent to which the target perceives that an agent

could have foreseen potential harm to the target, the greater the extent to

which the tease will be perceived as aversive.

It should be noted here that the theory of grievances predicts the same outcome
for both perceived intent to cause harm as well as perceived foreseeability of harm. For
this reason, no interaction between the two variables is hypothesized, but rather are
regarded as components of intentionality that both directly impact perceptions of
aversiveness. Thus:

H9: Perceived intent to do harm and perceived foreseeability of harm will,

together, predict perceptions of the aversiveness of a tease.

Finally, the question remains as to which of the component model of teasing or
the theory of grievances model provide a better explanation for the perception of
aversiveness in response to teasing. Thus:

RQI : do the data fit the component model of teasing or the theory of

grievances model better?
Summary and Hypotheses

A set of hypotheses designed to assess the extent to which identity confrontation,
aggression, ambiguity and humor are related to the extent to which a tease is perceived as
aversive. Further, different types of messages are expected to systematically vary along

these message components, suggesting that a tease is characterized not only by a specific

25

set of characteristics but particular levels of those characteristics. Further, an additional
hypothesis regarding the perception of agent intent to cause harm is explored as a means
of explaining the perception of the aversiveness of a tease. Finally, a research question is
posed inquiring as to whether the aversiveness of a tease is better explained by a set of
definitional components or target perception of agent intent to do harm. The set of
hypotheses as stated above are re-presented here for the reader’s convenience:

HI : The greater the extent to which a tease is perceived as identity confronting, the
greater the extent to which that tease will be perceived as aversive.

H2: The greater the extent to which a tease is perceived as aggressive, the greater the
extent to which the tease will be perceived as aversive.

H3: The greater the extent to which a tease is perceived as ambiguous, the greater the
extent to which the tease will be perceived as aversive.

H4: The greater the extent to which a tease is regarded as humorous, the lesser the extent
to which the tease will be perceived as aversive.

H5 : The combination of identity confi'ontation, aggression, ambiguity and humor will,
together, predict perceptions of the aversiveness of a tease.

H60: 0n identity confrontation, insults will score highest, jokes will score lowest, and
teases will score between the insults and jokes.

H6b: 0n aggression, insults will score highest, jokes will score lowest, and teases will
score between insults and jokes.

H6c: 0n humor, insults will score the lowest, jokes will score the highest, and teases will

score between insults and jokes.

26

H7: The greater the extent to which a target perceives that an agent intended to do harm,
the greater the extent to which the tease will be perceived as aversive, and
H8: The greater the extent to which the target perceives that an agent could have
foreseen potential harm to the target, the greater the extent to which the tease will be
perceived as aversive.
H9: Perceived intent to do harm and perceived foreseeability of harm will, together,
predict perceptions of the aversiveness of a tease.
RQI .' Do the data fit the component model of teasing or the theory of grievances model
better?

The methodology for a survey study designed to test these hypotheses is
described. The rationale for a survey study, a description of the survey, and a description

of all measures is provided in the following chapter.

27

METHOD

Participants

Participants were recruited from introductory communication classes at a large
mid-westem university using an electronic subject pool. Participants were offered course
or extra credit for participation in the study. The first page of the survey was the informed
consent form that participants read and acknowledged by advancing through fi'om the
consent form to the first page of the survey. The sampling method, survey, and informed
consent form were approved by the university’s institutional review board prior to
participant recruitment.
Procedure

Data collection for this non-experimental study occurred via online survey. The
survey was accessed by participants by way of a link which was provided upon signing
up for participation for the study via the online subject pool. Upon accessing the link and
consenting to participate, participants were advised that they would be completing a
survey about a time within the previous two days that they had been teased by a close
friend. The first section of the survey asked participants to provide information about the
agent of the tease, including the agent’s first initial, sex. and approximate age. Further,
participants were asked to provide information about the type of relationship they had
with their respective agents, including the duration of the relationship in months and
years, as well as the type of relationship (ranging from strangers to romantic partners)
that existed between the agent and the target. Following the provision of information
regarding the agent and the relationship, information about the context in which the tease

occurred was obtained. Participants were asked to report the date, time, and location in

28

which the tease happened. Participants were also asked to report whether anyone else was
present at the time of the tease.

Following the provision of contextual information, participants were asked to
report in as much detail as possible the content of a tease. Participants were advised that
the tease had to have occurred at some point within the two days prior to the completion
of the survey. Participants were asked to report the content of the tease itself, as well as
any conversation or action immediately preceding the tease, as well as anything they said
in response to the tease. Following the completion of the description of the tease,
participants were asked to evaluate the tease on a series of scales designed to measure the
perceptions of identity confrontation, aggression, ambiguity and humor, perceptions of
agent intent, and perceptions of aversiveness. Appendix C provides all of the survey
questions provided to the participants, and instructions for reporting the tease.

In order to test hypotheses 5a-c, a set of researcher-provided remarks were
included on the survey following the participants’ reports of their own teases. Participants
were advised that the remarks occurred in conversation between close friends in response
to some situation in which the friends were interacting. Participants were asked to
classify the each remark as a tease, ajoke or an insult, as well as evaluate the remark
using the same measures used to asses their own teases. Subject provided teases would
not have been suitable for this test because participants were prompted to provide teases,
and would likely have influenced the message classification participants made of their
own incidents.

Here, a note about the construction of the remarks might be useful. The remarks

were short, single-line utterances said in response to some scenario in which two friends,

29

Chris and Terry, were interacting. Initially, the goal of constructing these remarks was to
create a set of three remarks, one of which could be objectively classified as a tease, one
of which could be classified as ajoke, and the last of which could be classified as an
insult. Numerous unsuccessful attempts were made at creating definitive message types.
A set of messages would be constructed and several of the researcher’s colleagues were
consulted to see if a pure message type had been achieved. Inevitably, there was
variability in the message classifications — with no message being evaluated as one of the
three message types 100 percent of the time.

Due to the failure of all attempts to create definitive and clear message prototypes,
the goal became the creation of a set of different messages that could each be classified as
a joke, a tease, or an insult, and then each of the within-remark message types would be
compared to one another for the purposes of hypothesis testing. Despite the conciseness
of the design that would have resulted from the creation of message prototypes, the use of
three different messages allows for a cautious statement about the generalizability of the
pattern the different message types are hypothesized to form. If the hypothesized pattern
is observed across all three messages, then the tentative conclusion that the pattern of
results is not confined strictly to a single instance of each of the message types. Each of
the remarks is provided in Appendix C.

Rationale for Survey Stuafv

Initially, an event-contingent diary was going to be used, but due to the length of
multiple measures being assessed, it is likely that a diary study would be too demanding
of the participants which would result in decreased response rate. Retrospective studies,

on the other hand, are imperfect, being plagued by problems with participants having

30

incomplete or inaccurate memory of the event, or changes in the perception of the event
over time. In order to attempt to address the problems of both the diary and retrospective
approaches, participants were asked to complete the survey within two days of the
occurrence of a tease. This method reduces some of the problems of retrospective studies
which allow for a lengthy lapse of time between the time of the incident and the time of
reporting, yet reduces the demands on the participants by asking for a single report of
data.

Despite the attempt to minimize altered memory or perceptions of the teasing
incident by restricting the length of time between the tease and survey completion, it is
unlikely that participants responded to the instrument immediately following the incident.
Doing so may not have been convenient, and may make the agent self-conscious about
his/her behavior. This is a nearly unavoidable consequence of collecting retrospective
data. Despite there being very little to be done about memory errors in retrospective
recall, it was the hope that having participants provide a detailed narrative of that tease
would increase the salience of a specific teasing incident so as to direct responses on
scale items toward a specific episode rather than at some vague notion or memory of
having been teased in the past.

Of course, narrative accounts provide a uniquely rich data and allow for
interesting content analyses at some future point in time. Narratives coded by objective,
independent raters for the teasing components proposed by the definition of teasing, or
matched with participant ratings of the same tease to see the extent to which perceptions

of the tease are a function of being the target of the tease or characteristics of the content

31

of the tease. For the purposes of the current study, however, the narratives serve purely to
focus participant responses on a specific incident.
Instrumentation

The following is a presentation of descriptions of the scales used in assessing
responses to the tease. Instructions were provided on the survey that the participant
respond to the scales regarding the way they felt at the time of the event rather than how
they felt about it in retrospect.

Identity confrontation. Perceived intention to confiont the identity of the
participant was measured using a six-item Likert-type scale with a seven-point response
format. Higher scores indicate greater perceptions of identity confrontation. This scale
was developed for the purpose of this study due to the lack of an existing scale designed
to measure identity confrontation. The items were created based on the descriptions of
identity confiontation and face-threat as described by Goffinan’s (I959). The scale
produced a Cronbach’s or = .92.

Aggression of the tease. A five-item Likert-type scale with a seven-point response
format was used to measure aggression. Higher scores reflect greater perceptions of
aggression in the tease. The items were developed using Infante and Wigley’s (1986)
conceptual definition of verbal aggressiveness. Infante and Wigley (1986) indicate that
verbal aggressiveness is defined as messages which attack an individual’s self-concept to
make the target feel less favorably about the self. A set of five items were generated with
the intent of measuring the construct based on this definition. The original five-item scale
produced a reliability coefficient of or = .84. A review of the estimates of the scale

reliability with each item deleted indicated that the removal of one item would improve

32

the scale’s overall reliability, and so that item was dropped from the scale. The resulting
four items resulted in a Cronbach’s a = .87.

Ambiguity of the tease. Tease ambiguity was measured using a five-item Likert-
type scale with a seven-point response format with higher scores indicating greater
perception of ambiguity. There are currently no scales in existence that assess ambiguity
of a message. As such, the five-item measure employed here was developed for the
purposes of this study. The scale produced a reliability estimate of or = .74. An
examination of the scale reliability estimates with each item deleted indicated that the
removal of one item would improve the scale’s reliability. As a result, the problematic
item was removed, and the revised 4-item scale produced a Cronbach’s or = .79.

Humorousness of the tease. The degree to which the tease is interpreted as
humorous will be measured using a six-item Likert-type scale with a seven-point
response format. Higher scores indicate greater perceptions of humorousness. There are
several humor scales in existence which measure either the type of humor being enacted
or an individual’s dispositional sense of humor (I 993). These assessments, while ill-
suited for measuring the humorousness of the utterance, provided a conceptual definition
of humor that informed the creation of the measure used in this study. The scale produced
a Cronbach’s or = .95.

Intent to harm. Perception of the agent’s intent to harm the target was assessed
using a four-item Likert-type scale with a seven-point response format. These items are
designed to assess the extent to which the target believed that the agent meant to cause

hurt or harm in the course of the tease. No such scale previously existed to assess this

33

construct, and as such the scale was developed for use in this study. The scale produced a
Cronbach’s or = .96.

Foreseeable harm. Perceived foreseeability of harm was measured using a four-
item Likert-type scale with a seven-point response format. These items assess the extent
to which the target of the tease perceived the agent’s lack of intent to cause harm, but the
belief that the agent could have anticipated that harm would result from the tease. Again,
this scale was developed for use in this study due to a lack of an existing scale. The scale
produced a Cronbach’s ct = .91.

Tease aversiveness. Perceived aversiveness of the tease will be measured using a
six-item Likert-type scale with a seven-point response format. The original six-item scale
produced a reliability coefficient of or = .92. An examination of estimates of reliability
with each item deleted indicated that the removal of one item would improve the scale
reliability. As such, that one item was dropped from the scale. The resulting five-item
scale produced a Cronbach’s ct = .95.

Two additional notes are necessary with regard to the aforementioned reliability
assessments. First, although the scales were used a total of four times (once for the
assessment of the subject-provided tease, and once each for the three researcher-provided
remarks), the above reliabilities are for the scales as they were used in the evaluation of
the subject-reported teases. Some of the scales were modified for the evaluation of the
researcher-provided remarks, as some of the items made sense only in the context of the
target’s response to the tease rather than that of an observer. Reliabilities were calculated,

however, for each of the scales in each of the contexts in which they were used. Scale

34

reliabilities and descriptive statistics are provided for all instances of each scale in Table
l.

Secondly, items which systematically decreased the reliability of any of the scales
reviewed above were omitted from the scale when composites of the items on the scale
were calculated. This reduced the number of items assessing the ambiguity, aggression
and aversiveness scales, but the resulting improvement of reliability justifies such
exclusions. It should also be noted that the items that were problematic in the
measurement of subject-reported scales were consistently problematic across all of the
instances of the use of the scale across the assessments of the researcher-produced
remarks. This suggests that these items were fundamentally problematic in terms of
measuring these variables, thereby lending further support for the exclusion of these
items. Table 8 provides a list of all items (including an indication of which items were
deleted for use in the assessment of researcher-provided remarks and due to low
reliability), and the reliability coefficients and scale descriptive statistics for each use of
the scale.

Confirmatory factor analysis. All scales were subject to confirmatory factor
analysis. Table 8 presents all scale items and factor loadings for each item. In the cases of
ambiguity, humorousness, aversiveness, and perceived intent to do harm, all items on
those scales loaded highest on the expected factor. In the case of perceived foreseeability
of harm, half of the items ranked highest on the expected factor, and the other two items
ranked highest on the aggression factor. Factor loadings for items on the aggression and
identity confrontation scales indicate possible content validity problems with both scales.

Of the five items measuring aggression, two loaded highest on the foreseeable harm

35

factor, and one loaded highest on the identity confrontation factor. Of the four items
measuring identity confrontation, two of them loaded highest on the aggression
dimension.

Incorrect loading of seven of 33 items across the seven scales might be partially
explained by the substantial correlation among all of the variables. The construct validity
of the measures on which items did not factor correctly may be reasonably called into
question. This is especially true for the aggression and identity confrontation items, as
three of the seven items that did not factor as expected belonged to one of these two
scales, and factored on the other. This might suggest that these two constructs in

particular are not distinct from one another.

36

RESULTS

Participant Characteristics

A total of 157 people accessed an online survey. Upon examining the data, 11
participants failed to provide useable data by either failing to complete the survey or by
providing dishonest responses. The removal of these subjects from the final sample
yielded an N = 146. The sample was comprised of 47 men and 99 women. The mean age
of participants was 21.01 years (sd = 1.77). Participants reported a mean relationship
length with the teaser of 57.49 months (sd = 61.61), and a median of 36.00 months.
Participants were further asked to report about a teasing episode that occurred with a
close friend. Twenty-five participants indicated that the teasing episode happened with
either an acquaintance or a romantic partner. Analyses of variance were performed using
relationship type as the independent variable and each of the scales as dependent
measures. Relationship type did not yield significant differences on any of the measures
in the study, and as such; those participants indicating that they were teased by someone
other than a friend or close friend were retained in the final sample.
Tests of Linear Associations between Teasing Components and Aversiveness

Hypotheses I through 3 predict direct associations between perceptions of identity
confrontation, aggression, and ambiguity and perceptions of aversiveness. Table 2
provides the correlation matrix between aversiveness and each of the predictor variables.
Hypothesis I predicted that perceptions that the tease was identity confronting would be
positively associated with the perception of the tease as aversive. The calculation of the

Pearson product-moment correlation indicates a substantial and statistically significant

37

positive linear association between identity confrontation and aversiveness, where r =
.71 . p < .01.

Hypothesis 2 predicted that there would be a positive linear association between
perceptions of the aggression of a tease and perceptions of the tease as aversive. The data
indicate a substantial and statistically significant positive linear association between
aggression and aversiveness, where r = .81 , p < .01.

Hypothesis 3 predicted that there would be a positive linear association between
perceptions of the ambiguity of a subject-provided tease and perceptions of the tease as
aversive. The data suggest that ambiguity is unrelated to perceptions of aversiveness (r
=.l6, p = ns).

Hypothesis 4 predicted that perceptions of humor would be inversely related to
the perception of subject-provided teases as aversive. The data indicate a substantial, and
statistically significant negative linear relation between humor and aversiveness, where r
= -.73,p < .01.

Test of a Model of Teasing

Hypothesis 5 proposes a model of teasing that hypothesizes that each of the
components of the definition of teasing will, together, predict the extent to which a tease
is perceived as aversive. The expectation is that each of the components of the definition
of teasing will, individually, contribute meaningfully to perceptions of the aversiveness
of a tease. Further, the group of four predictors is expected to be related to a substantial
degree to perceptions of aversiveness. This hypothesis was tested via the calculation of a
multiple regression which included each of the four teasing components of teasing as

predictors and aversiveness of the tease as the dependent variable. To reduce the impact

38

 

of the high correlations among the predictors, each of the predictors was centered around
its own mean (Cohen, Cohen, West, & Aiken, 2003). Results of the regression analysis
showed that the set of predictors indicated above accounted for a substantial and
significant amount of the variance in the dependent variable (F (7, 138) = 55.6, p < .01,
adjusted R2 = .72). It was not the case, however, that each of the individual predictors
contributed meaningfully to perceptions of aversiveness. Aggression (B = .59, t(143) =
5.92, p < .01) and humor (B = -.34, t(143) = -5.85, p < .01). Neither identity confrontation
nor ambiguity accounted for a significant proportion of variance in perceptions of
aversiveness. Beta coefficients, t-values, and part correlations for each of the predictors
are provided on Table 3.
T est of Pattern of Different Message Types along Measures of Teasing Components

Hypotheses 6a, 6b, and 6c predicted that messages classified as teases, insults and
jokes would score in a predictable pattern along assessments for identity confrontation,
aggression and humor. Each of three different remarks provided to the participant by the
researcher were classified as a tease, a joke, or an insult. Each remark was evaluated
separately, and the classification of one remark as a certain message type did not
influence the classification of subsequent remarks. Table 4 provides the frequency
distribution for the number of times each remark was classified as an insult, a tease and a
joke.

It should be noted here that while participants’ message classifications were not
restricted as a matter of instruction, the order in which the remarks were presented to
participants was not varied. Further, all participants classified and evaluated each of the

researcher-provided remarks after the presentation and evaluation of the tease they

39

experienced. It is possible that the order in which teases were evaluated systematically
influenced the type of message classifications. Further, evaluations of each tease may
have been made relative to the remarks that preceded it. As such, it is possible that the
validity of the findings of the tests for hypotheses 6a-c is compromised due to order
effects.

To test hypotheses 6a-c, an a priori contrast test was conducted on each remark
for each message dimension. Each remark had three categories: insult, tease, and joke,
and each was assigned a contrast weight —- the specifics of which will be discussed for
each hypothesis. Omnibus F -tests were unnecessary because of the a priori hypothesizing
of specific differences between message types, and as a result were not conducted. The
means, standard deviations and value of all contrast tests can be found in Table 5.

Hypothesis 6a predicted that teases, jokes and insults would fall into a pattern
along a measure of identity confrontation such that insults would score highest, jokes
would score the lowest and teases would fall somewhere between jokes and teases.
Because of the hypothesized pattern, insults were given the contrast weight of l, teases
were weighted 0, and jokes were weighted -1. This test was conducted for all three of the
remarks provided, producing a total of three different contrast tests for this hypothesis. In

all three cases, the w contract coefficient was significant beyond the .01 level, and the

means fell into the predicted pattern.

Hypothesis 6b predicted that teases, jokes and insults would fall into a pattern
along aggression such that insults would be highest, jokes would be lowest and teases
would fall between insults and jokes. Because the message types were expected to fall

into the same pattern as specified in hypothesis 5a, the same contrast weights of l, 0 and

40

-l were given to insults, teases and jokes, respectively. This test was conducted for each
of the three remarks participants classified into message types, producing a set of three
tests across three remarks to test this hypothesis. In all three cases, the w contrast
coefficient was significant beyond the .01 level, and the means fell into the predicted
pattern.

Hypothesis 5c predicted that teases, jokes, and insults would fall into a pattern
along an index of the perceived humor in such way thatjokes would be highest, insults
would be lowest, and teases would fall between jokes and insults. The contrast weights in
this case were held in the same pattern as in the contrasts run for identity confrontation
and aggression. As such, if the remarks fell into the hypothesized pattern with insults
ranking lowest on humor and jokes ranking highest, the contrast coefficient should be
negative. A contrast test was conducted for each of the three remarks, again producing a
set of three contrast tests for this hypothesis. In all three cases, the t]! contrast coefficient
negative and significant beyond the .01 level, and the means fell into the predicted
pattern.

Test of a Model of Perceived lntentionality

Hypothesis 7 predicted that there would be a positive linear association between
perceptions that an agent intended to cause harm by teasing the target. The Pearson
correlation between perceived intent to harm and perceived aversiveness was r = .71, p <
.01. Hypothesis 8 similarly predicted that there would be a positive linear association
between the perception that the agent foresaw potential harm to the target and perceived
aversiveness. The Pearson correlation between perceived foreseen harm and perceived

aversiveness is r = .71, p < .01. Hypothesis 9 predicted that intent to harm and

41

foreseeable harm would predict perceptions of aversiveness. This hypothesis was tested
via the calculation of a multiple regression with both intent to harm and foreseen harm as
predictors of aversiveness. Because no interaction was predicted between intended and

foreseen harm, they are the only two predictors in the model with perceived aversiveness
as the outcome variable. Both perceived intent to harm (B=.48, t(143)=8.83, p<.01) and
perceived foreseen harm (B=.49, t(143)=9.04, p<.01) were significant predictors of
perceived aversiveness. Further, combined, these two predictors accounted for a
substantial and significant amount of variance in the dependent variable (F (2,
l43)=154.27, p < .01, R2 = .68, adjustedRZ = .68).

Research Question: Which model to the data fit better?

Considering just the size of each model’s adjusted R), it appears that the teasing
components model is a better predictor of tease aversiveness. The teasing components
model produced an adjusted R2 = .72. Removal of the non-significant predictors and a
recalculation of the regression resulted again in an adjusted R2 = .72. The theory of
grievances model, on the other hand, produced an adjusted R2 = .68. These numbers
might suggest that the teasing components model is a better predictor of tease
aversiveness than is the theory of grievances model. On the other hand, the differences
between the adjusted squared multiple correlations is small, so it is difficult to assert that
one model is definitively better than the other. More interesting, perhaps, might be to
combine the significant predictors from both models and create a comprehensive model
of teasing including both message components and perceptions of intent to cause harm.

This will be elaborated upon in the discussion section of the paper.

42

DISCUSSION
Interpretation of Findings

Mixed support was found for the component model of teasing. In terms of
bivariate correlations, three of the definitional components of the tease —- identity
confrontation, aggression, and humor -- were substantially and significantly related to the
perception of a tease as aversive. When the same message characteristics were entered
into a regression equation as predictors of aversiveness, however, only aggression and
humor remain as significant predictors of aversiveness. Aggression was positively related
to aversiveness, and humor was negatively related to aversiveness, as expected.

These findings seem to suggest that the feeling of being attacked in a general,
non-specific way is more important than feeling as though one’s identity has been
confronted. Identity confrontation, even as described by identity scholars, is a rather
complex cognitive phenomenon. It involves awareness and activation of a particular
identity for in a particular situation as well as being aware of the extent to which any
communicative instance contradicts or confronts that identity. Responses to aggression or
attack, on the other hand, are likely connected to the more primal and instinctual parts of
the mind. Defending one’s self from attack is an innate survival response. As such, it
makes sense that an attack, in the broadest of senses, is easier to identify and react to than
is the more complex process of identifying and responding to an identity confi'ontation.
As such, it makes sense that the effect of feeling attacked supersedes the effect of having
one’s identity confi'onted when reacting to a tease. Arguably, identity confrontation might

be conceptually subsumed under the broader conceptual definition of attack.

43

Humor, as expected, was negatively related to the perception of a tease as
aversive. This might suggest that humor has an important role in managing
uncomfortable or difficult situations. The existing literature seems to suggest that the
presence of humor in challenging situations such as stressfiIl life situations or interactions
between employers and subordinates eases the negative effects of such situations. The
same process appears to be occurring within a tease. Because both aggression and humor
are significant predictors of the aversiveness of a tease, it seems that the absence of
humor in a tease is likely to result in higher evaluations of aversiveness. This lends
support to both the inclusion of humor in the conceptual definition of teasing as well as
the notion that different levels of the components of a tease combine to create different
message types. This conclusion is further supported by the data’s consistency with
hypotheses 6a-c which predicted that different types of messages would systematically
differ from one another on these message components.

Ambiguity was entirely unrelated to perceptions of aversiveness of a tease, both
in terms of bivariate correlations and as a predictor of aversiveness. This seems to
suggest that ambiguity does not have a systematic impact on the perception of message
aversiveness. This makes sense when considering that the hypothesized effect of
ambiguity is variable depending on the kind of message with which it is associated. In the
case of flirting, as described in the first chapter, ambiguity enables the saving of face in
potentially embarrassing situations. On the other hand, ambiguity was argued to
contribute to perceptions of aversiveness in the case of teasing because an ambiguous
tease might be construed as featuring off-record markers just to make a legitimate insult

seem less offensive. These findings suggest that ambiguity as a message characteristic

44

appears not to be a reliable indicator of message aversiveness. It is important to note,
however, that unreliability in the measurement of ambiguity may be, at least in part,
attributable to the low correlations between it and the other variables in the model. In
reconsidering the role of ambiguity in a tease, it makes sense that ambiguity should have
no relationship to aversiveness. It is possible that in some cases, ambiguity of a tease
increased perceptions of aversiveness if the target of the tease perceived the presence of
off-record cues as a means of disguising a legitimate criticism. In other cases, however,
participants may have thought that ambiguity contributed to the diminishing of
aversiveness in that it provided the face-saving benefits that accompany ambiguity, such
as in the case of flirting. Future investigations might benefit fi'om asking participants to
specifically assess the perception they made of the off-record markers in a tease and
assess the perception of the function of ambiguity as a moderator of aversiveness.

Support was found for the hypothesis that the components of teasing vary in their
relative amounts as a fimction of the type of message the components comprise. The data
indicate that teases are perceived as moderately identity confronting, aggressive, and
fiInny, where insults are high on the two negative components and low on humor and
jokes are high on humor and low on the two negative components. This lends credibility
to Kowalski’s previously unsupported contention that messages vary in systematic ways
with respect to the relative amounts of negative components and humor.

As noted in the results section, these findings should be taken somewhat
tentatively in that an order effect might be partially responsible for the pattern of
responses provided by participants. In all cases, participants presented their own tease

and then responded to the researcher-provided teases. The selection of the particular

45

teasing instance that participants selected may have influenced their perception of the
researcher-provided remarks based on their similarity or difference to their own reported
tease. Further, the order of the researcher-provided remarks was the same for all
participants. Again, the degree to which each of these remarks was rated along the teasing
components might have been influenced by the relative levels of each of the components
as compared to the other remarks. Although the data seem to suggest the reliability of a
pattern of message types along the components of a tease, potential order effects must be
corrected and findings replicated before any strong conclusions are drawn about the
relative levels of identity confrontation, aggression and humor across different message
types.

Finally, the theory of grievances-inspired model of perceived intent as a predictor
of aversiveness was supported both in terms of the bivariate relationships between
intended harm and foreseeable harm with perceived aversiveness Further individual
regression coefficients for each predictor suggest that both perceived intent to harm and
foreseeable harm individually account for meaningful proportions of variance in
perceived aversiveness.

It is important to note, however, that both the theory of grievances model and the
teasing components model account for substantial variance in perceptions of the
aversiveness of a tease. Based on these results, it is apparent that a model of aversiveness
must include a set of predictors comprised of message components as well as perceived
agent intent to do harm. Based on the data provided in this sample, a revised model of
teasing was tested. The predictors in the revised model used aggression and humor from

the teasing components model and intent to harm and foreseeable harm from the theory

46

of grievances model. The revised model maintains each of the predictors as significantly
accounting for variance in the revised model, with the overall model producing an
adjusted R2 = .78 (F (4, 141) = 132.34, p < .01). The association of the combined set of
predictors in this new model with aversiveness is a substantial improvement over either
of the models alone. These findings suggest that the aversiveness of a tease is a firnction
of both message attributes and perceptions of agent intent to harm. This is a substantial
improvement over the lack of conceptual or theoretical models currently in existence in
the literature. Future research should seek to replicate these findings both with
participant-provided teases, as well as with independent raters’ evaluations of teases. This
model will have significant theoretical contribution if it holds true in the objective
evaluation of teases by trained independent coders as it will suggest that the evaluation of
the aversiveness of a tease is meaningful even when the tease is not personally relevant to
the evaluator.
Limitations of the Study

In addition to the limitations specified in the discussion of the results, one
additional major limitation must be addressed. The motivation for this investigation was
largely to differentiate teasing from bullying on the grounds that, unlike bullying, teasing
has the potential to yield positive relational outcomes. This contention, however, was not
directly tested in this investigation. Instead, aversiveness — that is the negativity and
negative impact — of a tease was the focus of the paper. This precludes making the claim
that teasing and bullying are different as a result of the potential positive outcomes of
teasing. While this does not provide compelling evidence for the fact that anti-bullying

and safe-school initiatives should avoid eradicating teasing from campuses, the current

47

investigation does clearly allow for the conclusion that tease aversiveness is variable.
This finding dispels the belief that teasing is always a negative experience for its targets.

While this is not adequate reason to avoid punishing teasing in the name of
creating safe schools, it should cause policy-makers to pause and consider that teasing
and bullying are different phenomena and that the elimination of one does not necessarily
necessitate the elimination of the other. Future research should focus specifically on
investigating the positive relational outcomes claimed by some teasing scholars.
Specifically, solidarity, relational closeness should be investigated as relational outcomes
that should be influenced by teasing. Perhaps more importantly, however, individual
characteristics such as self-esteem, anxiety and depression should be examined as
outcomes of teasing. If scholars who assert that teasing can lead to closeness and
solidarity, and if the need to belong is a basic fundamental need, it might be the case that
being teased actually improves self-esteem and depression as a result of feeling like a part
of a group. Clearly, however, the negative bias in this study does not do what it should to
inform safe-school policies, but it does illuminate a path toward being able to do so.
Future Research

The findings of the current study provide a revised and comprehensive model of
teasing, specifically as it relates to perceptions of aversiveness. Assuming the findings of
this study can be replicated fiIture research might benefit from the examination of other
factors that are arguably related to aversiveness. Relational type, relational intimacy, and
the repetitiveness of a tease have all been argued to contribute to aversiveness. An
empirical test of the relationship between each of these variables and aversiveness would

assist in identifying as complete a model of teasing as possible.

48

Further, individual difference variables such as neuroticism, self-esteem, and
perception of risk in intimacy might be specifically examined as moderating the
experience of teasing. It makes sense that individual characteristics that make a person
prone to negative emotions (such as depression or low self-esteem) might exacerbate the
extent to which teases are perceived as negative, as well as eliminate any potential that
teasing might have to diminish aversive effects of aggression and intent to harm.

It may also be informative for fiIture research to focus on the extent to which the
topic of a tease influences aversive outcomes. For example, there is some evidence to
suggest that being teased about a personal attribute over which the target has no control is
a particularly negative experience. There does not, however, appear to be any empirical
support for this contention. It makes intuitive sense that being teased, even in a
lighthearted manner about a personal characteristic that cannot be readily changed (say,
the size of one’s nose, or socioeconomic status) might be particularly distressing. It is
possible that such teases might be subsumed under the foreseeable harm component of
the revised model of teasing in that despite a lack of malicious intent, the agent should
expect that being teased about an unchangeable characteristic might be especially hurtfirl.
Whether the topic of a tease has a direct influence on aversiveness or whether it is simply
a particular instance of foreseeable harm, is an empirical and answerable question. Safe-
school initiatives would benefit by knowing that particular topics are especially hurtfiJl
and can then implement specific anti-teasing policy based on such information.

Finally, future research could benefit from fiIrther examining the role of humor in
teases and perceptions of aversiveness. While evidence exists suggesting that humor is

negatively related to experiences of aversiveness, it is unclear as to what exactly

49

constitutes humor. Not only is humor variable at the individual level, but it is variable at
the cultural level as well. Take the case of ajoke, for instance. A joke is considered funny
when an unexpected punchline is successfully reframed (Coulson & Kutas, 2001). A joke
is funny when the audience of the joke is part of a shared language community with the
teller such that the new meaning (and therefore the fimniness) of the joke is accessible to
the audience. If the audience and the teller are from different language communities or
different social or cultural communities, it is possible that the content of the joke that
makes it funny will not be accessible to the audience. This has interesting implications
for teasing across language, social or cultural communities. While the teller of ajoke or
the agent of a tease may be using what she considers conventionally funny content, the
inability of the agent who does not share the same expectations of funniness may find the
tease offensive. Future research might focus on whether there are certain teasing topics
that might have more significant potential to be perceived as aversive. In an increasingly
diverse population — both in terms of the country as well as in schools — understanding
the possibility for offense because of a lack of shared language codes is an important
consideration in the future research of teasing.
Summary

The purpose of this investigation was to develop a conceptual model of teasing.
Based on previous research, message factors as well as agent intent to harm were tested
as predictors of the aversiveness of a tease. Results indicate that a comprehensive model
of teasing includes both message characteristics as well as perceptions of agent intent to
harm. These findings suggest that a) teases are not fundamentally aversive, but that the

experience of aversiveness is variable, b) that neither the characteristics of a message nor

50

agent intent alone are adequate explanations for the aversiveness of a tease, and c) that a
conceptual and theoretically driven model of teasing can be developed and empirically
tested.

A major implication of these findings is that teasing is a variably aversive
experience, unlike bullying which is always negative. The current study provided
information about the factors that increase and diminish the experience of aversiveness in
response to teasing. The findings of this study in conjunction with future research that
specifically addresses positive outcomes of teasing has the potential to inform policy-
making about anti-bullying programs in schools. A tentative recommendation to policy-
makers is that teasing should not be classified as bullying for the sake of convenience
because there is some evidence to suggest that teasing and bullying are different social
phenomena. Future research would benefit policy-makers by specifically addressing the
potential for teasing to have beneficial relational and personal outcomes. Future research
should also focus on individual, relational and additional message variables as potential

moderators affecting the experience of aversiveness in response to teasing.

51

APPENDICES

52

APPENDIX A

STATISTICAL INFORMATION

53

Table 1

Scale Reliabilities and Descriptive Statistics across Four Teasing Contexts

 

 

 

Reported Tease Scenario 1 Scenario 2 Scenario 3

a M s3 g u s.’ a M i g M s3
Ambiguity .79 3.42 2.70 .78 4.53 2.32 .84 4.22 2.73 .80 .45 2.22
Humor .95 4.32 3.54 .93 3.35 2.23 .96 3.17 3.01 .95 3.50 2.27
Id .92 3.11 3.10 .94 4.63 2.16 .94 5.30 1.77 .94 4.20 1.92
Confrontation
Aggression .87 3.34 3.39 .82 4.81 2.02 .92 5.36 1.74 .87 4.32 2.33
Aversiveness .95 3.24 3.33 .94 4.60 2.13 .95 5.30 2.11 .95 4.16 2.38
Intent to .95 2.61 2.65
harm
Foreseeable .91 3.5 3.2
harm
Table2
Bivariate Correlation Matrix for Subject-Provided T eases

l 3 4 5 6 7

1 Identity Confrontation 1.00
2 Aggression .88" 1.00
3 Ambiguity .11 .11 1.00
4 Humor -.57** -.65** -.08 1.00
5 Intent to Harm .55" .65" .02 -.49** 1.00
6 ForeseeableHarm .62" .71" .I9* -.59** .48" 1.00
7 Aversiveness .71" .81" .16 -.73""'I .71” .72" 1.00

 

* Indicates that the correlation is significant at the .05 level.
** Indicates that the correlation is significant at the .01 level.

Table 3

Beta Coeflicients for Teasing Components as Predictors of A versiveness

 

[3 SE sr
Id Confrontation -.01** 0.10 -.00
Aggression .59 0.1 l .26
Ambiguity .07" 0.06 .07
Humor -.34** 0.06 -.26

Model: F(4, 141) = 94.73, p < .01; R2 = .73; AdjustedRz = .72
** Indicates that the correlation is significant at the .01 level.

54

Table 4

Frequency of Message Classifications for Researcher-Provided Remarks
Insult Tease Joke
Remark 1 58 43 45
Remark 2 83 3 8 23
Remark 3 48 50 47

 

 

 

 

 

Table 5
Means and Contrast Values for Researcher-Provided Remarks for Teasing Components
ID Confrontation Aggression Humor
M sd M sd M sd
Insult] 5.18 1.20 5.51 0.97 4.11 1.01
Teasel 4.41 1.34 4.50 1.23 3.88 1.16
Jokel 4.23 1.21 4.18 1.00 2.46 1.05
w = 0.96, ((143) = 3.86, \I’ = 1.33, ((143) = 6.32, w = -l.65, t(143) = -7.76,
p<.001 p<.OOl p<.001
Insult2 5.68 1.1 l 5.89 0.99 2.37 1.24
Tease2 4.97 0.92 4.92 0.85 4.00 1.27
Joke2 4.51 1.18 4.29 1.47 4.76 1.33
w=l.16,t(141)=4.58, \y=1.60,t(14l)=6.44, w=-2.39,t(141)=-8.05.
p<.001 p<.001 p<.001
lnsult3 5.07 0.94 5.23 1.06 2.75 1.07
Tease3 3.93 1.21 3.82 1.30 3.82 1.38
Joke3 3.59 1.03 3.87 1.22 3.94 1.29
tu= l.51,t(142)=6.88, w=l.36,t(l42)=5.54, w=-1.19,t(l42)=-4.61,
p<.001 p<.001 p<.001

 

 

Table 6
Beta C oeflicients for Intent Variables as Predictors of Aversiveness
B SE sr
Perceived Intent to Harm .48* * 0.06 .42
Perceived F oreseeability of Harm .49“ 0 06 .43

Model: F(2, 143) = 154.27, p < .01; R2 = .68; AdjustedRZ = .68
** Indicates that the correlation is significant at the .01 level.

 

55

Table 7

Beta Coeflicients for Revised Model of Teasing

 

B SE sr
Aggression .30" 0.07 .17
Humor -.27** 0.05 -.20
Perceived Intent to Harm .28* * 0.06 .21
Perceived F oreseeability of Harm .22" 0 06 .15

Model: F(4, 141) = 132.34, p < .01; R2 = .79; AdjustedRz = .78

** Indicates that the correlation is significant at the .01 level.

56

APPENDIX B

FIGURES

57

 

Identity
Confiontation

 

 

 

Aggression

Aversiveness

 

 

 

 

 

Ambiguity

 

 

Humor

 

 

 

Figure 1. Model of the aversiveness as a function of the components of a tease.

 

Perceived Intent to Do
Harm

 

 

 

 

Aversiveness

 

Perceived
F oreseeability of Harm

 

 

 

 

 

 

Figure 2. Model of aversiveness as afitnction of perceptions of intentional or foreseeable
harm.

58

 

Aggression

 

 

Humor

 

 

Perceived Intent
to Do Harm

 

 

Perceived
F oreseeability of
Harm

 

 

 

Figure 3. Revised model of teasing.

59

 

 

Aversiveness

 

 

APPENDIX C

SURVEY ITEMS

60

Table 8

Scale Items and Factor Loadings
Scale Items
Tease Identity Confrontation Items (Factor 1)
The tease made me self—conscious.
The tease made me feel negatively about myself
The tease caused me to feel ashamed of myself
The tease caused me to question my self-worth.
The tease interfered with my ability to present myself
in the manner in which I wanted to be seen. *
The tease coryS'onted my identity.
Tease Aggression Items (Factor 2)
The tease attacked my self-concept.
The tease was insulting.
The tease poked fun at me. **
The tease was a personal attack
The tease made me feel defensive.
Tease Ambiguity Items (Factor 3)
The tease could have been interpreted in more than
one way.
The tease had multiple meanings.
I understood the tease in the only way possible. (R) *
There was one clear interpretation of the tease. (R)
The tease was ambiguous. **
Tease Humorousness Items (Factor 4)
I found the tease fimny.
I was amused by the tease.
The tease made me laugh.
The tease made me smile.
I thought the tease was witty.
I thought the tease was playful.
Intent to Do Harm Items (Factor 5)
My partner purposefully said something that would
hurt me.
My partner knew that the tease would make me feel
bad about myself
My partner intended to hurt my feelings by teasing
me.
My partner was purposefitlly unkind to me.

61

l

.47
.60
.87
.82

.71

.87

.73

Factor Loadings

2

.78

.75
.90

.70
.82

3

.66

.54
.81
.81

4

.94
.92
.95
.95
.67
.78

5 6

.94

.85

.93

.87

7
. 76
.83

Table 8, Continued

Scale Items and Factor Loadings

 

Foreseeable Harm Items (Factor 6)
While I do not believe my partner intended to hurt my feelings,
I fielt that s/he should have known that teasing me would
hurt me.
I believe my partner could have anticipated that I would feel .83
negatively about myself as a result of the tease even if
s/he didn’t mean to make me feel badly.
I do not believe that my partner had malicious intentions when
s/he teased me, but s/he should have known that the
tease would make me feel self-conscious.
I feel that my partner was careless when s/he teased me, but . 78
was not purposefiily unkind.
Tease Aversiveness Items (Factor 7)
Ifelt this tease was mean.
I felt this tease was kind. (R **
I felt this tease was insensitive.
I felt this tease was harsh.
I felt this tease was hurtful.
My partner was wrongto tease me this way.

* Indicates that the item was deleted for the assessment of researcher-provided remarks

.99

.81

.8]

.73

.87

.89
.91
.95
.87

"' * Indicates that the item was deleted from all instances of the scale due to compromised

reliability as a result of the item’s inclusion in the scale.

Bold values indicate that this was the factor on which the item had the highest factor

loading when the highest loading was not on the expected factor.

62

L)

10.

11.

Survey Items Related to the Subject-Provided Teases

. Teasing prompt: “Please recall a time within the last 2 days in which a close

friend teased you. A tease is a remark made to you by another person that was
probably meant to be funny, but at the same time called you out on something
about yourself (your ability, some aspect of your personality or appearance, etc.)
What is the first initial of the person who teased you?

What is the sex of the person who teased you?

What is the approximate age of the person who teased you?

Which of the following best describes the type of relationship you have with the
person who teased you (subjects chose from stranger, acquaintance, friend, close

friend and romantic partner)

Approximately how long have you known the person who teased you (subjects
indicated years and months of the duration of the relationship).

Please indicate the date on which the tease occurred.
At approximately what time did the tease occur?
Please specify the location in which the tease occurred.

Were other people present at the time of the teasing episode (subject selected yes
or no from a drop-down menu).

Describe, in as much detail as you can, the actual content of the tease. Include, to
the best of your memory, events, actions or conversation that preceded the tease.
Also, try as hard as possible to report exactly how the tease was worded. Finally,
include anything you said in response to the tease.

Survey Items Related to Participant Demographic Characteristics

. Please indicate your age (participants selected their age in years from a drop-

down menu).

Please indicate your sex (participants selected either male or female from a drop-
down menu).

63

 

Researcher-Provided Remarks

Remarkl:

Terry and Chris are close friends. They are sitting outside on a summer day having a
casual conversation. The fi'iends were reminiscing about the beginning of their
relationship. Terry says to Chris, "You know, when I first met you, I thought you were
an annoying, pretentious snob. "

Remark 2:

Terry and Chris are close friends talking about T erry's upcoming date with a new person.
Chris says to Terry, "You should probably work out pretty hard the next few days - you
know nobody likes a fatty. "

Remark 3:

Terry and Chris are discussing how Chris has been single and dateless for several
months in a row. Chris complains that there are no good dating options in the area.
Terry replies by saying "Maybe you should lower your standards - you might be aiming
a little too high. "

64

REFERENCES

Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal
attachments as a fundamental human motivation. Psychological Bulletin, 11 7,
497-529.

Baxter, L. A. (1992). Forms and functions of intimate play in personal relationships.
Human Communication Research, 18(3), 336-363.

Cohen, J ., Cohen, R, West. S. G., & Aiken, L. S. (2003). Applied Multiple
Regression/Correlation Analysis for the Behavioral Sciences (3rd ed.). Mahwah,
NJ: Lawrence Erlbaum Associates.

Coker, D. A., & Burgoon, J. K. (1987). The nature of conversational involvement and
nonverbal encoding patterns. Human Communication Research, 13, 463-494.

Coulson, S., & Kutas, M. (2001). Getting it: Human event-related brain response tojokes
in good and poor comprehenders. Neuroscience Letters, 316, 71-74.

Forrest, S., Eatough, V., & Shelvin, M. (2005). Measuring adult indirect aggression: The
development and psychometric assessment of the indirect aggression scales.
Aggressive Behavior, 31, 84-97.

Gleason, J. H., Alexander, A. M., & Somers, C. L. (2000). Later adolescents' reactions to
three types of childhood teasing: Relations with self-esteem and body image.
Social Behavior and Personality, 28(5), 471 -480.

Goffman, E. (1 95 9). The Presentation of Self in Everyday Life. New York: Doubleday.

Goffman, E. (1967). Interaction Ritual: Essays in Face-to-Face Behavior. Chicago:
Aldine.

Horowitz, J. A., Vessey, J. A., Carlson, K. L., Bradley, J. F ., Montoya, C., McCollough.
B., & David, J. (2004). Teasing and bullying experiences of middle school
students. Journal of the American Psychiatric Nurses Association, 10, 165-172.

65

Hurren, B. L. (2006). The effects of principals' humor on teachers' job satisfaction.
Educational Studies, 32, 373-385.

lmahori, T. T., & Cupach, W. R. (2005). Identity Management Theory: Facework in
Intercultural Relationships. In W. B. Gudykunst (Ed.), T heorizing about
Intercultural Communication. California: Sage.

Infante, D. A., & Wigley, C. J., 111. (1986). Verbal aggressiveness: An interpersonal
model and measure. Communication Monographs, 53, 63-69.

Keery, H., Boutelle, K., van den Berg, P., & Thompson, J. K. (2005). The impact of
appearance-related teasing by family members. Journal of Adolescent Health, 37.
120-127.

Kelly, W. E. (2002). An investigation of worry and sense of humor. The Journal of
Psychology, 136, 657-666.

Keltner, D., Capps, L., Kring, A. M., Young, R. C., & Heerey, E. A. (2001). Just teasing:
A conceptual analysis and empirical review. Psychological Bulletin, 127(2), 229-
248.

Keltner, D., Young, R. C., Oemig, C., & Monarch, N. D. (1998). Teasing in hierarchical
and intimate relations. Journal of Personality and Social Psychology, 75(5),
1231-1247.

Kowalski, R. M. (2001). Aversive interpersonal behaviors: On being annoying,
thoughtless, and mean. In R. M. Kowalski (Ed.), Behaving Badly: Aversive
Behaviors in Interpersonal Relationships. Washington, DC: American
Psychological Association.

Kowalski, R. M. (2003). Complaining, Teasing, and Other Annoying Behavior. New
Haven: Yale University Press.

Kowalski, R. M. (2004). Proneness to, perceptions of, and responses to teasing: The
influence of both intrapersonal and interpersonal factors. European Journal of
Psychology, 18, 331-349.

Kowalski, R. M., Howerton, E., & McKenzie, M. (2001). Permitted disrespect: Teasing
in interpersonal Interactions. In R. M. Kowalski (Ed.), Behaving Badly: A versive

66

Behaviors in Interpersonal Relationships. Washington. DC: American
Psychological Association.

Larsen, D., & Dehle, C. (2007). Rural adolescent aggression and parental emotional
support. Adolescence, 42(25-50).

Leary, M. R., Kowalski, R. M., Smith, L., & Phillips, S. (2003). Teasing, rejection, and
violence: Case studies of the school shootings. Aggressive Behavior, 29, 202-214.

Leary, M. R., & Springer, C. A. (2001). Hurt feelings: The neglected emotion. In R. M.
Kowalski (Ed.), Behaving Badly: Aversive Behaviors in Interpersonal
Relationships. Washington, DC: American Psychological Association.

Lunner, K., Werthem, E. H., Thompson, J. K., Paxton, S. J., McDonald, F., &
Halvaarson, K. S. (2000). A cross-cultural examination of weight-related teasing,
body image, and eating disturbance, in Swedish and Australian samples.
International Journal of Eating Disorders, 28, 430-435.

McCabe, R. E., Antony, M. M., Summerfeldt, L. J., Liss, A., & Swinson, R. P. (2003).
Preliminary examination of the relationship between anxiety disorders in adults
and self-reported history of teasing or bullying experiences. Cognitive Behaviour
Therapy, 32(4), 187-193.

Miller, R. S. (2001). Breaches of propriety. In R. M. Kowalski (Ed.), Behaving Badly:
Aversive Behaviors in Interpersonal Relationships. Washington, DC: American
Psychological Association.

Simmons, R. (2003). Odd Girl Out: The Hidden Culture of Aggression in Girls:
Harcourt.

Smith, P. K., & Shu, S. (2000). What good schools can do about bullying: Findings from
a survey in English schools after a decade of research and action. Childhood, 7,
193-212.

Solberg, M. E., Olweus, D., & Endresen, I. M. (2007). Bullies and victims at school: Are
they the same pupils. British Journal of Psychology, 77, 441-464.

Spitzberg, B. H., & Cupach, W. R. (1984). Interpersonal Communication Competence.
Beverly Hills: Sage.

67

Tedeschi, J. T., & Nesler, M. S. (1993). Grievances: Development and reactions. In R. B.
Felson & J. T. Tedeschi (Eds), Aggression and Violence: Social Interactionist
Perspectives. Washington, DC: American Psychological Association.

Ting-Toomey, S. (2005). The Matrix of Face: An Updated Face Negotiation Theory. In
W. B. Gudykunst (Ed.), Theorizing about Intercultural Communication.
California: Sage.

Warkentin, J. B., & Gidycz, C. A. (2007). The use and acceptance of sexually aggressive
tactics in college men. Journal of Interpersonal Violence, 22, 829-850.

Yip, J. A., & Martin, R. A. (2006). Sense of humor, emotional intelligence, and social
competence. Journal of Research in Personality, 40, 1202-1208.

68