|
The Good, the Bad and the Ugly of How the General Public
Views PR Practitioners: The Results of a Hybrid Survey/Experiment
Employing a Nationwide Sample
by
Coy Callison
Assistant Professor
Texas Tech University
School of Mass Communications
Box 43082
Lubbock, TX 79409-3082
(806) 742-3385
[log in to unmask]
Paper submitted to the Public Relations Division at the 2003 annual meeting
of the Association for Education in Journalism and Mass Communication in
Kansas City, Missouri.
The Good, the Bad and the Ugly of How the General Public
Views PR Practitioners: The Results of a Hybrid Survey/Experiment
Employing a Nationwide Sample
Abstract
Public perception of PR practitioners was measured using a telephone survey
with a source manipulation experiment interwoven into a standard opinion
poll. The nationwide sample (N=593) revealed that sources affiliated with
the organization on whose behalf they speak are viewed more negatively than
unaffiliated sources. Additionally, PR practitioners were judged no more
critically than other affiliated sources. Finally, a multi-item measure of
public relations in general demonstrated that perceptions of practitioners
are stable across demographics.
The Good, the Bad and the Ugly of How the General Public
Views PR Practitioners: The Results of a Hybrid Survey/Experiment
Employing a Nationwide Sample
Introduction
Recent public opinion surveys and source credibility experiments have not
painted a flattering picture of public relations. Both have labeled
practitioners as spokespersons that the public highly doubts and have
ranked them behind almost every other information source imaginable,
including pollsters, student activists, and funeral directors. As a result
of these findings, researchers have been quick to relegate practitioners to
working in the wings, out of the public eye, and speaking through other
company mouthpieces. These conclusions, however, have been tempered by
claims of awkward research procedures, poor measurement of constructs, and
research participant pools not accurately representing the overall U.S.
population. This study attempts to speak to these methodological
shortcomings by including in the research population members of all
households in the contiguous United States and employing a hybrid telephone
survey/experiment that allows for a spokesperson manipulation to more
accurately gauge perceptions or public relations practitioners and their
ability to serve as quality information sources. Additionally, general
perceptions of practitioners are garnered using multiple items capable of
being scales.
Literature Review
Credibility is the cornerstone attribute public relations practitioners
must posses if they have any hopes of influencing an audience's perceptions
or attitudes (Budd, 2000; Judd, 1989). From Hovland and Weiss' work (1951)
during the middle of the last century to Priester and Petty's research
(1995) as the millennium ended, scholars have established credibility and
its counterpart, trustworthiness, as they key source and message attributes
necessary in communicating persuasive messages. Without credibility, the
annals of research would suggest that any hopes of effectively garnering
support for an espoused position are slight if present at all (O'Keefe,
2002; Perloff, 1993).
The volumes of research labeling the importance of credibility has not been
disregarded by the public relations industry especially considering the
evidence that exists demonstrating the public's lack of faith in
practitioners being fully forthright (PR deserves its low credibility
marks, 1999.) Perhaps the most revealing of the studies damning the
industry is one conducted by the Public Relations Society of America
itself. In September 1998, PRSA sponsored, along with the help of the
Rockefeller Foundation, a telephone survey where 1,000 respondents were
asked to rate the credibility of sources of information in general on a
four-point scale anchored by "very credible" and "not at all credible"(The
National Credibility Index, 1999). Of the 44 information providers rated by
each respondent, "public relations specialist" finished third from the
bottom, just above "famous entertainer" and "TV or radio talk show host"
but behind "famous athlete," "pollster," and "student activist" among
others. Incidentally, "Supreme Court justice" and "teacher" finished first
and second respectively.
Research published two years later echoed the findings concerning negative
public perception of practitioners. In this study, experimental methods
were employed using teachers, whose demographics closely match those of the
average newspaper reader, as research participants. These participants read
one of two news stories with information presented attributed to either a
public relations practitioner or a company spokesperson whose job title was
not revealed (Callison, 2001). With all text in the articles held constant
across message condition other than identification of the information
source as either a public relations specialist or a more non-descript
company spokesperson, analyses revealed that participants were much more
critical of the PR source and the organization employing the source than
his/her unlabeled counterpart and accompanying organization. More
precisely, the public relations source was perceived as less likely to be
telling the truth, more dishonest, and less trustworthy. The author
ultimately suggested that the negative perception of public practitioners
was due, in part, to a perceived reporting bias (see Eagly, Wood, &
Chaiken, 1978), where audiences see information sources as limited by
situational constraints (i.e. organization affiliation) from taking an
unbiased stance on an issue.
No doubt, the public and public relations practitioners themselves
understand that company spokesperson are, to some extent, paid advocates
for an organization, and as such are mandated to posses a reporting bias to
some extent (Murphy, 2001). It would seem none fault public relations
practitioners for taking their employing organization's side in a
discussion; but it should come as no surprise that spokespersons who are
paid to present their employers in the best possible light are not always
seen as stalwarts of honesty, which often leads to motives being
questioned. In fact, Sallot (2002), using a mixed participant pool of
students and mall shoppers found that perceptions of motives were the key
indicator of how people evaluate public relations and its practitioners.
Through experimental manipulations, four brief news articles were created
detailing a press conference held by a fictitious manufacturer of laser
printers that had launched a recycling program. Manipulations centered on
the motives of the organization involved with one manipulation employing a
"selfish" backgrounder that explained the organization planned a recycling
subsidiary and the program would ensure raw materials. Another
backgrounder, labeled by the researchers as "altruistic," stated that
company executives personally supported conservation and that the company's
sole objective with the program was to make a positive contribution to
society. A "mixed" backgrounder explained that good corporate citizenship
results in favorable publicity and profits while serving social interests.
A final "control" backgrounder detailed no motive. The results suggested
that motives were questioned and participants were critical of information
and its source when a "personal" benefit to the organization seemed to be
behind an activity and its accompanying communications. The author
concluded, as could be expected, that in order to enhance credibility and
trustworthiness, purely altruistic motives should be announced—and other
motives, if present, concealed.
If audiences distrust PR practitioners and their employers when possible
company benefits are assumed to result from their communications (Durham
1997; Sallot, 2002), it would seem that the very nature of the public
relations industry, which involves guarding organization reputation, is at
the root of its credibility problems. How can a practitioner, whose job it
is to position her company in the best possible light, do so when the very
public she is trying to influence doubt her honesty any time she speaks
well of her employer? One remedy that has been suggested is filtering
organizational positive information through third-party sources (Callison,
2001; Lamons, 2002; Murphy, 2001) eliminating any easy attribution of bias.
Research investigating the effects of organizational affiliation on
public perception of spokespersons has indeed revealed that when it comes
to communicating company positive news, any source is a good source as long
as it is not viewed as working solely on behalf of the benefiting
organization. In fact, experimental research employing sources of various
organizational affiliation refuting negative claims has shown that sources
identified as a company spokesperson are viewed, by a student subject pool,
as less credible than unaffiliated sources providing the same refutational
information (Callison & Zillmann, 2002). Equally interesting, sources
labeled as outside experts hired by the organization to investigate
company-negative claims were viewed as equally credible as governmental
agency experts investigating negative claims independently.
It would seem that affiliation must be clear and company ties must be
absolute in order for the perceptions of bias to come into play and taint
attitudes toward communications and communicators. Again, this points to
public relations practitioners working in a no-win situation. Any
organization-positive communication stemming from an organization itself,
and especially from a PR practitioner employed by the organization, would
seem to be unquestionably perceived as lacking credibility. As it stands,
public opinion polls have shown that the public rates practitioners poorly
in credibility indexes, and experimental research has suggested that the
use of the term "public relations" harms an information source.
Experimental research has also suggested that espousing positive news
calls motives into question, especially when the source is seen as employed
by the organization that stands to gain the most from communicative
efforts. Limitations in these studies, however, have to be taken into
account before drawing conclusions about public relations. The Public
Relations Society of America's National Credibility Index used a
single-item measure to gauge credibility, a construct that research has
shown is multidimensional (Berlo, Lemert, & Mertz, 1970; McCroskey, 1966),
and the measure was taken free from any communicative context. No scenario
was provided so as to allow respondents the opportunity to rate information
sources in a "real life" situation. The Callison 2001 study employed only
two possible sources, a public relations specialist and a generic company
spokesperson. If as stated previously company affiliation influences
perceptions, no manipulation was made to test internal sources against
external, presumably less biased, sources. The Sallot 2002 research varied
potential motives behind an announcement, but the company spokesperson was
always labeled as a public relations practitioner. With no variation in
source occupation and affiliation performed, any conclusion about the
effectiveness of public relations practitioners as opposed to other
spokespersons is ungrounded. Finally, the Callison and Zillmann 2002 study
manipulated the affiliation of the information sources employed but it did
not clearly label the internal source as a public relations practitioner.
The public relations position of the company spokespersons used in the
article may have been assumed, but without direct testing of internal
spokesperson titles, no data can be presented to bolster any claim that
certain company sources are more credible than others (Durham, 1997).
The present study attempts to address these limitations by employing a
hybrid of both methodologies used in the previous research and by gathering
data using a "real world" context within which various sources—company
affiliated and not; public relations labeled and not—communicate.
Additionally, the present study employs a multidimensional scale to record
perceptions of public relations practitioners. Lastly, it should be noted
that the present study gauges perceptions held by a sample drawn from the
overall general public and does not rely on subgroups to represent the
typical information consumer.
Hypotheses and Research Question
Credibility of Affiliated Spokespersons Hypotheses
Past research has suggested that perceived bias influences audience
perception of spokesperson credibility and trustworthiness (Eagly, Wood &
Chaiken, 1978). In research focusing on the PR industry, Sallot (2002)
stated, "Audience knowledge of selfish…motives on the part of… advocates
may…lead to greater distrust of public relations" (p. 152). This claim has
been partially supported in experimental investigations of practitioner
credibility, but researchers have yet to distinguish how sources with clear
ties are perceived in contrast to sources clearly not working on behalf of
an organization. In fact, the previously described experiment (Callison &
Zillmann, 2002) that pitted sources ranging in affiliation against each
other blurred affiliation by employing one fictitious source who worked for
an independent research firm that happened to have been hired by an accused
organization. In this case, source motivation is not clear. Would a hired
researcher be loyal to his craft or to the organization that had contracted
his work? To further test how affiliation influences public perception, the
following hypothesis directed the research:
H1a: A source directly affiliated with an organization will be viewed
more negatively than a source not directly affiliated with an organization.
As a result of affiliated sources being viewed more negatively, these
sources should be less persuasive in their communicative efforts. To
determine how messages attributed to company-affiliated sources are
perceived in contrast to messages attributed to non-affiliated sources, the
following hypothesis directed the research:
H1b: Company-positive information stemming from a source directly
affiliated with an accused organization will be less persuasive than the
same information stemming from a source not directly affiliated with the
organization.
Credibility of Internal Spokespersons Hypotheses
When facing crises that potentially could have a profound impact on an
organization and its various communities, crisis management teams often
include public relations officials, legal counselors and members of top
management (Lee, Jares, & Heath, 1999). Past research has suggested that
attaching the words "public relations" to a spokesperson ruins any chance
of a positive audience reaction (Callison, 2001). Previously published
research, however, has not investigated the influence of different internal
organizational affiliations against one another. The present research pits
various company-affiliated sources against each other in an effort to
determine if reporting bias and the resulting negative perceptions are
attributed equally across all internal spokespersons. To determine if
clearly labeled public relations practitioners are viewed more negatively
than other internal, company-affiliated sources, the following hypothesis
directed the research:
H2a: A clearly labeled public relations source will be viewed more
negatively than a source not clearly labeled as a public relations source.
Again as in H1b, negatively viewed sources should also be less persuasive
in their communicative efforts. To determine if public relations sources
are less persuasive than other internal sources, the following hypothesis
directed the research:
H2b: Company-positive information stemming from a public relations source
will be less persuasive than the same information stemming from a source
not clearly labeled as a public relations source.
Perception Scale Research Question
Data that result in claims that public relations is "a tainted term"
(Holmes, 2003, pg. 9) are often taken using single-item measures (National
Credibility Index, 1999). Because subtle differences in perception are
better measured using indexes composed of multiple components, the
following research question guided the research:
RQ1: How will the general public indicate it perceives public relations
practitioners on a multi-item scale and the individual items composing the
scale?
Influence of Demographics on Perceptions Research Question
Additionally, research has suggested that respondent demographics affect
perceptions of credibility across all information sources (National
Credibility Index, 1999). As public relations practitioners are asked to
influence audiences ranging widely in demographic characteristics, the
following research question guided the research:
RQ2: Do demographics such as gender, age, and level of education influence
perceptions of public relations practitioner credibility as measured on a
multi-item scale?
Method
Overview
US households were contacted through a telephone survey center, and
qualifying individuals were asked to respond to a series of questions
concerning perceptions of public relations practitioners and other
professionals who may speak for companies during times of crisis. In
addition to a common set of questions that all respondents were asked, one
of four spokesperson conditions was randomly assigned to respondents, and
measures related to the condition were employed.
Respondents
English speaking persons over the age of 18 were eligible respondents. A
purchased list of 3890 contiguous-US household telephone numbers served as
the initial sample. Of these numbers, 662 were not called, and 597 were
business numbers, fax numbers or numbers that had been disconnected. A
total of 2595 numbers were deemed to represent households, and of these 919
eligible respondents were contacted. Ineligible non-English speaking
respondents answered 36 calls. From the households where an eligible person
answered the telephone, callers completed 293 interviews. As the goal was
to record 500 interviews, two other number lists were generated from the
first. Two randomly-generated digits were added to all numbers from the
initial list resulting in an additional 7780 telephone numbers. For
example, the random digits 5 and 7 were added to all initial numbers
transforming (768) 639-4412 to (768) 639-4417 and (768) 639-4419. These
newly generated numbers had a higher problem incidence but nonetheless
resulted in 262 additional completed interviews from the 2117 numbers
actually used. In these additional calls, 1206 calls resulted in an
eligible person actually answering the telephone. Ultimately, the rate of
eligible, answering households who responded to the survey was 26.1
percent. This cooperation rate seems to be in line with recent survey
research and reflects the declining response rates resulting from increased
use of answering machines, caller identification systems, and the increased
magnitude of telephone solicitation (Massey, O'Connor, & Krotki, 1997).
Procedure
Trained student callers conducted the telephone survey from a 15-station
telephone bank housed at a large Southwestern university. Calls were made
Feb. 17 – 20, Feb. 24 – 27, and March 3 – 5 nightly from 5:30 p.m. CDT to 9
p.m. CDT. States in the Eastern Time Zone were called from 5:30 p.m. CDT to
8 p.m. CDT; states in the Central Time Zone from 6 p.m. CDT to 9 p.m. CST;
states in the Mountain Time Zone from 6:30 p.m. CDT to 9 p.m. CDT; and
states in the Pacific Time Zone from 7:30 p.m. CDT to 9 p.m. CDT. Callers
recorded call disposition noting completed interviews, no answers,
answering machines, refusals, language barrier, disconnects, fax numbers,
and business numbers. Numbers that resulted in answering machines or no
answers were called a maximum of seven times.
Callers introduced themselves to potential respondents by giving their name
and university affiliation, providing a statement concerning the academic
purpose of the call, and promising no attempt to sell goods or services.
Callers then asked to speak to the person at least 18 years of age who
experienced the most recent birthday. This method of selecting respondents
has proven effective at ensuring broad representation as demographics are
equally distributed across the calendar year (Salmon & Nichols, 1983).
Data Collection and Measures
A computer assisted data entry system prompted callers through the
questionnaire, and responses were entered directly into a computer data
file hosted on a main server. The measures were designed to collect two
types of data through a joining of experimental and survey methods. The
questionnaire was composed of three parts: a section on respondent
demographics, a section on attributes of public relations practitioners,
and a section that required respondents to listen to and respond to a
randomly assigned fictitious scenario involving either a company president,
a lawyer, a public relations practitioner or a newspaper journalist denying
charges against a company. The demographics section always concluded the
survey, but the sections on practitioner attributes and spokesperson
perception were rotated between first and second positions. The reasoning
behind this decision centered on an effort to keep responses to one section
of the survey from systematically tainting responses to another. For
example if questions concerning the fictitious scenario always followed the
practitioner attribute section, which asked questions about ethics and
honesty, the possibility exists that respondents would be prompted by
previous questions to respond in a biased manner to questions that follow.
In an attempt to overcome limitations of previous public relations
perception surveys that did not require respondents to view practitioners
in light of an actual communication situation, the study contained a
section, as described above, that varied spokespersons across a fixed
condition. Following the rules of between-subject design, computer software
randomly assigned one of four spokesperson conditions to respondents. In
all cases, callers read a brief scenario to respondents with the
spokesperson mentioned in the scenario varying between a company president,
a lawyer representing the company, a public relations specialist working
for the company, or a newspaper reporter responding to accusations. The
text of the scenario was as follows: "A company named Chapman Enterprises
has been accused of polluting a stream that flows through your town. (The
president of Chapman) (The public relations specialist for Chapman) (A
lawyer representing Chapman) (A reporter for the local newspaper) is
interviewed on television and says that Chapman did not pollute the
stream." The three questions that accompanied the scenario asked if the
spokesperson was telling the truth, if the spokesperson was acting
ethically, and if the respondent believed it would be safe to swim in the
stream. Each question was accompanied by a 11-point scale anchored by
"Definitely No" at 0 and "Definitely Yes" at 10.
For the public relations practitioner perception section of the survey,
eight statements borrowed from past credibility research (Berlo, 1970;
Callison, 2001; McCroskey, 1966) were adapted for the telephone survey
environment. On an 11-point scale anchored by "Strongly Disagree" at 0 and
"Strongly Agree" at 10, callers gathered responses to the statements
"Public relations practitioners are (intelligent), (trustworthy),
(uniformed), (dishonest), (credible sources of information), (unethical),
(virtuous), and (biased sources of information).
Results
Description of the Sample
A cleaning of all data resulted in the responses of 593 individuals being
left for analysis. Of these, 57.7% were female, and 42.3% male.
Additionally, the mean respondent age was 42.7 years (SD = 15.7) while the
modal age was 45 and the median 42. In terms of highest level of education
achieved, 6.9% attended high school but did not graduate, 17.9% possessed a
high school degree; 30.4% attended college but did not earn a degree, 30.5%
graduated from college, and 14.3% held graduate degrees.
Credibility of Affiliated Spokespersons Hypotheses
H1a and H1b predicted that sources closely affiliated with an organization
would be less credible than sources not directly employed by the
organization. For data analysis, the company public relations practitioner,
the company lawyer and the company CEO were combined into a single
"affiliated source" group; the local newspaper journalist served as the
"unaffiliated source." Additionally, education level was collapsed into a
"high school degree or less" category and a "college experience or more"
category, and age was collapsed using a median split where all respondents
42 years old and younger were grouped as were all respondents 43 years old
and older. These groupings allowed for a 2 (affiliation level) X 2
(education level) X 2 (age level) X 2 (gender) full factorial design.
Demographics were included in analyses to help produce a more in-depth
investigation of public perception of affiliation. Because initial
statistical tests uncovered no main effects or interactions involving age
level or gender, they were excluded from subsequent analyses.
Both H1a and H1b were supported. Analyses revealed a main effect for level
of affiliation on perceptions of the source telling the truth (F(1, 589) =
10.06, p
< .01) and on perceptions of the source acting ethically (F(1, 589) = 4.07, p
< .05.) Overall, respondents rated affiliated sources as less likely to be
telling the truth (M = 3.40, SD = 2.78) compared to unaffiliated sources (M
= 4.79, SD = 3.06) and affiliated sources as less ethical (M = 3.76, SD =
3.07) than unaffiliated sources (M = 5.00, SD = 3.15).
Analyses also revealed significant interaction effects involving level of
education and affiliation for both measures. Affiliated sources fared
better in terms of perceptions of telling the truth with lower educated
respondents (M = 3.65, SD = 3.02) than they did with higher educated
respondents (M = 3.32, SD = 2.69). The finding reversed itself, however, on
unaffiliated sources with lower educated respondents rating the
spokesperson more negatively (M = 4.03, SD = 3.18) than did higher educated
respondents (M = 5.01, SD = 3.00), F (1, 589) = 3.97, p < .05. This
interaction effect was consistent across the measure of ethical behavior.
Lower educated respondents rated affiliated sources as more ethical (M =
4.02, SD = 3.36) than did higher educated respondents (M = 3.67, SD =
2.96), but nonaffiliated sources rated more positively with higher educated
respondents (M = 5.33, SD = 3.15) than they did with lower educated
respondents (M = 3.81, SD = 3.57), F (1, 589) = 6.79, p < .01.
H1b predicted that information attributed to affiliated sources would be
less persuasive than information attributed to sources not seen to be
working directly for the organization. While neither source type was able
to achieve scores above a 3 on a 0 to 10 scale with a higher number
indicating more faith placed in the spokesperson, affiliated sources were
less persuasive (M = 1.68, SD = 2.75) than were unaffiliated sources (M =
2.58, SD = 3.27), (F (1, 589) = 5.77, p < .05) at convincing respondents
that the stream a company has been accused of polluting was safe for
swimming. Analyses revealed no significant interactions between education
level and the measure of persuasiveness.
Credibility of Internal Spokespersons Hypotheses
H2a and H2b predicted that public relations practitioners would be less
credible than other internal, affiliated sources. As outlined in the
discussion of results relating to H1a and H1b, education level, age level,
and gender were included in initial analysis. Mimicking the previous
findings, initial statistical investigations revealed no simple main
effects or interactions involving age level or gender, and the two factors
were excluded from subsequent analyses.
Neither H2a nor H2b was supported. Respondents rated a company's public
relations practitioner (M = 3.60, SD = 2.71), lawyer (M = 3.05, SD = 2.73),
and president (M = 3.65, SD = 2.87) as equally unlikely to be telling the
truth as compared to a local newspaper reporter (M = 4.79, SD = 3.06, F (3,
585) = 3.98, p < .01 (post hoc tests conducted using Student-Newman-Keuls).
Incidentally, analyses also revealed a significant interaction effect
between job title and education level, F (3, 585) = 4.53, p < .01.
Respondents in the low education group rated PR practitioners and lawyers
more favorably (M = 3.82, SD = 2.93; M = 4.00 , SD = 3.13 respectively)
than did their more educated counterparts (M = 3.53, SD = 2.64; M = 2.68,
SD = 2.47 respectively). The other sources, a company president and a local
newspaper reporter, fared better with higher educated respondents (M =
3.87, SD = 2.84; M = 5.01 , SD = 3.00 respectively) than they did with the
lower educated respondents (M = 2.94, SD = 2.90; M = 4.03, SD = 3.12
respectively).
On the question of acting ethically, respondents rated a company's public
relations practitioner (M = 3.97, SD = 3.15), lawyer (M = 3.78, SD = 3.21),
and president (M = 3.52, SD = 2.80) as equally ethical as a local newspaper
reporter (M = 4.99, SD = 3.29, F (3, 585) = 2.15, p = .09. Analyses did
reveal, however, a significant interaction effect between job title and
education level on the ethical behavior measure, F (3, 585) = 2.84, p <
.05. These findings mirrored those on the previous measure of "telling the
truth." Respondents in the low education group rated PR practitioners and
lawyers more favorably (M = 4.24, SD = 3.86; M = 4.35 , SD = 3.24
respectively) than did their more educated counterparts (M = 3.89, SD =
2.91; M = 3.56, SD = 3.19 respectively). Also as earlier, company president
and local newspaper reporter fared better with higher educated respondents
(M = 3.59, SD = 2.77; M = 5.33 , SD = 3.15 respectively) than they did with
the lower educated respondents (M = 3.30, SD = 2.95; M = 3.81, SD = 3.57
respectively).
H2b predicted that information attributed to public relations sources
would be less persuasive than information attributed to other source types.
Data analyses revealed that the company's public relations practitioner (M
= 1.87, SD = 3.07), lawyer (M = 1.60, SD = 2.52), and president (M = 1.61,
SD = 2.69) were equally unpersuasive in communicating the safety of
swimming in the stream when compared to the local newspaper reporter (M =
2.58, SD = 3.27), F (3, 585) = 2.64, p < .05 (post hoc tests conducted
using Student-Newman-Keuls). No significant interaction effects were revealed.
Perception Scale Research Question
The descriptive statistics associated with the eight individual items
employed to measure public relations practitioner credibility are provided
in Table 1. As can be seen in the Overall Mean column of the table, scores
grouped around the measure mean (5). In fact, taking into account that a
lower score is more positive on the measures of uninformed, dishonest,
unethical, and biased, public relations practitioners scored better than
the midpoint on all items other than virtuous and biased. Practitioners
received the highest score on the intelligent measure and the lowest on the
biased measure.
In order to group the single-item measures into a more complex multi-item
scales, the eight individual ratings were subjected to a principal
component analysis without rotation of components. With the ratings of
uninformed, dishonest, unethical, and biased reverse coded, three factors
emerged accounting for 65.0 % of the variance (30.5%, 21.1%, and 13.4 %
respectively).
The first factor, labeled truthful, showed high loadings on three traits,
namely trustworthy (.79), credible (.73), and virtuous (.61). A correlation
analysis revealed that the traits showed a high degree of inter-item
consistency (_ =.74) indicating that the items were measuring a similar,
single trait. This warranted the construction of a composite measure of
truthful by averaging the ratings across all items.
The second factor, labeled moral, showed high loadings on the three traits
– reverse-coded dishonest (.57), reverse-coded unethical (.63), and
reverse-coded biased (.541). Averaging the ratings on the three items
created a composite measure of moral. Inter-item consistency proved
sufficient for this combination (_ =.62).
The third factor, labeled smart, showed high loadings on intelligent (.50)
and the reverse-coded uninformed (.68). Analysis revealed a low level of
inter-item consistency between the two items (_ =.24). Due to this low
value, no composite score could be created for the third factor.
To answer the question of how the general public perceives PR
practitioners as indicated on a multi-item scale as posited in RQ1, simple
means of the indexes were calculated. The overall mean score on the
truthful measure was 5.02 (SD = 2.05), slightly above the scale midpoint.
The overall mean score on the moral measure was 4.64 (SD = 2.02), slightly
below the scale midpoint.
Influence of Demographics on Perceptions Research Question
The first research question allowed for the creation of a scale that could
be employed to measure public perceptions of PR practitioners. With a scale
in place, the influence of demographic characteristics on views of public
relations could be examined. In fact, researchers surveying the general
public concerning public relations in the past have stated that
demographics such as age and education influence perceptions of credibility
in sources. (The National Credibility Index, 1999). These researchers,
however, did not statistically validate their claims.
Data analysis in the present study revealed that perceptions of
practitioners were fairly stable across demographics. The only significant
difference within gender, age or educational level groups appeared on the
single-item measure of virtuous with men viewing PR practitioners as less
virtuous than women viewed them, t(591) = 2.02, p < .05. All means are
provided in Table 1. Approaching significance was the difference between
the genders on practitioners being intelligent, trustworthy, and the
single-item measure of credible, t(591) = 1.82, p = .07, t(591) = 1.86, p =
.06, and t(591) = 1.80, p < .07 respectively. In each case, men were more
critical of practitioners than were women. Additionally, the difference
between low educated respondents and high educated respondents on
perceptions of practitioner bias approached significance, t(591) = 1.95, p
= .052, with those more highly educated seeing practitioners as more biased
as compared to the less educated counterparts. No differences approached
significance across the individual measures between young and old.
Employing the composite scores created through factor analysis, it was
revealed that males view practitioners as less truthful (M = 4.80, SD =
1.95) than females view them (M = 5.19, SD = 2.11), t(591) = 2.32, p < .05.
There was no difference between the genders on perceptions of practitioners
being moral with men indicating a mean score of 4.56 (SD = 1.86) and women
indicating a mean score of 4.69 (SD = 2.14), t(591) = 0.79, p = .43.
Age had no impact on perceptions of practitioners being truthful. Those 42
years old and younger rated practitioners with a mean score of 5.03 (SD =
1.94) while those older than 42 gave practitioners a mean rating of 5.02
(SD = 2.11), t(566) = 0.37, p = .97. Likewise, age did not influence
perceptions of practitioners being moral as the younger group's mean score
was 4.64 (SD = 1.95) and the more senior's was 4.61 (SD = 2.06), t(566) =
0.19, p = .85.
Finally, high and low educated respondents viewed practitioners as equally
truthful (M = 4.98, SD = 1.95; M = 5.15, SD = 2.31 respectively), t(591) =
.85, p = .39. Likewise, the more educated did not attribute morality to
practitioners any more than the less educated did (M = 4.62, SD = 1.97; M =
4.69, SD = 2.17 respectively), t(591) = .37, p = .71.
Discussion
The goal of this study was to determine how the general public views
public relations practitioners. This in itself does not make this study
noteworthy. Numerous research projects have been undertaken with that goal
in mind. What does set this particular study apart, however, is that it
attempted to measure perceptions of public relations through a combination
survey/experiment where research participants not only could rate PR
practitioners in a communicative setting and a real-world context, but that
they could also register their opinions using a multiple-item measure in an
effort to gauge the complete construct of credibility. On this level, the
study has shown that experiment and survey can be married and that
composites provide richer data than single items. Most important, however,
is the fact that a more complete understanding of perceptions of
practitioners results from the study—and the news for the PR industry is
some good, some bad and some ugly.
Sources directly affiliated with an organization on whose behalf they
speak were, as predicted, more negatively perceived and less effective than
an independent source. This finding mimics the findings of other
credibility research (Callison & Zillmann, 2002 in particular) but paints
an even direr picture for public relations practitioners. Because ethically
and practically a company spokesperson cannot disavow any connection to the
company represented, it seems that any attempt to communicate on an
organization's behalf from the organization's own pulpit is doomed.
Callison and Zillmann 2002 suggested that a company being accused of
negative actions would do best to hope for an independent agent to take its
side in public discourse. The present study supports that this method of
crisis communication would, in fact, be most effective. What has yet to be
uncovered is how an organization should pass its time in a firestorm
waiting for an avenger to show up on the doorstep.
In particular, data suggested that affiliated sources are viewed as less
likely to be telling the truth and less ethical than their independent
counterparts. The finding in terms of telling the truth seems explicable in
light of the research suggesting that company motives are often doubted,
and audiences often doubt organizational intentions (Sallot, 2002). The
ethics finding, however, is more troubling. Not detailing all motivations
could be viewed, probably wrongly so, as a part of a savvy media plan;
there is no way though to positively spin perceptions of unethical behavior.
Equally interesting is how level of education influenced perceptions of
sources. College-educated respondents were more critical of affiliated
sources and, on the other hand, more supportive of independent sources than
their less educated counterparts. It would seem that this educated public
is especially wary of source affiliation. The data does not offer an
explanation for this effect or detail how the information should be used
when planning communications, but assuming that a college-educated
demographic serves as an important public, problems surrounding source
affiliation may be even more pronounced than detailed here.
Finally, in terms of affiliation, sources viewed as married to the
organization were less persuasive than the outside source at convincing the
public that accusations against the organization were false. Past
experimental research (Callison 2001; Callison & Zillmann, 2002) has not
linked negative perceptions of a spokesperson to lack of persuasive
ability. The present study not only suggests that company spokespersons are
saddled with a negative reputation, but it also suggests that their
communications are not effective, providing even more convincing evidence
that their use should be limited if possible.
It has been posited that a quality spokesperson can be selected from
various personnel, including PR practitioners, company lawyers, CEOs and
researchers (Budd, 2000; Callison, 2001; Durham, 1997; Reber, Cropp, &
Cameron, 2001; Wisenblit, 1989). This study was particularly interested in
determining how public relations sources rated in comparison to other
internal sources. The findings in this regard could be offered up in terms
of good news/bad news. The good news is that public relations were rated no
more negatively or less persuasive than other affiliated sources; the bad
news is that all internal sources fared poorly in comparison to an
independent source. While it was not unexpected that the newspaper reporter
source employed in the study would be seen most positively (see Budd, 2000,
Lamons, 2003), it went against predictions that all those affiliated with
the organization would rank so similarly, especially considering
speculation that CEOs have taken a credibility hit due to recent corporate
scandals (Dobrow, 2002) and the fact that lawyers seem to generally rate at
the bottom of America's most trusted professionals (Trust in Priests,
2002). The good news is public relations practitioners are equally capable
as other known company sources of representing an organization.
Nonetheless, it is doubtful that PR practitioners want to be thought of as
"the information source that is no worse than whatever else you could push
behind a microphone."
In parceling out how demographics influenced perceptions of internal
sources, data revealed that better educated respondents were more critical
of the public relations practitioner and company lawyer speaking on an
organization's behalf than they were of the CEO or journalist. As before,
this interaction effect reveals that the consumer group that may present
the biggest obstacle in terms of communicating corporate news effectively
is the highly educated demographic, one that is surely targeted in many
communication campaigns.
The second component of the present study centered on measuring overall
public perception of public relations practitioners independent of the
experimental setting. Specifically, the use of multiple items to measure
credibility that could later be combined into a composite index drove the
research. On the single items, perceptions of public relations sources
ranged around neutral to slightly positive. The only ratings that dipped
into the negative tail of the scale were for biased and virtuous. The poor
rating on biased comes as no surprise as research has suggested that
audiences regularly question the motives behind organizational
communications (Sallot, 2002) and that public relations practitioners are,
in the end, employed to help an organization maintain a positive
perception. The poor rating on virtuous, like the previous finding on
organizational sources being unethical, presents more of a point of concern
for the industry. Again, this study offers no remedy for these negative
perceptions, but perhaps by outlining their existence along specific
attributes, steps can be taken to correct misconceptions or improper
actions by practitioners.
In gauging overall perceptions of practitioners, the fact that two
composite measures of credibility could be created indicates that single
item measures of credibility may be invalid. In particular, credible
grouping with the measure for trustworthy and virtuous, but not with
intelligent, the reverse-coded dishonest, the reverse-coded unethical, the
reverse-coded biased, and the reverse-coded uninformed demonstrate the
complexity of the credibility construct. Any attempt to portray public
perception of PR practitioners must note that simply asking whether a
source is credible or not does not reveal the subtle attributes that
influence source effectiveness.
Finally, the comparison of perceptions across demographic groups revealed
the consistency of the attitude toward PR practitioners. Comparing
respondents of different genders, respondents of different age groups and
respondents of different education levels, the only difference on the
individual items that emerged was the males viewed practitioners as less
virtuous than females viewed them. On the composite measures of truthful
and moral, the only difference that surfaced was that males perceived
practitioners as less truthful than did their female counterparts. While
these differences should not be overlooked, it should be noted that of the
30 comparisons made, only two demonstrated significant discrepancy between
perceptions. This finding indicates that no one group, of those
investigated, is any more critical of the profession than the others.
However, the fact that education level seemed to play such an important
role in measures taken during the experimental portion of the study
suggests two conclusions: 1) attitudes taken using a real-world context
differ from attitudes taken using simple recall and recount measures, and
2) more research is needed into how different demographics react to
organizational messages and their sources.
In conclusion, the present study revealed the good (PR perceptions are
fairly stable across demographics), the bad (PR practitioners are no better
than other company sources), and the ugly (All company affiliated sources
are negatively viewed) as it relates to public perception of the industry
and those working within it. No doubt other good and bad can be found in
the results. Overall, however, like other similar studies this one does not
place practitioners in an overly positive light. Also like other similar
studies, this one does little to offer solutions to the problem.
Ultimately, the value in this study may lie in the fact that it confirms
the findings of past experimental research in a much broader population,
and it outlines the importance of multi-item measures in future surveys of
practitioner credibility.
References
Berlo, D. K., Lemert, J. B., & Mertz, R. J. (1970). Dimensions for
evaluating the acceptability of message sources. Public Opinion Quarterly,
33, 563-576.
Budd, J. F., Jr. (2000). The incredible credibility dilemma. Public
Relations Quarterly, 45 (3), 22-26.
Callison, C. (2001). Do PR practitioners have a PR problem?: The effect of
associating a source with public relations and client-negative news on
audience perception of credibility. Journal of Public Relations Research,
13, 219-234.
Callison, C., & Zillmann, D. (2002). Company affiliation and communicative
ability: How perceived organizational ties influence source persuasiveness
in a company-negative news environment. Journal of Public Relations
Research, 14, 85-102.
Dobrow, L. (2002, September 30). End of the CEO cult? Corporate PR heads
rethink how much a CEO should personify a company. PR Week, 5 (38), 17.
Durham, D. (1997). How to biggest the bang out of your next spokesperson
campaign. Public Relations Quarterly, 42 (2), 38-41.
Eagly, A. H., Wood, W., & Chaiken, S. (1978). Causal inferences about
communicators and their effect on public opinion. Journal of Personality
and Social Psychology, 36, 424-435.
Hovland, C. I., & Weiss, W. (1951). The influence of source credibility on
communication effectiveness. Public Opinion Quarterly, 15, 633-650.
Judd, L. R. (1989). Credibility, public relations and social
responsibility. Public Relations Review, 15 (2), 34-40.
Lamons, B. (2002, September 16). Advertising drops ball, new champ is PR.
Marketing News, 8.
Lee, J., Jares, S. M., Heath, R. L. (1999). Decision-making encroachment
and cooperative relationships between public relations and legal counselors
in the management of organizational crisis. Journal of Public Relations
Research, 11, 243-270.
Massey, J. T., O'Connor, D, & Krotki, K. (1997). Response rates in random
digit dialing (RDD) telephone surveys. Proceedings of the American
Statistical Association Survey Research Method Section, 707-712.
McCroskey, J. C. (1966). Scales for the measurement of ethos. Speech
Monographs, 33, 65-72.
Murphy, P. (2001). Affiliation bias and expert disagreement in framing the
nicotine addiction debate. Science, Technology, & Human Values, 26, 278-299.
The National Credibility Index. (1999).
http://www.prsa.org/_about/prsafoundation/nciIndex.asp.
O'Keefe, D. J. (2002). Persuasion: Theory and research (2nd ed.). Newbury
Park, CA: Sage.
Perloff, R. M. (1993). The dynamics of persuasion. Hillsdale, NJ: Lawrence
Erlbaum Associates.
PR deserves its low credibility marks. (1999, August). O'Dwyer's PR Service
Report, 6.
Priester, J. R., & Petty, R. E. (1995). Source attributions and persuasion:
Perceived honesty as a determinant of message scrutiny. Personality and
Social Psychology Bulletin, 21, 637-654.
Reber, B. H, Cropp, F., Cameron, G. T. (2001). Mythic battles: Examining
the lawyer-public relations counselor dynamic. Journal of Public Relations
Research, 13, 187-218.
Sallot, L. M. (2002). What the public thinks about public relations: An
impression management experiment. Journalism and Mass Communication
Quarterly, 79, 150-171.
Salmon, C. T., & Nichols, J. S. (1983). The next-birthday method of
respondent selection. Public Opinion Quarterly, 47, 270-276.
Trust in priests and clergy falls 26 points in twelve months. (2002,
November 27). www.harrisinteractive.com/harris_poll/index.asp?PID=342.
Wisenblit, J. (1989). Crisis management planning among U.S. corporations:
Empirical evidence and a proposed framework. Advanced Management Journal,
54, 31-41.
Table 1. Mean Responses to Individual Items Employed to Measure
Practitioner Credibility
Gender Age Education Level
Overall Male Female 42 & 43 & H.S. Grad. Coll. Exp.
Item Mean Under Over & Less & More
Intelligent 7.10 6.91a 7.25b 7.22a 7.08a 7.02a 7.13a
Trustworthy 5.04 4.81a 5.20b 5.06a 4.98a 5.04a 5.03a
Uninformed* 4.21 4.38a 4.08a 4.10a 4.34a 4.56a 4.13a
Dishonest* 4.67 4.74a 4.62a 4.67a 4.68a 4.80a 4.63a
Credible 5.29 5.05a 5.46b 5.23a 5.35a 5.57a 5.19a
Unethical* 4.65 4.69a 4.61a 4.65a 4.65a 4.76a 4.61a
Virtuous 4.75 4.52a* 4.91b* 4.79a 4.73a 4.83a 4.72a
Biased* 6.77 6.88a 6.68a 6.75a 6.83a 6.37a 6.90b
Intelligent (2.56) (2.08) (2.37) (2.09) (2.32) (2.71) (2.09)
Trustworthy (2.51) (2.31) (2.63) (2.42) (2.56) (2.82) (2.40)
Uninformed* (2.95) (2.85) (3.03) (3.00) (2.94) (3.14) (2.89)
Dishonest* (2.59) (2.48) (2.66) (2.55) (2.63) (2.75) (2.53)
Credible (2.71) (2.70) (2.71) (2.65) (2.80) (2.98) (2.62)
Unethical* (2.61) (2.48) (2.70) (2.54) (2.67) (2.94) (2.49)
Virtuous (2.33) (2.23) (2.40) (2.14) (2.47) (2.59) (2.24)
Biased* (2.88) (2.92) (2.85) (2.84) (2.91) (3.18) (2.76)
Sample Size 593 251 342 288 280 147 446
Note. All comparisons are horizontal within demographic grouping. Means are
compared using t tests. Means within demographic grouping not sharing a
subscript are significantly different at p < .10; superscripts marked with
an asterisk are significantly different at p < .05.
Items marked with an asterisk denote an item reverse coded for factor analysis.
Standard deviations are within parentheses.
All items were measured on an 11-point scale anchored by Strongly Disagree
at 0 and Strongly Agree at 10. Items accompanied the following question:
Public relations practitioners are….
|
