$3.. i. r.

‘vv .

 

 

SITY LIBRARIES

IIIIIII IIIII I I

 

 

 

 

 

 

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIII

3 1293 0155

1

This is to certify that the

dissertation entitled

TESTING POWER AND RELATIONSHIP ASSUMPTIONS
IN SAFETY REQUESTS

presented by

James Mulcmne

has been accepted towards fulfillment
of the requirements for

 

PhD- degree in Lannmnication ;

Dr. Charles Atkin

Major professor

Date W915.—

MSU i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

 

 

LIBRARY
M'chlgan State
University

 

 

 

PLACE II RETURN BOX to roman this chockout from your rocord.

1'0 AVOID FINES rotum on or baton dot. duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Afflrmotlvo Action/Equal Opportunity lnotltutlon
Wanna-91

 

TESTING POWER AND RELATIONSHIP ASSUMPTIONS IN SAFETY REQUESTS

By

James Mulcrone

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Communication

1 996

ABSTRACT

TESTING POWER AND RELATIONSHIP ASSUMPTIONS IN SAFETY REQUESTS

A substantial portion of persuasion research focuses on request formation,
specifically the types and forms requests take in response to particular situations. This study
examines the directness, politeness, and elaboration for manipulated conditions of low and
high situational power and low and high relational intimacy, for initial requests and in
response to rejection of the initial requests, with requests made of unbuckled persons to wear
their automotive safety belt. Additionally, the link between mass media messages and
interpersonal communication is explored by tests of slogans that may be used in mass media
campaigns encouraging seat belt wearers to ask unbuckled others to wear the seat belt.

Results showed no hypothesized main effects for power and relationship on the
politeness, directness, and elaboration of initial requests. However, additional analyses
showed an interaction effect of power and relationship on directness that suggests subjects
consider both power and relationship in framing these requests. Surprisingly, subjects in the
low power and non-intimate relationship condition made the most direct requests. Gender
differences also appeared in additional analyses.

In response to rejection, subjects’ requests did not support the changes hypothesized.
Situational power did not affect the directness of requests and had an effect opposite what
was hypothesized on politeness. Requests afier rejection were less polite, and more
elaborate, than the initial request for those in the low power condition. Coders did identify
qualitative differences in second requests from initial requests that showed some change.

The altruistic type of request or a weak manipulation is offered as explanations for the lack

of significant findings.

Slogan evaluations generally favored a theme connected to Good Health for broad
campaign use. A slogan reminding unbuckled persons about the state seat belt Law was
preferred by passengers in requests to drivers. The Law substitutes for high power in this
situation. The data suggest different slogans are applicable in certain conditions and that
targeting messages to conditions will be effective. Self-report data support the overall
approach; subjects report a willingness to ask others, say they want to be reminded, and

think their requests generate a high sense of obligation to buckle up.

Cepyright by
JAMES MULCRONE

1996

Dedicated, with love, to my wife, Susan M. Ball, and children,

James Ryan, Patrick Louis, and Lauren Marie.

ACKNOWLEDGMENTS

Completion of my Ph.D. program would not have been possible without the help and
support of many people. Primarily, I want to thank my wife, Susan M. Ball, D.V.M.,
Ph.D.(MSU 1981; 1993), and children, Ryan, Patrick, and Lauren. They encouraged,
supported and suffered through the process over the years. I also want to thank my extended
family for their unwavering support. I thank God for the gifts he has given me to complete
this work.

My committee is composed of faculty that I admired as instructors and researchers.
They have been patient, supportive, and available when I needed their assistance. My
committee chairperson, Dr. Charles Atkin, stands out among the faculty in the Department
of Communication for his length and breadth of service to MSU. He is an outstanding
instructor and committee chair. He can get to the heart of issues, grasp key ideas, and
always make the work seem easier. I will never forget the discussions we had over the years
about communication, advertising, research and sports. Thank you for everything, Chuck.

I want to thank my other internal committee members, Dr. Bill Donohue and Dr. Ron
Tamborini. They brought a balance in personality and style to both the department and my
Ph.D. program. A special thank you to my external committee member, Dr. Stan Kaplowitz,
from the Department of Sociology at MSU. He made himself available and always had time
for my work throughout the years. Dr. Kaplowitz is both an outstanding instructor and

researcher who has a great rapport with Communication students and departmental faculty.

vi

I also wish to recognize the other faculty in the Department of Communication at
MSU for providing an outstanding and quality education. The department faculty was
particularly strong in the interpersonal, mass media and required course studies. I also wish
acknowledge my classmates. We had a great group and I am relieved to complete the
program and be counted among them - graduates of the Ph.D. program in the Department of
Communication at MSU. Finally as I continue to work and meet others personally and
professionally, I am amazed at the network of people from the MSU Department of

Communication; I intend to be actively involved as an alumnus.

vii

Table of Contents

Testing Power and Relationship Assumptions in Safety Requests .................. 1
Request Form ..................................................... S
Intimacy and Requests ............................................. 10
Power and Requests ............................................... 12

Automotive Safety Belt Use .............................................. 15
Previous Research ................................................ 18

Response to Rejection ........................... ‘ ........................ 24

Communication Campaigns ............................................... 26
Application Data ................................................. 30
Background Data ................................................. 34

Method ............................................................... 38
Design ......................................................... 38
Manipulations ................................................... 39
Subjects ........................................................ 41
Procedure ....................................................... 41
Instrumentation .................................................. 42
Coding Procedures ................................................ 43

viii

Results ............................................................... 46

Sample Description ............................................... 46
Independent Variables ............................................. 48
Slogan Evaluation Scales ........................................... 49
Media Choices ................................................... 50
Findings .............................................................. 51
Discussion ............................................................ 68
Initial Requests .................................................. 68
Response to Rejection ............................................. 71
Slogan Evaluations ................................................ 74
Appendix A ........................................................... 83
Appendix B ........................................................... 86
Appendix C ........................................................... 99
List of References ..................................................... 101

ix

List of Tables

Table 1 ......................................................... 52
Table 2 ......................................................... 53
Table 3 ......................................................... 58
Table 4 ......................................................... 59
Table 5 ......................................................... 60
Table 6 ......................................................... 62
Table 7 ......................................................... 65
Table 8 ......................................................... 66
Table A-1 ....................................................... 83
Table A-2a ...................................................... 84
Table A-2b ...................................................... 85

Testing Power and Relationship Assumptions in Safety Requests

Much of hmnan communication can be viewed as persuasive in nature as persons
seek control in their environment to realize various physical, economic, and social rewards
(Miller and Steinberg, 1975). Many of these persuasive communications are requests that
are judged to be successful if the other person complies with the request. One way scholars
have advanced the study of communication is to manipulate variables expected or known to
affect communication behavior in predictable ways and then examine the dependent
communication variables to determine if the predictions are fulfilled (Berger, 1985). Power
and relationship are two of the most studied variables in communication research. Recently
they have been used in a series of studies by Roloff and colleagues (Jordan and Roloff, 1990;
Roloff and Janiszewski, 1989; Roloff, Janiszewski, McGrath, Burns and Monrai, 1988 )
examining communication outcomes in manipulated situations involving persuasive
communication. These studies indicate that power and relationship differences are important
in the formation of persuasive requests. The testing of various situational, relational, and
categorical variables, can provide more data on how specific variables affect communication
request formation in interpersonal compliance gaining. With a greater understanding of
interpersonal compliance gaining requests may come direction for media campaigns to
activate persuasive communication on issues that benefit society or individuals. Health and
safety issues are important content areas for coordinating mass media and interpersonal
communication strategies and worthy of additional research.

Understanding communication choices in interpersonal requests can help provide

2

direction to make mass media campaigns more effective in encouraging interpersonal
communication that is congruent with health and safety conscious behaviors. Several studies
have decumented the success of mass media stimulated interpersonal communication in
health studies ( Maccoby, Farquhar, Wood, and Alexander, 1977; Maccoby and Solomon,
1981). Examples of current high profile health and safety issues include: smoking, drinking
and driving, AIDS and IV drug use, safe sex and AIDS, other disease control, and automotive
safety belt use. The focus of this research is on the manipulated variables of power,
relationship, and rejection of the initial request and how they affect communication strategies
in requests.

In this study, power and relationship are manipulated in the context of a safety
request to buckle up the automotive safety belt. It is hypothesized that power and
relationship differences will influence the directness, politeness, and elaboration of the
persuasive message requests generated by the subjects. Tests of the change in request
formation after being told of a refusal to comply with the initial request are also made. The
study also will comparatively evaluate some mass media campaign slogans that are suggested
by the literature as potentially effective in encouraging interpersonal communication
regarding wearing seat belts.

This research has specific objectives. First, to test assumptions about persuasive
requests due to power and relationship differences in the content area of automotive safety
belt use and develop theoretical arguments about the persuasive strategies employed. This
provides a theoretical link to the application of requests to wear the seat belt among the
subject population. Second, to test expected changes in requests when there is no

compliance with the initial request; this segment tests theoretical assumptions about

3
response to rejection. Third, to test different mass media slogans that might be effective
in activating interpersonal communication aimed at increasing automotive safety belt use and
provide some evaluation about their perceived effectiveness. The approach suggested here,
using personal communication from users to nonusers to increase seat belt use, has not been
widely tested.

The rationale for this study is based on the belief that a campaign with a focus on
stimulating personal (or interpersonal) requests to "Buckle Up" can be a successful strategy
for increasing automotive safety belt use. For a campaign to be successful it has to have a
message that will activate and remind the target audience to ask others to buckle up. Media
messages can be evaluated on various components: easy to remember, clear, informative,
positive, likable, effective, and self or other directed, that may be linked to message
effectiveness because of the attractiveness, recall, clarity or comprehension of the message
(McGuire, 1989). McGuire (1989) reviews the theoretical foundations of public
communication campaigns and presents a persuasive model that details input (independent
variables) and output (dependent variables) important to the communication process. The
research proposed here follows many of McGuire’s (1989) campaign guidelines, including:
a realistic problem that can be suitably addressed by the proposed research; a fair amount
of existing research on the subject; testing out the most promising themes; constructing the
communication, and evaluating the effectiveness (this is only done in the context of the study
and not a campaign application evaluation). This research meets these criteria and expands
on the existing data in some of the categories. The broad assumptions being tested are that
a mass media campaign could, first, encourage communication from among peers about

safety belt use as people remind each other to buckle up. Second, the media messages can

4
assist in the activation and formulation of the persuasive messages because they can focus
on activating the peer influences. In this respect the research follows many of the guidelines
of a formative evaluation research campaign outlined by Atkin and F reimuth (1989). This
type of campaign approach may provide a cost-effective way to increase safety belt use
through the combination of mass media and interpersonal communication messages.

It appears that the potential for changing seat belt behavior from nonuse to use, and
making the change long lasting, is good with a strategy that includes interpersonal and mass
media communication. The use of mass media messages to influence interpersonal
communication can change perceptions about seat belt use in the target group. Miller (1986:
1987a: 1987b) has argued that we have very limited knowledge of the interconnectedness
between macro societal levels of persuasion and the inter- and intrapersonal. It is expected
that the ability to analyze and then individually construct a persuasive message will result not
only in high compliance with the seat belt request, but in long term behavior change at a rate
much greater than a mass media message would generate alone. Generating interpersonal
communication addresses several problem areas with the seat belt issue: perceptions of use
among peers, personal reinforcement for long-term change, and involving those who already
"Buckle Up,” that have not been included and therefore limited the success of previous
automotive safety belt campaigns. The study can help clarify the potential of a campaign
using both mass media and interpersonal communication and test media messages to activate
persuasive communication in a one-to-one setting.

This approach may be particularly appropriate for "laissez-faire" or secondary
enforcement areas, such as the State of Michigan, where methods for increasing safety belt

use without primary enforcement of the law is a critical issue. If the overall campaign

5
strategy suggested here is successful for drivers and right front-seat passengers, it could be
applied to rear-seat occupants whose seat belt use is not governed by the law. Belt use for
rear-seat occupants is much lower at 35.6%, compared to 50.6% for front-seat occupants
(Streff and Molnar, 1990). The benefits, both from a personal and monetary standpoint, from
a campaign strategy that can improve automotive seat belt use are great. This study provides
test data to determine if the mass media and interpersonal communication approach can be
effective and some direction on how to proceed. Evaluation of the communication request

is key to this research.

RequestEQrm

As noted earlier, persons communicate to control their environment as they seek
certain desirable outcomes over other, less desirable, outcomes (Miller and Steinberg, 1975).
This means that all communication is strategic as communicators assess the concerns of
textual coherence, interaction efficiency, and face presentation and protection present in a
communication situation (O'Keefe and Delia, 1982) before framing a request. Some
researchers (Gibbs, 1986; F rancik and Clark, 1985) have proposed an "obstacle hypothesis"
that suggests that individuals frame compliance gaining requests "to overcome the greatest
potential obstacle they see to getting the information they wan " (1985, p. 560). Several
studies have focused on how persons perceive and attack the "obstacles to compliance" in
very specific situations (Jordan and Roloff, 1990; Roloff and Janiszewski, 1989; Gibbs,
1986; Francik and Clark, 1985). These investigators found that the type of request and the

level of intimacy in the relationship are two key variables. Each situation has obstacles,

6

anything that stands in the way of compliance with the request, that the communicator has
to take into consideration in forming the request. Requestors have a wide variety of choices
available when making a request of another. What then determines or limits the options of
a requestor? The way a requester formulates the request indicates how she or he has assessed
the components of the interaction, both situational and relational, and what judgements are
made about formulating a request that is consistent with the assessment of the situation.
While compliance with the request is a primary concern, secondary issues, like time
constraints, and situational and relational considerations, are often present. These affect
request formation. Often there is a need to quickly determine whether or not you need to
seek compliance in another way or from another person.

Since compliance with a request cannot take place without the listener making correct
inferences about the meaning, shared meaning is important to achieving successful
communication and probably most important in requests (Gibbs, 1985). The request must
be in a form that is interpretable to the receiver and appropriate for the situation. The
communication request itself shows how the individual has assessed the relational and
situational concerns present in trying to achieve compliance with the request. Request forms
have a direct or literal component and an indirect or implied component (Gibbs, 1985).
Overall, information about the social setting, and broader knowledge of the roles of the
individuals and their relationship, can determine if a request type is conventional, or not
conventional, for that situation (Gibbs, 1985). Conventional, or appropriate, requests meet
the needs of making, understanding, and remembering requests in a particular situation
(Gibbs, 1985). Conventional requests focus on a shared knowledge of the interactants in the

particular social situation that goes beyond just a linguistic interpretation to a shared meaning

7
(Gibbs, 1985). This makes the request more easily understood.

Because requestors are aware of obstacles, research has indicated they often frame
compliance gaining requests in an indirect or conditional manner (Ervin-Tripp, 1976;
Kemper and Thissen, 1981; F rancik and Clark, 1985). These indirect requests, implications
of what is wanted as opposed to a direct statement, have been found to be conventional and
easily interpretable in numerous situations and have often allowed the recipient of the request
to save face (Gibbs, 1986) in the situation. The indirect requests are indicative of the
perception of the obstacle and consideration of face, relational history, and power (Gibbs,
1986). Researchers have begun to test hypotheses in order to clarify the conditional
differences in requests. The measurement of the directness of the request is an important
indication of the recognition of situational considerations.

One effective method of communication in requests is the use of indirect requests.
Indirect requests, which utilize inferences to obtain the desired response, can substitute for
direct requests which are generally more conventional. Direct requests focus on the specific
behavior expected of the other (Rake the leaves! Please loan me $5.) and not on any
introduction or softening of the request (I think you need you to rake the leaves. Do you
think you can loan me $5?) Gibbs (1979, 1981a, 1981b, 1982, 1983, and 1985) has
researched the conventional nature of indirect requests and concludes that whether the
request is conventional or not depends on the social context. He also found that, in general,
indirect requests are conventional and vary little in "convention of form" across various
situations but do vary in "convention of means" in response to situational (relational, request
type) differences. For example, in many routine circumstances such as service requests, the

request is easily planned (mentally automatic) and easily comprehended by the listener; these

8

routine situations allow the receiver of the request to understand the meaning of many
different forms of indirect requests without taking the time to analyze the literal meaning of
each request (Gibbs, 1985). To summarize, in most common situations it takes less time for
the speaker to formulate the indirect request and less time for the listener to recognize,
process, and respond to an indirect request (Gibbs, 1985). As the situation necessitating a
request becomes more unusual or complex, more time and effort would be expected in the
processing and formulation of both the request and the understanding of the request. The
uses of direct or indirect literal, grammatical, or surface forms of requests are important
because they affect the understanding and remembering of the requests (Gibbs, 1985).
Understanding and remembering the request are important to the direction that the interaction
will take. Indirect requests are good examples of a communication approach by the
requestor because the request is being framed in an indirect way due to particular situational
perceptions.

Roloff and Janiszewski (1989) hypothesized that the perceptions of the situation and
the obstacles to compliance are determined by the type of help being requested and the nature
of the relationship between the communicators. Thus, Roloff and colleagues (Jordan and
Roloff, 1990; Roloff and Janiszewski, 1989; Roloff, Janiszewski, McGrath, Burns and
Monrai, 1988) have examined persuasive requests in help seeking situations. The type of
help being requested has been categorized as borrowing a resource or asking a favor. The
nature of the relationship with the requestor was varied between an intimate fiiend, a fiiend,
or an acquaintance. In brief, borrowing generally involves some concrete resource which is
physically transferred for use by the requestor for a specific amount of time and under

specific conditions, such as maintaining the resource in its present condition or replacing it

9
if it is consumed (Roloff and Janiszewski, 1989). In contrast, favors are generally services
which are performed by the listener and involve some expenditure of time and energy (Roloff
and Janiszewski, 1989). Favors can take a wide variety of forms for various requests that do
not involve concrete resources, such as asking for a ride or typing a paper; these favors were
used in a previous study (Roloff and Janiszewski, 1989).

Favors vary in the degree of benefit to the listener. Some compliance gaining research
(Hunter and Boster, 1979; Williams and Boster, 1981; Boster and Stiff, 1984) has
concluded that the benefit to the listener is an important variable. And, that requestors will
make more of an effort at gaining compliance and will perceive more compliance messages
as acceptable when the benefit to the listener is high. Requests for compliance involving
health and safety issues, such as "Please don't smoke." and "Please wear your seat belt. ", are
more closely related to the favor (no concrete resources involved) category then to the
borrowing category. However, these particular safety requests have not been researched as
a separate category of favors and examining the requests used in this particular situation.
These and other health and safety issues may also be categorized as high in benefit to the
listener and therefore likely to generate many strategies and substantial effort by the requestor
to attain compliance. Safety requests may be a unique type of request, with a high altruistic
component, not yet clearly identified in the existing research but worthy of specific study.

Other research (Maccoby and Solomon, 1981) has indicated that the media can
positively stimulate communication regarding healthy behavior. If, applying the current
categories of requests, safety and health requests may be expected to resemble favors, they
are more likely to be indirect and indicative of the greatest obstacle to compliance given the

perception of specific relational, situational, and face considerations.

10
One type of approach which utilizes an indirect request is a "conditional request.”
The request is conditional on the elimination or the absence of the greatest obstacle as
perceived by the requestor (Clark, 1979). F rancik and Clark (1985) state that requestors
evaluate the preconditions necessary for obtaining the information they desire from others
and then focus their attention on the conditional aspect of the request and not on the request
itself (if, or because, p, do q). An example of this form of request is, "Did you read in the
paper what time the game is tonight?" which means "Did you read about the game, and if you
did, what time does it start?". Here, the important preconditions belong to the categories of
ability (is the receiver of the request able to give the information) and willingness (is the
receiver of the request willing to give the information) to respond to the request (Francik and
Clark, 1985). The requestor evaluates the current situation making decisions based on
information such as the relationship with the receiver, the receiver's habits, and general
knowledge about where and how people acquire information (Francik and Clark, 1985).
Such "conditional requests" allow for a high degree of politeness and flexibility in framing
requests and can assist the person in searching for the requested information in his or her
memory (F rancik and Clark, 1985). These examples indicate how a requestor evaluates the
situation and decides a communication approach that is appropriate. These request
considerations are important in categorizing requests on a scale of directness. Situational
power and relational intimacy are major concerns in planning and evaluating request
formation, and as power and intimacy are varied, request communication is expected to

change in a predictable manner.

11

NW

The level of intimacy with another is both a result of previous communication and
a predictor of differences in future communication. Miller and Steinberg (1975) defined
interpersonal and noninterpersonal relationships on the basis of qualitative and quantitative
differences in information. Personal information obtained from another person helps plan
future communication and affects the expectations of outcomes from this communication
(Miller and Steinberg, 1975). Knapp (1984) describes intimacy as the bond between two
individuals that involves a self-fulfilling attachment and a self-surrendering caring. In
examining compliance gaining situations, intimate knowledge helps direct persuasive
strategies based on the increased level of personal information available. Conceptually, these
individuals have greater knowledge and relational history with each other than with a friend
or acquaintance. This relational history makes maintenance of the relationship a high priority
and this, in turn, affects the formulation of compliance gaining requests. One way that
messages differ between intimate individuals and non-intimate others is in the elaboration
of the request itself; intimate individuals require less elaboration than non-intimate
individuals. Research reveals that not only do we, as communicators, have greater
knowledge about intimate individuals, which assists in the formation of compliance requests,
but we also have greater expectations that these relationally close individuals will comply
with the request. Studies have found that individuals are more likely to seek assistance from
a friend than a stranger and that these individuals expect the fiiend to be more likely to
comply with the request (Shapiro, 1980). Other researchers concluded that fiiends are more

obligated than strangers to comply with requests (Bar-Ta], Bar-Zohar, Greenberg, and

12
Hermon, 1977). Relational history and knowledge affect the choice of who and how one
seeks assistance from another, as well as the expectation of compliance. It also suggests
different problems (obstacles) in the evaluation of the situation for initial requests and if
compliance with the request is rejected. Thus relational considerations result in more
individual planning in the formation of the request itself and also increase use of conditional
requests which are "face-saving" for the listener. This leads to the increased use of indirect
requests which help maintain the relationship. So, one of the effects of relational intimacy
is to expect requests that are more sensitive to the other person. That generally means more
polite and indirect requests because of the importance of the relationship. Relational intimacy
is an important consideration in the formulation of compliance gaining requests and the
evaluation of the response. Situational power is another variable expected to vary the

communication strategy in formulating a request.

Wars

Power has been an important concept in social science research and continues as a
key variable in communication studies. Shaw (1981) defines power as "control of
reinforcers" (p.294); where one person has the ability to affect behavior in another. In his
review of persuasion, Bostrom (1983) discusses power as an influencing agent in
communication and equates it with credibility in the communicator. Power is a central
construct in persuasion relating to the sources of influence. In an assessment of the social
bases of power that has guided much of the ensuing research, French and Raven (1959)
identified 5 kinds of social power used as ways to control or influence another person:

attraction power, reward power, coercive power, legitimate power, and expert power. In

13

social relationships that move through various contexts, it is possible that at some point any
of the five bases of social power may dominate. There may also be situational considerations
where multiple types of power are present based on relational and situational considerations.
Noting the reciprocity of power and relationships, Berger (1985, p. 440) stated, "Persons can
actualize power through their communicative conduct, and communicative conduct can serve
for making inferences about a given individual's ability to exercise power." Upon
examining the request, inferences can be made about perceived obstacles to compliance and
how power may have affected the compliance gaining strategy.

Power, then, is important in the flaming of requests to other individuals. Kemper and
Thissen (1981) use examples of superiors and subordinates in detailing the situational uses
of direct and indirect requests. This research provides examples of the use of different
request forms due to power differences. High power, when an individual controls
reinforcers, can lead to directness and ordering with little elaboration in a request; these
direct requests may also be perceived as impolite and self-directed (Kemper and Thissen,
1981). In the case of the driver of an automobile, he or she might be more self directed in
requests due to power differences. Direct requests are appropriate when comprehension
cannot be assumed but, as in a high power case, compliance can be assumed (Kemper and
Thissen, 1981). Those in positions of power ("high power") can use direct forms to ensure
compliance and comprehension. Conversely, when the level of social or situational power
is low ("low power"), the requests tend to be high in elaboration, socially appropriate (polite),
and indirect (Kemper and Thissen, 1981). These indirect requests give the listener more
options when forming his or her response (Kemper and Thissen, 1981). Indirect requests are

seen as more polite and less face-threatening (Brown and Levinson, 1987) and beneficial to

14
maintaining a focus on the relationship. In summary, low power situations create high
obstacles and generally limit the requestor to indirect forms, deferring to the higher status of
the other; while high power situations allow for choice on the part of the requestor in the use
of direct or indirect requests based on a different perception of the obstacles (Kemper and
Thissen, 198 1).

Research on different types of facework, closely related to politeness, indicates power
is not likely to be used to assure compliance between fiiends (Lirn and Bowers, 1991), since
friends are much more concerned with relational maintenance and liking by the other.
Acquaintance relationships do not share the same high level of maintenance and liking
concerns. Given these conditions, a requestor must select a communication form for the
situation on the dimensions of politeness and directness (conventions of conversation) that
are appropriate for the particular situation (Kemper and Thissen, 1981).

Elaboration is monitored as a separate variable. It may be expected to have a positive
correlation with indirect requests because by their nature indirect requests may be longer.
However, elaboration should provide more than just the difference between a simple indirect
request (Don’t you think that you ...) and a direct request (You should ...). Subjects are
expected to offer some explanation and perhaps conversation differences as part of a request.
Any increased explanation, conversation, or associated rationale will increase the
elaboration in the request. Elaboration is expected to differ due to specific situational
considerations of power and relationship and not just as specifically related to the directness
of the request. Elaboration provides another dependent variable to track request variations
on the issue of automotive safety belt use.

The request can be evaluated and categorized on the dimensions of directness,

|I

1 5
politeness, and elaboration previously discussed. In a scaling study of "request forms" on
the dimensions of directness and politeness, Kemper and Thissen (1981) provide
comparative ratings, using a multidimensional scaling approach, of different request forms
across these two dimensions. The study revealed that the most direct and least polite request
was a simple action imperative sentence (Rake the leaves); while the use of "Please" before
the imperative was rated the most polite and "I think...", "Do you think...", and "Don't you
think..." (The leaves need to be raked), three forms incorporating a need assertion, were
rated the least direct (Kemper and Thissen, 1981). Other polite request forms included:
"Should you...", "Did you...", "Do you think you...", and "You should..." followed by the
imperative (Kemper and Thissen, 1981). Other direct forms included: "You should...",
"Why don't you...", "I think you...", "Don't you think you should...", and "Please..." followed
by the imperative (Kemper and Thissen, 1981). The rank ordering of request forms on the
basis of politeness and directness provides a coding scheme for analyzing these dimensions
in the flaming of requests. The Kemper and Thissen (1981) coding scheme, applied in this
study, is detailed in Appendix C. The issue of requests to wear automotive safety belts when
riding with another person provides a safety issue which is applicable to testing compliance
gaining strategies by examining directness and politeness conventions in different intimacy

and power situations.

WSW

Automotive safety belt use is a key health and safety issue addressed both personally

and publicly in today's society. Automobile accidents are the leading cause of fatalities for

16

all age groups flom 1 to 34 years old (Baker, O'Neill, and Karpf, 1984). Auto accident are
a large cause of lost work time costing billions of dollars in insurance costs, medical care,
and lost productivity (American Demographics, 1987). Research estimates (Evans, 1987)
indicate that these costs can be lowered significantly by the simple act of using the
automotive safety belt correctly. Automotive crash data have shown that the use of the
automotive safety belt is the single most effective method for lowering fatalities and reducing
the seriousness of crash injuries. In response to these facts, 33 states, the District of
Columbia, and several Western nations have legislated some form of mandatory automotive
safety belt use. While these laws have clearly been successful in increasing use and lowering
fatalities and serious injuries, maximum benefits flom safety belt use are not being realized
due to noncompliance with the laws governing usage (Evans, 1987). The search for a cost
effective method or campaign to increase automotive safety belt use continues.

The automotive safety belt issue continues to be controversial due to differences in
laws (primary versus secondary type of enforcement) and the advancement of other safety
devices, such as passive restraint seat belt systems and air bags. Not all passive or automatic
safety belt systems provide the same amount of protection (Keebler and Pinto, 1991 ). The
automatic system used by General Motors has come under criticism and continues to undergo
further study by the National Highway Traffic Safety Administration (NHTSA) (Keebler and
Pinto, 1991) and development by GM. Problems with the GM system include: separate
shoulder and lap belts (the shoulder belt is motorized and the lap belt is manual), anchoring
of the belts on the door, and the ability to disconnect the belts (Keebler and Pinto, 1991).
The last problem directly negates the original purpose of passive restraint systems. A study

that examined safety belt use among automatic systems found lower use for this system

17
(74.3%) than for other motorized systems (79.2%) (Streff and Molnar, 1990) in the float seat
of the vehicle.

The air bag is mandatory on the driver's side of automobiles, and increasingly more
common on the right flont seat passenger's side in many models. While air bags are a safety
advancement, the potential exists for individuals to get a false sense of security flom air bags
and neglect the use of safety belts. Safety belts are designed to be used regardless of whether
or not an automobile is equipped with air bags, since air bags are only effective in head-on
collisions. As the presence of airbags increases in all types of vehicles, it may become even
more difficult to get people to "Buckle Up", especially if individuals do not have a strong
pattern of safety belt use or view the airbag as a substitute safety device replacing the safety
belt. One study (Williams, Wells, and Lund, 1990) found no difference in the use of manual
safety belts in cars equipped with the safety belts and safety belts and air bags. Belt use in
vehicles with automatic systems was examined separately in another study (Streff and
Molnar, 1990), showing an increase in use for drivers and right flont seat passengers (over
70% use) over the state average (about 50%). However, it will be years before a majority of
vehicles have these advanced safety features since they are currently available only on newer
vehicles; and the seat belt is still the most important safety device.

The passive restraint systems attempt to increase safety belt use, while air bags are
an additional safety device designed to be used in conjunction with seat belts. Both
advancements confound the safety belt issue because they do not necessarily always
contribute to the correct or increased use of safety belts. Therefore, the inability to maximize
safety benefits, regardless of new advancements in safety devices, is mainly due to a lack of

seat belt use, making the issue of buckling up still relevant for a persuasive campaign study.

18
A campaign involving interpersonal (personal) communication and supportive media
messages, targeted at a particular group, aimed at peer encouragement for others to "Buckle

Up," provides a viable way to approach this safety issue.

W

In examining the historical development of existing research on the automotive safety
belt issue, it is usefirl to separate the findings into two categories: 1) research which came
before the mandatory laws requiring belt use and 2) the findings published since mandatory
laws have been in effect. The early studies, those done prior to the implementation of
mandatory laws, generally categorized persons as users or nonusers (F haner and Hane, 1973);
this being the most important distinction. Since the implementation of mandatory laws it has
become more useful to distinguish use across different groups and to examine use under
different conditions, such as differences in positions in the car, size of the vehicle, weather
and road conditions, time of day and the laws governing seat belt use. These changes in
categorization are partially the result of mandatory safety belt laws. The State of Michigan
enacted a mandatory seat belt use law in July of 1985 and continues to monitor seat belt use
(Streff and Molnar, 1990). Research studies conducted in Michigan since the mandatory law
was enacted consistently show the lowest percentage of use is in the young adult (ages 16-29
years) age group (Streff and Molnar, 1990).

Both pre- and post-implementation research has focused on some situational and
personal predictors of use. Documented situations which result in increased use are:
inclement weather, highway versus surface road driving (Wells, Williams, and Lund, 1990),

higher speed driving, driving longer distances (Fhaner and Hane, 1973), and driving compact

1 9
cars and/or newer cars (Lund, 1986). Safety belt use by drivers is also found to be positively
correlated with age (Wagenaar, Molnar, and Businski, 1988), i.e. older drivers are more
likely to use the safety belt. These finding indicate the need to target the young adult age
group for a seat belt campaign. The situational characteristics detail conditions that predict
use. This information can contribute to outlining a situational manipulation for a study that
is realistic and consistent with existing data. A persuasive situation focused on "Buckling
Up" in a situation where belt use is low (i.e. Friday night, city driving, young adult age
group) would not only be realistic but generate further insight into persuasive strategies for
buckling up in a situation when belt use is low. The situation is structured to be realistic so
that the explanation in the instructions that the other person is not buckled up is not
inconsistent with the circumstances being described. This also makes the request to buckle
up a logical request. In conjunction with situational factors which determine use, reported
reasons for use and nonuse have been examined and reported. The major reasons given for
nonuse are: forgetfulness, laziness, inconvenience, discomfort, fear of entrapment or wanting
to be thrown flee in an accident (F haner and Hane, 1973; 1974). In addition to these factors,
many drivers recognize the low accident risk per trip statistic (Slovic, F ishoff, and
Lichtenstein, 197 8) and therefore "putting on the seat-belt means recognizing the risk of an
accident" (Fhaner and Hane, 1973 p.34), thus lowering use. L00 (1985) also found
significant correlations between seat belt use and four fear of death and dying subscales that
indicated that seat belt use was related to greater fear of death and dying in one’s self and
others. The sample for this study was a young adult (18-24 years old) undergraduate group.
It appears that, consciously or subconsciously, one can avoid the recognition of this risk by

not using the seat belt. More recent data indicate that persons wish to exercise flee choice

20

about the safety belt use and the accident risk. Donohue (1988) concluded that flee choice
and effectiveness are important issues for usage in the state of Michigan and are actually a
deterrent to moving toward primary enforcement. Other data (Jonah and Dawson, 1982)
indicate that in regions where safety belt use is mandated by law, the attitude toward the law
becomes an important predictor of use. In a study in Canada with a similar student sample
(18-24 year old undergraduates), L00 (1985) found support for mandatory seat belt laws
before they were very wide spread. Females (78%) were more supportive of the laws then
were males (66%). Where the mandatory laws have been passed a pattern in usage has been
observed. Shortly after passage of the law there is an initial time of high compliance then
usage tapers off (Robertson, 1978; Fhaner and Hane, 1979; Wagenaar et al., 1988; Streff
and Molnar, 1990) indicating the need for campaigns that continue to encourage use.

Geller (1989) addresses the issue of use through existing organizations using a safety
triad application or ABC model. Geller (1989) posits that an activator (A) is needed to elicit
behavior (B) and a consequence (C) is also effective. Here the personal and environmental
conditions determine the behavior (or lack of behavior) which may be safe or unsafe.
Therefore, Geller's research has focused on "activators" such as reminders (Thyer, Geller,
Williams, and Pursell, 1987; Weinstein, Grubb, and Vautier, 1986), and incentives (Geller,
1983; Geller, Johnson, and Pelton, 1982; Ellman and Killebrew, 1978). Generally, these
campaigns have raised usage during the campaign, but usage tapers off when the "activators"
are removed. The limited success, short term change in behavior at relatively high costs in
personnel and incentives, suggests that a better match of strategies and target audience needs
to be explored.

In another study (Lund, Stuster, and Fleming, 1989) a special law enforcement and

21
publicity campaign was developed for Modesto, California to test techniques of primary
enforcement in a secondary enforcement state. Primary enforcement is where police can
pullover and ticket for not using a seat belt while driving; while in a secondary enforcement
state, police can ticket for noncompliance with the law but the car must be pulled over for
a violation other than the seat belt not being in use. In this study (Lund, Stuster, and
Fleming, 1989), a similar sized city in a separate media market was used as a control group.
With the special law enforcement and publicity campaign, belt use increased in the test area
among all subgroups but comparisons showed increases were lower for pickup trucks and
vans than cars and station wagons, on Friday nights than weekdays, among men compared
to women, among passengers compared to drivers, and among younger compared to older
occupants (Lund, Stuster, and Fleming, 1989). These finding support much of the previous
data and further indicate conditions of high versus low belt use and situational conditions that
can be used in structuring a realistic scenario regarding belt use. This study (Lund, Stuster,
and Fleming, 1989) also indicates how perceptions of primary enforcement and a supportive
mass media campaign can increase seat belt use. More recent work on safety belt use and
effectiveness has focused on identifying personality traits, such as sensation seeking, that
result in at-risk behaviors like drinking and driving (Gregersen and Berg, 1994; Pfefferbaum
and Wood, 1994; Vingilis and Stoduto, 1994). Other work has tested education campaigns
targeted at the at-risk behaviors (Garvin and Alcom, 1990; Kooler and Bruvold, 1992).
Some safety belt campaign work has focused on information dissemination (Barokas, 1995)
and multiple strategy campaigns (Streff and Molnar, 1991) to address the at-risk groups and
personality traits related to risky behaviors. See Donohue, Hildalgo, and Gathman (1995)

for a review of the literature. This paper proposes generating peer communication about

22
buckling up, that may change perceptions of use, as the strategy for increasing use. The
target area in this study is a secondary enforcement state where there is opposition to primary
enforcement.

In review, it seems to this researcher that 100% compliance with automotive safety
belt use, even in mandatory and primary enforcement settings, is unrealistic. There appears
to be a segment of the driving population that will not use automotive safety belts under any
circumstances. A realistic target in the secondary enforcement states in the United States
appears to be at the 70% to 80% compliance rate currently obtained in heavily enforced
mandatory jurisdiction areas. New Zealand, Australia, and Ontario, Canada (Jonah and
Grant, 1985) come near or exceed these rates. Overall, compliance patterns show a wide
variance flom around 80% (Jonah, Dawson, and Smith, 1982), in heavily enforced areas, to
50% (Streff and Molnar, 1990) in a secondary enforcement state, Michigan. Much of the
variance in use for different areas appears to be due to the level of enforcement, secondary
vs. primary. The secondary enforcement is generally a "laissez-faire", or "hands off",
approach by the state. A reluctance to change to primary enforcement eliminates the heavy
enforcement strategy that has been successful in raising belt use in New Zealand, Australia,
and Canada. As previously noted, use varies by age also. The compliance pattern observed
in Michigan provides further justification for exploring other persuasive approaches that may
be useful in encouraging persons to "Buckle Up" especially in secondary enforcement states.
Studying the persuasive strategies that can be used in a combination, interpersonal and mass
media campaign, approach extends this area of research.

Given the previous discussion of the obstacles to seat belt use and the effect of power

and intimacy on the elaboration, politeness, and directness of requests, the following

23
research hypotheses are offered as tests for a study of compliance gaining messages about
safety belt use. The initial requests should reflect the subject’s evaluation of the conditions
for obtaining compliance. Intimacy and situational power manipulations in the instructions
to the subjects should make the persuasive approaches predictable. In general, more intimate
persons should be more efficient in their communication requests because of increased
knowledge of the other (Miller and Steinberg, 1975) resulting in less elaborate requests.
Also, those low in power will employ strategies that are less direct, more explanatory or
elaborate, and more polite to supplant a lack of control of reinforcers enjoyed by those high
in situational power. Subjects are randomly assigned to one of four conditions (intimacy by
situational power, each high or low). Subjects are not aware of the specifics, or the number
of other conditions in the study. The following hypotheses are proposed for initial requests

to “Buckle Up.”

H1 Requestors in the low intimacy condition will use more elaborate

requests than persons in the high intimacy condition.
H2 Requestors high in situational power will use more direct requests
while those low in situational power will use more indirect requests.
H3 Requestors high in situational power will be less elaborate in their

requests than those low in situational power.

24

H4 Requestors high in situational power will be less polite in their requests than

those low in situational power.

After the initial request, subjects are told the other did not comply.

In addition to examining persuasive strategies, Roloff et a1. (1988) examined
responses to rejection for a compliance gaining request for borrowing a resource. The
rejection of a request by an intimate other was found to be very disconfirrning of the high
expectations for compliance generally associated with relational intimacy (Roloff et al.,
1988). Research also showed that rejection by an intimate resulted in greater feelings of
resentment for the requestor than if the rejection was by a non-intimate (Bar-Tal, Bar-Zohar,
Greenberg, and Hermon, 1977). Roloff et al. (1988) found no differences in response
elaboration due to intimacy but there was a main effect for flequency of counter persuasion.
Differences were found for politeness in the initial request and the response to rejection, such
that initial requests were perceived as more polite than the responses, for fliends (intimates)
but not for acquaintances (Roloff et al., 1988). These findings are limited to borrowing a
resource and do not include a situational focus on power differences because that was not
manipulated in the study. One obvious method for a person with high situational power to
address an initial rejection is to exert his or her power. While the use of power is perceived
as less polite, it may be the preferred response to operationalize whatever feelings of

rejection are generated by an intimate or acquaintance rejection of a request. Given the

25
greater feelings of resentment for intimates reported by Bar-T01 et a1. (1977), it might be
expected that those intimates in positions of power may exert their power in the follow up
request due to the increased resentment generated by the initial non-compliance. If power
is the control of reinforcers, then those with that control can be expected to more strongly
exert that influence after having the initial request rejected.

Non-compliance with an initial request may be a common response when the
requestor asks someone else to "Buckle Up". Therefore it is valuable to determine how the
requestor uses directness, politeness, and elaboration after the initial request and how these
perceptions are operationalized into a subsequent request. The previously cited data allows
for generating testable hypotheses regarding second requests. The following hypotheses are
offered for a test of situational power differences in second requests, after rejection,

compared to initial requests.

H 5 After rejection, requestors in positions of power will use a more direct

request compared to the initial request.

Conversely, those low in situational power might be expected to change or adjust the
request for compliance differently in a second request. In responding to rejection or non-
compliance with the initial request, without the option of high situational power, low power
individuals may further accent low power strategies along the dimensions of politeness and

elaboration.

26
H 6 After rejection, requestors low in situational power will use more

polite requests compared to the initial request for compliance.

In addition, intimacy has been shown to effect differences in elaboration, such that
non- intimates generally require more explanation in the requests. These effects are expected
to be accentuated in second requests as those low in power add explanations in an attempt

to gain compliance with the second request.

H7 After rejection, requestors in the low intimacy condition will use more

elaborate requests than requestors in the high intimacy condition.

Tests of these seven hypotheses for requests to use an automotive safety belt under
the detailed relational and power conditions will provide clear data on how the
communication variables of directness, politeness, and elaboration differ under the test
conditions. The assumptions are that power and relationship, manipulated in this study, are
key independent variables in the formation of persuasive communication on this important
safety issue. In order to firlly develop and structure a communication campaign designed to
increase seat belt use among the target group, additional data on proposed campaign
approaches through slogans and media vehicles are gathered. This will allow for conclusions
about how to apply this information to a media campaign aimed at encouraging users of seat

belts to remind non-users to "Buckle Up".

 

27

C "2'

Communication researchers have studied and evaluated many types of public
campaigns and offer some guidelines on the organization and components of a successful
campaign. Mendelsohn (1973) began laying the framework for successful campaigns by
developing the "limited effects" model of media effects and delineating the conditions
necessary for media influence. Historically, the "limited effects" model followed the "null
effect" (no effects of media) and "hypodermic needle" (all powerful media) theories of media
influence and effectiveness. The "limited effects" approach provides a much more realistic
perspective of what media can and cannot do to influence individuals. Since these early
approaches, communication scholars have broadened and delineated the application of
successful mass media campaigns to include information models (Atkin, 1981), political
models (Chaffee, 1981), and prevention campaigns (Maccoby and Solomon, 1981;
McNamara, Kurth, and Hansen, 1981). In a review of communication campaigns, Rogers
and Storey (1987) describe the importance of campaigns in bridging individual and societal
level objectives. The individual and mass media components observed and researched allow
for a multi-level approach to the study of communication behavior (Rogers and Storey,
1987). These multiple-levels are activated by numerous channels of communication, both
through the mass media and interpersonally (Rogers and Storey, 1987). Other work indicates
the need to have interpersonal support messages for mass media campaigns aimed at
changing individual behavior (Roloff, 1980). These findings by communication researchers
indicate key campaign components and support the mass media and interpersonal

communication idea for addressing the issue of activating safety belt usage more effectively.

28

In planning a persuasive campaign one should consider a person's self-awareness
regarding the subject. Roloff (1980) argues that persons self-awareness is extremely low for
many daily, ritualistic activities and certainly much lower than many researchers give
subjects credit for. A habitual activity, like a driving routine and seat belt use, takes little
cognitive activity until the habit is changed. Theoretically then, automotive safety belt use
can be categorized as requiring little cognitive activity, falling into the "mindless" (Langer,
1978) or "scrip " (Schank and Abelson, 1977) realm. Langer (1 97 8) offers suggestions on
how to approach changing these habitual or "mindless" activities. These strategies are
applicable for increasing compliance with the automotive safety belt law: 1) have a
campaign that increases the amount of interpersonal communication about the safety belt
issue, 2) increase objective self-awareness about the safety belt issue, and 3) relate
automotive safety belt use to another, more highly cognitive, issue. Similarly, Roloff(1980)
suggests increased interpersonal communication, objective self-awareness, and situational
differences in self-awareness as campaign methods for increasing the cognitive awareness
and thus potentially increasing success at changing behavior. Mass media messages can
increase awareness of the automotive safety belt issue both as it pertains to personal use and
asking others. These suggestions about campaigns can be operationalized in slogans about
seat belt use and asking others to buckle up.

A research plan under the advertising paradigm should provide valuable data on the
potential of a campaign to be successful at increasing automotive safety belt use in a
legislated, but not heavily enforced, environment. The strategy tested here includes both
mass media and interpersonal components, which should result in a change in behavior that

is not temporary but actually becomes integrated into the personal habits of the individual

29
(Flay and Cook, 1981). The evaluation of this research should clarify the potential of a
combined mass media and interpersonal communication campaign approach.

The current automotive safety law in Michigan has, in some ways, components of
both mass media and interpersonal communication. Mass media information flow because
enactment of the law and its stipulations requires dissemination of information.
Interpersonal information flow because the law involves others in the vehicle (children) for
which the driver is legally responsible. This makes interpersonal communication about law
abiding behavior likely. The law requiring that children under the age of 4 be appropriately
restrained no matter where they sit in the vehicle was enacted in 1982 and children 0-3 years
old are the highest user group at 78.4% (Streff and Molnar, 1990). Because of the inability
of the child to buckle him or herself up at this age, the act of "Buckling Up" often requires
parent/child communication, verbally or non-verbally, which in turn reinforces compliance
with the law. Personal observation indicates that young children who have developed a
pattern of seat belt use can remind parents to "Buckle Up". Mass media campaign messages,
or a change in existing public service messages, that stimulate law compliant communication
among persons in a vehicle may be a logical strategy to encourage safety belt use. Assessing
the obstacles a person perceives as existing in requesting that another person wear a safety
belt can provide guidelines for creative approaches. The research approach in this work
addresses the problems associated with changing behaviors regarding automobile seat belt
use by incorporating mass media and interpersonal communication components.

Communication researchers have been interested in the impact of mass media and
interpersonal communication on social change. A review of the literature on automotive

safety belts and public communication campaigns and persuasion suggests improvements in

30
safety belt campaigns as a viable strategy to increased belt use. This research will gather
pertinent data on the attitudes and behaviors regarding automotive safety belts by college
students, a subset of the lowest using group in the state, and will evaluate the persuasive
requests of others in the vehicle buckle up. A second part will test campaign slogans
utilizing different strategies aimed at increasing interpersonal communication about wearing
seat belts to nonusers. Generating an increase in interpersonal communication about
wearing seat belts is seen as a viable campaign strategy for promoting seat belt use among
a young adult sample. The data flom the survey responses will also indicate the media
choices the sample group rates as best for delivering messages encouraging seat belt use to
the target audience. Conclusions flom the study will include media and message choices that
should provide the most appealing campaign strategy to reach the target group with messages

that promote requests for non-users to buckle up.

51"]:

The synthesis of mass media and interpersonal communication strategies offered by
this paper tests theoretical and practical applications. Research questions on theme
evaluations and methods of delivering the information to the target audience extends this
study by exploring ways to successfully implement a campaign. These data provide an
indication of how to successfully apply the suggestions in the study and also some indication
of how successful this application may be. Four slogans that may be used to prompt the
suggested communication about buckling up the automotive safety belt have been developed

for this research. Multiple slogans were chosen because the study is not testing a specific

31
slogan as a theoretical variable. Additionally, the literature on seat belts suggests slogan
applications that might be useful at initiating the behavior change desired.

One purpose of the study is to compare the four slogans chosen to see if there is a
preference by the target audience that suggests differences in effectiveness for the slogan.
Two of the slogans use the phrasing “Remind Others to Buckle Up” as the encouragement
to request use of the safety belt. The other two slogan use the phrase, “Make Sure Everyone
is Buckled Up” as the wording to encourage discussion about belt use. Each slogan then has
a different motivational link or practical theme. One message is a reminder of the safety belt
law: Make Sure Everyone is Buckled Up - It's the Law. Free choice has previously been
identified as an important issue in Michigan (Donohue, 198 8) so the reminders about the law
are not expected to be evaluated favorably; however it is a viable theme for testing given the
results of the study in California (Lund, Stuster, and Fleming, 1989) and the high general
knowledge about the law. The existing data about moving the law to one of primary
enforcement suggests the Law slogan message will be viewed negatively in evaluations and
comparison to other messages. The other messages take different approaches to behavior
change suggested in the literature review (other directed, habitual nature, and link seat belt
use to other healthy behaviors).

The second slogan focuses on an altruistic concern for others: Show You Care -
Remind Others to Buckle Up. Previous research on requests flom fliends or acquaintances
have focused on borrowing a resource or asking a favor (Roloff and Janiszewski, 1989); this
type of altruistic concern for others is in direct contrast to this previous work. Some
compliance gaining research (Hunter and Boster, 1979; Williams and Boster, 1981; Boster

and Stiff, 1984) has concluded that requestors will make more of an effort at gaining

32
compliance, and will perceive more compliance messages as acceptable, when the benefit
to the listener is high. These findings make a message focused on caring for the other a
viable slogan to test. This slogan is expected to be neutral compared to the Law slogan.

The third slogan addresses the habitual nature of belt use and encourages raising
awareness to change the habit: Share a Good Habit - Remind Others to Buckle Up. The
message strategy is raise cognitive awareness of the habitual nature of safety belt use to enact
the behavioral change of reminding others or buckling up if you are a not using the safety
belt. This addresses some of the suggestions made about changing behaviors in routine, or
mindless, activities (Langer, 1978; Roloff, 1980). This slogan is expected to be evaluated
positively. Survey data shows most people view the use of seat belts favorably and the term
“Good Habit” should have positive connotations.

Finally, the fourth slogan is aimed at promoting safety belt use as a healthy behavior:
Protect Good Health - Make Sure Everyone is Buckled Up. This may have the benefit of
moving the automotive safety belt issue away flom just a safety issue and into the broader
spectrum of general health. The potential is to take advantage of the overall focus on
personal health many Americans have and that is prominent in other media messages (i.e.
athletic shoes, clothes, and equipment, as well as dieting and eating right). This slogan is
also expected to be evaluated positively due to the reference of the broader issue of good
health. The literature on persuasion and safety belt use suggests that concern for others,
good habits, and healthy behaviors may all be good themes for promoting safety belt use.
Data for evaluation and comparison purposes can provide direction on which of these slogans
may be the most effective under certain conditions (i.e. driver, passenger, friend, or

acquaintance). Evaluations of the themes and responses regarding effective media delivery

33
of the messages provide data on how to construct a successful campaign following the
guidelines suggested in this paper.

Each of the four campaign slogans will be evaluated on a seven point semantic
differential scale with seven bi-polar adjectives as the anchors. The anchor pairs include:
Easy to remember - Not easy to remember, Dislike - Like, Clear - Unclear, Negative -
Positive, Self-directed - Other-directed, Informative - Uninforrnative, and Effective -
Ineffective. A factor analysis will test if the evaluations factor into two components of
advertising messages: a motivational component (other directed, like, positive, and
effective) and a clarity component (easy to remember, informational, and clear). These
factors will be used to evaluate the slogans because they are indicators of the slogan’s
potential effectiveness (i.e. clear, positive, informational, effective, easy to remember, ...)
as a message that can lead to behavior change. Testable hypotheses about the slogans are
offered.

Consistent with the information that indicates the law is well known and is viewed
by most people as a good law, the slogan reminding subjects about the law should be the
highest rated on the clarity component (easy to remember, informational, clear). Because it
is well know and the other themes are not, the law message should stand out for its clarity
when compared to the other messages. However, the law approach should be perceived
much differently as a motivator and should be rated low on the motivational components,
based on Donohue’s (1988) findings about flee choice. Using the law as an advertising
appeal to request others to buckle up is not expected to be highly motivational, especially
when compared with other strategies. The state of Michigan, and other jurisdictions, have

purposely remained secondary enforcement domains due to the flee choice issue and any

34
expected backlash flom movement to primary enforcement. For all these reasons, the law
approach to reminders about buckling up should be perceived differently on the factors of

motivation and clarity. This leads to the following hypotheses.

H8 The campaign slogan reminding subjects about the law will be the rated

the highest on the clarity components.

H9 The campaign slogan reminding subjects about the law will be the rated

the lowest on motivational components.

No testable hypotheses are offered about the other slogans. Research evaluation of
these slogans is exploratory; however, they offer viable, realistic approaches for campaign
slogans. Tests of the data will provide information on how the other slogans are received
compared to the Law slogan. Statistical comparisons of all four slogans on the two factor
criteria will determine which may be the most effective for use in a broad campaign
approach, where the data indicate specific situations favor a particular thematic approach,
these will be explained. The data will provide an indication of the strategy needed for

increasing use of automotive safety belts among the target age group.

W

Each individual participating in the study will complete a questionnaire that provides

personal background information regarding the automotive safety belt issue under

35

investigation in this study. These data serve two purposes. One is to validate that the sample
is representative of the college age target group with regard to their attitudes and behaviors
about automotive safety belt use. Having said this, the group should be licensed drivers who
are aware of the law but not in favor of stricter enforcement (secondary vs. primary
enforcement), and aware that seat belts are effective safety devices. A second purpose for
the background information is to indicate there is currently little communication about the
seat belt issue within the target group, but there may be a willingness to communicate about
buckling up (i.e. to determine if there is much potential for the proposed approach).
Wagenaar and Webster (1986) report that young children have high rates of seat belt use
because of the mandatory child restraint legislation and exert an upward influence on overall
use rates with other groups except for the young adult target group in this study. While no
rationale is provided for this influence, it might be inferred that the child serves as example
to the adult. Many young children who can talk can also be expected to remind adults in the
vehicle to buckle up, if they observe adults who are not buckled. This example points to the
potential of an approach that involves communication flom others in the vehicle and the
expectation that it can work with peer influence.

It is expected that subjects will report that a minimal amount of the communication
about other wearing seat belts is currently taking place. Low levels of communication
suggest that this is perhaps an untapped method to promote belt use, with a supportive
advertising campaign. While the situational characteristics and power variables manipulated
in this study differ greatly flom communication and relationships of parents with children,
where some form of awareness or verbal or non-verbal communication has helped increase

seat belt use (Wagenaar and Webster, 1986), the data gathered here should confirm the

36
potential of this approach.

The self report data allows for categorizing subjects into groups. Analysis will be
conducted by segmenting the groups into those who indicate a high willingness to ask others
to buckle-up and those who do not. These existing attitudes might be expected to influence
subjects self reports in the study on affective components such as how they feel about the
sense of obligation and likelihood of compliance generated by their request.

The sense of obligation can be viewed in two different ways. One is the self report
of the sense of obligation the requestor thinks is generated by the request. The second is the
sense of obligation to comply with the request flom the perspective of the person who is
asked. Previous research (Roloff et al., 1988) indicates obligations increase with levels of
intimacy. This relationship is so strong that, in some instances, intimates expect the other
to anticipate needs before requests are made. For these reasons, requestors will oflen seek
out particular others when making requests based on the level of intimacy. This sense of
obligation has been examined in requests for resources and may be applicable in subsequent
research on safety issues. However, in this study it is important to get some baseline data on
the sense of obligation, with self report measures, to determine how requestors view this
component of the request. Similarly, baseline measures of the likelihood of compliance
provide an estimate of the requestors perception of the response to his/her request. These
ratings are expected to vary based on existing attitudes about likelihood of asking. Those
who report a higher likelihood of asking may feel that their message is stronger in generating

the feelings of obligation and likelihood of compliance.

37
H1 0 Subjects who report they are more likely to ask someone to buckle up, will rate

their requests generate a greater sense of obligation and a higher likelihood of compliance

in the other .

These hypotheses reflect a thorough test of the automotive safety belt problem as
outlined and defined. Primarily, the study will test and evaluate persuasive requests to put
on the safety belt based on predictions about power and relationship differences and
predictions based on rejection of the initial request. Also, the data will test campaign
slogan applications and define the conditions under which these might be the most
successful. This includes existing attitudes and behaviors as well as the best slogan and
media delivery methods. As previously noted, almost any success from the approach
suggested here at increasing seat belt use with drivers and front-seat passengers, without
changing the existing secondary enforcement law, should also be applicable to rear-seat
passengers. The benefits to individuals and society in increased automobile travel safety are

potentially great.

Method

This study is a 2 X 2 independent groups design. Two levels of the variable intimacy
(friend or acquaintance/high or low) were crossed with two levels of the variable situational
power (driver or passenger/high or low). An initial request was acquired flom the subjects
and they were told the other had not complied and they were to ask one more time. This
allows for tests of initial and follow up request differences (time one and time two measures).
Coders were trained to rate the dependent variables politeness and directness of the requests
to buckle up on scales ranging flom one to five (high to low). The elaboration of each
request was also coded by counting the number of words in each request. Subject ratings of
the sense of obligation and likelihood of compliance by the other are tested using self reports
of feelings of influence and likelihood of asking another to buckle up as covariates. Finally,
the slogan or campaign messages are compared on responses to the items that form the clarity
and motivational factors used to evaluate message effectiveness and applicability to the
persuasive approach outlined in this research. This segment tests a reminder about the law
and three other approaches suggested in the literature. These messages were scored on seven

individual items that are hypothesized to form the clarity and motivational factors.

38

39

”.1.

The variables intimacy and situational power were manipulated in the
instructions. The high intimacy condition used the word fliend in the following example.
The low intimacy condition used the work acquaintance. The following details the
manipulation for situational power, where driver is the high power condition and
passenger the low power condition. The text for the passenger (low power) situation is as

follows:

Imagine that you are enrolled in a Humanities course that counts toward your
general studies requirement. One way to pick-up extra credit for the course

is to view, and write a short critique, of a foreign film shown at a theater
across town. The instructor encourages viewing and discussing the film with

a classmate as a way to enhance understanding and generate ideas. With this

in mind, you make arrangements with a same sex acquaintance/friend flom

class to see the film.

The day of the film arrives and your acquaintance/friend arrives to pick you up.
You get in the flont seat and buckle your seat belt. Your
acquaintance/friend says “Let’s go!” but he/she does not have on a seat belt.

At this time you are going to request that he/she “Buckle Up!”. In the space
below, write the exact words you would say to request that your

acquaintance/friend put on the seat belt.

40
The manipulation for the driver condition had the same first paragraph and the

following second paragraph.

The day of the film arrives and you drive over to pick-up your acquaintance/friend.
You are buckled in and your classmate gets in and says “Let’s go!” but does
not put on his/her seat belt. At this time you are going to request that he/she
“Buckle Up!”. In the space below, write the exact words you would say to

request that your acquaintance/friend put on the seat belt.

After subjects wrote down their initial request to buckle up, they were asked to
evaluate the sense of obligation and likelihood of compliance generated by their request.
Following these evaluations, the manipulation for rejection of the initial request was inserted.
The rejection message was purposely left without a particular reason so that the subjects
would focus on the obstacle of persuading the other to buckle up, and how they perceived

it as different than the initial request. Instructions to the subject were as follows:

For whatever reason your acquaintance/friend did not put on his/her seat belt
when you asked. You are willing to try again to persuade him/her to “Buckle
Up!”. In the space below, write the exact words you would say now to

request that your acquaintance/friend put on the seat belt.

No manipulation check was built into the study. The bold face emphasis on the key

words, and their repetition, is expected to be enough to ensure an adequate manipulation.

41

Subjects were not aware of the number or the specifics of the other conditions in the study.

511182915

Subjects were 350 students flom an introductory communication course at a large
Midwestern university. They were all of legal driving age in the state and participated in the
study willingly. Participants received extra credit for participation, at the discretion of the
individual instructor, and were randomly assigned to treatments. None of these subjects had
participated in a pre-test of situational descriptions explained in the next section. Neither had

these subjects been previously exposed to the research materials.

The instructor allocated 15 minutes at the end of class for the experiment. A research
assistant was introduced and instructed the participants. A brief introduction to the research
was provided by the research assistant before the materials were delivered. Each participant
signed a departmental research consent form and then was handed a packet containing the
study materials. The research consent forms were provided to the instructor for assigning
participation credit. Each packet of materials contained: a questionnaire of demographic and
background information on attitudes toward automotive seat belt use, one of the four
scenario descriptions (e. g. driver with acquaintance) with the appropriate response forms, and
all four campaign slogans and their evaluation forms sorted randomly. Subjects were

randomly assigned to conditions. Anonymity of response was maintained. Subjects were

42
allocated 15 minutes to complete the research packet. Upon completion of the research

packet, subjects were excused.

The questionnaire package contained three sections. Section order was maintained
for each version of the package; however, random ordering of materials occurred within a
section, if necessary. Section one contained questions regarding background information and
attitudes and behaviors toward automotive seat belt use. (See Appendix B for the
questionnaire.) These answers serve two purposes in the analysis: 1) to verify that the
sample is representative of the larger population on some demographic and attitudinal
variables, and 2) to measure current attitudes toward seat belts and the likelihood of asking
others to buckle them, to utilize as covariates in the analysis. The second section of the
questionnaire packet contained the scenario (i.e. driver or passenger by fliend or
acquaintance) and the compliance gaining request form for writing in the persuasive request
in the situation. The scenario description was pre-tested on a class of introductory
communication students (n=21). Rated on 7 point semantic differential scales (flom clear
=1 to unclear = 7), the description of the situation was rated as clear (M = 2.5, SD = 1.2),
and (realistic =1 to unrealistic = 7) realistic (M = 2.9, SD = 1.7). The pre-test participants
also indicated an interest (interested = 1 to uninterested = 7; M = 3.0, SD = 1.7) and a high
likelihood (likely =1 to unlikely =7; M = 2.8, SD = 2.0) of participating in the extra credit
situation described. For the study, each participant received one of the four situations

manipulated in this section (driver or passenger by friend or acquaintance). The situations

 

 

43
were randomly distributed among participants. After the scenario description and the subject
writing in an initial request, the questionnaire indicated the receiver of the message had not
complied with the request and subjects were instructed to write a second request. This
completes the second section of the questionnaire.

The third section of the questionnaire package contained two parts. Part one
contained the four separate pages with campaign slogans at the top of each page in oversized
type, randomly rotated, with 7-point semantic differential scales for rating. The verbal
anchors for the scales were easy to remember/not easy to remember, like/dislike,
clear/unclear, informative/uninformative, negative/positive, other-directed/self—directed, and
effective/ineffective for slogan evaluation. The polarity of some adjective pairs were
reversed. This focuses the subject’s attention to each pair and helps avoid the problem of
systematic response to all items. Part two of this section consisted of a single page asking
the participants to comparatively rate different media for delivering messages about seat belt
use to a target audience of their peers. This section provides data about campaign slogans

and approaches that may help generate interpersonal communication about seat belt use.

Qadinzhasedm

Seat belt use requests were examined and coded using a five-category scheme for
politeness and directness based on the findings of Kemper and Thiessen (1981) described
earlier. Two coders were trained to identify the level of politeness and directness in each
request. In addition, an elaboration variable was calculated by counting the number of words

in each request.

44

Coders were trained to rate politeness on a 1-5 scale flom the most polite request of
"Please + imperative" to the least polite request of the Imperative only. Coders were trained
to recognize other introductory words and phrases, and their variations, as indicators of
politeness (see Appendix C for examples and category differentiation) so they could
adequately classify the request on the politeness scale.

Coders were also trained to rate directness on a scale of 1-5 flom the most direct, an
lrnperative only, to the least direct, "I think + need assertion". In general, requests including
an imperative are more direct than those including need assertions. Coders were trained to
recognize other introductory words and phrases, and their variations, as indicators of
directness (see Appendix C for examples and category differentiation) so they could
adequately classify the request on the directness scale.

It was expected that some subjects would provide rationale or examples to persuade
others to "Buckle Up" and these would show up in more elaborate requests. Occasionally
these examples were provided without restating the original request to wear the seat belt.
Coders recognized a portion (11%) of the responses in the second request were so indirect
as to not mention the request to buckle up again. After reviewing the data, a sixth category
was added for directness in the second request only. The sixth category, other request, was
added to indicate the subject had continued the request with a supporting story or example
without restating the need to "Buckle Up". All previously coded date were reviewed and, if
necessary, recoded after this additional category was added. The request took the form of
comments as an extension of the previous request. Some were stories about personal
experiences with seat belts saving the requestor or someone they know, these were very

serious. However, some stories or examples were very humorous in their persuasive appeal.

45
Overall, this category is viewed and coded as the least direct because the subject did not
specifically restate the request to buckle up.

Elaboration was simply coded as the number of words in each subject's request. This
measure is appropriate for the situation based on the hypotheses generated by the situational
considerations and the common nature of the seat belt issue. Counting words allows for
expanded variance, precise measurement and the ability to detect changes in requests
suggested by the differences in the situation or the relationship. Larger units of measure
(phrases, clauses, ...) would not provide the necessary precision for this study. Since the
requests could be very short (Buckle Up!) or very elaborate, if subjects provided supporting
stories or examples, counting words was deemed the best way to measure elaboration.

Applying these coding schemes, coders reached a global agreement rate, calculated
by Cohen’s Kappa, of .69 for politeness and directness; based on a sample of 220 statements
out of the universe of 700. More precisely, reliability tests for politeness were Cohen’s
Kappa = .62 for the initial request and Cohen’s Kappa = .76, for the request after the refusal.
Reliability tests for directness were Cohen’s Kappa = .66, for the initial request and Cohen’s
Kappa = .73, for the request after the refusal. Elaboration reliability scores were higher at
Cohen’s Kappa = .86 for all elaboration. For the initial request the reliability is Cohen’s
Kappa = .89 and Cohen’s Kappa = .84 for the request after the refusal. Some discrepancies
in elaboration existed due to the use of the term seat belt. Coders were instructed to count
the term as one or two words depending on how the subject wrote the term. Even with these
instructions the coders did not totally agree because of difficulty reading the handwriting or
determining if the words were connected. Overall, these reliability levels are viewed as

satisfactory for this study.

Results

A total of 350 subjects participated in the research. Characteristics of the sample
population are provided in Table A-1 (Appendix A). The age range of participants is flom
17 to 26, with 93% flom 18 to 21 years old. The gender distribution is 53% female and 47%
male. Other important demographic characteristics show: 97% are licensed drivers, 90%
are licensed in the state of Michigan, and 99% of the drivers have licenses flom within the -
United States. Other information about the subjects was obtained in the initial section of the
questionnaire. A majority of subjects "very much" or "somewhat" favor the safety belt law
(87%). In addition, 64% think the law should be enforced more strictly than it is; while 29%
said it should be enforced as it is now. On another item, 92% answered "strongly agree" or
"agree" to the item "seat belts significantly reduce the severity of injuries in accidents.”
These responses parallel earlier findings that show general support for seat belt laws and
nearly universal recognition of the benefits of use, even if overall use in less than 100%.

Several other items assessed the reasons for wearing seat belts among the subject
group. Responses of "very important" or "important" were given for: feeling safer (86%),
worry about an accident (81%), used to buckling up (70%), and the law requires use (58%).
Only 31% showed that "others expect me" to buckle up as an important reason. These data
are consistent with the law not being named as a highly significant factor and the low

recognition of use by others with this age group. These responses support the rationale of

46

47
this study to examine peer pressure as a strategy for increasing seat belt use. This researcher
believes in the potential of peer pressure as a viable and under utilized strategy to increase
use with the target audience. All indications are that the target group is more likely to
respond to peer pressure than to a reminder message about the law.

Subjects provided an indication of what they think about the issue of asking others
by answers to questions about whether other drivers and passengers want to be reminded to
buckle up. Asking about other drivers wanting to be reminded, 41% said "most" or "some"
unbuckled drivers want to be reminded. Similarly, 45% reported "most" or "some"
unbuckled passengers want to be reminded. About themselves, 64% answered "always" and
"usually" to the item about wanting someone to remind you if you are unbuckled. This
response may be a little inflated due to the nature of the previous items (questions about seat
belts and the law) and the subjects providing the "socially acceptable response"; however,
they are higher than the 31% that said they wear seat belts because “others expect me” to
buckle up. This gap (33%) between the “want” to be reminded and what others currently
expect suggests the potential of this approach because it shows that peer expectation or peer
example is an underutilized reason for buckling up. The desire to be reminded is not being
met by current practices of others for reminding unbuckled peers to wear a seat belt. There
is also a gap (10% or more) between the perception that other drivers (41%) and passengers
(45%) want to be reminded and current response of the importance of wearing a seat belt
because others expect me to (31%). While the response scales are not directly comparable,
there appears to be a response gap between the current behavior of asking others to buckle
up and the perception that others want to be reminded.

The study is directed to be generalizable to the 16 to 29-year-old population that uses

48
seat belts less than any other group in Michigan. The sample provides some generalizability
since it represents, age wise, a large portion of the low using target group. The education
level of this sample limits generalizability because it is a group more educated than the
general population and education is positively correlated with seat belt use (Blomquist, 1979;
Leigh, 1990). However, the study is successful at gathering data about automotive seat belt
use and potential campaigns flom a large sample of young adult, licensed drivers, in the state
of Michigan. On the key issue of the current behavior of asking others to buckle up, the study
sample may not differ greatly flom the target population of young adults; both are probably
low in asking others. Therefore conclusions about persuasive approaches, the willingness
to ask peers to use seat belts, and some of the results of campaign slogan evaluations should

be generalizable to the target group of young adults.

Independentlatiahlas

Situational power (driver or passenger) and relationship (friend or acquaintance) are
manipulated in the instructions and randomly distributed among subjects. Hypothesis ten
posits differences in likelihood of compliance based on reported attitudes about asking
others to buckle up. Several items in the initial section of the questionnaire addressed these
attitudes. A total of 53% of the subjects indicated they "always" or "often" ask unbuckled
passengers to buckle up when driving. This variable will be split into dichotomous ask or
do not ask. On asking unbuckled drivers, 35% reported they "always" or "often" ask. Again
this variable splits the sample into those who ask and those who do not ask.

Responses to these items on driver and passenger behaviors support the expected

49
differences in communication based on the situational manipulations in the study. Subjects
in this study reported that passengers are much less likely to ask drivers to wear their seat
belts then drivers are to ask passengers (18% difference). These reported differences in
requesting safety belt use by passengers to drivers compared to drivers to passengers may
largely reflect power differences based on the situation in the vehicle. Follow up items
asked about the perceived influence of drivers over passengers’ safety and passengers over
drivers’ safety. Clear differences exist as 80% of subjects report drivers have "a lot" or
"some" influence over passenger safety, compared with passengers 30% report of influence
over driver safety. While this difference may include other thoughts regarding safety than
just the seat belt, it points out a difference in driver and passenger power that suggests
different situations exist and therefore different obstacles to persuasion are present. These
responses support the driver/passenger manipulation and the expected differences in
directness, politeness, and elaboration of requests to persuasion and the change between the

initial and second requests.

Slcgaaflxaluaticnficales

Seven semantic differential items were used to evaluate the suggested seat belt
campaign slogans. The verbal anchors of the paired items were: easy to remember/not easy
to remember, like/dislike, clear/unclear, inforrnative/mrinformative, negative/positive, other
directed/self directed, and effective/ ineffective. The researcher posited that the items will
factor into two components of advertising messages: a motivational component and an

clarity component.

50

A confirmatory factor analysis verifies a two-factor solution; however, the items did
not load on the factors as was hypothesized, based on the factor criteria established. The
criteria for the factor loadings were that items needed to load greater than .60 on a single
factor and less than .50 on other factors. The first factor consisted of five items: easy to
remember, like, clear, informative, and effective, with an Alpha reliability of .80 (see
Appendix A, Table A-2B). This factor is interpreted as a message potency scale for this
research because it contains three clarity factor items (easy to remember, informative, and
clear) as well as two other items (effective and like) that were hypothesized to load on the
motivational factor. The factor is labeled Potency because it contains advertising and
motivational items that suggest a potent message (informative, clear, as well as liked, easy
to remember, and effective). The mixture of these items point to a useful campaign message
that will be influential with the target audience. The other two items, self or other directed
and positive, did not load within the criteria established or with each other. The item
positive was dropped flom the initial analysis; while the self or other-directed item is
analyzed as a single item because of its importance to the study. It is floated as a measure
of other- directedness (higher scores = other directed). These two factors explain 60% of the

variance.

I l 1' CI .
To provide a complete campaign overview and recommendation, subjects were asked
about which media should be used to promote reminding others to “Buckle Up.” Subjects

gave primary and secondary media choices. They were asked to think about the best way to

51
reach people like themselves with a message about encouraging others to wear their seat
belts. Television was the overwhehning primary choice (56%), followed by radio (14%),
billboards (13%), and in car stickers (11%). Radio was the most popular secondary choice
(32%), followed by billboards (23%), television (14%), bumper stickers (11%), and in car
stickers (10%). This information provides suggestions on the right media, flom the target
audience, to deliver this message about automotive seat belt use. Depending on the
conditions of a campaign (financial, available media channels, etc.), the appropriate media
can be chosen flom the options suggested by this study sample as the best media to
disseminate messages about seat belt use to the target group. The appropriate media choice

and a targeted and tested message offer the best combination for a successful campaign.

Hypothesis one predicts requestors in the acquaintance condition will use more
elaborate requests than those in the fliend condition. These differences in the elaboration of
the request are based on expected differences due to the manipulation of the relationship
(intimacy). Tests of the means for elaboration of the initial request in the friend (M=13.01,
SD=8.72) and acquaintance (M=l3.90, SD=8.28) conditions did not differ significantly as
posited and provide no support for the research hypotheses regarding elaboration differences
due to relationship. These results indicate no main effect for relationship; acquaintances did
not elaborate more than fliends in their persuasive requests.

Hypotheses two, three, and four predict subjects in the higher power (driver)

condition will be more direct, but less polite and less elaborate, in their requests than subjects

52
in the low power (passenger) condition. Tests of the data show no difference in the ratings
of directness, politeness, or in elaboration of the request in the high versus low power
condition. The results for directness and politeness are detailed in Tables 1 and 2.
Elaboration by drivers (M=12.92, SD=7.89) was lower, but not significantly lower than

passengers (M=l4.03, SD=9.23) (F=2.45, df=342, p=.12).

 

 

TABLE 1
Means and Standard Deviations for Directness
by Power and Relationship
Friend Acquaintance
n = 187 n = 157
Driver 1.59 1.73 1.66
n = 189 (.70) (.75) (.71)
Passenger 1 .84 l .48 1 .70
n = 155 (1.00) (.73) (.92)
1.72 1.64
(.84) (.74)

 

NOTE: No significant main effects. An interaction effect was significant at the
p< .05 level. Standard deviations are in parentheses.
(coding flom most direct =1 to least direct =5)

Though no interaction effect for power and relationship on directness is
hypothesized, an AN OVA tested for the interaction. The ANOVA yielded a significant
interaction (F = 8.15, df=l , p < .01) for power and relationship. The data indicate subjects

in the passenger and acquaintance condition were the most direct overall. The friend and

53
driver condition was the next most direct. The passenger and fliend condition generated the

least direct requests. The pattern is that passengers were more direct with acquaintances then

with fliends. The opposite effect is found for drivers who were more direct with fiiends than

 

 

with acquaintances.
TABLE 2
Means and Standard Deviations for Politeness
by Power and Relationship

Friend Acquaintance

n = 187 n = 157
Driver 2.83 2.58 2.70
n = 189 (1.84) (1.83) (1.84)
Passenger 2.84 3.10 2.94
n=155 (1.61) (1.78) (1.68)

2.83 2.78

(1.72) (1.74)

 

NOTE: No significant differences at the p< .05 level. Standard deviations are in
parentheses (coding flom most polite =1 to least polite =5).

Though no interaction effect for power and relationship on politeness is hypothesized,
an AN OVA tested for the interaction. The AN OVA showed no significant interaction for
power and relationship on the politeness of requests to buckle up.

Additional analyses probed for support of the assumptions among subgroups of the

sample. Male and female subject data were separated into groups and AN OVA tests were

54

run for the dependent variables of politeness, directness, and elaboration in the initial
requests. No significant findings emerged flom male subjects for any of the dependent
variables. However, female subjects showed a pattern of means similar to the overall
findings but slightly more pronounced. The sample contains slightly more female (53%) than
male (47%) subjects. The study was constructed to focus on the manipulations of power and
relationship and not any particular gender differences. The instructions were carefully
written to focus on a same sex fiiend in the situation because of the expected complications
in power and relationship mixed gender pairs would involve. However, these data show that
the some gender differences in persuasive strategies. The differences may also be due to a
weak low power manipulation in this study. That would explain why the findings for
directness are opposite what was hypothesized in Hypothesis Two. Similarly, the findings
for politeness are opposite what was hypothesized in Hypothesis Three. It also possible that
the findings are due to chance since they where not in the hypothesized direction. No
differences emerged for elaboration of the requests.

The fifth hypothesis predicts a more direct request, than the initial request, will be
used by drivers after they are told the passenger is not buckled up following their initial
request. A paired sample t-test shows no significant difference in the directness flom the
initial (M=1.66, SD=.71) to the follow up (M=1.78, SD=1.48) request (t=-.100, df=184,
p>.10), indicating there was no change in the directness of the request due to the failure to
comply initially.

Hypothesis six predicts more polite requests will be used by passengers (low
power) after they are told the driver is not buckled up following their initial request. A paired

sample t-test shows an effect opposite what is hypothesized occurred. Differences in

55

politeness flom the initial request to the second request, for those in the passenger condition,
moved flom more to less polite. Results of the test for 147 pairs were M=2.91, SD=1.68 in
the initial request and M=3.91, SD=1.69 (F596, df=146, p<.000)for the second request.
Additional paired t-tests show second requests were more elaborate (M=15.87, SD=9.08)
compared with the initial requests (M=14.24, SD=9.27; t=-l .75, df=146, p<.05), and that the
requests moved flom more direct initially (M=1.70, SD=.91) to less direct (M=2.00, SD=1.7)
for the second request (t=-1.98, dfi146, p<.03). These data show the strategy by passengers
in the second request to be more elaborate but less polite and less direct in their request to
get the driver to buckle up. Parts of this approach, increased elaboration and more indirect
requests, to persuasion are consistent with other findings for a low power strategy (Kemper
and Thissen, 1981).

The differences found in this study for politeness and directness support measurement
as separate items. Additional evidence for the separate measurement of politeness and
directness is shown in the correlation data. The correlation between politeness and directness
in the initial request is r= -.52, p=.00, a strong negative correlation. In the second requests,
no significant correlation exists between politeness and directness (r= -.08, p>.10). This
supports continued measurement of politeness and directness as separate items.

Hypothesis seven predicts that acquaintances (low intimacy) will use more
elaboration in second requests compared to fliends (high intimacy). A paired t-test found no
difference in the elaboration of second requests between subjects in the fliend (M=15.30,
SD=8.98) versus acquaintance (M=15.02, SD=8.82) condition (df=338, p>.05). An
additional t-test was run where the most indirect cases (new directness category six) were

removed; still no difference exists in elaboration between the first and second requests for

56
the relational intimacy condition. Though coders recognized a qualitative difference in 11%
of the cases, no statistically significant difference emerged from the analyses whether these
cases were included or not.

Additional analyses for the male and female groups provides insight to the lack of
overall findings and show some differences based on gender of the subject. The overall
finding that passengers (low power condition) were less polite in second requests, compared
to the initial request, was true for both male and female subjects. This was opposite what
was predicted (Hypothesis six). Other findings examining differences due to the power
manipulation showed that females used more polite requests (M=2.30, SD=1.77) when
higher in power (driver) than they did in the low power (M=2.99, SD=1.70), passenger,
condition (t=2.69, df=184, p<.01). Overall no differences were found for elaboration and
directness, when power was manipulated; but upon additional analyses, female subjects
were more elaborate and less direct in second requests compared to the initial requests. Male
subjects showed no differences in elaboration and directness flom the first to second request
in the analysis of high versus low power. Similar results were found in tests of differences
due to the relationship manipulation. Female subjects were more elaborate and less polite
in second requests compared to the first Male subjects showed no differences and the
overall findings were not significant. After the initial request, female subjects provided more
elaboration in the second request. This is consistent with predictions about low intimacy and
low power increasing elaboration.

Hypotheses eight and nine predict differences in the evaluations of campaign slogans
on clarity and motivational components constructed flom the previously described factor

analysis of item evaluations of the slogans. However, the factor structure resulting flom the

57

data is different flom what was predicted. The resulting factor structure, with an Potency
factor and a Self or Other-directed item, has previously been detailed in the section on
Independent Variables. However, while this factor can be labeled and constructed flom the
data, it remains uninterpretable because it contains a mixture of items hypothesized to load
differently. The disconfirmation of the clarity and motivational factors expected to be
components of advertising messages makes the resulting factor unclear in relationship to the
study. Tests of Potency factor are conducted and reported. Hypothesis eight and nine cannot
be tested as hypothesized and are only referenced in relation to the findings of the described
factor and item analysis. The tests of the data are still reported though interpretation of the
results is difficult. Examination of Potency factor and individual items rated in the study
provide relevant data for examination.

Tests of the mean ratings of the Potency factor for the four slogans shows a
significant difference in the ratings (see Table 3, higher ratings are more postive evaluations).
Subject ratings showed a clear preference for the slogan, “Protect Good Health - Make Sure
Everyone is Buckled Up.” The “Share a Good Habit” slogan was the next most highly rated.
The Law slogan, hypothesized to rate the highest on the expected advertising factor, was
rated the lowest on the resulting Potency factor. While the law is well know, focusing on the
law in a campaign message is not highly rated on the Potency factor that emerged as
important in this study. It appears that linking the message to good health or habits, or the
originality of these messages, is stronger than referring to the law. The data suggest that
citing or referring to the law as a strategy for increasing seat belt use is not an approach that
is likely to be effective, as an overall campaign message, with the target audience. Given that

the data show higher rated alternatives, the messages about the law would be least effective

58
overall. There may be some conditions where these responses may differ. Additional
statistical tests, based on the manipulated conditions in the study, were conducted to

determine if group preferences differed by condition.

TABLE 3

Potency Factor Means

 

 

Factor Mean Standard Deviation
Good Health 20.70'l 7.30

Habit 17 .83b 6.66

Care 14.00° 5.48

Law 13 .20d 6. 1 3

n=335

 

Values with different superscripts differ at the p <.05 level in paired T-test.

Tests were conducted to examine differences for power and relationship for the
Potency factor. Differences were found for ratings on the Law slogan only. Passengers,
those lower in power, gave higher Potency scores (M=13.93, SD=5.4) then drivers

(M=12.58, SD=5.5) on an Independent-samples T-Test (t=2.26, p<.03). This suggests that

59
the condition, low power, may affect the rating of the slogan and these low power individuals

rate the Law slogan higher because it may be one they want to use, given their condition.

TABLE 4

Potency Factor Means by Condition

 

 

IV Mean Standard Deviation
Power Law Slogan

Low (Passenger) 13.93" 5.4

High (Driver) 12.58b 5.5

n=335

 

Values with different superscripts differ at the p <.05 level in paired T-test.

For the independent variable relationship, differences exist only in the ratings for the
Caring slogan. Friends gave higher recall ratings (M=14.66, SD=6.1) then acquaintances
(M=13.21, SD=6.1) on an Independent-samples T-Test (t=2.18, p=.03). This suggests that
relational intimacy may affect the recall ratings for the Caring slogan, which is the most
directed at the other person. While no causality can be inferred flom the data, intuitively it
appears that the manipulated condition description can affect slogan ratings. Subjects would
be predisposed to give higher ratings to a particular slogan because it may be one they are

likely to use given the condition description in the study.

60
Additional support for this assumption was found by analyzing the single item Like
for the slogans and conditions. The pattern and significance follow the findings for the
Potency factor Passengers liked the Law slogan (M=3.99, SD=1.94) more then drivers
(M=3.45, SD=1.79; t=2.64, p<.01).
TABLE 5

Potency Factor Means by Condition

 

 

IV Mean Standard Deviation
Relationship Caring Slogan

Low (Acquaintance) 13.21“ 6.1

High (Friend) 14.66b 6.1

n=335

 

Values with different superscripts differ at the p <.05 level in paired T—test.

For tests of differences in relational intimacy, friends (M=3.27, SD=1.85) like the
Caring slogan better then acquaintances (M=2.93, SD=1.72; t=1.76, p<.05). Again these
results indicate that subjects may give higher ratings on liking a particular slogan based on
the condition described in the study. The single item of self or Other-directedness was also
tested; no differences were found for either the conditions of power or relationship for this
item on the four slogans used in this study.

Hypothesis nine predicts that the slogan reminding subjects about the seat belt law

61
will score the lowest on the motivational factor. No clear motivational factor emerged, as
was predicted, flom the data. The second factor of particular interest in this study was the
single item rating of self or Other-directedness of the slogan. While the means for this item
hover around the midpoint of the scale, significant differences occur (see Table 6). The Law
slogan is rated significantly more selfish (higher ratings are Other-directed and lower ratings
are Self-directed) than two other slogans related to Good Health and forming a Good Habit
about buckling up. No difference exits in the ratings of the Law and Show you Care slogans
on Other-directedness. The Good Health slogan was rated the highest on this item as well
as the previous Potency factor. Since the resulting factor structure did not support what was
hypothesized, a review of the ratings on individual items is appropriate. All item means are

presented in Table 6.

62

 

 

TABLE 6
Item/ Slogan Mean Ratings

Item Slogan
(mean/ Good Good
standard deviation) Law Habit Caring Health
Easy to 2.25‘1 3.62" 2.59c 4.26‘
remember (1.53) (1.86) (1.63) (1.94)
Like 3.70c 4.15" 3.12" 4.78“

(1.88) (1.87) (1.80) (1.88)
Clear 1.86d 2.79" 2.146 3.49‘

(1.35) (1.73) (1.43) (1.96)
Positive 3 .23'l 2.48c 1.946 2.83"

(1.84) (1.48) (1.17) (1.50)
Other- 3.98c 4.09" 3.906 4.38“
directed (2.01) (2.01) (2.22) (1.86)
Informative 2.44d 3.61" 3.280 3.90“

(1.65) (1.83) (1.84) (1.89)
Effective 2.93c 3.65" 2.88c 4.25'

(1.77) (1.72) (1.68) (1.85)
n=338

 

Row values with different superscripts differ at the p <.05 level in paired T-test.

(Results have been recoded so that higher scores are ratings

of other directed and the positive adjective. )

63

The Good Health slogan scores the highest on most items, supporting this as the
preferred slogan. The ratings for the Positive item are particularly interesting. The Law
slogan scores highest although the overall means for this item are comparatively low (all
below the midpoint of the scale). The low positive scores suggest that none of the slogans
were viewed that positively or that the subject matter (reminding others about automotive
safety belt use) in general is not seen very positively by the target audience. Other data,
previously reported and in the self reports by the subjects, indicate the high regard most
people have for safety belt issue. The data in this study shows that the ratings are not as
favorable for asking people to request others buckle up. This perceived change in the
positive nature of this issue must be recognized because it could effect any campaign using
this two-step flow approach.

The differences in the rating of the potential slogans point to a campaign direction
aimed at getting the attention of the target audience and serving as a catalyst for increasing
discussion about seat belt use. Overall, the item scores for the Good Health slogan are the
most positive, indicating that as the best slogan for a campaign to generate requests to buckle
the safety belt. The Good Health slogan was most highly rated on the clear, like, easy to
remember, informative, effective, and other-directed items. With comparatively high ratings
on these characteristics, a slogan tied to overall good health can be an effective campaign
message to encourage requests for automotive safety belt use. The Good Habit slogan was
the next highest rated; while, not surprisingly , the Law slogan was the most poorly rated and
only appropriate in certain conditions. Overall, these findings point to a campaign approach
linking reminders of seat belt use to other, positive, health and habit issues.

Hypothesis ten predicts that subjects’ willingness to ask others will affect scores on

64
the likelihood of compliance and sense of obligation ratings of the requests. Subjects were
categorized as high or low based on the summed scores on two five point Likert-type items
about the willingness to ask other drivers and passengers to buckle up, creating a high-low
dichotomy on this independent variable. A median split was used to create the high and low
groups. Results of the statistical tests for likelihood of compliance are reported in Table 7
and show no significant differences in the perceived likelihood of compliance based on
existing attitudes about willingness to ask. In general, likelihood of compliance was judged

to be high in all conditions.

65
TABLE 7

Means and Standard Deviations for Likelihood of Compliance
in Initial and Second Requests by Ratings of Likelihood of Asking Others

 

 

Likelihood of Asking
Low High
Initial Request 5.46 5.68
n= 321 (1.37) (1.35)
Second Request 5.74 5.94
n= 312 (1.34) (1.37)

 

NOTE: No significant differences at the p<.05 level. Standard deviations are in

parentheses.

Additional statistical tests examined the sense of obligation in others as perceived
by the subjects. The subjects who rate their likelihood of asking as high report generating
a higher sense of obligation rating in the second request (Table 8) compared to the initial
request. These data show only a change in second requests flom the view of the requestor.
The requestors more likely to ask think they generated a request stronger in sense of
obligation for the other to comply than those less likely to ask. This suggests that those

likely to ask believe they can generate a strong sense of obligation and may be successful

66
in increasing seat belt use. If the requestor can generate a request that he or she thinks
creates a strong sense of obligation in the other, it should result in a high rate of compliance
especially given the other directed benefits of this request. The fact that the sense of
obligation increases in the second request indicates a perceived change flom the perception
of the requestor; though no other statistical differences exist, they must think that they have
done something to different that makes them believe they have created a greater sense of

obligation with the second request.

TABLE 8
Means and Standard Deviations for Sense of Obligation

in Initial and Second Requests by Ratings of Likelihood of Asking Others

 

 

Likelihood of Asking
Low High
Initial Request 4.94 5.36
n= 321 (1.46) (1.43)
Second Request 562'' 6.01"
n= 312 (1.37) (1.26)

 

NOTE: Rows with different superscripts differ at the p<.05 level. Standard

deviations are in parentheses.

These data show general perceptions by requestors that they can generate a high

67
likelihood of compliance and sense of obligation in others through their requests. The
likelihood of compliance does not differ across the situations presented in this study. The
sense of compliance is perceived as slightly higher by those more likely to ask in the second
request condition. Overall these data support the utility of this peer pressure approach for

raising seat belt use with the target group.

Discussion

I . . l B

The data in this study do not support the predictions about the main effects of power
and relationship on compliance gaining strategies in the test situations. Specifically, the
results flom tests of hypotheses one, two, three, and four show subjects did not respond as
the hypotheses predicted in initial requests. Although the seat belt situation is a routine and
daily issue for the users, much like others that have been reported by Roloff and colleagues
(Jordan and Roloff, 1990; Roloff and Janiszewski, 1989; Roloff, Janiszewski, McGrath,
Burns and Monrai, 1988), this study goes beyond one’s personal use and probes for
persuasive requests in asking others to buckle up. This changes the characteristics of the
request and the data suggest the other directed nature of this situation accounts for the lack
of findings in the hypothesized direction. It is likely that the lack of support for the
hypotheses is due mainly to the situational differences in asking others to wear seat belt
compared to previous work. The result is that the predictions of the persuasive request
strategies for directness, politeness, and elaboration used in the manipulated power and
relationship situations, were not supported by the data for the automotive seat belt situations
tested in this study.

While no main effects were found for politeness, directness, or elaboration of
requests for the manipulated obstacles of power and relationship, the data show an
interaction effect of power and relationship on directness. This indicates subjects weigh
both power and relationship heavily in planning a request strategy in the described situation.

This interaction effect for power and relationship shows opposite patterns for drivers and

68

69

passengers with friends and acquaintances on the measure of directness. Drivers are more
direct with acquaintances and more indirect with fliends. The passengers are more direct
with fiiends then with acquaintances. The pattern shows how the two independent variables
interact to cause differences in choices of directness of the request. The findings are
consistent with earlier explanations of power and intimacy and request type. If drivers are
more concerned with the relationship of a fliend then an acquaintance, they will be less direct
in requests to their fiiend. If passengers want to be clear to the driver regarding their request,
they will be more direct. Subsequent tests show female subjects account for this difference
and responded opposite the hypothesized direction. There was no interaction effect for
politeness or elaboration found in the analysis. These differences found in the dependent
variables support directness and politeness as separate constructs worthy of independent
measurement and reporting, but do not support the hypotheses generated in this study.

The nature of this request, a safety request highly beneficial to the other, may be the
reason for this focus on relationship and power. The requestor has much less of a personal,
direct, physical benefit in the outcome than in other requests tested in the literature. It
appears that with the situation (seat belt request) and the coding scheme focusing on the
elaboration, directness, and politeness of requests, differences did not exist except among
female subjects. In this type of a safety request several of the potential obstacles noted by
Gibbs (1986), possession, ability, and permission, are removed or not evident and these
might affect the elaboration, politeness, or directness of the requests. These safety requests
may require other categories to better reflect the obstacles requestors perceive in this
particular situation of a safety request with a high benefit to the other.

The gender differences found in the data were also surprising. The study was

70
constructed to minimize any gender differences, by specifying same sex pairs in the
manipulation, so the focus would be on the obstacles to compliance with the seat belt
request. The result of the study reinforce gender differences in communication.

The violations of the predictions for power and relationship and the interaction effect
for directness also can be interpreted as showing that subjects either are not using one
obstacle or do not perceive the obstacles as they have been manipulated. The stronger case,
based on the interaction effect found in the AN OVA, suggests perceptions of both power and
relationship in developing a persuasive strategy for female subjects. The data also suggest
that the focus on benefit to the other person in this study may explain some of the lack of
support for the hypotheses. The overall lack of support for the hypotheses suggest that the
theoretical perspective of the obstacle hypothesis, requests are flamed to address the single
greatest obstacle, is not applicable to the problem of requesting unbuckled persons to wear
their automobile safety belt.

This study examined requests of an altruistic nature, with benefits mostly to the other
person not the requestor (law compliance in seat belt use benefits the driver in a certain way),
different then what has previously been researched; the exchange of more tangible goods or
services that benefit the requestor. Without a directly tangible physical benefit for the
requestor, the relationship itself may take on additional importance. This would explain the
maintenance of politeness and directness flom the first to second requests; if the requestor
was focused on maintaining the relationship, he or she might hesitate to be more direct and
less polite in a second request. Females were even more polite in second requests, a low
power strategy application. The description of a situations with a physical presence of

another person in a vehicle may also work to put an elevated focus on the relationship.

71
While the goal of the study was not to focus on the physical proximity of the interactants, the
description of the situation did indicate the subject would be in the car for the time it took
to get to the theater. Further study should examine the degree to which subjects are in fact
weighing both power and relationship considerations in this type of request, with high
benefits to the other person and close physical proximity, compared to other requests that

generally benefit the self.

Another key finding is the lack of a significant difference in the method and form of
the request flom the initial request to the second request. Subjects were told the other had
not complied with the initial request and they had to ask again. This information did not
result in a change in the elaboration, directness, or politeness forms of their follow up
request. An explanation for the lack of a significant difference between the initial request
and the second request is that subjects consider the situations more similar then they do
distinct, despite refusal to comply with the initial request. The study expectation was that
a change in the situation would change the perceived obstacles and the type of request likely
to be used by telling subjects the first request had been ignored. It was expected this
manipulation would alter the subject’s view of the problem and result in predictable change
in the persuasive approach. However, no manipulation check was built into the study to
check this view of rejection of the initial request and to ask subjects about their perceptions
of the obstacles at various points in the study. Other studies (Roloff, et al., 1988) have

examined response to rejection as separate and different flom the initial request. Further

72
study is needed to determine if this explanation for the lack of a change in persuasive strategy
between the initial request and the second request is correct.

Another interpretation is that the manipulations in the study were weak. With a weak
manipulation, subjects might not perceive any great change in the situation which would
require any change in persuasive strategy. This type of request may not require many
changes in strategy. The focus in this study is mostly altruistic, or other-directed, requests
without an exchange of goods or services. Consequently, subjects may not apply a large
amount of cognitive effort in this situation as compared with others studied (ask for money,
to borrow class notes: Roloff, et al., 1988). However, observation during data gathering
indicates a wide range of time spent forming the requests and the qualitative data show a
broad range of responses. Future studies would do well to monitor the time generating the
request as a dependent variable perhaps indicative of the amount of cognitive effort put into
the situation. The manipulations may be weak, even though situation specifications were
repeated in the instructions, and could be improved upon in any additional work.

While the lack of significant findings may be easily explained by saying that
situational differences did not cause any change in communication strategy, or that subjects
view the problem as a simple issue, the qualitative information on the persuasive problem
provides some additional data to consider. The fact that a new coding category had to be
added to the second request data because so many subjects (11%) had examples or stories
to tell about this issue shows some depth to the problem that would not be gleaned just flom
a review of the quantitative results. Subjects used personal information that included graphic
details and fear appeals, to persuade the other to buckle up. The depth of qualitative

information presented by subjects in the second requests varied flom very serious and

73
personal, one’s own experience with seat belts in an accident, to more humorous and
personal, e.g., “I’m a bad driver and I won’t pay your hospital bills if anything happens.” It
was clear flom reviewing the text that these subjects did not feel the need to restate the
request to buckle up but focused more on the strategy that would lead to compliance. In
second requests that follow immediately flom initial requests, subjects may not feel the need
to restate a request. The conditions where this takes place may include a request with a high
benefit to the other in a situation that is simple and direct. If there is a high benefit to the
self, the requestor probably restates the request for clarity and to ensure that there is not a
misunderstanding of the request. This restatement issue may reflect a strategy difference
based on benefit to the self or the other. Other qualitative information flom this study, the
increased use of expletives in second requests, supports this interpretation. The requests
were reviewed by a coder for the use of expletives and a change flom the initial to the second
request. In initial requests, only five contained expletives. That number increased by nearly
four times to nineteen in the second requests suggesting a strategic change in persuasive
strategy flom time one to time two. This may be indicative of another consideration, such
as force or urgency, the requestor is applying to the situation. While these changes were
not evident in the statistical tests of directness, politeness, or elaboration, it is important to
note and provides some considerations for interpreting the results of this study. This shows
subjects made some adjustments in their requests that are better tracked through semantic and
perhaps non-verbal methods then in the coding of text on the variables of politeness,
directness, and elaboration used in this study. Reviewing the qualitative and
quantitative results suggests the other-directness of this persuasive situation and the lack of

a physical, tangible exchange of goods or services between the requestor and the other to be

74
the likely explanation for deviation flom hypothesis predictions. Compared with what has
been studied in the past, there is little benefit for the requestor in the seat belt situation, but
there is also little cost to the other person. This makes this a very direct and perhaps simple
situation. This may limit the number of request options that a requestor considers. The lack
of change in directness of requests in the manipulated conditions shows that these mostly
indirect requests are conventional in the described situation. This supports Gibbs (1985)
position that indirect requests are conventional in certain contexts. The manipulations in this

study did not change the context enough to cause a change in the directness of the request.

SleaanExaluaticns

Although the hypothesis predictions were not supported by the data, the study holds
some other valuable information. The data show some gaps in attitudes and current practices
that can be reinforced to change the behaviors. Specifically, the data about persuasive
approaches, campaigns, and slogan evaluation show the potential of peer pressure with the
target audience. Subjects reported a high willingness to ask others (definitely/probably ask -
61% passenger and 41% driver). They also suggested that they want to be reminded by others
if they are not buckled up (64% always/usually want to be reminded by others). The
questionnaire data also indicates subjects believe they can influence others and think their
requests generate a high likelihood of compliance. The high scores given for the sense of
obligation and likelihood of compliance generated by the request suggest that subjects think
this approach will work on the target group, their peers. Perhaps the requestors think that

because there is so much benefit to the other, they feel a high sense of obligation and/or

75
likelihood of compliance are generated and there is little strategy needed by them in this
situation.

I The slogan evaluation ratings themselves show which appeals (Good Health and
Habit) will work the best with the target audience, overall, and in which condition the Law
approach may be relevant and fitting ( passenger, low power). This finding about a Law
reference is important because it shows a niche where the Law approach will work.
Referencing the law can be effective for passengers to get drivers to buckle up. These data
also show the difference between referencing the law when asking someone else to buckle
up and forcing widespread compliance by strengthening the existing law. This study found
people will use law to get others to buckle up but prefer the current state of secondary
enforcement. The Caring slogan was liked best by fliends, showing it too has value in
encouraging safety belt use in a particular condition. In constructing campaign messages for
nondrivers, it is acceptable to suggest they reference the safety belt use law to get drivers to
buckle up. Messages for drivers encouraging others should take different approaches, such
as indicating you care for a fiiend. Broader campaign messages, aimed at the young adult
target group, should reference safety belt use in the context of overall healthful behavior or
changing habits. These themes provide the most universal messages to generate requests to
buckle up.

Subjects also suggested which medium, television, is best for delivering messages
aimed at encouraging discussion about seat belt use. However other, more precise,
secondary media may address particular findings in this study. Print media, through
billboards, bumper stickers, flyers, and interior stickers, can provide group targeting. Flyers

or billboards with the Good Health theme in or near health clubs, sporting events, or athletic

76

fields match the theme with the audience. Interior stickers on the driver’s or passenger’s side
of vehicles can remind person’s of the appropriate strategy for that position (Caring for driver
and Law for passenger). Message targeting is suggested by the data in this study and there
is sufficient information flom which to direct a campaign aimed at increasing peer reminders
about seat belt use among young adults.

One goal of the study was to assess the viability of a campaign to increase automotive
safety belt use; the data point to the utility of this approach. One is the gap identified (33%)
between the number of subjects who indicate they want to be reminded to buckle up (64%)
and the number who currently say they ask others (31%). A media campaign can be directed
at raising the number of people who ask others given the expectation, flom the data in this
study, that there will be high compliance with the request because they are many others who
want to be reminded to buckle up. This gap alone provides validation of the approach
suggested in this study. Additionally, the sense of obligation and likelihood of compliance
the requestors report suggests they believe that this approach will work. Although the sense
of obligation and likelihood of compliance were highly correlated (r=.51, p< .01), they
provide value as separate measures. Subjects who gave the highest ratings on sense of
obligation (n=149) were direct in their requests (90% in the two most direct categories), but
used different levels of politeness. One third used the least polite request form, while 62%
fell into the two most polite categories. Elaboration of these requests averaged 13 words,
near the mean for the entire study. In examining the strategies of the subjects who gave the
highest ratings of likelihood of compliance (n=203), they showed a similar pattern of
directness, with 91% in the two most direct categories, and elaboration (13 word average).

The politeness of the requests changed with this group as 5% more of the requests were in

77

the least polite category, with the requests shifting flom the two most polite categories to the
least polite category. Subjects who gave their requests a high rating of sense of obligation
or likelihood of compliance (or both), were direct in their request, average in elaboration and
generally at either of the extremes of politeness. This provides a “profile” of the type of
request that subjects rated with a high potential for getting the other to buckle up. This
information can be included in advanced campaigns (“When asking another to buckle up, be
direct but polite in your request”) to further enhance compliance. Another part of the study
provides data on mass media messages that support encouraging others to wear their seat
belts.

The slogan evaluations provide some clear direction on the attractiveness of the
campaign slogans to the sample population and hopefully to the target population as well.
These differences are important for several reasons. For a slogan to be effective at generating
behavioral change, it must be attention getting and memorable. Recall that making the issue
of seat belt use more salient for the target group is a way to elicit behavior change (Langer,
1978; Abelson, 1977; Roloff, 1980). Also recall that Boster and Stiff (1984) pointed out the
importance of benefit to other in persuasive communication and that the requestor is more
likely to attempt different strategies and ask again, if the benefit to the other is high. The
focus of other—directedness of this request makes this study stand out flom previous work.
The results of these slogan ratings support the benefit to the other component in the
campaign and tie the seat belt behavior to other life goals (good health). The Good Health
slogan is clearly preferred overall by the sample population and this slogan fits with the data
on campaigns that are successful in leading to behavior change, in this case increased use

of the seat belt.

78

The approach suggested in this study also parallels Geller’s (1989) ABC model for
seat belt behavior change. An activator is needed, here it is the mass media message or
campaign slogan. This activates the behavior; one person asking another to buckle up. The
consequence in this situation is either a consequence or a reward. It’s a consequence if the
unbuckled person does not comply with the request because they will be asked again
(possibly more than twice). The second request contains a much greater likelihood of stories,
personal experiences, fear appeals, and the use of expletives to assure compliance. Several
of these can be viewed as consequences. However, there can be a reward if the person
complies. The reward is the building of the relationship between the two people. This
problem can result in a win-win situation without any personal (physical) cost and in a
situation in a vehicle where the interactants are likely to be together, in close proxirrrity, for
a least a short period of time. The benefit of the approach outlined here compared to Geller’s
(1989) is the continued reinforcement flom others that, over the long run, is more likely to
result in long term behavior change at a lower cost then with the external rewards offered in
his program. Even with the increased appearance of advanced safety and automatic systems,
putting on the safety belt is still the greatest safety feature in a vehicle and the maximum
benefits of the seat belt will not be realized until use is increased beyond the current plateau.

The high sense of obligation and likelihood of compliance found in this study, while
measured flom the perspective of the requestor, support this approach to persuasion. These
requestors think that their request will result in compliance by the other. This really is the
ultimate goal of the process and in this condition leads to the safety conscious behavior. The
next step would be to test various persuasive messages generated to request someone buckle

up and acquire rating of sense of obligation and likelihood of compliance flom the person

79
being asked. This additional work would clarify the kind of messages that would make the
target audience “Buckle Up” and provide data on the best persuasive approach flom the point
of view of the other. This would be important data because of the previously identified
other-directed nature of this research.

A different study design, with a manipulation check, measures of cognitive
complexity in generating the request, and tracking some expected non-verbal differences,
could provide additional data on persuasive strategies and conditions for success in
persuading the other. The ultimate campaign would go beyond the Good Habit slogan
suggested here to a campaign actually reminding people to ask and suggesting successful
strategies for the requests. A more realistic study that involves role playing or audio tapes
of actual requests could provide valuable data on the non-verbal changes in requests not
gleaned flom the current study. This could lead to providing additional request information
to maximize compliance. The lack of a significant difference found in the initial and second
requests for politeness and directness could be a reflection of no verbal change but in reality
non verbal aspects may change between the initial and second requests.

The data from this study support previous information about the safety belt law being
well known and viewed as a good law that most people support. However, reminding people
in the low use target audience about the law is not, generally, a good way to promote
buckling up by those who currently don’t use their seat belts. However, when the requestor
is the passenger, a low power position, the law approach may be beneficial in providing
clarity and directness to the request. The reference to the seat belt law may also help give
credibility or equalize the perception of situational power flom the view of the passenger

making the request. The only finding for relationship showed that subjects in the friend

80
condition preferred the Show you Care- Remind Others slogan more than acquaintances.
This is not surprising given the focus on other and the level of the relationship. These
findings show message preference can be affected by the condition, in certain instances, and
this may be valuable information for targeting messages in the future. Overall, the data in
this study show that connecting safety belt use to other healthy behaviors and good habits is
much preferred as a campaign strategy.

These slogans may indirectly address some of the reasons for not using seat belts:
discomfort, forgetfulness, fear of entrapment, and the risk of an accident (F haner and Hane,
1973; 1974) by putting a positive spin on seat belt use and connecting use to other larger
issues like overall good health. Given the other, self-report, data on wanting to be reminded
to buckle up, it seems reasonable to expect that the right campaign message could positively
affect seat belt use through a “two-step flow” approach flom media to individual
communication. This approach also lets the state maintain a secondary type of enforcement
with some assurances that belt use will increase but the choice will be made by the
individual. Keeping secondary enforcement averts any criticism or negative affect toward
the seat belt issue that can result flom moving to primary enforcement.

Is there a broader application to other health and safety issues which benefit society,
the self, or the other person? The automotive safety belt issue and the campaign approach
outlined in this study may provide an enhancement and catalyst for behavioral change. The
data gathered and the approach suggested here follow guidelines of a formative research
project outlined by Atkin and Freimuth (1991) that lead to a successful media campaign.
Formative research should outline the audience characteristics and provide some pre-

production data to guide the campaign (Atkin and F reirnuth, 1991). This study tested

81

persuasive requests within a theoretical flarnework, provided audience characteristics,
profiled subject request styles, and tested some slogans to determine a preference. Subjects
also provided suggestions for media choices in delivering the message. Although the data
do not support hypotheses generated by the theory, the findings contribute to the literature
and provide much of the detail of a formative research project. The study addresses issues
aimed at changing individual behavior: increased interpersonal communication and self-
awareness that increases cognitive awareness (Roloff, 1980), and specifically targets a group
of low users. The data identifies a gap in the perceptions that others (peers) expect me to
buckle up. The campaign approach suggested here can generate communication that may
change the perceptions of peer use and raise expectations of seat belt use among the target
group. All these make for a viable, cost effective, approach to increased seat belt use with
the expectation that the change will be long lasting. The peer reminder approach suggested
here takes advantage of the two-step flow of mass media and interpersonal communication
and provides the format applicable for increasing automotive safety belt use. Also recall that
children exert an upward influence (Wagenaar and Webster, 1986) on overall use rates
except with the target group, young adults. Think how effective adult peer pressure and use
might be as a persuasive strategy with a low user group. If the perception of use among
college students, a subset of the young adult target audience, can be raised it could filter into
the larger population of the young adult target group encouraging and reinforcing seat belt
use. The data in this study suggests this possibility and supports the generalizability of the
data to the larger target audience of young adults.

Other media campaigns, see Rice and Atkin (1991) for a review, have documented

positive changes in affect, education, and behavior among targeted groups. These campaigns

82
are examples of the potential of mass media, and in some cases interpersonal
communication, to affect change. The situation outlined in this study, high awareness of the
law, a gap between self-report of wants and current behaviors, several viable message slogan
choices, media choices recommended by the target audience, and close proximity of the
interactants, should lead to high communication and enhance the success of this “two step
flow” approach. This may be the ideal condition for mass media and interpersonal
communication to work together to affect positive change. The approach presented here can
be used to guide a campaign, and is worthy of a field test, in a controlled experiment of
comparable groups or in a pre-test and post-test situation, to quantify results. This would be

the next logical step based on the findings in this work.

APPENDIX A

83

TABLE A-l

Characteristics of the Subject Sample

 

Gender

Age (X =19.25) -
Percent of Sample

Licensed driver

State

Female - 53% Male - 47%

18-33% 19-29% 20- 13%

Yes - 97%

Michigan - 90% Other - 10%

21-12%

84

TABLE A-2a

 

Factor Analysis
Correlation Matrix

 

CLEAR EASY EFF INFO LIKE POS OTHER

CLEAR 1.00000

EASY .52436 1 .00000

EFF .421 18 .53094 1.00000

INFO .36694 .38550 .59072 1.00000

LIKE .25948 .48161 .55155 .37554 1.00000

POS .36792 .27366 .35243 .32135 .44635 1.00000
OTHER -.25331 -.33567 -.30784 —.20030 -.22708 -.01539 1.00000

85

TABLE A-2b

 

Factor Analysis Extraction
Principal Components

 

Variable Communality * Factor Eigenvalue Pet of Var Cum Pct

CLEAR 1.00000 * 1 3.24435 46.3 46.3
EASY 1.00000 * 2 1.02672 14.7 61.0
EFF 1.00000 * 3 .76438 10.9 71.9
INFO 1.00000 * 4 .69737 10.0 81.9
LIKE 1.00000 * 5 .57847 8.3 90.2
POS 1.00000 * 6 .36081 5.2 95.3
OTHER 1.00000 "' 7 .32790 4.7 100.0

PC extracted 2 factors.
Factor Matrix:

Factor 1 Factor 2

EFF .81932 -.01441

EASY .76253 -.20579
LIKE .71971 .19351
INFO .69953 .08616
CLEAR .67418 -.06775
POS .57848 .60284

OTHER -.44173 .75583
Final Statistics:

Variable Communality * Factor Eigenvalue Pct of Var Cum Pct
*

CLEAR .45911 * 1 3.24435 46.3 46.3
EASY .62381 * 2 1.02672 14.7 61.0
EFF .67149 *
INFO .49676 *
LIKE .55543 *
POS .69806 *

OTHER .76641 *

APPENDIX B

 

86
COMMUNICATION RESEARCH QUESTIONNAIRE

INSTRUCTIONS

The attached survey asks you for your responses to questions and several
situations. Please answer all questions as you believe you would respond in the described
context. Your responses will remain confidential and will be analyzed as a representative
college-age group. Additional items ask for information and some general opinions on
related issues.

Please circle, check, or write in the your answers on the forms. There are no right
or wrong answers, just your opinions. For all references to automotive safety belt use
assume full and proper use of the safety system (lap and shoulder belts if applicable) as
instructed in the owners manual.

Your instructor has gratefully offered class time for your participation and will
provide you with extra credit. The entire survey should take 15 to 20 rrrinutes. Thank
You for your participation. You can turn your survey over when completed and someone
will pick it up.

87

BACKGROUND INFORMATION
1. Age __
2. Sex: Female [] Male []
3. Do you currently have a valid driver's license? YES NO

3a. From which state or nation?

 

4. In 1985, Michigan passed a law requiring safety belt use. What's your opinion
of this law?

VERY MUCH SOMEWHAT SOMEWHAT STRONGLY
IN FAVOR IN FAVOR IN BETWEEN OPPOSE OPPOSE

5. How strictly do you think the safety belt law should be enforced?

VERY SOMEWHAT AS IT IS LESS NOT AT
STRICTLY STRICTLY NOW STRICTLY ALL

6. When driving, how often do you ask unbuckled passengers to buckle up?
ALWAYS OFTEN SOMETIMES SELDOM NEVER

7. When you're a passenger, how often do you ask an unbuckled driver to buckle up?
ALWAYS OFTEN SOMETIMES SELDOM NEVER

8. Do you think unbuckled passengers want to be reminded to buckle up?

MOST OF THEM SOME OF THEM A FEW OF THEM NONE OF THEM

9. Do you think unbuckled drivers want to be reminded to buckle up?

MOST OF THEM SOME OF THEM A FEW OF THEM NONE OF THEM

10. If you are unbuckled, do you want someone to remind you to buckle up?

ALWAYS USUALLY SOMETIMES SELDOM NEVER

88

Please respond to the following statements.
11. Safety belts significantly reduce the severity of injuries in accidents.

STRONGLY AGREE NEUTRAL DISAGREE STRONGLY
AGREE DISAGREE

12. How important are the following reasons for wearing a safety belt?
Circle (1) Very Important, (2) Important, (3) Somewhat Important or (4) Not
Important:

Because I feel safer when buckled.

Because I'm used to buckling up now.

Because the law requires seat belt use.

Because I'm worried about being in an accident.
Because others expect me to buckle up.

sue-99‘s»
~O-‘HI-‘H
NNNNN
wwwww
hA-h-h-h

13. Would you ever ask a passenger in the car you're driving to buckle up his/her seat
belt before you drive away?

DEFINITELY PROBABLY POSSIBLY NO

14. Would you ever ask a driver in the car you're riding in to buckle up his/her safety
belt before driving away?

DEFINITELY PROBABLY POSSIBLY NO

15. How much influence over a passenger's safety belt use do you have as the driver of
the car?

A LOT OF SOME A LITTLE NOT MUCH NO
INFLUENCE INFLUENCE INFLUENCE INFLUENCE INFLUENCE

16. How much influence over the driver's safety belt use do you have as the passenger
of the car? '

ALOT OF SOME A LITTLE NOT MUCH NO
INFLUENCE INFLUENCE INFLUENCE INFLUENCE INFLUENCE

17. If you were going to use seat belts more often, how helpful would it be to have
reminders in the media you see daily?

VERY HELPFUL SOMEWHAT HELPFUL A LITTLE HELPFUL NOT HELPFUL

89

Imagine that you are enrolled in a Humanities course that counts toward your
general studies requirement. One way to pick-up extra credit for the course is to view,
and write a short critique, of a foreign film shown at a theater across town. The instructor
encourages viewing and discussing the film with a classmate as a way to enhance
understanding and generate ideas. With this in mind, you make arrangements with a
same sex WW flom class to see the film.

The day of the film arrives and you drive over to pick-up your
MW. You are buckled in and your classmate gets in and says "Let's go!"
but does not put on his/her seat belt. At this time you are going to request that he/she

"Buckle Up!". In the space below, write the exact words you would say to request that

your Wm put on the seat belt.

 

 

 

 

 

 

 

 

 

 

(After completing this part, go on to the next page.)

90

Thinking about how you asked your mm to wear his/her seat
belt, rate your request on the following scales.

How much do you think your request generates a sense of obligation in the other

 

person?
No Strong
obligation obligation

How likely is it that this person would comply with your request to wear a seat

belt in this situation?
No Certain
likelihood likelihood

of compliance of compliance

91

Imagine that you are enrolled in a Humanities course that counts toward your
general studies requirement. One way to pick-up extra credit for the course is to view,
and write a short critique, of a foreign film shown at a theater across town. The instructor
encourages viewing and discussing the film with a classmate as a way to enhance
understanding and generate ideas. With this in mind, you make arrangements with a
same sex Wm flom class to see the film.

The day of the film arrives and your We; arrives to pick
you up. You get in the flont seat and buckle your seat belt. Your W
says "Let's go!" but he/she does not have on a seat belt. At this time you are going to

request that he/she "Buckle Up!". In the space below, write the exact words you would
say to request that your W put on the seat belt.

 

 

 

 

 

 

 

 

 

 

(After completing this part, go on to the next page.)

92

For whatever reason your WW did not put on his/her seat belt
when you asked. You are willing to try again to persuade him/her to "Buckle Up!". In the

space below, write the exact words you would say now to request that your

friendlaganaintance put on the seat belt.

 

 

 

 

 

 

 

 

 

 

(After completing this part, go on to the next page.)

93

Thinking about your second request to your Wm to wear his/her

seat belt, rate your request on the following scales.

How much do you think your request generates a sense of obligation in the other

person?
N 0 Strong
obligation obligation

 

How likely is it that this person would comply with your request to wear a seat

belt in this situation?
No Certain
likelihood likelihood

of compliance of compliance

94

The next several pages will ask for your evaluations of various messages that can
be used to encourage people to request that others who are not buckled up wear their seat
belts. Please read each message and evaluate it on the scales provided then move to the

nextpage.

MAKE SURE EVERYONE IS BUCKLED UP - IT'S
THE LAW

 

 

 

 

 

Easy to Not easy
remember to remember
Dislike Like

Clear Unclear
Negative Positive
Self— Other—
directed directed
Informative Uninformative

 

Effective Ineffective

 

95

SHOW YOU CARE - REMIND OTHERS TO BUCKLE
UP

 

 

 

 

 

Easy to Not easy
remember to remember
Dislike Like

Clear Unclear
Negative Positive
Self— Other-
directed directed
Informative Uninformative

 

Effective Ineffective

 

96

SHARE A GOOD HABIT -
REMIND OTHERS TO BUCKLE UP

 

 

 

 

 

 

 

 

 

 

Easy to Not easy
remember to remember
Dislike Like

Clear Unclear
Negative Positive
Self— Other-
directed directed
Informative Uninformative

 

 

Effective Ineffective

 

 

97

PROTECT GOOD HEALTH -
MAKE SURE EVERYONE IS BUCKLED UP

 

 

 

 

 

Easy to Not easy
remember to remember
Dislike Like

Clear Unclear
Negative Positive
Self— Other—
directed directed
Informative Uninformative

 

Effective Ineffective

 

98

Finally, messages about seat belt use targeted at
people like you can be delivered through a variety of media.
If the goal is to increase seat belt use, finding the media
that best allows for encouraging interpersonal messages
about buckling up is important. Below are listed a variety
of ways messages about seat belt use can be delivered.
Please check a pmimary and secondary media that you think
will be most effective for messages about seat belt use to

people like you.

Primary Media (check one)

television ads radio ads bumper stickers
billboards buttons newspaper ads
magazine ads posters

reminder stickers inside the car

Secondary Media (check one)

television ads radio ads bumper stickers
billboards buttons newspaper ads
magazine ads posters

reminder stickers inside the car

This completes the questionnaire. Thank you for your time!
Please turn. your' completed. questionnaire into ‘your

instructor.

APPENDIX C

99

POLITENESS

(Comparative ratings from most to least polite)

Please + imperative (most polite)
Should you + imperative

 

Did you + imperative
Do you think that you + imperative
You should + imperative
Should + need assertion

 

Don't you think you + need assertion (Neutral)
Need assertion
Why don't you + imperative
Don't you think + imperative

 

I think + need assertion
I think you + imperative

 

Do + need assertion
Do you think + need assertion
Imperative only (least polite)

100

DIRECTNESS

(Comparative ratings from most to least direct)

Imperative only (most direct)
You should + imperative

 

Why don't you + imperative
Don't you think you + imperative
Do you think that you + imperative
Please + imperative
I think you + imperative

 

(neutral)
Did you + imperative
Should + need assertion
Should you + imperative

 

Don't you think + need assertion
Do + need assertion
Need assertion only

 

Do you think + need assertion
I think + need assertion
(From Kemper & Thissen, 1981)

 

_— ..___. —.__. —- 7 ...-

LIST OF REFERENCES

LIST OF REFERENCES

Abelson, R. (1976). Script processing in attitude formation and decision making. In J.
Carroll and T. Payne (Eds.). Cognition and Social Behavior (pp. 33-45). Hillsdale,
NJ: Lawrence Erlbaurn.

American Demographics. (March, 1987). Buckle up for savings. 9, 18.

Atkin, CK. (1981). Mass media information campaign effectiveness In R.E. Rice and
WJ. Paisley (Eds), Public Campaigns (pp. 265-280). Beverly Hills, CA:Sage.

Atkin, OK. and Freimuth, V. (1989). Formative evaluation research in campaign design.
In R.E. Rice and CK. Atkin (Eds), Public Campaigns, 2nd ed. (pp. 131-150).
Beverly Hills, CA:Sage.

Baker, S.P., O'Neill, B. and Karpf, RS. (1984). The Injury Fact Book. Health and
Company; Lexington, MA.

Barokas, J. (1995). Lessons learned from public health campaigns and applied to anti-
DWI norms development. The US. Department of Transportation, the National
Highway Traffic Safety Administration.

Bar-Ta], D., Bar-Zohar, Y., Greenberg, MS, and Hermon, M. (1977). Reciprocity
behavior in the relationship between donor and recipient and between harm-doer and
victim. Sociometry, 40, 293-298.

Berger, CR. (1985). Social power and interpersonal communication. In M.L. Knapp
and GR. Miller (Eds), Handbook of interpersonal communication (pp. 439-499).
Beverly Hills, CA: Sage.

Berger, CR. and Chaffee, S. (1988). On bridging the communication gap. Human
Communication Research, 15(2) 311-318.

Bloomquist, G. (1979). Value of life saving: Implications of consumption activity.
Journal of Political Economy, June, 540-558.

Bostrom, RN. (1983). Persuasion. New Jersey: Prentice-Hall, Inc.

101

102

Boster, F.J. and Stiff, JR. (1984). Compliance-gaining message selection behavior.
Human Communication Research, 10, 539-556.

Brown, P. and Levinson, S. (1978). Universals in language usage: Politeness
phenomena. In E.N. Goody (Ed.), Questions and Politeness: Strategies in social
Interaction (pp. 56-289). London: Cambridge University Press.

Clark, H.H. (1979). Responding to indirect speech acts. Cognitive Psychology. 11,
430-477.

Cody, M.J. and McLaughlin, ML. (1980). Perceptions of compliance-gaining situations:
A dimensional analysis. Communication Monographs, 47, 132-148.

Dillard, J .P., and Burgoon, M. (1985). Situational influences on the selection of
compliance-gaining messages: Two tests of the predictive utility of the Cody-
McLaughlin topology. Communication Monographs, 52, 289-394.

Donohue, WA. (1988). Issues of risk in adult and teen safety. Alcohol, Drugs and
Driving. 4, 297-304.

Donohue, W.A., Hildalgo, R. and Gathman, S. (1995). Literature review for the OHSP
1997 highway safety plan. Department of Communication, Michigan State
University.

Elman, D. and Killebrew, T.J. (1978). Incentives and seat belts - Changing a resistant
behavior through extrinsic motivation. Journal of Applied Social Psychology, 8, 72-
83.

Ervin-Tripp, S. (1976). Speech acts and social learning. In K. Basso and H. Selby
(Eds), Meaning in Anthropology. Albuquerque: University of New Mexico Press.

Evans, L. (1987). Fatality risk reduction from safety belt use. Journal of Trauma. July,
746-749.

Fhaner, G. and Hane, M. (1973). Seat belts: The importance of situational factors.
Accident Analysis and Prevention. December, 5(4), 267-285.

Fhaner, G. and Hane, M. (1974). Seat belts: Contextual factors and bias of reported use.
Journal of Safety Research. December, 6(4), 166-170.

Fhaner, G. and Hane, M. (1979). Seat belts - Opinion effects of law-induced use.
Journal of Applied Psychology, 64, 205-212.

103

Folger, J.P., Hewes, DE, and Poole, MS. (1984). Coding social interaction. In B.
Dervin and M.J. Voight (Eds) Progress in Communication Sciences Volume IV (pp.
115-159). Ablex Publishing: Norwood, NJ.

Francik, ER and Clark, H.H. (1985). How to make requests that overcome obstacles to
compliance. Journal of Memory and Language, 24, 560-568.

French, J .R.P., Jr. and Raven, B. (1959). The bases of social power. In D. Cartwright
(Ed.), Studies in Social Power. Ann Arbor: ISR (pp.150-167).

Garvin, RB. and Alcom, J .D. (1990). Behavioral strategies for alcohol abuse prevention
with high risk college males. Journal of Alcohol and Drug Education.

Geller, E. S. (1989). Intervening to increase children’s use of safety belts. Alcohol,
Drugs, and Driving, 5, 37-59.

Geller, ES. (1983). Rewarding safety belt usage at an industrial setting: Tests of
treatment generality and response maintenance. Journal of Applied Behavior
Analysis, 16: 189-202.

Geller, ES. and Hahn, H.A. (1984). Promoting safety belt use at industrial sites: An
effective program for blue collar employees. Professional Psychology: Research and
Practice, 15, 553-564.

Geller, E.S., Johnson, RP, and Pelton, S.L. (1982a). Community-based interventions

for encouraging safety belt use. American Journal of Community Psychology, 10:
183-195.

Geller, E.S., Kalsher, M.J., Rudd, J .R., and Lehman, GR. (1989). Promoting safety belt
use on a university campus: An integration of commitment and incentive strategies.
Journal of Applied Social Psychology, 19(1), 3-19.

Gibbs, R.W. (1979). Contextual effects in understanding indirect requests. Discourse
Processes, 2, 1-10.

Gibbs, R.W. (1981). Your wish is my command: Convention and co in interpreting
indirect requests. Journal of Verbal Learning and Verbal Behavior, 20, 431-444.

Gibbs, R.W. (1982). A critical examination of the contribution of literal meaning to
understanding nonliteral discourse. Text, 2, 9-27.

Gibbs, R.W. (1983). Do people always process the literal meanings of indirect requests?
Journal of Experimental Psychology: Learning, Memory, and Cognition, 9, 524-533.

104

Gibbs, R.W. (1985). Situational conventions and requests. In J. Forgas (Ed.), Language
and Social Situations (pp. 97-110). New York: Springer-Verlag.

Gibbs, R.W. (1986). What makes some indirect speech acts conventional? Journal of
Memory and Language, 25, 181-196.

Gregersen, NP. and Berg, H.Y. (1994). Lifestyle and accidents among young drivers.
Accident Analysis and Prevention.

Grice, HP. (1975). Logic and conversation. In P. Cole and J. Morgan (Eds), Syntax
and semantics: Speech Acts (V 01. 3). New York: Academic Press.

Hunter, J. and Boster, F.J. (1979). Situational differences in the selection of compliance-
gaining messages. Paper presented at the annual meeting of the Speech
Communication Association, San Antonio, TX.

Jonah, BA. (1984). Legislation and the prediction of reported seat-belt use. Journal of
Applied Psychology, 69, 401-407.

Jonah, BA. and Dawson, NE. (1982). Predicting reported seat-belt use from attitudinal
and normative factors. Accident Analysis and Prevention, 14, pp305-310.

Jonah, B.A. and Grant, BA. (1985). Long-term effectiveness of selective traffic
enforcement programs for increasing seat-belt use. Journal of Applied Psychology,
70, 257-263.

Jordan, J .M. and Roloff, ME. (1990). Acquiring assistance from others: The effect of
indirect requests and relational intimacy on verbal compliance. Human
Communication Research, 16, 51 - 555.

Journal of American Medical Association. (November, 1989). Safety-restraint
assessment - Iowa, 1987-1988. p. 2804.

Kasher, M.J., Geller, E.S., Clarke, S.W., and Lehman, GR. (1989). Safety belt
promotion on a naval base: A comparison of incentives vs. disincentives. Journal of
Safety Research, 30(3), 103-113.

Keebler, J. and Pinto, L. (1991). Tardy NHTSA to expose failure of 'passive' belts.
Automotive News. July, 29 (pp. 1 and 32).

Kello, J .E., Geller, E.S., Rice, J .C., and Bryant, S.L. (1988). Motivating auto safety belt
wearing in industrial settings: From awareness to behavior change. Journal of

Organizational Behavior Management, 9(2) 7-21.

105

Kemper, S. and Thissen, D. (1981). Memory for the dimensions of requests. Journal of
Verbal Leaming and Verbal Behavior. 20, 552-563.

Knapp, ML. (1984). Interpersonal communication and human relationships. Boston:
Allyn and Bacon.

Kooler, J .M. and Bruvold, W.H. (1992). Evaluation of an educational intervention upon
knowledge, attitudes, and behavior concerning drinking/drugged driving. Journal of
Drug Education.

Lakoff, R. (1975). Language and the woman’s place. New York: Harper and Row.

Langer, E.J. (197 8). Rethinking the role of thought in social interaction. In J .H.
Harvey, W.I. Ickes, R.F. Kidd (Eds), New Directions in Attribution Research (V 01.
2, pp. 35-58). Hillsdale, NJ: Erlbaum.

Lirn, TS. and Bowers, J .W. (1990). Politeness: Solidarity, approbation, and tact. Paper

presented at the annual meeting of the International Communication Association,
Dublin.

Loo, R. (1984). Correlates of reported attitude towards and use of seat belts. Accident
Analysis and Prevention, 16, pp. 417-421.

Lund, AK. (1986). Voluntary seat-belt use among United States drivers - Geographic,

socioeconomic and demographic variation. Accident Analysis and Prevention, 18,
pp. 43-50.

Maccoby, N., and F araquhar, J. (1975). Communication for health: Unselling heart
disease. Journal of Communication, 23(3), 114-126.

Maccoby, N., F araquhar, J ., Wood, R, and Alexander, J. (1977). Reducing the risk of
cardiovascular disease: Effects of a community-abased campaign on knowledge and
behavior. Journal of Community Health, 3, 100-114.

Marwell, G., and Schmidt, DR. (1967). Dimensions of compliance-gaining behavior:
An empirical analysis. Sociometry, 30, 350-364.

McGuire, W. (1989). Theoretical foundations of campaigns. In R. Rice and CK. Atkin
(Eds); Public Communication Campaigns, 2nd ed. (pp. 43-65). Beverly Hills, CA:

Sage.

McLaughlin, M.L., Cody, M.J., and Robey, CS. (1980). Situational influences on the
selection of strategies to resist compliance gaining attempts. Human Communication
Research, 7, 14-36.

106

McNamara, E., Kurth, T. and Hansen, D. (1981). Communication efforts to prevent
wildfires. In R.E. Rice and WP. Paisley (Eds) Public Communication Campaigns.
(pp.143-160). Sage.

Mendelsohn, H. (1973). Some reasons why information campaigns can succeed. Public
Opinion Quarterly, 37: 50-61.

Miller, G.R. (1987a). Media message and information processing in interpersonal
communication: "Generally speaking ..." Information and Behavior.

Miller, G.R. (1987b). Persuasion. In C.R. Berger and SH. Chafee (Eds) Handbook of
Communication Science. pp.446-483. Sage.

Miller, GR. (1986). A neglected connection: Mass media exposure and interpersonal
communication competence. In G. Gumpert and R. Cathcart (Eds) Inter/media:
Interpersonal communication in a media world. (3rd edition, pp. 132-139). New
York: Oxford Press.

Miller, G.R., Boster, F.J., Roloff, M., and Seibold, D. (1977). Compliance-gaining
message strategies: A topology and some findings concerning effects of situational
differences. Communication Monographs, 44, 37-51.

Miller, GR. and Steinberg, M. (1975). Between People. Chicago: Science Research
Associates.

O'Keefe, D. and Delia, J. (1982). Impression formation and message production. In M.E.
Roloff and CR. Berger (Eds), Social cognition and communication (pp. 33-72).
Beverly Hills, CA: Sage.

Petty, R.E., Cacioppo, J .T., and Schumann, D. (1983). Central and peripheral routes to
advertising effectiveness: The moderating role of involvement. Journal of Consumer
Research, 10, 135-146.

Pfefferbaum, B. and Wood, RB. (1994). Self-report study of impulsive and delinquent
behavior in college students. Journal of Adolescent Health.

Robertson, L. (1976). The great seat belt campaign flop. Journal of Communication, 26,
41-46.

Robertson, LS. (1978). Automobile seat-belt use in selected countries, states, and
provinces with and without laws requiring belt use. Accident Analysis and
Prevention, 10, 05-10.

107

Rogers, EM. and Story, D. (1987). Communication campaigns. In C. Berger and S.
Chafee (Eds), Handbook of communication science (pp. 817-846). Newbury Park,
CA: Sage.

Roloff, ME. (1980). Self-awareness and the persuasion process: Do we really know
what we’re doing? In M.R. Roloff and GR. Miller (Eds), Persuasion: New
Directions in Theory and Research (pp.29-66). Beverly Hills: Sage.

Roloff, ME. and Janiszewski, CA. (1989). Overcoming obstacles to interpersonal

compliance: A principle of message construction. Human Communication
Research, 16, 33-61.

Roloff, M.E., Janiszewski, C.A., McGrath, M.A., Burns, CS, and Manrai, LA. (1988).
Acquiring Resources from intimates: When obligation substitutes for persuasion.
Human Communication Research, 14, 364-396.

Schank, RC. and Abelson, RP. (1977). Scripts, Plans, Goals, and Understanding: An
inquiry into Human Knowledge Structures. Hillsdale, NJ: Erlbaum.

Searle, J .R. (1969). Speech Acts. London/New York: Cambridge University Press.

Serle, J .R. (1975). Indirect speech acts. In P. Cole and J .L. Morgan (Eds), Syntax and
Sematics. Volume 3. Speech Acts. New York: Academic Press.

Shapiro, E.G. (1980). Is seeking help from a friend like seeking help fiom a stranger?
Social Psychology Quarterly, 43, 259-263.

Shaw, ME. (1981). Group Dynamics: The Psychology of Small Group Behavior.
McGraw-Hill.

Siebold, D.G., Cantrill, J .G., and Meyers, R.A. (1985). Communication and
interpersonal influence. In M.L. Knapp and GR. Miller (Eds) Handbook of
Interpersonal Communication. Beverly Hills: Sage.

Slovic, P., Fishoff, B., and Lichtenstein, S. (197 8). Accident probabilities and seat-belt
usage - Psychological perspective. Accident Analysis and Prevention, 10, 281-285.

SPSS Release 6.0 (1993). SPSS, Inc. Chicago, IL.

Streff, F.M., Molnar, L.J., and Christoff, C. (1994). Automatic safety belt use in
Michigan: A two-year follow-up. Journal of Safety Research. Winter, 25, 215-220.

Streff, RM. and Molnar, L.J. (1990). Direct observation of Safety belt use in Michigan:
Spring 1990. Technical Report 90-33. The University of Michigan Transportation
Research Institute.

108

Streff, RM. and Molnar, L.J. (1991). Evaluation of the effectiveness of innovative
occupant protection measures: US 31 shoreline area vehicle enforcement program.
US. Department of Transportation.

Vingilis, E. And Stoduto, G. (1994). Psychosocial characteristics of alcohol-involved
and nonalcohol involved seriously injured drivers. Accident Analysis and
Prevention.

Wagenaar, A.C., Molnar, L.J., and Bussinski, KL. (1988). Direct Observation of Seat
Belt Use in Michigan: Spring 1988. Technical Report UMTRI-88-24. The
University of Michigan, Ann Arbor.

Weinstein, N.D., Grubb, PD. and Vantier, J .S. (1986). Incresing Automobile seat-belt
use - An intervention emphasizing risk susceptibility. Journal of Applied
Psychology, 71, 285-290.

Wells, J.K., Williams, A.F., and Lund, AK. (1990). Seat belt use on interstate
highways. American Journal of Public Health, June, 80, 741-742.

Wells, J.K. and Williams, A.F. (1993). Seat belt use in air-bag-equipped cars: 1986-92
models. Journal of Safety Research, 24(2) 73.

Williams, D.L. and Boster, F.J. (1981). The effect of beneficial situational
characteristics, negativism, and dogmatism on compliance-gaining message
selection. Paper presented at the annual convention of the International
Communication Association, Minneapolis, MN.

Williams, A.F., Wells, J .K., and Lund, AK. (1990). Seat belt use in cars with air bags.
American Journal of Public Health, December, 80, 1514-1516.

Williams, A.F., Wells, J .K., Lund, A.K., and Teed, N. (1989). Observed use of
automatic seat belts in 1987 cars. Accident Analysis and Prevention, 21(5) 427-433.

Wittenbraker, J ., Gibbs, BL, and Kahle, LR. (1983). Seat-belt attitudes, habits, and
behaviors - An adaptive amendment to the Fishbein model. Journal of Applied
Social Psychology, 13, 406-421.

 

"’illlllllllilllllllis