PREFERENCE FOR HOMOPHILY, CREDIBILITY, AND THE WORD-OF-MOUTH
PROCESS
By
Caitlin Michelle McLaughlin

A DISSERTATION
Submitted To
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Communication Arts and Sciences – Media and Information Studies
2012

ABSTRACT
PREFERENCE FOR HOMOPHILY, CREDIBILITY, AND THE WORD-OF-MOUTH
PROCESS
By
Caitlin Michelle McLaughlin

This paper studied the influence of source credibility on the word-of-mouth process, and
the effect of involvement on this relationship. Source credibility’s three components: expertise,
trustworthiness, and likeability, were found to have different roles such that sources were
consistently trustworthy and likeable, while expertise was only considered as involvement in the
purchase increased. The utilization of source cues in highly involving purchase situations is an
expansion of the traditional elaboration likelihood model research. Homophily was explored as a
component of source credibility, such that similar sources were considered to be more likeable.
Homophily played a significant role in word-of-mouth source selection due to its link to source
likeability. A new scale for homophily was developed based on participants’ own words and
perceptions, in contrast to most homophily research that studied homophily based on
characteristics that researchers believed would be relevant to the homophily process. By
allowing the participants to describe homophily in their own words, a more accurate picture of
how perceived homophily influences actions and attitudes can be drawn than when researchers
utilize their own language and perceptions to form the measurement scale.

Copyright by
CAITLIN MICHELLE MCLAUGHLIN
2012

ACKNOWLEDGEMENTS
Thank you to my two advisors, Keith and Bob – you were both extraordinarily supportive and
helpful throughout this process, and I truly appreciate it. Thank you to my committee members
Lucinda and Bruce – your participation on this committee and your contributions to the rest of
my time at Michigan State are truly appreciated. Thank you to my family and friends for getting
me through the rough spots, and for always believing in me. I wouldn’t have made it without
each and every one of you.

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................................ vii
LIST OF FIGURES ..................................................................................................................... viii
INTRODUCTION .......................................................................................................................... 1
LITERATURE REVIEW ............................................................................................................... 5
Word-of-Mouth Advertising ....................................................................................................... 5
Homophily ................................................................................................................................ 11
Source Credibility ..................................................................................................................... 20
Elaboration Likelihood Model .................................................................................................. 26
RESEARCH QUESTIONS AND HYPOTHESES ...................................................................... 29
METHODOLOGY ....................................................................................................................... 37
Study 1 ...................................................................................................................................... 38
Overview ............................................................................................................................... 38
Procedure .............................................................................................................................. 38
Study Two ................................................................................................................................. 40
Overview ............................................................................................................................... 40
Measurement ......................................................................................................................... 41
RESULTS ..................................................................................................................................... 44
Study One Phase One................................................................................................................ 44
Participants ............................................................................................................................ 44
Research Question 1 ............................................................................................................. 44
Research Question 2 ............................................................................................................. 45
Research Question 3 ............................................................................................................. 46
Research Question 4 ............................................................................................................. 48
Study One Phase Two ............................................................................................................... 53
Participants ............................................................................................................................ 53
Scale Development ............................................................................................................... 55
Study One Phase Three ............................................................................................................. 58
Participants ............................................................................................................................ 58
Scale Development ............................................................................................................... 60
Study Two ................................................................................................................................. 63
Participants ............................................................................................................................ 63
Hypothesis 1.......................................................................................................................... 65
Hypothesis 2.......................................................................................................................... 66
Hypothesis 3.......................................................................................................................... 66
Hypothesis 4.......................................................................................................................... 66
Hypothesis 5.......................................................................................................................... 66

v

Hypothesis 6.......................................................................................................................... 68
Hypothesis 7.......................................................................................................................... 68
Hypothesis 8.......................................................................................................................... 68
Hypothesis 9.......................................................................................................................... 69
Hypothesis 10........................................................................................................................ 69
DISCUSSION ............................................................................................................................... 71
Study One.................................................................................................................................. 71
Study Two ................................................................................................................................. 76
CONCLUSION ............................................................................................................................. 80
APPENDICES .............................................................................................................................. 88
Appendix A: Study 1 Phase 1 Interview Protocol and Consent Form ...................................... 89
Appendix B: Study 1 Phase 2 Questionnaire and Consent Form ............................................. 93
Appendix C: Study 1 Phase 3 Questionnaire and Consent Form ........................................... 107
Appendix D: Study 2 Questionnaire and Consent Form ........................................................ 122
Appendix E: Figure 1 - Hypothesized Path Models ............................................................... 138
Appendix F: Figure 2 - Distribution of Purchase Involvement in Watches ........................... 139
Appendix G: Figure 3 - Study 1 Phase 3 Path Diagram ......................................................... 140
BIBLIOGRAPHY ....................................................................................................................... 142

vi

LIST OF TABLES

Table 1: Percentage of People Who Utilize Word-of-Mouth Advertising Versus
Traditional Advertising When Making Purchase Decisions ……….……..……………. 7
Table 2: Study 1 Phase 2 Participant Demographics…………………………………………… 53
Table 3: Study 1 Phase 2 Item Factor Loadings, Means, and Standard Deviations……............. 57
Table 4: Study 1 Phase 3 Participant Demographics…………………………………................ 59
Table 5: Study 1 Phase 3 Item Path Estimates, Means, and Standard Deviations ……............... 61
Table 6: Study 1 Phase 3 Error Covariances …………………………………………………... 62
Table 7: Study 2 Participant Demographics……………………………………………………. 64
Table 8: H5, H7, H9 Regression Coefficients: Predictors of Likelihood of Following
Recommendation ………………………………………………………….………….. 67
Table 9: H6, H8, H10 Regression Coefficients: Predictors of Likelihood of Engaging
in Additional Search ………………………………………………………................. 67
Table 10: Side-by-Side Comparison of McCroskey, Richmond, and Daly’s (1975)
Homophily Scale and Current Study Homophily Scale ……………………………... 71

vii

LIST OF FIGURES

Figure 1: Hypothesized Path Model…………………………………..……………. 141
Figure 2: Distribution of Purchase Involvement in Watches ……………………… 142
Figure 3: Study 1 Phase 3 Path Diagram ………………………………………….. 143

viii

INTRODUCTION

Word-of-mouth advertising, defined as the process of communicating information about
a product to another person without commercial (e.g., financial) reasons to do so (Arndt, 1967)
was a very popular topic in the 1950’s and 1960’s. However, due to measurement issues, wordof-mouth was abandoned by most academic researchers until recently. Since the advent of Web
2.0, which includes online brand communities, social networking sites, and blogs, word-ofmouth communication is more easily measurable, and a topic of interest for researchers once
again. Furthermore, these online forums and review sites have put power in the hands of
consumers that is unprecedented by allowing them to reach a much larger audience than ever
before (Vaynerchuk, 2011).
Source credibility literature suggests that individuals find similar, more attractive sources
more credible (e.g., Britt, 1978; Gotlieb & Sarel, 1991; Jones, Moore, Stanaland, & Rosalind,
1998; Lichtenstein & Bearden, 1989; Patzer, 1983). Although word-of-mouth sources tend to be
deemed more trustworthy than traditional marketing sources of information (Schiffman, Kanuk,
& Wisenblit 2010; Solomon, Zaichkowsky, & Polegato, 2002), expertise and similarity /
likeability are more variable. Thus, these are two variables that may be more or less important to
word-of-mouth source selection depending upon the situation.
Furthermore, research has demonstrated that individuals preferred to receive advice and
recommendations from others who were similar to them (e.g., Brown & Reingen, 1987; Gilly,
Graham, Wolfinbarger, & Yale, 1998; Wang, Walther, Pingree, & Hawkins, 2008; Wright, 2000).
This is likely a component of source likeability, as homophily (similarity) has been linked to

1

attractiveness in several situations (e.g., Allgeier & Byrne, 1973; Byrne, 1961a; Duck, 1975;
Eyal & Rubin, 2003; Singh, 1973; Steele & McGlynn, 1979; Tan & Singh, 1995).
While research has shown that individuals demonstrate homophilous preferences when
forming friendships and choosing information sources, the characteristics that researchers
measure similarity upon have tended to be based upon the researcher’s hypotheses rather than
what participants actually consider when judging similarity. In other words, rather than asking an
individual which characteristics they prefer their friends to share with them, most research has
simply chosen to measure the similarity of friends or sources based on demographic
characteristics (e.g., Crosnoe, Frank, & Mueller, 2008; Huston & Levinger, 1978; Reagans, 2005;
Weare, Musso, & Jun, 2009) or based upon some other psychographic variable of interest to the
researcher (e.g., Allgeier & Byrne, 1973; Byrne & Griffitt, 1966; Geen & Stonner, 1974; Hanish,
Martin, Fabes, Leonard, & Herzog, 2005; Neimeyer & Mitchell, 1988). McCroskey, Richmond,
and Daly (1975) created a general homophily scale, but still failed to elicit open-ended responses
from individuals regarding which characteristics were important for them when they were
making similarity judgments, as recommended by grounded research (e.g., Charmaz, 2006;
Harmon & Boeringer, 2004; Morse, 1994). Instead, these researchers simply created a large set
of statements based on previous research, which was limited based on what was readily
measurable. To rectify these measurement issues, one of the goals of this paper is to generate a
homophily scale based on the characteristics that participants generate themselves in a series of
qualitative interviews. These interviews will also attempt to better understand the reasons why
individuals demonstrate this preference to spend time with and heed advice from similar others.
The elaboration likelihood model suggests that involvement increases motivation to
elaborate on a persuasion episode, and thus decreases the importance of source attractiveness

2

cues (Petty, Cacioppo, & Goldman, 1981). However, Petty and Cacioppo (1984) found that
source cues may be utilized as a central cue when involvement and motivation to process were
high while traditionally, the elaboration likelihood model suggested that source credibility was a
peripheral cue to aid in the formation of attitudes and intentions when involvement was low
(Petty, Cacioppo, & Goldman, 1981). Thus, source credibility should be further explored to
determine the relationship between source credibility factors (trustworthiness, expertise, and
likeability) and involvement. Although certain facets of source credibility may have a negligible
effect in high involvement situations, other source cues may be important.
This research paper is focused on a better understanding of the word-of-mouth process
and the role of source credibility and preference for homophily in the selection of word-of-mouth
sources. As such, the paper begins with a review of the literature regarding word-of-mouth and
the resurgence of this literature in the digital age. A preference for homophily has been
demonstrated many times in the persuasion (e.g., Britt, 1978; Gotlieb & Sarel, 1991; Jones,
Moore, Stanaland, & Rosalind, 1998; Lichtenstein & Bearden, 1989; Patzer, 1983) and
relationship formation literature (e.g., Allgeier & Byrne, 1973; Byrne, 1961a; Duck, 1975; Eyal
& Rubin, 2003; Singh, 1973; Steele & McGlynn, 1979; Tan & Singh, 1995), but it is still not
understood when this preference is present and why it occurs. Measurement and manipulation of
this construct have been inconsistent and, because of this, its effect on interpersonal relationships
and persuasion has been varied in the research (e.g., Dohanos, 2003; Glascock & Ruggiero, 2006;
Morris, Gorham, Cohen, & Huffman, 1996; Rocca & McCroskey, 1999; Wright, 2004). A
review of the current homophily literature, including the literature that attempts to understand
why this tendency for homophily is so prevalent, is considered next. Preference for homophily in
sources and source credibility are inextricably linked, due to the consistent link between

3

homophily and the likeability facet of source credibility (e.g., Gotlieb & Sarel, 1991; Jones,
Moore, Stanaland, & Rosalind, 1998; Lichtenstein & Bearden, 1989; Patzer, 1983; Pornpitakpan,
2003). As such, source credibility is discussed next, and its role in the word-of-mouth process.
Finally, the elaboration likelihood model predicted that cues such as source credibility would be
utilized in low involvement situations when central processing of more detailed cues was
undesirable or not possible (Petty, Cacioppo, & Goldman, 1981). However, the literature
suggests that this relationship is more complex than originally thought (e.g., Eisend, 2007; Petty
& Cacioppo, 1984; Wu & Schaffer, 1987). A discussion of these conflicting findings and their
implications for source credibility and involvement’s role in word-of-mouth follows. After this
review of the literature a series of research questions aimed at a better understanding of the
homophily phenomenon are proposed, in addition to hypotheses concerning the role of
homophily and source credibility in the word-of-mouth process at varying levels of involvement.

4

LITERATURE REVIEW

Word-of-Mouth Advertising
Bloch, Sherrell, and Ridgway (1986) discussed the external information search process,
which included word-of-mouth solicitation and consideration, for individuals seeking
information before and after a purchase. For both pre and post-purchase information searches,
the extent of the information search process was determined by involvement in the purchase
process, the market environment (e.g., the availability of information, how many products were
on the market), and situational factors (e.g., time constraints). Involvement has been
demonstrated to increase the external search process in multiple studies (e.g., Arndt, 1966;
Katona & Mueller, 1954).
Arndt (1967) defined word-of-mouth advertising as “oral, person-to-person
communication between a receiver and a communicator whom the receiver perceives as noncommercial, concerning a brand, a product, or a service” (p. 3). Word-of-mouth advertising as a
topic of discussion for scholars and laypeople enjoyed a certain amount of popularity in the
1950’s and 1960’s, with newspapers reporting the proliferation of ‘whisper campaigns’ and ‘slur
campaigns’ in which companies would hire a private firm to promote their own brand and
disparage their competitors’ brands (Arndt, 1967). Jacobson (1958) noted that there were several
companies available to supply these whisper campaigns, involving disparagement of competitors
and promotion of the company’s own brand.
Several studies have demonstrated the effectiveness of word-of-mouth advertising. For
example, Wilke (1934) found that in-person speeches resulted in the most drastic changes in
attitudes when compared to radio and printed presentations. Eisenstadt (1952) found that while
5

impersonal contacts converted 10% of immigrants to Israel, word-of-mouth communications
converted 65%. Mueller (1958) found that more than 50% of respondents had turned to word-ofmouth sources for information regarding durable goods purchases, while Rich (1963) found that
41%-48% of respondents utilized store ads to find out information before making a purchase
versus the 57%-62% who utilized word-of-mouth sources.
Arndt (1967) noted repeatedly, however, that while researchers discovered that
individuals preferred to obtain information from word-of-mouth sources, very little was known
about the content of these communications. This was a major issue with traditional word-ofmouth studies that could be addressed with the advent of electronic word-of-mouth. Thus, a new
set of word-of-mouth studies began to appear in the literature (e.g., Beck, 2007; Harrison-Walker,
2001; Kempf & Palan, 2006; Wangenheim & Bayon, 2004; Zhang & Daugherty, 2009) and a
new definition of word-of-mouth needed to emerge in order to address the complexities of oneto-many communication that was now possible. Hennig-Thurau, Gwinner, Walsh, and Gremler
(2004) defined electronic word-of-mouth communication as “any positive or negative statement
made by potential, actual, or former customers about a product or company, which is made
available to a multitude of people and institutions via the Internet” (p. 39). Recent, Internet-based
word-of-mouth studies were able to measure the actual effect of word-of-mouth advertising that
could be tracked or manipulated, and thus provided fresh insight into the phenomenon. Despite
the advances in technology that have allowed researchers to better understand the word-of-mouth
process, marketers are still fascinated with determining how much influence word-of-mouth
advertising has on final purchase decisions. For example, Keller and Berry (2003) found that
personal sources were used much more often than advertising sources to glean information about

6

purchase decisions. Refer to Table 1 for more information, a table adapted from Keller and
Berry’s (2003) book.
Table 1: Percentage of People who Utilize Word-of-Mouth Advertising Versus Traditional
Advertising When Making Purchase Decisions
People Advertising
Restaurants

83%

35%

Places

71%

33%

Prescription drugs

71%

21%

Hotels

63%

27%

Health tips

61%

19%

Movies

61%

67%

Best brands

60%

33%

Retirement planning

58%

9%

Automobiles

58%

36%

Clothes

50%

59%

Computer equipment

40%

18%

Websites to visit

37%

12%

Opinion leaders are an integral part of the word-of-mouth process. Furthermore, they
have certain characteristics in common across product categories. Katz and Lazarsfeld (1955)
found that opinion leaders were more likely to seek out media related to their area of expertise.
For example, leaders who specialized in cars read more car magazines and watched more car
shows. Furthermore, opinion leaders tended to be more interested in their topic of expertise than
other individuals (Lazarsfeld, Berelson, & Gaudet, 1948). Opinion leaders were also more likely
to engage in information-seeking behaviours from interpersonal sources than other individuals,
indicating that the two-step flow process may, in fact, be more complex than originally thought

7

(Arndt, 1967). Specifically, information may flow from media sources to a key group of
interested individuals, who then seek out information from similarly interested individuals before
passing it along to other, less interested or informed individuals. On the web, information may
flow from the company to a set of opinion leaders and then on to a much larger audience on the
Internet, expanding the reach of opinion leaders. These opinion leaders spend a great deal of time
learning all that they can about their topic of interest (Lazarsfeld, Berelson, & Gaudet, 1948). As
such, they are highly expert when discussing their topic area. This lends them one dimension of
source credibility. Is this expertise relevant to the people they are advising, or is its relevance
dependent upon the situation?
In order for word-of-mouth communication to occur, there must be a shared interest in
the product. In support of this, Back et al. (1950) and Festinger et al. (1948) found that rumours
reached individuals for whom the information was of interest, but nobody else. Thus, it is not
only the opinion leader who must be interested in the communication, but also the recipient of
information. On a related note, Peterson and Gist (1951) found that, when rumours were less
relevant to the individuals who were recipients of the rumours, they tended to become distorted
over time. Thus, information was more accurate when it was passed through interested parties.
However, Caplow (1947) and Back et al. (1950) found that distortion was unlikely when
rumours were relevant. In the digital age, inaccuracies may prevail if the original information is
false, but the source material is readily available so that the information is less likely to be
modified over time between exchanges. As such, both consumers and companies have access to
the original message rather than a distorted version – something that is not possible for orally
transmitted information. This allows for a more thorough understanding of the dimensions of

8

word-of-mouth communication that increase or decrease its effectiveness while also allowing
researchers to study the language and content used in these recommendations.
Although electronic word-of-mouth is increasingly popular, its effects might be less
dramatic than traditional word-of-mouth. Herr, Kardes, and Kim (1991) found that face-to-face
communication was more persuasive than printed word-of-mouth communication. Thus,
although electronic word-of-mouth is more trackable than traditional word-of-mouth
communications, applying findings across these two mediums should be undertaken with caution.
Park and Kim (2008) found that, for online product reviews, experts were better able to
sift through the cues to find the most relevant information, whereas novices were likely to form
an opinion based on a consensus in reviews rather than message content. Involvement levels may
have influenced which cues were utilized to evaluate the product, as experts were likely more
motivated to process the detailed information, while novices did not have the ability or
motivation to do more than interpret the overall tone of the reviews. In support of this, Lee, Park,
and Han (2008) found that low involvement consumers based their attitudes solely on the
proportion of positive / negative reviews, while high involvement consumers based their
attitudes on the quality of the reviews as well as the proportion of positive to negative reviews.
Furthermore, Chiou and Cheng (2003) found that message favourableness and number of
messages had an effect on brand evaluations, but only when the consumer did not have a strong,
pre-existing attitude toward the brand. This data suggests that, overall, individual online sources
may have less of an impact on attitudes and intentions, but the volume of information available
from other consumers may influence word-of-mouth effects as well (at least for novices).
A preference for homophily, defined as “a tendency for friendships to form between
those who are alike in some designated respect” (Lazarsfeld & Merton, 1954, p. 23) has been

9

demonstrated to be an important variable in friendship formation (e.g., Allgeier & Byrne, 1973;
Batchelor & Tesser, 1971; Byrne, 1961a; 1965; Farmer & Farmer, 1996; Huston & Levinger,
1978; Neimeyer & Mitchell, 1988; Posavac & Pasko, 1974; Rosenbaum, 1986; Tan & Singh,
1995). Furthermore, it has been demonstrated to be an important variable in word-of-mouth
processes (Brown & Reingen, 1987; Bruyn & Lilien, 2004; Dellande, Gilly, & Graham, 2004;
Gilly, Graham, Wolfinbarger, & Yale, 1998; Steffes & Burgee, 2009, Wang, Walther, Pingree, &
Hawkins, 2008; Wright, 2000). Both opinion leaders, defined as “those people who are
knowledgeable about products and who are frequently able to influence others’ attitudes or
behaviours with regard to a product category”, (Solomon, Zaichowsky, & Polegato, 2002, p. 390)
and reference groups, defined as “an actual or imaginary individual or group that has a
significant effect upon an individual’s evaluations, aspirations or behaviour” (Solomon,
Zaichowsky, & Polegato, 2002, p. 368), are integral to the word-of-mouth process, and draw part
of their influence from perceived similarity (French & Raven, 1968; Torrance & Mason, 1956).
This relationship between similarity and word-of-mouth success has been demonstrated multiple
times in the literature since French and Raven’s (1968) discussion of the bases of power (Brown
& Reingen, 1987; Bruyn & Lilien, 2004; Dellande, Gilly, & Graham, 2004; Gilly, Graham,
Wolfinbarger, & Yale, 1998; Steffes & Burgee, 2009, Wang, Walther, Pingree, & Hawkins,
2008; Wright, 2000). Since homophily has been linked to source persuasion in word-of-mouth
situations and interpersonal attractiveness consistently in the literature, homophily’s effect on
relationship formation and persuasion is discussed next, followed by a discussion of why people
prefer others who are similar to them.

10

Homophily
Homophily has been discussed several times in the relationship formation literature. Most
of the homophily research to date has measured homophilous tendencies, rather than preferences.
In other words, researchers have tended to measure homophilous relationships (how similar
friends are to each other, etc.), rather than the construct behind these behaviours. It is important
to understand both behaviours and the reasons behind them – including personal preferences and
environmental factors - and thus the focus of this paper will be homophilous preferences and the
reasons why people prefer to spend time with others who are similar to them grounded in
literature regarding homophilous tendencies. Lazarsfeld and Merton’s (1954) original concept of
homophily discussed a preference for friends who were similar on two dimensions: social status
within the population and values. However, since its inception, homophily research has
discussed individuals’ preferences to form homophilous groups based on many different
characteristics, such as a preference to form ties with individuals who are similar regarding
gender (e.g., Farmer & Farmer, 1996; Huston & Levinger, 1978; Roth, 2004a; 2004b; Turban,
Dougherty, & Lee, 2002), age (e.g., Saiki & DeLong, 2006; Thelwall, 2008; Zenger & Lawrence,
1989), and race (e.g., Doyle & Kao, 2007; Farmer & Farmer, 1996; Joyner & Kao, 2000; Turban,
Dougherty, & Lee, 2002), as well as other, less easily identified characteristics such as
personality (e.g., Eyal & Rubin, 2003; Neimeyer & Mitchell, 1988; Posavac & Pasko, 1974;
Singh, 1973; Steele & McGlynn, 1979), and attitudes (e.g., Allgeier & Byrne, 1973; Batchelor &
Tesser, 1971; Byrne, 1961a; 1961b; Feren, Carroll, & Olian, 1988; Rosenbaum, 1986; Tan &
Singh, 1995).
Although Lazarsfeld and Merton (1954) focused on value homophily, very few studies
that cite this original article study homophily based on psychographic or attitudinal variables. As

11

demonstrated by McPherson, Smith-Lovin and Cook’s (2001) literature review, many
researchers have come to believe that value-homophily is an artefact of demographic homophily.
In other words, researchers believe that those with similar backgrounds will have similar
experiences, and will therefore have similar values and attitudes. This is a view that Lazarsfeld
and Merton (1954) suggested themselves. However, there is plenty of research demonstrating
that homophilous attitudes are important in predicting relationship formation in their own right
(e.g., Allgeier & Byrne, 1973; Batchelor & Tesser, 1971; Byrne, 1961a; 1961b; Feren, Carroll, &
Olian, 1988; Rosenbaum, 1986; Tan & Singh, 1995). Although demographic similarity is easier
to measure, and therefore study, attitude and personality similarity may in fact contribute
significant explained variance to the traditional demography homophily studies. However, there
is a rich literature from the social sciences that studies attitude homophily without referring to
the homophily construct. Byrne and colleagues have studied attitude homophily and its
antecedents (e.g., Byrne, 1961a; 1961b; 1965; Byrne, Clore, & Smeaton, 1986; Byrne & Griffitt,
1966; Byrne, Griffitt, & Stefaniak, 1967; Byrne & Nelson, 1965; Byrne & Rhamey, 1965).
Specifically, these researchers studied the effects of attitude similarity in an experimental setting
without demographic information available, and found that attitude similarity positively
predicted attraction ratings to a described individual. Furthermore, Duck and associates have
extensively studied the positive relationship between homophilous personalities and relationship
formation (e.g., Duck, 1973; 1975; Duck & Spencer, 1972).
While the Internet provides a large amount of anonymity in some forums, web surfers
have developed new ways to evaluate homophily based on the information available. In
relatively anonymous ratings sites focused on electronic word-of-mouth communications, rating
objects in a similar fashion led to feelings of similarity and trust in the other rater’s judgment

12

(Goldbeck, 2009). In other forums, the expression of similar attitudes led to perceived similarity
and trust in the participants (Meijnders et al., 2009). Once again, similarity on characteristics
other than demographics were instrumental in interpersonal attraction and persuasion.
Fiore and Donath (2005) explored online dating sites, which provide a wealth of
information about their members, and found that members were more likely to contact
individuals who indicated similar aspirations for the future, habits, education, and physical
appearance. Similarly, Schrock (2007) found that online personals tended to express a desire for
similarity regarding demographics and vices –a tendency that was more prevalent in women than
men. Overall, it appears that in both electronic word-of-mouth situations and interpersonal
relationships on the Internet, there is a preference for homophily on characteristics that are
deemed relevant to the situation at hand, and homophily is intertwined with attractiveness.
To demonstrate the power of the relationship between attractiveness and similarity, Moss,
Byrne, Baskett, and Sachs (1975) asked participants to fill out attitudes of a stranger based only
on attractiveness ratings. Responses suggested that the more attractive a stranger was supposed
to be, the more participants perceived them to be similar to themselves. This, however, raises a
question: does attraction come first, or perceived similarity? The author of the attractionsimilarity hypothesis (Morry, 2004; 2005; 2007) argued that attraction came before perceived
similarity. This hypothesis also posited that perceived similarity was the key to relationship
formation, and not actual similarity, a finding supported by Strauss, Barrick, and Connerley’s
(2001) work regarding actual vs. perceived personality homophily.
Further supporting the theory that perceived similarity is of more importance than actual
similarity, Werner and Parmalee (1979) found evidence that although participants believed they
were similar to their friends on both attitudes and activities, these friends were only actually

13

similar to each other in their activity choices. Furthermore, Acitelli, Douvan, and Veroff (1993)
found that newlywed couples believed that their partners were more similar to themselves in
their communication behaviours than they actually were. Finally, Levinger and Breedlove (1966)
found that the perception of similarity regarding the importance of communication was more
predictive of marital satisfaction than actual similarity. Based on these studies, it would seem
that perceived similarity is more instrumental than actual similarity in relationship formation.
In an attempt to answer the question regarding the importance of actual versus perceived
similarity, Montoya, Horton, and Kirchner (2008) conducted a meta-analysis which compared
effect sizes for perceived and actual similarity. Results suggested that actual similarity was only
important when respondents had not interacted with each other. Perceived similarity, however,
was a positive predictor of attraction for no-interaction studies, short-interaction studies, and
even long-interaction studies. These results suggest that perceived similarity is, on average, a
better predictor of attractiveness than actual similarity. Since attractive sources are more
persuasive than unattractive sources (e.g., Gotlieb & Sarel, 1991; Jones, Moore, Stanaland, &
Rosalind, 1998; Lichtenstein & Bearden, 1989; Patzer, 1983; Pornpitakpan, 2003), this perceived
similarity will have a positive effect on persuasion. Therefore, future research should focus on
perceived homophily rather than actual homophily, and should attempt to better understand
under which circumstances actual homophily is of importance.
Morry, Kito, and Ortiz (2011) explored this tendency, and found that couples reported
high degrees of perceived similarity between partners regarding dimensions of both moderate
and high relevance to the relationship. However, accuracy was higher for moderately important
characteristics than low importance characteristics regarding relationship success. As such, these

14

results suggest that perception is more important than reality, but on characteristics that really
matter perception is closer to reality than on those characteristics of little to no importance.
A related study conducted by Lutz-Zois, Bradley, Mihalik, and Moorman-Eavers (2006)
found that, for those characteristics that were important to participants, a high degree of
perceived similarity was positively related to relationship satisfaction. Characteristics that were
unimportant to the participant, however, were not related to relationship satisfaction. In support
of this, research has demonstrated that similarity is more impactful on attraction ratings for
topics that were important to an individual or to the situation than topics that were irrelevant
(Byrne, 1961a; Byrne, Bond, & Diamond, 1969; Clore & Baldridge, 1968; Michinov & Monteil,
2002). In other words, similarity regarding education may not always be important, but when an
individual is seeking out advice regarding whether or not they should attend graduate school,
similarity will be desirable. These findings further demonstrate the importance of a homophily
scale comprised of items of importance to participants, rather than the current state of homophily
research which has been extremely varied in its approach to how homophily is measured.
Without consistency in the manipulation and measurement of homophily in the literature,
homophily research will be inconsistent and incapable of predicting behaviour.
A more complete scale that includes only those characteristics that are relevant to
individuals and speaks about homophily in a language that makes sense to participants will lead
to a greater understanding of when homophily is perceived to be important. Until one scale is
utilized in the research, it is unlikely that researchers will be able to fully understand the role of
homophily in society, as differences will too often be due to changes in the scale and not actual
differences in preferences. A consistently utilized scale based on the characteristics of
importance to those filling out the surveys, rather than those writing them, will allow for

15

predictions of behaviour to be made across situations, explain the existing conflicting findings in
the literature regarding the circumstances in which homophilous tendencies are present, and
prevent future confusion.
McCroskey, Richmond, and Daly (1975) created a homophily scale by asking about
several different characteristics, including similarity regarding likes, dislikes, morals, values, and
emotions. This scale was later revised by McCroskey, McCroskey, and Richmond (2006).
Although this scale was a step in the right direction, the authors still failed to ask, in an openended format, how individuals evaluated similarity, and what characteristics were considered
when making similarity judgments. Thus, researchers may have missed important components of
similarity, and included items that were unimportant to individuals when they were evaluating
liking based upon homophily. McCroskey, Richmond, and Daly (1975) created their pool of
items by searching through the literature for previous research on similarity and including these
previously-utilized items. Therefore, if researchers had not thought of a characteristic before this
study, it was not included. Furthermore, items may have been included in the scale that were of
no importance to participants when making judgments regarding homophily with a target. This is
likely, as the appearance factor had a mean below the midpoint (4) of the Likert-type scale in the
original study (McCroskey, Richmond, & Daly, 1975).
Previous work that has utilized McCroskey, Richmond, and Daly’s (1975) scale has often
determined that only certain elements of homophily were significant predictors of relationship
formation (e.g., Glascock & Ruggiero, 2006; Rocca & McCroskey, 1999; Wright, 2004). In these
three studies that utilized McCroskey, Richmond, and Daly’s (1975) scale attitude homophily,
background homophily, or both of these emerged as significant predictors of interpersonal
attractiveness. However, the other two factors (values and appearance) were either not mentioned

16

at all or were not significant. Why, then, should they be included in the scale? Furthermore,
effect sizes have sometimes been null (e.g., Dohanos, 2003) or small when statistically
significant relationships were found between homophily and attractiveness ratings (e.g., Morris,
Gorham, Cohen, & Huffman, 1996). By focusing on only those characteristics that are relevant
to similarity judgments in natural settings, the current scale will allow future researchers to find
out the impact of homophily in real-life situations.. A scale based on characteristics that are
actually evaluated when making similarity judgments would provide researchers with a more
accurate perception of how homophily functions in relationship formation and the choice of
word-of-mouth advisors. To continue, however, utilizing various scales and manipulations to
measure homophily will result in varying effect sizes, conflicting null and positive results, and a
confusing literature that is unable to shed light on how homophily works outside of the lab.
Methodologically, there are other problems with much of the homophily research that
does not focus on demographics. Although some research has explored individuals’ tendency or
preference for homophily in real-life friendships and marriages (e.g., Bentler & Newcomb, 1978;
Carli, Ganley, & Pierce-Otay, 1991; Fiedler, Warrington, & Blaisdell, 1952; Fiore & Donath,
2005; Lutz-Zois, Bradley, Mihalik, & Moorman-Eavers, 2006; Morry, Kito, & Ortiz, 2011;
Murstein & Beck, 1972; Schrock, 2007; Strauss, Barrick, & Connerley, 2001; Zenger &
Lawrence, 1989), much of the attitude and personality homophily research has been based on
experimental procedures that merely describe an individual on one or two characteristics and ask
participants to evaluate their interpersonal attractiveness (e.g., Byrne, 1961a; 1961b; 1965; Byrne,
Bond, & Diamond, 1969; Byrne, Clore, & Smeaton, 1986; Byrne, Griffitt, & Stefaniak, 1967;
Byrne & Nelson, 1965), which maximizes internal validity by ensuring that all effects are due to

17

the manipulation, but also minimizes external validity as this limited information does not
resemble relationships outside of a laboratory setting.
Furthermore, there is a lack of theory in the homophily literature. In other words, despite
numerous studies regarding the relationship between similarity and attraction, there has been
little done to understand why individuals display this homophilous tendency in their relationships
and information-seeking processes.
This research is focused not only in the measurement and application of homophily, but
also its roots – why is this phenomenon so prevalent in human interactions? Homophily research
has traditionally attempted to manipulate or measure perceived similarity in order to predict
attraction ratings. However, there is some research that has studied what demographic,
psychological, and sociological factors lead to preferences for homophily in the first place. Ibarra
(1992) and Levinson, McCollum, and Kutner (1984) found that women’s networks were less
gender homophilous than men’s networks. Furthermore, Shrum, Cheek, and Hunter (1988) found
that age influenced racial and gender homophily, while other researchers found that racial
minorities’ networks were more likely than racial majorities’ networks to be homophilous
(Mollica, Gray, & Trevino, 2003). Lazarsfeld and Merton (1954) demonstrated that religious
homophily varied depending upon the community that an individual had grown up in. As such,
there is evidence that certain demographic groups have a greater preference for homophily – at
least on certain characteristics.
While the previous discussion has focused on demographic variables that influence
preference for homophily, psychological variables influence homophilous preferences as well.
Snyder and Morris (1978) found that perceptions of judgmental ability influenced preference for
homophily. Furthermore, cognitive complexity (Malhotra, 1988), self-actualization (Olczak &

18

Goldman, 1975), extroversion (Singh & Teoh, 1999), dogmatism (Palmer & Kalin, 1985),
authoritarianism (Mitchell & Byrne, 1973), self-monitoring (Jamieson, Lydon, & Zanna, 1987),
need for approval (Posavac, 1971), and need for affiliation (Byrne, 1961b) had positive
correlations with homophilous tendencies.
These studies are a first step in understanding why preference for homophily is so
prevalent in our society. The findings regarding need for approval (Posavac, 1971) and need for
affiliation (Byrne, 1961b) support the findings of Erwin (1981), who found that the perceived
likelihood that a target would like a participant mediated the relationship between similarity and
attraction. Thus, for people who are highly motivated to be liked, homophily serves as a cue to
how likely another individual is to like them. This hypothesis is further supported by the studies
that found that targets who evaluated the respondent positively were better liked than targets who
evaluated the respondent negatively (Arrowood & Short, 1973; Byrne & Rhamey, 1965; Condon
& Crano, 1988; Erwin, 1982; Insko, Thompson, Stroebe, Shaud, Pinner, & Layton, 1973). This
is one possible reason why homophilous preferences are common in society – an ever-present
need for approval and validation from others. However, further research is required in order to
better understand how this need for approval influences preference for homophily.
Perceived homophily with a source, however, is only one consideration when individuals
are evaluating a prospective advisor. Source credibility literature states that there are three main
attributes of a desirable information source: likeability, trustworthiness, and expertise
(Pornpitakpan, 2003). Homophily is one component of likeability, but when considering the
selection of word-of-mouth sources it is prudent to examine all three of these facets of source
credibility. As such, source credibility’s effect on the word-of-mouth process is discussed next,
and the different situations in which it is utilized to aid in attitude formation.

19

Source Credibility
Source credibility refers to the characteristic of a source that makes him or her
trustworthy of accurate and helpful information. At first, source credibility was divided into two
main components: expertise (the degree to which a source was perceived to have the knowledge
to make assertions about a stimulus object) and trustworthiness (the degree to which the source
was perceived to be making assertions which they themselves perceived to be valid) (Hovland,
Janis, & Kelley, 1953). In a study of source credibility’s two dimensions, it was found that a
source that was both trustworthy and expert resulted in the greatest opinion change, but results
also indicated that trustworthiness was a more salient factor than expertise (McGinnies & Ward,
1980). Wiener and Mowen (1986) confirmed that both trustworthiness and expertise were
positively related to agreement with the viewpoints of the source, while Pornpitakpan (2003)
added a dimension to the source credibility literature: source likeability. This is the dimension
related to homophily, as interpersonal attractiveness and homophily are linked (e.g., Allgeier &
Byrne, 1973; Batchelor & Tesser, 1971; Byrne, 1961a; 1965; Farmer & Farmer, 1996; Huston &
Levinger, 1978; Neimeyer & Mitchell, 1988; Posavac & Pasko, 1974; Rosenbaum, 1986; Tan &
Singh, 1995).
In general, a highly credible source has been found to result in greater persuasion than a
source low in credibility (Cheung, Luo, Sia, & Chen, 2009; Horai, Naccari, & Fatoullah, 1974;
Hovland & Weiss, 1951; Johnson & Izzett, 1969; Johnson, Torvicia, & Poprick, 1968; Jones,
Sinclair, & Courtneya, 2003; Kelman & Hovland, 1953; Lirtzman & Shuv-Ami, 1986; Maddux
& Rogers, 1980; Miller & Baseheart, 1969; Watts & McGuire, 1964; Whittaker & Meade, 1968).
Furthermore, credibility is correlated with intentions to follow through with advice (Campbell &
Wright, 2002). As such, it is a topic of great interest to advertising and marketing researchers.

20

Several researchers have found that, when sources were perceived to be more trustworthy,
participants were more likely to be persuaded by the source’s arguments (e.g., Haas & Grady,
1975; Kiesler & Kiesler, 1964; Walster & Festinger, 1962). This is one of the key benefits of
word-of-mouth communication – sources are perceived to be less biased and are therefore more
impactful on attitudes and purchase intentions. This effect was reversed, however, when a biased
source appeared to be primarily concerned with the welfare of the audience. When an altruistic
motivation was primed, persuasion was enhanced when persuasive intent was made salient
(Mills, 1966). Thus, trustworthiness is a powerful factor when considering the information
search process. Furthermore, trustworthiness is based not simply on how biased the source is, but
instead how likely it is that the source is looking out for the advisee’s best interest - bias seems to
be a heuristic cue to how likely it is that a source is looking out for somebody else’s best
interests.
However, trustworthiness alone is not enough to create a substantial attitude change. In
one study of source credibility it was found that a source that was perceived to be biased was
deemed to be less fair and produced less attitude change than unbiased sources, but this
difference only occurred when the unbiased source was also an expert (Hovland & Weiss, 1951;
Kelman & Hovland, 1953). Although source credibility literature has largely focused on
expertise and trustworthiness, other research has demonstrated that the likeability and
attractiveness of a source also play a role in source credibility (e.g., Gotlieb & Sarel, 1991; Jones,
Moore, Stanaland, & Rosalind, 1998; Lichtenstein & Bearden, 1989; Patzer, 1983). Thus, more
recent literature has recognized trustworthiness, expertise, and source likeability as the three
bases of source credibility. Since perceived similarity has been related to attractiveness (e.g.,
Allgeier & Byrne, 1973; Batchelor & Tesser, 1971; Byrne, 1961a; 1965; Farmer & Farmer, 1996;

21

Huston & Levinger, 1978; Neimeyer & Mitchell, 1988; Posavac & Pasko, 1974; Rosenbaum,
1986; Tan & Singh, 1995), homophily can be linked to source credibility based on this
dimension.
Markham (1968) found that, contrary to typical source credibility studies, showmanism /
entertainment and trustworthiness were the factors that contributed to credibility of newscasters.
Thus, likeability and trustworthiness were important for newscasters, but expertise was
unimportant. However, the authors hypothesized that this may have been due to a perception that
all newscasters were experts. In this case, two dimensions of source credibility were influential
in evaluating the source while expertise was not. As such, this research indicated that the
dimensions of source credibility relevant to source selection vary depending upon the context
and require further study in order to determine when each dimension is relevant to recipients of
information.. What causes this variation?
Advertisers have often chosen attractive models for their ads under the assumption that
‘what is beautiful is good’ (Friedman & Friedman, 1979; Singer, 1983). However, other studies
have indicated that this assumption is not necessarily always correct (Kamins, 1990). Results of
this study indicated that an attractive source was significantly related to credibility only when the
product being sold was also related to attractiveness (e.g., make-up, hair dye). For a product
which was not related to attractiveness, there was no difference in credibility between an
attractive and an unattractive celebrity endorser. Thus, the relevance of attractiveness seemed to
play a role in the effectiveness of celebrity endorsers.
For different types of products, then, it makes sense that there are different groups that
individuals turn to for advice on buying that product. Selection system theory was one theory
which described three ideal types of product selection: market, expert, and peer selection (Priem,

22

2007; Wijnberg, 2004; Wijnberg & Gemser, 2000). In a study of how these different groups
impacted consumers’ attitudes toward seeing a new movie it was found that for independent
films, an award given by experts was more effective than other types of awards at changing
ticket sales (Gemser, Leenmdersr, & Wijnberg, 2008). Authors postulated that because these
films were unlikely to be reviewed by an individual’s peers, they often turned to professional,
expert critics to decide which films to see. As such, the expertise facet of source credibility
became salient for this product due to the nature of the purchase.
Recently, a preference for credible sources has been encouraged in some environments
where credibility is particularly low, such as the Internet. When the Internet was first introduced,
individuals were encouraged to determine the credibility of the website by first evaluating the
author of the work (Warnick, 2004). Individuals were supposed to judge the author’s motives,
expertise, and associations in order to make a decision regarding the credibility of the
information contained on the site. In a study of actual respondents that frequently used the
Internet, it was determined that different types of websites (e.g., online retailers versus online
information sources for topics such as sports and entertainment) resulted in different criteria
being utilized to evaluate the site’s credibility. Thus, the situation influenced which dimensions
of source credibility were the most salient.
Golbeck (2009) found that, in an online context, similar attitudes in reviews could lead to
higher credibility ratings for individual raters. Similarly, Bhuiyan (2010) found that similar
interests led to a higher level of trust in recommendation systems. Furthermore, similar linguistic
styles led to higher levels of trust in a source (Scissors, Gill, Geraghty, & Gergle, 2009). Overall,
the research supports the notion that, in the online world, there are several cues to similarity such
as similar expressed attitudes (e.g., Golbeck, 2009), similar interests (e.g., Bhuiyan, 2010), and

23

similar linguistic styles (e.g., Scissors, Gill, Geraghty, & Gergle, 2009) utilized to determine
perceived homophily, leading to higher source credibility perceptions and persuasion. Even
though the cues available are different from those that are utilized in face-to-face interactions,
users’ tendencies to favour similar individuals indicate that homophilous judgments are still
being made on the Internet.
The importance of cues related to similarity and credibility within the online context
when determining the credence to give to online word-of-mouth communication require further
exploration, as the findings of studies utilizing the elaboration likelihood model (Petty, Cacioppo,
& Goldman, 1981) are inconclusive regarding the situations and involvement levels in which
source credibility cues are utilized to form attitudes and intentions (e.g., Eisend, 2007; Petty &
Cacioppo, 1984; Wu & Schaffer, 1987).
Multiple studies on the effects of source credibility and decision involvement have been
done over the years. It has been found that when an individual processed a persuasive message
on the periphery of their consciousness, heuristic cues such as source credibility were often
utilized to draw conclusions about the featured brand (Petty, Cacioppo, & Goldman, 1981), a
prediction of the elaboration likelihood model. However, Petty and Cacioppo (1984) found that
source credibility could be utilized as a central cue when it was relevant to the evaluation of a
message. As such, the dimensions of source credibility and their relationship with involvement
call for further exploration in order to determine when source credibility cues are relevant to
attitude and intention formation. Source credibility has been organized into three dimensions:
trustworthiness, likeability, and expertise. Are certain aspects utilized only in certain
involvement situations (high or low)? The current study attempted to better understand this
relationship.

24

Many other researchers have attempted to understand how involvement influences the
influence that credibility has on information processing and attitude formation. Andrews and
Shimp (1990) found that involvement was significantly related to the effect of source credibility
on attitude formation. Respondents were more affected by a favourable source than an
unfavourable source for a low involvement situation, but this effect was not present for a high
involvement situation. These findings were in accordance with the elaboration likelihood model
(Petty, Cacioppo, & Goldman, 1981), as source credibility is typically a peripheral heuristic cue.
Mondak (1990) found that when message elaboration was minimal, credibility acted as a
heuristic that influenced attitudes. However, when elaboration increased, source credibility’s
effect on attitude change diminished – but didn’t disappear.
Priester and Petty’s (2003) research suggested that participants who read an
advertisement endorsed by an untrustworthy endorser were more influenced by argument quality
than participants who were exposed to an advertisement endorsed by a trustworthy endorser –
suggesting that message content may have an effect on the role of source credibility. Heesacker,
Petty, and Cacioppo (1983) found that, for highly credible sources, attitudes in response to a
strong message were more favourable than attitudes in response to a weak message. However,
for low credibility sources, message strength did not influence attitudes. Slater and Rouner (1996)
demonstrated that there was a relationship between message quality and source credibility.
Furthermore, Beach, Mitchell, Deaton, and Prothero (1978) found that information was
discounted more as its relevance and source credibility decreased.
Eisend (2007) found that source credibility enhanced consumers’ attention and
motivation to process a message (message involvement), improved attitudes toward the ad, and
improved attitude toward the brand – yet another way that involvement and credibility influence

25

each other. Because involvement has been discussed in the source credibility literature on
multiple occasions, it is relevant to discuss the elaboration likelihood model and how its
predictions can contribute to the understanding of the relationship between involvement, source
credibility, and preference for similarity.
As demonstrated by Petty and Cacioppo (1984), source credibility’s impact on the
decision-making process is influenced by more than the information processing route utilized
(central or peripheral). The elaboration likelihood model (Petty, Cacioppo, & Goldman, 1981)
suggested that cues to source credibility would be utilized in low involvement situations only.
However, Petty and Cacioppo’s (1984) work suggested a more complex relationship between
involvement and source credibility’s role in a decision. Further research is needed to better
understand the relationship between source credibility, information elaboration and processing,
and attitude formation.
Elaboration Likelihood Model
The elaboration likelihood model stated that as involvement with an issue increased,
more central or elaborate processing would be utilized to make a decision or evaluate any
information provided (Petty, Cacioppo, & Goldman, 1981). This model distinguished between
central processing, whereby messages were carefully processed and evaluated, and peripheral
processing, in which heuristic cues (such as likeability and similarity) were utilized to evaluate a
message rather than in-depth deliberation.
When involvement in an issue is low, it is more likely that peripheral processing will be
utilized to make decisions (Petty, Cacioppo, & Goldman, 1981). Several studies have
demonstrated this tendency (e.g., Andrews & Shimp, 1990; Bhattacherjee & Sanford, 2006;
Petty & Cacioppo, 1984; Yang, Hung, Sungh, & Farn, 2006). Common examples of peripheral

26

processing include evaluating the credentials of the author of a work or website or evaluating a
site based on the attractiveness of a page rather than the actual content. Researchers have related
involvement to external search effort in several studies. For example, Hugstad, Taylor, and
Bruce (1987) found that involvement was positively related to external search effort, such that
interpersonal sources and consumer guides were more influential in these cases. The current
study attempted to understand source factors that would influence external search effort, in
addition to the likelihood of following the source’s advice.
Larsen and Phillips (2002) suggested that source cues may be more important in
determining attitudes in highly stressful situations than in less stressful situations. Therefore,
once again, source credibility’s role in attitude formation and information processing is not as
simple as the elaboration likelihood model (Petty, Cacioppo, & Goldman, 1981) proposed.
Although source characteristics such as expertise and likeability have been traditionally
thought of as peripheral cues (Chaiken, 1980; Petty, Cacioppo, & Goldman, 1981), research has
demonstrated that source cues can be used to evaluate ability and bias in word-of-mouth
situations (Petty & Cacioppo, 1984). Furthermore, these authors found that when involvement
was moderate, source cues were utilized to determine how much additional effort should be
made in evaluating the persuasion episode. Overall, this research suggests that involvement will
affect the degree to which source cues are processed. When involvement is low, it is more likely
that heuristic cues such as source likeability and similarity will be utilized to evaluate word-ofmouth sources. When involvement is high, however, expertise will be utilized as a more central
cue to evaluate word-of-mouth sources.
Word-of-mouth sources are members of an individual’s own personal network and social
networks tend to be homophilous (e.g., Duck, 1975; Fiore & Donath, 2005; Morry, 2005;

27

Thelwall, 2008). Furthermore, homophily tends to be related to attractiveness (e.g., Allgeier &
Byrne, 1973; Batchelor & Tesser, 1971; Byrne, 1961a; 1965; Farmer & Farmer, 1996; Huston &
Levinger, 1978; Neimeyer & Mitchell, 1988; Posavac & Pasko, 1974; Rosenbaum, 1986; Tan &
Singh, 1995), another element of source credibility. As such, homophily is likely to play a role in
all word-of-mouth situations – but to varying degrees. This tendency for homophily in
interpersonal relationships and information sources support the notion that word-of-mouth
sources will be homophilous. However, since source credibility involves expertise (Hovland,
Janis, & Kelly, 1953), which may necessitate differences between the word-of-mouth
communicator and receiver (at least regarding knowledge and experience, if not interests as well),
homophily between the source and receiver may be less prevalent in situations in which source
expertise is required by the purchaser. Despite the common assumption that source credibility is
primarily utilized as a peripheral cue in low involvement situations (e.g., Andrews and Shimp,
1990; Chaiken, 1980; Petty, Cacioppo, & Goldman, 1981) there is evidence that source
credibility is utilized at vary levels of involvement (e.g., e.g., Eisend, 2007; Petty & Cacioppo,
1984; Wu & Schaffer, 1987). As such, further research is needed in order to better understand
how involvement influences the relationship between source credibility, information processing,
and attitude formation.

28

RESEARCH QUESTIONS AND HYPOTHESES

Since homophily is important to the understanding of source credibility and its
dimensions, a better understanding of preference for homophily was the first step to this study,
followed by a study of the relationship between homophily, source credibility, and involvement.
Up to this point, homophily research has been done using researcher-defined categories of
homophily. Specifically, researchers have determined which characteristics to study based upon
whatever characteristics were the easiest to study or the most theoretically relevant for the
researcher. Although this focus on theory building has added to scientific knowledge in many
ways it has diminished the ability of researchers to focus on those dimensions of homophily that
were the most relevant to similarity judgments for participants. Furthermore, the lack of
consistency in the research has reduced the ability of researchers to compare results between
studies. For example, Byrne and colleagues (e.g., Byrne, 1961a; 1961b; 1965; Byrne, Bond, &
Diamond, 1969; Byrne, Clore, & Smeaton, 1986; Byrne, Griffitt, & Stefaniak, 1967; Byrne &
Nelson, 1965) have focused on how similarity regarding attitudes influences attraction to an
individual. Although these studies have been successful in determining attraction to a target,
there is little evidence that these characteristics are the most relevant to the participants. In these
studies, the participants had no information about the target other than attitudes – as such,
generalizability and external validity is limited. Future research should utilize the most relevant
characteristics to measure or manipulate homophily rather than simply focusing on what has
worked in the past, which would allow researchers to better understand how homophily
influences interactions between individuals.

29

Because a tendency for homophily has been demonstrated so many times in the literature
(e.g., Bentler & Newcomb, 1978; Byrne, 1961a; 1961b; Byrne & Nelson, 1965; Carli, Ganley, &
Pierce-Otay, 1991; Duck, 1975; Fiore & Donath, 2005; Lutz-Zois, Bradley, Mihalik, &
Moorman-Eavers, 2006; Morry, 2004; 2005; 2007; Schrock, 2007; Thelwall, 2008), it is useful
to create a scale to measure preference for homophily based upon characteristics that are relevant
to the participants. Previous research may have over or underestimated homophilous preferences
and tendencies simply because the researchers have been measuring homophily based upon
characteristics that are irrelevant to homophily judgments. We know that similarity is preferred
over dissimilarity, but researchers are still unaware of which characteristics everyday similarity
judgments are based upon in natural settings, and for which characteristics individuals display a
preference for homophily.
In order to fully understand homophilous preferences, researchers need to have a valid
scale with which to conduct research and experiments. A scale to measure preference for
homophily will be based upon conscious (rather than subconscious) preferences, and will create
an interval-level variable that will be precise enough to measure situation-specific preferences as
well as consistent preferences individuals hold across situations. Individuals have been
demonstrated to be inherently irrational in describing their preferences (Kahneman & Tversky,
1982; Tversky, 1977; Tversky & Kahneman, 1974), but since perceptions of homophily tend to
be more accurate in predicting relationship formations than actual homophily (Morry, 2004;
2005; 2007; Strauss, Barrick, & Connerley, 2001), this is an important first step in understanding
homophilous tendencies.
McCroskey, Richmond, and Daly (1975) and McCroskey, McCroskey, and Richmond
(2006) constructed a general homophily scale. In a study of preference for homophily, however,

30

it is important to study homophily based on characteristics that are relevant to homophily
judgments in natural settings. McCroskey, Richmond, and Daly (1975) produced a set of items
based on previous literature which eliminated any items of relevance that had yet to be studied.
However, when studying a phenomenon that is not yet understood, it is important to discover
how participants talk about it and its influence on their daily lives, rather than assuming that the
previous research has already discovered all of the dimensions of the construct that are relevant
(e.g., Charmaz, 2006; Harmon & Boeringer, 2004; Morse, 1994). Thus, a general scale of
preference for homophily that is based on what participants believe they utilize to make
homophily judgments, rather than what researchers think is utilized to make homophily
judgments, is important in order to accurately measure preference for homophily.
Furthermore, although the research suggests that participants prefer homophilous others
in friendships (e.g., Atkinson, Brady, & Casas, 1981; Griffitt & Veitch, 1974), romantic
relationships (e.g., Bentler & Newcomb, 1978; Murstein & Beck, 1972), and word-of-mouth
sources (e.g., Brown & Reingen, 1987; Gilly, Graham, Wolfinbarger, & Yale, 1998; Wang,
Walther, Pingree, & Hawkins, 2008; Wright, 2000), are the characteristics that individuals prefer
a target to share with themselves dependent upon the situation? If preference for homophily is
situationally defined, under which circumstances is it most desired by participants?
However, it is not enough to only understand the circumstances of preference for
homophily. In order to understand preference for homophily, researchers must understand why it
occurs. Thus, we need to understand the motivation behind this preference for homophily. Is it
out of a desire for congruity between activities, as suggested by Warner and Parmalee’s (1979)
findings that friends were actually homophilous regarding activities, but not on other factors? Is

31

the preference for homophily based on a desire for approval (Byrne, 1961b; Erwin, 1981;
Posavac, 1971) ?
The current study focuses on the conscious aspects of preference for homophily because
participants are unable to describe their unconscious desires. Furthermore, the research has
suggested that perceived homophily (which is a conscious evaluation of homophily) is more
important in predicting tie strength and relationship formation than actual homophily (Montoya,
Horton, & Kirchner, 2008), and thus the conscious aspects of homophily are particularly relevant
to this study. Thus, we need to understand whether or not individuals are aware of their
homophilous tendencies, when (situationally) individuals demonstrate a preference for
homophily, and how they make their homophilous judgments. The following research questions
are proposed to help better understand these issues.
RQ1: Do actors believe they have a preference for similarity in selecting friends,
associates, word-of-mouth sources, or important others?
RQ2: Does preference for similarity vary depending upon the situation, or stay constant?
RQ3: What do actors mean by similarity when they refer to friends and associates?
To date there has been a paucity of research regarding why homophily is prevalent in
interpersonal relationships. Researchers have been able to determine that homophilous
preferences are prevalent in society across several different types of relationships (e.g., Bentler &
Newcomb, 1978; Byrne, 1961a; 1961b; Byrne & Nelson, 1965; Carli, Ganley, & Pierce-Otay,
1991; Duck, 1975; Fiore & Donath, 2005; Lutz-Zois, Bradley, Mihalik, & Moorman-Eavers,
2006; Morry, 2004; 2005; 2007; Schrock, 2007; Thelwall, 2008). However, researchers have not
addressed the issue of why. Thus, the following research question is proposed:

32

RQ4: Why do individuals prefer to spend time with and get information from homophilous
rather than heterophilous sources?
This study measures involvement in order to understand situational preferences for
similarity. However, there may be stable differences between individuals that affect preference
for similarity across all situations. Thus, to explore this relationship between individual
characteristics regarding preference for similarity and situational preference for similarity, the
following hypothesis is proposed:
H1: An individual’s preference for homophily will be positively related to homophilous
tendencies when selecting word-of-mouth sources.
Much research has demonstrated that individuals are more interpersonally attracted to
similar others (e.g., Allgeier & Byrne, 1973; Byrne & Griffitt, 1966; Crosnoe, Frank, & Mueller,
2008; Neimeyer & Mitchell, 1988; Weare, Musso, & Jun, 2009) and prefer to receive advice
from similar others (e.g., Brown & Reingen, 1987; Wang, Walther, Pingree, & Hawkins, 2008).
Thus, the following hypothesis is proposed:
H2: Perceived similarity will be positively related to source likeability
Word-of-mouth communication is a form of external information search. Specifically,
individuals seek out more information from media, advertising, friends, or other sources when
they do not have enough information to make a decision on their own (Hawkins, Mothersbaugh,
& Best, 2007; Solomon, Zaichkowsky, & Polegato, 2003). Furthermore, in low involvement
situations in which individuals are less likely to exert effort in order to seek out information from
word-of-mouth sources, they will pay more attention to the information sources that are already
available to them (Kotler, Keller, Sivaramakrishnan, & Cunningham, 2013).

33

The extent of an information search is influenced by many factors: product class
knowledge, time availability, purchase involvement, and attitudes toward shopping (Beatty &
Smith, 1987). Furthermore, solicitation of advice from interpersonal sources is positively
correlated with purchase involvement and attitudes toward shopping, while personal knowledge
is negatively correlated with solicitation of advice from interpersonal sources.
An extensive external search is a process that delays the purchase decision. Greenleaf and
Lehmann (1995) explored the reasons why individuals delay making a purchase decision, and the
reasons why individuals end their search and purchase the product. Results indicated that, among
other reasons, many people delayed their product choice because they needed advice or consent
from another individual before making a purchasing decision, or because they did not have
sufficient information in order to make an informed decision. Delay was also correlated with
financial or social risk, such that individuals tended to delay the purchase of items when they
wanted to be certain that they were making the right decision. Word-of-mouth advice is sought
out and solicited in order to aid in gathering information, and therefore word-of-mouth advice
that ends or shortens an individual’s purchase delay is considered successful. Thus, word-ofmouth success can be measured by intentions to follow the advice of the individual. Furthermore,
word-of-mouth is truly successful at changing an individual’s opinions when they end the
external search following communication, indicating that consumers will intend to purchase the
product recommended without engaging in additional research. Therefore, a more effective
source will result in higher intentions to purchase the recommended product and lower intentions
to continue in an external search. Since similar sources tend to be more persuasive than
dissimilar sources (e.g., Brown & Reingen, 1987; Gilly, Graham, Wolfinbarger, & Yale, 1998;

34

Wang, Walther, Pingree, & Hawkins, 2008; Wright, 2000), the following hypotheses are
proposed:
H3: Perceived similarity will be positively related to likelihood of following advice
H4: Perceived similarity will be negatively related to additional search processes.
Source credibility (Hovland & Weiss, 1951; Kelman & Hovland, 1953) and its three
facets: expertise, trustworthiness, and likeability, has been utilized in the word-of-mouth
literature to explore word-of-mouth source selection. The source credibility literature suggests
how word-of-mouth sources are evaluated, but some research indicates that source characteristics
are not always relevant when evaluating the quality of a communication. The elaboration
likelihood model (Petty, Cacioppo, & Goldman, 1981) discusses the effect involvement has on
word-of-mouth source evaluation. Involvement increases the likelihood that an individual will
utilize more cognitive resources to evaluate a source, thus resulting in central cues to source
credibility being analyzed. Thus, source likeability (a peripheral cue to credibility) will be less
instrumental in evaluating source credibility in high, rather than low, involvement situations.
However, source credibility can be evaluated centrally when source credibility is relevant to the
decision (Petty & Cacioppo, 1984). Source expertise is likely to be a more central source
credibility cue, and will thus be more important in high involvement situations than low
involvement situations. High involvement will likely lead to a practice of optimizing (choosing
the best alternative) rather than satisficing (choosing an alternative that is good enough) (Simon,
1990), and thus word-of-mouth sources will be evaluated in a more stringent manner. In order to
test how involvement influences source credibility’s role in the decision-making process, the
following hypotheses are proposed:

35

H5: Preference for similarity will be higher in low, rather than high involvement
situations when predicting likelihood of following advice.
H6: Preference for similarity will be higher in low, rather than high involvement
situations when predicting additional information search processes.
H7: Likeability will be more important in low, rather than high involvement situations
when predicting likelihood of following advice.
H8: Likeability will be more important in low, rather than high involvement situations
when predicting additional information search processes.
H9: Expertise will be more important in high, rather than low involvement situations
when predicting likelihood of following advice
H10: Expertise will be more important in high, rather than low involvement situations
when predicting additional information search processes.

36

METHODOLOGY

As stated previously, there is no scale to measure preference for homophily in the
existing literature. As such, before engaging in any research regarding preference for homophily,
a measurement tool needed to be created. Therefore, per the recommendations of grounded
theory (e.g., Charmaz, 2006; Harmon & Boeringer, 2004; Morse, 1994), the first step of this
study was to create a measurement tool. In order to create a set of items for testing, a number of
open-ended interviews were conducted. This stage of the research gave the researchers a set of
items to be tested, rather than McCroskey’s (McCroskey, McCroskey, & Richmond, 2006;
McCroskey, Richmond, & Daly, 1975) method of simply listing variables previously utilized in
research or of theoretical relevance to proposed research. Furthermore, these open-ended
interviews aided in determining when and why preference for homophily existed in friendships
and in the decision-making process.
After creating a set of items based on the open-ended interviews, a questionnaire format
was utilized to determine which items needed to remain in the preference for homophily scale,
and the factor structure of the scale. This was undertaken in an iterative process, such that a
preliminary survey informed the factor structure, and a second survey validated the structure of
the scale.
After the instrument had been created, a survey was undertaken to determine the effect of
purchase involvement on preference for homophily and source credibility of the advisor. Data
was collected online utilizing a sample of the same demographic characteristics as the samples
for the instrument creation – American residents between the ages of 25 and 35, with a fairly
even gender split.

37

Study 1
Overview
Study 1 was focused on developing a set of scale items to measure preference for
homophily based on items generated in qualitative interviews, determining whether or not
individuals are aware of their homophilous tendencies and preferences, and understanding why
homophilous preferences were present in interpersonal relationships and word-of-mouth
information searches. This study involved three phases. The first phase consisted of a set of
open-ended qualitative interviews to investigate the situations in which homophilous preferences
were activated, the characteristics that individuals evaluated when making homophilous
judgments, and the reasons why homophily was important. These interviews were transcribed
and analyzed for themes regarding the four research questions related to homophilous
preferences. These interviews were the basis for the development of a preference for homophily
scale, which was validated in phases two and three of this study. The next phase involved the
development of a quantitative questionnaire to test the scale items created in phase one. Analysis
of this scale involved an analysis of the alpha change statistics for each scale item and an
exploratory factor analysis. Following this analysis, an analysis of means was utilized to
determine which items were included in the final scale that was tested in phase three of study one.
Phase three involved a validation of the preference for homophily scale utilizing alpha change
statistics and a confirmatory factor analysis.
Procedure
Open-ended interviews were conducted in person and on the phone. Consent forms were
emailed to all phone participants before the interview, and presented to all face-to-face

38

participants upon arrival. As stipulated in the consent form (Appendix A), participation in the
interview indicated consent to the terms. The interviews took between 8 and 22 minutes each.
All participants were assigned a pseudonym, and no identifiable data was stored for the
participants. See Appendix A for a copy of the interview protocol.
Questions utilized to create the preference for homophily scale were designed to elicit a
set of characteristics based on the similarity measurement procedure set out by Tversky (1977),
which is a conscious measure of similarity. Specifically, Tversky (1977) stated that evaluations
of similarity were based upon a weighted sum of the similarities and differences that individuals
shared. Characteristics that individuals shared were weighted and added to the overall similarity
score, and characteristics on which individuals were different were weighted and subtracted from
the overall similarity score. Thus, Tversky (1977) asserted that perceptions of similarity were
formed by judging similarity and dissimilarity across a wide variety of characteristics, and
therefore this study’s questions asked participants to describe their friends and sources based
upon not only those characteristics that they shared, but also those characteristics that they did
not share. In order to build a list of characteristics to create scale items, all interview transcripts
were analyzed for themes related to the reasons for homophilous preferences, the characteristics
participants utilized to describe homophily, and the situations in which preference for homophily
was more or less prevalent for participants. Instances where individuals discussed characteristics
of importance in judging similarity / dissimilarity and instances in which individuals discussed
why they thought homophily was important were identified and tagged. Following this process,
two independent coders coded each tagged interview excerpt according to a codebook that
included all of the themes identified in the analysis of the transcripts, with 85% intercoder

39

agreement. Specifically, coders agreed on the theme of the tagged instance 85% of the time.
Scale items were then developed based on the characteristics identified throughout this process.
Data for phases 2 and 3 of the scale development was collected online. In phase two,
participants were asked to fill out a set of likert-type scale items based upon the results of the
individual open-ended interviews in order to form the scale. See Appendix B for a copy of the
questionnaire utilized. Data was analyzed utilizing inter-item correlations, alpha change statistics,
and an exploratory factor analysis. Phase three validated the scale by asking participants to fill
out the scale (see Appendix C for a copy of the questionnaire) and, once again, analyzing interitem correlations, alpha change statistics, and the alpha reliability statistic. This scale was aimed
at measuring preference for homophily and thus individuals were asked not to describe a source,
but to describe how important it was to them for a source to share each characteristic with them.
Study Two
Overview
Study two was focused on determining how source credibility, preference for homophily,
and perceived homophily influenced how individuals chose who to go to for advice about a
purchase. Furthermore, it was utilized as a further test of the preference for homophily scale
developed in study one. Involvement was studied as a moderating variable to determine its
influence on which characteristics of source credibility would be important to consumers, and
how homophilous sources were to the consumer. In order to accomplish this, preference for
homophily was measured, in addition to the consumer’s perceptions of their chosen word of
mouth source’s perceived homophily, expertise, trustworthiness, and likeability.
All data for this study was collected utilizing an online survey application, Zoomerang.
The survey was hosted on the Zoomerang website and participants were randomly selected from
40

the Zoomerang panel. Before filling out the survey, participants were informed of their rights as
a research participant such that they had the right to refuse to participate without consequence,
and they had the right to refuse to answer any question. Furthermore, they were given contact
information for the researchers and the Michigan State University Human Research Protection
Program, should they have any questions or complaints. Participants were informed that by
clicking the ‘Submit’ button they were indicating their consent to participate in the study. The
first four hypotheses were tested utilizing simple linear regressions. The models proposed in
figure 2, which encompassed hypotheses 5 – 10, were tested utilizing two linear regression
analyses that included interaction terms. Significant interaction terms were then analyzed
individually utilizing a one-tailed t-test to compare the means of the dependent variable for the
high and low involvement groups.
Measurement
Involvement was measured utilizing Zaichkowsky’s (1985) 20-item Likert-type product
class involvement scale. This scale has been utilized in many academic studies (e.g., Bearden,
Netemeyer, & Teel, 1989; Beatty & Smith, 1987; Celsi & Olson, 1988; Mano & Oliver, 1993;
Spreng, MacKenzie, & Olshavsky, 1996). This scale was selected due to its excellent reliability
statistics and its common use in psychological, sociological, and communication studies.
Preference for homophily was measured based on the items identified in Study one of the
research. The items focused not on describing an actual word-of-mouth source, but asked
individuals to rate their level of preference for a source to be similar to them regarding each item
identified in study one. This was a measure of conscious preference for homophily, and was a
further test of the construct validity of this new scale.

41

Perceived homophily with a source was also measured based on the items identified in
study one of this paper. These questions asked them to describe how similar their source was to
themselves on each of the items identified for the preference for homophily scale in study one
(e.g., values, personality, sense of humour). This was a measure of perceived, not actual,
homophily with a word-of-mouth source. However, since perceived homophily has been
demonstrated to be a better predictor of relationship formation and attraction (Montoya, Horton,
& Kirchner, 2008) and preference for homophily was measured in a conscious manner,
perceived homophily was appropriate. This scale was adapted to measure perceived homophily
by asking participants to report how similar or dissimilar their sources were to themselves rather
than reporting how important it was to them for their sources to be similar to them on each
characteristic. This scale was utilized in order to ensure that this measure was also based on the
characteristics that were important to consumers, rather than based on what the researchers
thought was important. Measuring perceived homophily and preference for homophily also
helped to confirm the scale’s construct and predictive validity by testing its ability to predict the
relationship between perceived homophily and the influence of a source – which has been
demonstrated many times in previous literature (e.g., Brown & Reingen, 1987; Bruyn & Lilien,
2004; Dellande, Gilly, & Graham, 2004; Gilly, Graham, Wolfinbarger, & Yale, 1998; Steffes &
Burgee, 2009, Wang, Walther, Pingree, & Hawkins, 2008; Wright, 2000).
Intentions to engage in additional search and intentions to follow through with a source’s
recommendation were measured based upon Bearden, Lichtenstein, and Teel’s (1984) intentions
scale. This scale was selected due to its simplicity, short length, and its high reliability statistics
in previous studies (e.g., Li, Daugherty, & Biocca, 2002).

42

Source credibility was measured based upon its three individual dimensions:
trustworthiness, expertise, and likeability. In order to measure trustworthiness and expertise,
Ohanian’s (1990) 5-item Likert-type scales were utilized. This scale was utilized due to its short
length and its pervasiveness in the literature (e.g., Goldsmith, Lafferty, & Newell, 2000; Lafferty
& Goldsmith, 2004; Senecal & Nantel, 2004; Sun, Youn, Wu, & Kuntaraporn, 2006). Ohanian
(1990) also created an attractiveness scale, but the items were related to physical attractiveness
rather than interpersonal attractiveness which was less relevant to the current study. Thus, the
Reysen Likeability Scale was utilized to measure likeability (Reysen, 2005), as it measures
likeability on both sociability and physical attractiveness. It is less pervasive in the literature than
other likeability scales, but reliability statistics were good and the items were more relevant to
the current study than other likeability scales.

43

RESULTS

Study One Phase One
Participants
Twenty-one interviews were conducted in May of 2011 by the author. Interviews lasted
between 8 and 22 minutes each, and were based on the interview guide in appendix A. All
participants were between the ages of 25 and 35. Eleven of the interviewees were male.
Seventeen of the interviewees were Caucasian, 3 Asian, and 1 Black. Note that all participants’
names were changed to protect their anonymity.
Research Question 1: Do actors believe they have a preference for similarity in selecting friends,
associates, word-of-mouth sources, or important others?
Most participants admitted that their friends were quite similar to them, and that
similarity was at least somewhat important to them. However, they seemed to indicate that
similar interests were more important than similar personalities. For example, John said “It’s
important to have similar interests. It’s not important to be similar people. Because the
differences are what makes everyone unique and it’s generally differences that’ll catch
someone’s eye a lot quicker than similarities will.” He said that these differences caught a
person’s eye because they were exciting and offered fresh, interesting new perspectives and
activities to consider. Other participants often made similar assertions – they tended to prefer to
spend time with people who had similar interests to their own, but personality differences helped
offer fresh perspectives on situations, and different activity interests took them out of their own
‘comfort zone’. There were some participants who stated that similarity was unimportant to them,
44

but then admitted that they preferred spending time with people who had similar interests,
because it made choosing activities to do together simpler and it allowed them to have
conversations with their friends about things that they found interesting.
One participant, however, had spent a large amount of time in the military and had a
different perspective. As Andrew said “I don’t think it’s as important as we thought it was when
I was growing up … you don’t have to have similarities. Because in the end sometimes you
adopt some of the things that you never knew about so… I think no, especially if you have
people that are open to trying new stuff and going outside that box then no you don’t have to be
too similar.” He explained that, when he was growing up, his closest friends were all very similar
to him. However, since joining the military, he had been forced to befriend people with different
backgrounds, interests, and personalities. After this military experience, he tended to seek out
friends who were different from him because he had grown to appreciate those differences and
the new experiences and perspectives that they offered him. However, his closest friends were
still those friends from childhood who were extremely similar to him. Thus, preference for
homophily seems to be present in all individuals, even those who do not claim to have a
preference for similarity. However, experiences and other factors may influence how important
this characteristic is, and how consciously individuals admit to a preference for similarity.
Research Question 2: Does preference for similarity vary depending upon the situation, or stay
constant?
Regarding friendship, participants stated that during times of stress, similarity was more
important. Specifically, an individual who had a shared background, and who had been through a
similar experience, was better able to lend support to the participant. For example, Sheila said
that “Say there was a death in the family, for example. You would need somebody that is used to

45

that kind of situation and would be able to console you. Say you had a birth in the family, or you
were giving birth, you need somebody that’s close to that situation to help you out.” Furthermore,
John stated that “Similarities would be more important during uneasy times. Times of high stress.
Or times, fights or disagreements, generally the similarities will be what the group would fall
back on to stay together whereas the differences would be what forces it apart.” When probed
further, John said that this was because friends would be able to fall back upon similar interests
to share time together doing something they all enjoy, whereas disagreements would cause
discord regarding what to do and how they communicate. Thus, all participants who spoke about
situations that would influence the importance of similarity between friends discussed the
importance of similarity during times of stress, be it personal stress or disagreements between
friends.
Research Question 3: What do actors mean by similarity when they refer to friends and
associates?
When asked what sorts of things they had in common with their friends, and what sorts of
differences they had with their friends, there were several categories of responses. People either
spoke of personalities or tastes first. For example, people often stated that they liked similar
movies, television shows, and music as their friends. They also spoke of how outgoing or quiet
their friends were, and how temperamental they were. Belief systems were also discussed on
several occasions, with political and religious beliefs being discussed the most often.
Demographic variables were also discussed, particularly regarding gender and age.
When trying to categorize characteristics, a pattern emerged. In general, there were some
characteristics that were easy for participants to ascertain based on little or no interaction, and
other characteristics that participants would have been required to spend time with the other

46

person in order to determine similarity. Demographic characteristics and personal appearance
were the easiest for participants to identify, but were also discussed only peripherally –
indicating that individuals were aware of similarity on these characteristics, but it was not
important enough for them to discuss these characteristics in-depth. Alternatively, concerns
regarding the social desirability of admitting a preference for homophily on certain
characteristics may have prevented participants from discussing it.
Furthermore, there were some categories of characteristics that were easy to identify
given a short conversation, such as similar backgrounds, similar familial situations, and similar
social networks. These items were discussed slightly more often, and were often mentioned as a
way for individuals to meet and for individuals to spend time together. For example, having
similar friendship networks allowed for people to spend time in social gatherings, and having
similar familial statuses allowed for play dates with children and double dates with other couples.
As stated previously, the most common characteristics discussed had to do with interests.
Participants discussed what they had in common with their friends and information sources with
regards to activities, television shows, movies, and several other entertainment mediums. These
were also the characteristics that they seemed to find particularly important to be similar upon, as
it helped them engage in activities and discussions that were interesting to both parties. This set
of characteristics required a deeper level of knowledge of the individuals than previously
discussed characteristics, and it appeared to be of particular importance. However, these
characteristics are fairly easy to ascertain during a brief conversation.
These first three categories: physical and demographic, relational networks and familial
status, and activities were all listed as ways of aiding in an introduction. Thus, mutual friends,
similar interests, and similar physical appearance were all mentioned as a way to facilitate

47

introductions and meet new people. However, based on discussion of these traits, interests were
more likely to be important to maintaining a relationship than the other characteristics.
The fourth category of characteristics that was discussed almost as often as interests had
to do with personality and values. Participants often stated that they were similar to their friends
with regards to humour, temperament, and degree of extroversion, and regarding their attitudes
toward world issues such as gay marriage and war. This was the first set of characteristics that
did not facilitate an introduction, but it was discussed almost as often as similar activities to
facilitating the maintenance of a relationship. Participants indicated that similarity regarding
these characteristics made it a more comfortable interaction in which they felt that they could be
themselves, and in which they felt that their feelings and thoughts were not only understood but
accepted, because their interaction partners felt the same way. This particular set of
characteristics may facilitate individuals’ need for approval (Byrne, 1961; Posavac, 1971), as
having friends with similar opinions facilitates agreement and support for those values.
Research Question 4: Why do individuals prefer to spend time with / get information from
homophilous rather than heterophilous sources?
As previously stated, participants tended to emphasize the importance of similar interests
even more than similar personalities in their friends. The reasons why similarity was important
were reflective of this tendency. People stated that it was easier to find things to talk about and
easier to find things to do with friends who were similar to them. For example, Jane said that
“Well, [similarities] just give you something to talk about more than the weather.” Furthermore,
Samantha stated that “I think it helps if you guys are similar, you know, you have more to talk
about.” Tori also stated that people with similar interests required less background information in
order to understand what she was trying to say, and therefore the conversation tended to flow

48

more easily. For example, if an individual was interested in cooking and attempting to explain
his or her latest exploit, a person who was knowledgeable about cooking would require fewer
definitions and explanations, and the emphasis could be on the actual experience of cooking and
not on the technical aspects of the activity. Thus, similarities give people common ground to start
a discussion, and also help them share common experiences. As Tamara said “I figure if we’re
going to be spending time together we should be doing something I like to do. That we both like
to do.” Even Andrew, who strongly believed that similarity was unimportant, stated that it was
easier to find things to do together when they shared common interests.
Tamara also stated that it was easier to meet people who were similar to her “… say I’m
at the gym or something and I see someone who I already know, it’s easier to develop those
friendships when you’re in the same places.” Therefore, not only do similarities facilitate
conversation and activity choice, similarities facilitate introductions as well. As such, most
participants believed that friendships developed between similar people because these
similarities allowed them to meet each other, and share experiences and conversations that led to
more meaningful relationships.
Although most people stated that similar interests were the most important thing to share,
leading to shared experiences and conversations, Nick valued a similar sense of humour above
all else. He said that this allowed his friends to joke around and feed off the energy of each other,
and also let them be themselves without a fear of being judged. This is similar to the need for
approval (Posavac, 1971) and need for affiliation (Byrne, 1961b) findings regarding homophily,
which suggest that a desire to be similar to one’s friends is rooted in a desire to be well-liked or
accepted. As stated above, similar personality and world views facilitate this feeling of comfort
and acceptance, and thus deepen the relationship. Furthermore, these findings are in line with the

49

findings of Erwin (1981), who discovered that perceived preference of a target toward a
participant led to more homophilous and favourable perceptions toward the target.
Participants were also asked how they chose their shopping partners. While many
participants indicated that social intimacy and entertainment were factors when they went
shopping with other people, several still had comments on when similarity or dissimilarity would
be desirable in a shopping partner. When asked specifically about similarity regarding shopping
partners, most people said that similarity was desirable because these partners better understood
their needs, and were able to give better advice. As Amanda said “It’s great to have somebody
who likes the things you like.” Amanda was speaking of clothing shopping, and said that, for
these sorts of purchases, it was nice to have somebody with her who could give her advice about
what suited her own sense of fashion. She also said that it was nice that they liked the same sorts
of stores, and tried on similar styles to each other, making the shopping experience simpler.
Henry also said that, at times, he just wanted somebody to agree with him, and so
similarities were important to ensure that he would receive this agreement “If I take somebody
shopping with me I usually want the person to say ‘yes’. And usually it’s because I’m unsure.
Otherwise I would almost definitely go for it myself.” He said that this tended to happen for
high-involvement items, and thus this desire for reassurance was the “human condition” and that
we (humans) want assurance that we’re making the right choice when we make risky decisions.
Although to a lesser extent, other participants expressed a desire to have people with similar
interests so that the person would encourage them to purchase something that they really wanted.
This desire for reassurance is similar to the desire for acceptance mentioned earlier for friends,
which was mostly driven by similarities regarding values and personalities when discussing
friendships.

50

However, for big-ticket items and impulse buys, participants stated that differences were
often good to offer a different perspective. For example, Brent said that “Maybe slightly
differences [are] more [important] if I am making kind of an impulse buy and they are giving me
good reasons why I shouldn’t make an impulse buy.” When discussing a car purchase, Janet said
that she valued differences because she believed that helped her make a more informed decision
“I didn’t want somebody who’s just going to confirm what I think, I certainly want some
differences of opinion. I think, if I think somebody is just agreeing with me that’s not very
helpful. I’d rather have somebody kind of offer me some critique of what I’m thinking or offer
me counterpoints because I think I can then have a fuller picture of all my options and I think I
can make a more informed decision.” Thus, depending on the purchase situation, there is a
different value placed on similarities and differences. Impulse buys and big-ticket items inspire a
desire for dissimilarities in order to make a more informed decision, while smaller-ticket items
such as clothing inspire a preference for similarity so that people will make recommendations
that are appropriate for the individual.
Differences in knowledge and experience were also desirable for big-ticket items. In
these cases, people didn’t just want a different perspective - they wanted a more informed
perspective. Thus, the most common reason why participants chose a friend as an advisor for a
big-ticket item was because they had experience with the product, or because they had
knowledge that surpassed that of the participant. For example, Jane spoke of purchasing a
videogame console “I think in this case their differences are more important because I don’t have
that knowledge so I’m going to them because they have that knowledge.” Tori spoke of
purchasing her GPS, and said that although she relies on herself and Internet research for most
purchases, she did not have the information about which features were important when making a

51

new GPS purchasing decision and so she asked her friends who had a GPS which features they
thought were necessary. In this case, it was not her friends’ expertise that made them a desirable
advisor, but their experience with the product. James spoke of the strength of weak ties
(Grannovetter, 1973), stating that weak ties may have access to resources (deals) that he did not,
and thus they became ideal advisors in situations where their unique resources would help
accomplish a goal.
Although participants often spoke of differences being desirable in purchasing situations,
many participants also stated that differences between friends allowed them to experience new
things that they otherwise would not have tried. For example, Tamara said that her boyfriend’s
dissimilar interests had led to her finding some new interests regarding movies and television
shows “Differences are good with everyday things like going to movies cause there’d be some
movies that I would say that I would have never seen unless he said ‘hey let’s watch it’ and then
it turns out I actually like them.” Other friends spoke of similar experiences when their friends’
dissimilar preferences led to them discovering new interests (e.g., food, music, television) that
they very much enjoyed. However, it should be noted that these differences in experience and
taste eventually led to a shared interest, thus increasing homophilous interests between the two.
Although none of the participants stated that they preferred differences for this reason,
they stated that it was a possibility that some people would value differences in order to inspire a
healthy debate. Thus, in that case, differences in political, religious, and other belief systems
would allow them to be involved in a spirited discussion of their views. However, a mutual
enjoyment of debate would be required in order for a friendship to flourish in this situation.

52

Study One Phase Two
Participants
301 participants were recruited from a Zoomerang panel, which has been demonstrated to
be demographically similar to the United States population (Zoomerang, n. d.). All participants
were between the ages of 25 and 35, and were notified of the survey opportunity via email.
Zoomerang participants were rewarded 100 zoom points in return for survey participation. See
table 2 for a demographic description of the participants.
Table 2: Study 1 Phase 2 Participant Demographics
Percent

N

Gender
Male

43.1%

129

Female

56.9%

170

Caucasian

71.4%

212

9.4%

28

1%

3

Asian

8.1%

24

Pacific Islander

0.7%

2

Hispanic / Latino

6.4%

19

Multiracial

2%

6

Don’t want to disclose

1%

3

8.4%

25

Full-time

53.5%

159

Part-time

8.4%

25

22.2%

66

Student

7.4%

22

Elementary School

0.3%

1

Race
African American
Native American

Employment
Self-employed

Not Employed
Education

53

Table 2 (cont’d)
Some High School

2%

6

High School

16.2%

48

Some College

25.3%

75

Associate Degree

11.8%

35

Bachelor’s Degree

29.3%

87

Master’s Degree

11.4%

34

3%

9

0.7%

2

Under $10,000

27.2%

80

$10,000 - $20,000

10.9%

32

$20,000 - $35,000

21.1%

62

$35,000 - $50,000

22.4%

66

$50,000 - $75,000

10.2%

30

$75,000 - $100,000

5.4%

16

Over $100,000

2.7%

8

Under $10,000

11.5%

34

$10,000 - $20,000

6.4%

19

$20,000 - $35,000

16.9%

50

$35,000 - $50,000

20%

59

$50,000 - $75,000

23.4%

69

$75,000 - $100,000

12.2%

36

Over $100,000

9.5%

28

Now Married

49%

147

Divorced

4.7%

14

Separated

1%

3

45.3%

136

Professional Degree
Doctorate Degree
Personal Income

Household Income

Marital Status

Never Married

54

Scale Development
Based on the interviews outlined above, a set of 45 Likert-type scale items were created
to measure preference for homophily, with 50% of the items phrased such that they needed
reverse coding before analysis – a tactic utilized to reduce response bias (Kerlinger & Lee, 2000).
The selection of which items were reverse coded was done on a random basis utilizing a random
number generator. However, certain items made little sense after reverse coding, and thus other
items were selected instead utilizing a random number generator once again. Furthermore,
questions were asked about whom individuals would choose as friends as well as whom
individuals would go to for advice. The same procedure as that described above was utilized to
select which items would be written about friends, and which items would be written about
information sources, such that random assignment was used up to the point that items made
sense to the participants. Some items could not be reverse coded. For example “I tend to spend
more time with friends who have similar relationships styles to my own” would have made little
sense as a reverse coded item. All scale items were coded such that a higher score on the 7-point
Likert-type scale indicated a higher preference for homophily. See appendix B for a copy of the
questionnaire utilized, including the consent form, demographic questions and scale items.
After recoding of negatively phrased items, data analysis was performed. Before factor
analysis was done, all items were subjected to an analysis of skewness and variance to ensure
that they varied enough to be predictive of preference for homophily. All items had a skewness
of less than 1, answers ranging across the scale of 1 – 7, and a variance between 1 and 2.5.
Although responses tended to be clustered around the midpoint, 4, titled “Neither agree nor
disagree”, all of the items were deemed fit to be included in the factor analysis, based on the
recommendations of Clark and Watson (1995). After this was established, all items were

55

removed from the scale with item-total correlations of less than 0.2, eliminating 12 items from
the scale. Thus, all items that did not fit well within the scale were eliminated in this step.
Next, an exploratory factor analysis was conducted to analyze the structure of the
remaining items. Two factors resulted. The data were analyzed utilizing unrotated factors,
varimax rotation, and oblimin rotation. Each rotation was performed utilizing both standardized
and unstandardized scores. Furthermore, multiple data screening techniques were utilized to
eliminate cases, such as removing all cases in which more than five responses in a row were the
same. Although the items that remained in the scale after eliminating all items with crossloadings, each factor analysis resulted in only two factors – one factor for items phrased such
that similarity was preferable, and one factor for items phrased such that dissimilarity was
preferable. This bifurcation of positively and negatively worded items is common, however
(Schmitt & Stults, 1985). As such, it was determined that this was a unidimensional scale, and
that the wording (positive / negative) was the only difference between the two factors and a
decision was made to treat this variable as unidimensional. Per Torgerson (1958), this scale is
transitive such that a higher score on the preference for homophily scale denotes a higher degree
of preference, and a lower score denotes a lower degree of preference for homophily. This is the
main requirement and description of a unidimensional scale.
Once it was determined that this was a unidimensional scale, eliminating items due to
cross-loadings was deemed inappropriate, and these items were added back into the original
scale. Analysis utilizing t-tests indicated that some items were no different from the midpoint, or
had an average rating below the midpoint. Since these items were correlated with a low amount
of variance and were not of importance to participants when forming homophilous judgments

56

and relationships these items were eliminated from the scale before further analysis, leaving a
total of 22 items in the scale.
The unidimensional scale had an α = 0.845. Refer to table 3 for factor loadings for each
2

item in the scale. The factor analysis had a significant sphericity ( = 4010.023, p < 0.001)
indicating that the items have at least some level of correlation with the other items. Furthermore,
the Kaiser-Meyer-Olkin test of sampling adequacy indicated that correlations between items
could be explained by other variables (KMO = 0.927). Although not all items had loadings of 0.6
or higher, they contributed to the alpha reliability statistic and were greater than 0.45, and thus
remained in the scale. Some items had significant crossloadings (above 0.4), but due to the
unidimensional nature of the scale they were kept.
Table 3: Study 1 Phase 2 Item Factor Loadings, Means, and Standard Deviations
Factor 1

Factor 2

Item

Loading

Loading

Mean

SD

Sense of Humour
Personality
Attitude Towards World
Issues
Attitude Towards Drinking
Intelligence
Attitude Towards Gaming
Life Goals
Taste in Books
Taste in Movies
Family Values
Attitude Towards Education
Relationship Style
Way of Handling Money
Knowledge of Technology
Taste in Friends
Family Situation

0.44
0.566

0.233
0.134

5.39
4.76

1.223
1.302

0.765
0.594
0.737
0.729
0.698
0.693
0.755
0.702
0.703
0.72
0.603
0.699
0.766
0.595

0.140
0.211
0.211
-0.106
0.285
-0.025
0.118
0.386
0.246
0.109
-0.035
-0.124
0.142
0.121

4.76
4.98
4.88
4.47
4.84
4.39
4.69
5.12
4.98
4.71
4.58
4.44
4.59
4.58

1.206
1.154
1.22
1.224
1.15
1.215
1.207
1.236
1.195
1.234
1.238
1.088
1.131
1.277

57

Table 3 (cont’d)
Political
Attitude Towards Gay
Rights
Attitude Towards Children
Attitude Towards Sports
Values
Religion

-0.264

0.61

4.49

1.43

-0.245
-0.274
-0.296
-0.246
-0.329

0.754
0.772
0.688
0.776
0.731

4.52
4.37
4.31
4.56
4.42

1.525
1.333
1.267
1.421
1.462

It is worth noting that most of the items that remained in the scale were related to
activities, personality, and value systems – those items that were most talked of in the interviews.
As such, the items that remained in this scale tended to be the most universally discussed in the
interviews, and the most salient for interviewees. Furthermore, people tended to admit to
preferring similar individuals based on those characteristics that interviewees from phase one
were able to justify – specifically, people admitted a preference for homophily based on traits
that interviewees spoke of when they were asked why homophily was important.
Study One Phase Three
Participants
301 participants were recruited from a Zoomerang panel, which has been demonstrated to
be demographically similar to the United States population (Zoomerang, n. d.). All participants
were between the ages of 25 and 35, and were notified of the survey opportunity via email.
Zoomerang participants were rewarded 100 zoom points in return for survey participation. See
table 4 for a demographic description of the participants.

58

Table 4: Study 1 Phase 3 Participant Demographics
Percent

N

Gender
Male

53%

158

Female

47%

140

76.8%

228

African American

7.1%

21

Native American

0.3%

1

Asian

6.1%

18

Hispanic / Latino

6.7%

20

Multiracial

2%

6

Don’t want to disclose

1%

3

Self-employed

11.7%

34

Full-time

58.3%

169

Part-time

7.6%

22

16.9%

49

Student

5.2%

15

Retired

0.3%

1

Some High School

0.7%

2

High School

16.3%

48

Some College

18.3%

54

Associate Degree

11.2%

33

Bachelor’s Degree

30.8%

91

Master’s Degree

16.3%

48

Professional Degree

3.1%

9

Doctorate Degree

3.4%

10

Race
Caucasian

Employment

Not Employed

Education

Personal Income

59

Table 4 (cont’d)
Under $10,000

16.3%

47

$10,000 - $20,000

13.2%

38

$20,000 - $35,000

16%

46

$35,000 - $50,000

19.8%

57

$50,000 - $75,000

14.6%

42

$75,000 - $100,000 10.8%

31

Over $100,000

9.4%

27

Under $10,000

6.1%

18

$10,000 - $20,000

7.8%

23

$20,000 - $35,000

16.7%

49

$35,000 - $50,000

15.7%

46

$50,000 - $75,000

20.8%

61

16%

47

Over $100,000

16.7%

49

Now Married

51.%

153

Widowed

0.7%

2

Divorced

3.7%

11

Separated

1.7%

5

42.4%

126

Household Income

$75,000 - $100,000
Marital Status

Never Married

Scale Development
Based on the analysis in study one phase two, 22 items remained in the Likert-type scale.
All items were phrased in the same way that they were phrased in the original test of the items,
resulting in a 22-item scale with 6 items that needed to be reverse-coded prior to analysis. See
appendix D for a copy of the questionnaire utilized, including the consent form, demographic
questions and scale items.

60

After recoding of negatively phrased items, data analysis was performed. All items had a
skewness of less than 1, answers ranging across the scale of 1 – 7, and a variance between 1 and
2. Although responses tended to be clustered around the midpoint, 4, all of the items were
deemed fit to be included in the factor analysis based on the recommendations of Clark and
Watson (1995). After this was established, item correlations and alpha change statistics were
analyzed for each item. This was analyzed in two groups: positively phrased items and
negatively phrased items. Based on this analysis, all of the items contributed to the alpha
reliability statistic, and all items had inter-item correlations of 0.5 or higher. Thus, all 22 items
contributed to and remained in the final scale. The reliability of the positively phrased items was
α = 0.939, and the reliability of the negatively phrased items was α = 0.928. The overall
reliability of the scale was α = 0.794, which is acceptable (Kerlinger & Lee, 2000). A
confirmatory factor analysis was utilized to analyze the scale, performed utilizing AMOS
Graphics 20. Refer to Figure 3 for the factor structure of the scale and Table 6 for a summary of
the scale’s path estimates. All paths were significant (p < 0.001). Some error terms were covaried
in order to improve the fit of the model, and are reported in Table 5. The final model had a CFI
of 0.902 and an RMSEA of 0.08, indicating a good fit for the model (Byrne, 2010).
Table 5: Study 1 Phase 3 Item Path Estimates, Means, and Standard Deviations
Factor 1

Factor 2

Standardized Standardized
Item

Path Estimate Path Estimate

Mean

SD

Sense of Humour

0.501***

5.43

1.271

Personality

0.638***

5.03

1.288

Attitude Towards World Issues

0.653***

5.05

1.235

Attitude Towards Drinking

0.592***

5.02

1.327

61

Table 5 (cont’d)
Intelligence

0.589***

5.14

1.203

Attitude Towards Gaming

0.678***

4.8

1.284

Life Goals

0.754***

5.01

1.178

Taste in Books

0.674***

4.4

1.299

Taste in Movies

0.757***

4.84

1.211

Family Values

0.723***

5.13

1.185

Attitude Towards Education

0.712***

4.88

1.25

Relationship Style

0.791***

4.84

1.271

Way of Handling Money

0.677***

4.74

1.353

Knowledge of Technology

0.744***

4.59

1.359

Taste in Friends

0.804***

4.75

1.279

Family Situation

0.684***

4.71

1.347

Political

0.842***

4.1

1.53

Attitude Towards Gay Rights

0.770***

4.3

1.745

Attitude Towards Children

0.742***

3.98

1.576

Attitude Towards Sports

0.720***

3.86

1.462

Values

0.848***

4.2

1.621

Religion

0.797***

4.02

1.6

*** p < 0.001

Table 6: Study 1 Phase 3 Error Covariances
Item 1

Item 2

Covariance

Factor 1: Positively Phrased

Factor 2: Negatively Phrased

Items

Items

Attitudes Towards Gays

Attitudes Towards Children

0.546***

Sense of Humour

Attitude Towards World Issues

0.292***

Sense of Humour

Attitude Towards Drinking

0.283***

62

-0.363***

Table 6 (cont’d)
Attitude Towards World Issues Attitude Towards Drinking

0.312***

Personality

0.258***

Attitude Towards World Issues

*** = p < 0.001

A series of products were tested in this questionnaire in order to determine a product
category for study two that had varying levels of involvement among participants (the
independent variable). Six products were tested in this pretest: shoes, jackets, watches, DVD
players, jeans, and sunglasses. Watches were selected for the purchase situation because they had
the highest level of standard deviation on a 35-point scale (SD = 7.47) and the histogram
indicated that there were several people in each of the two tails of the normal curve (see Figure
2). Due to the varying levels of involvement reported regarding this purchase it is likely that, for
some people, it was an impulse purchase. However, for most participants it was at least
somewhat involving and required a certain amount of elaboration and thought. A watch is worn
publicly, and is therefore a part of the image portrayed by the wearer. As such this purchase was,
to a certain degree, a status-oriented product. This may have influenced the way that individuals
selected their word-of-mouth advisors and envisioned the shopping experience. Involvement in
the purchase was then measured in study two in order to ensure that this variability of responses
remained and in order to test correlations between involvement and source characteristics.
Study Two
Participants
150 participants were recruited from a Zoomerang panel, which has been demonstrated to
be demographically similar to the United States population (Zoomerang, n. d.). All participants

63

were between the ages of 25 and 35, and were notified of the survey opportunity via email.
Zoomerang participants were rewarded 150 zoom points in return for survey participation. See
table 7 for a demographic description of the participants.
Table 7: Study 2 Participant Demographics
Percent

N

Gender
Male

49.3%

74

Female

50.7%

76

Caucasian

68.7%

103

African American

9.3%

14

Native American

1.3%

2

8%

12

Hispanic / Latino

6.7%

10

Multiracial

2.7%

4

Don’t want to disclose

3.3%

5

Self-employed

10.1%

15

Full-time

49.3%

73

Part-time

10.8%

16

Not Employed

21.6%

32

Student

6.1%

9

Retired

2%

3

2.7%

4

High School

20.7%

31

Some College

18.7%

28

Associate Degree

12.7%

19

Bachelor’s Degree

26%

39

Master’s Degree

14%

21

Race

Asian

Employment

Education
Some High School

64

Table 7 (cont’d)
Professional Degree

4.7%

7

Doctorate Degree

0.7%

1

29.5%

43

$10,000 - $20,000

8.9%

13

$20,000 - $35,000

18.5%

27

$35,000 - $50,000

17.8%

26

$50,000 - $75,000

9.6%

14

$75,000 - $100,000

8.9%

13

Over $100,000

6.8%

10

Under $10,000

12.8%

19

$10,000 - $20,000

5.4%

8

$20,000 - $35,000

16.9%

25

$35,000 - $50,000

20.9%

31

$50,000 - $75,000

21.6%

32

$75,000 - $100,000

7.4%

11

Over $100,000

14.9%

22

Now Married

54.7%

81

Widowed

1.4%

2

Divorced

4.7%

7

Separated

2%

3

37.2%

55

Personal Income
Under $10,000

Household Income

Marital Status

Never Married

Hypothesis 1: An individual’s preference for homophily will be positively related to
homophilous tendencies when selecting word-of-mouth sources.
In order to test hypothesis 1, a linear regression was conducted, utilizing preference for
homophily as a predictor of perceived homophily between a word-of-mouth source and the

65

participant, as defined by those characteristics included in the preference for homophily scale
created in study one, included in Appendix D. Hypothesis 1 was supported. Preference for
homophily was positively related to perceived homophily with the source (β = 0.460, p < 0.001)
and explained 21.1% of the variance in perceived similarity.
Hypothesis 2: Perceived similarity will be positively related to source likeability.
In order to test hypothesis 2, a linear regression was conducted utilizing perceived
similarity as a predictor of source likeability. Hypothesis 2 was supported. Perceived similarity
was positively related to source likeability (β = 0.646, p < 0.001) and explained 41.7% of the
variance in source likeability.
Hypothesis 3: Perceived similarity will be positively related to likelihood of following advice.
In order to test hypothesis 3, a linear regression was conducted utilizing perceived
similarity as a predictor of intentions to follow advice. Hypothesis 3 was supported. Perceived
similarity was positively related to intentions to follow advice (β = 0.456, p < 0.001) and
explained 20.8% of the variance in intentions to follow advice.
Hypothesis 4: Perceived similarity will be negatively related to likelihood of additional search
processes.
In order to test hypothesis 4, a linear regression was conducted utilizing perceived
similarity as a predictor of intentions to engage in additional search processes. Hypothesis 4 was
supported. Perceived similarity was negatively related to intentions to engage in additional
search processes (β = 0.178, p < 0.05) and explained 3.2% of the variance in intentions to engage
in additional search processes.
Hypothesis 5: Preference for similarity will be higher in low, rather than high involvement
situations when predicting likelihood of following advice.

66

Refer to tables 8 and 9 for regression coefficients related to hypotheses 5-10.
Table 8: H5, H7, H9 Regression Coefficients: Predictors of Likelihood of Following
Recommendation

Variable
Perceived Homophily
Involvement
Expertise
Likeability
Involvement * Perceived
Homophily
Involvement * Likeability
Involvement * Expertise

Standardized
Coefficient
0.115
0.267***
-0.031
0.811**

R2= 0.501

0.001
-0.281
0.147

* p < 0.05
** p < 0.01
*** p < 0.001
Table 9: H6, H8, H10 Regression Coefficients: Predictors of Likelihood of Engaging in
Additional Search

Variable
Perceived Homophily
Involvement
Expertise
Likeability
Involvement * Perceived
Homophily
Involvement * Likeability
Involvement * Expertise

Standardized
Coefficient
-0.178
0.297**
-0.643*
0.263
0.337
-0.252
0.696*

* p < 0.05
** p < 0.01
*** p < 0.001
67

R2=0.153

In order to test hypothesis 5, a linear regression was conducted utilizing an interaction
term, such that involvement was centred and multiplied by the centred preference for similarity
variable and the product was utilized as a predictor. Hypothesis 5 was not supported.
Involvement had no impact on the relationship between preference for similarity and likelihood
of following advice.
Hypothesis 6: Preference for similarity will be higher in low, rather than high involvement
situations when predicting additional search processes.
In order to test hypothesis 6, a linear regression was conducted utilizing an interaction
term, such that involvement was centred and multiplied by the centred preference for similarity
variable and the product was utilized as a predictor. Hypothesis 6 was not supported.
Involvement had no impact on the relationship between preference for similarity and likelihood
of engaging in additional search behaviours after receiving advice.
Hypothesis 7: Likeability will be more important in low, rather than high involvement situations
when predicting likelihood of following advice.
In order to test hypothesis 7, a linear regression was conducted utilizing an interaction
term, such that involvement was centered and multiplied by the centered likeability variable and
the product was utilized as a predictor. Hypothesis 7 was not supported. Involvement had no
impact on the relationship between likeability and likelihood of following advice.
Hypothesis 8: Likeability will be more important in low, rather than high involvement situations
when predicting additional information search processes.
In order to test hypothesis 8, a linear regression was conducted utilizing an interaction
term, such that involvement was centred and multiplied by the centred likeability variable and

68

the product was utilized as a predictor. Hypothesis 8 was not supported. Involvement had no
impact on the relationship between likeability and additional search processes.
Hypothesis 9: Expertise will be more important in high, rather than low involvement situations
when predicting likelihood of following advice.
In order to test hypothesis 9, a linear regression was conducted utilizing an interaction
term, such that involvement was centred and multiplied by the centred expertise variable and the
product was utilized as a predictor. Hypothesis 9 was not supported. Involvement had no impact
on the relationship between expertise and likelihood of following advice.
Hypothesis 10: Expertise will be more important in high, rather than low involvement situations
when predicting additional information search processes.
In order to test hypothesis 10, a linear regression was conducted utilizing an interaction
term, such that involvement was centred and multiplied by the centred expertise variable and the
product was utilized as a predictor. The interaction term was significant ( = 0.258, p < 0.05),
which indicated that involvement did moderate the relationship between expertise and intentions
to engage in additional search processes. The data was then split into high and low involvement,
and two regressions were conducted under high and low involvement conditions. The results
suggested that expertise was not a significant predictor of additional search processes in low
involvement situations, but it was a significant predictor of additional search processes in high
involvement situations ( = 0.160, p < 0.05). As such, hypothesis 10 was supported.
Additional Analysis
Since trustworthiness was not tested in the hypotheses, further analysis was undertaken to
understand this facet of source credibility. Results indicated that trustworthiness was positively
related to intentions to follow through on recommended product choice ( = 0.480, p < 0.001),

69

and intentions to engage in additional search behaviours ( = 0.164, p < 0.05). Similarly,
likeability was positively related to intentions to follow the advice given (( = 0.277, p < 0.001)
and intentions to engage in further search ( = 0.126, p < 0.01) Furthermore, as indicated in
tables 5 and 6, involvement and perceived similarity are positively related to intentions to follow
through on advice and intentions to engage in additional search behaviours. As such, it would
seem that trustworthiness, likeability, similarity are uniformly deemed important source cues for
consideration when evaluating a word-of-mouth message. Contrarily, expertise is only important
in highly involving situations.

70

DISCUSSION

Study One
In developing the homophily scale, items related to personality, values, attitudes,
activities, and points of view on issues impacting the world were considered by participants
when making judgments of similarity. This new scale has several items in common with the
items identified by McCroskey, Richmond, and Daly (1975), although several overlapping items
tested in this study were eliminated. Both final scales included several items related to attitudes,
personality, and values, however – the key variables discussed in the interviews conducted in
phase one of the scale construction. See Table 10 for a side-by-side comparison of the two scales.
Table 10: Side-by-side Comparison of McCroskey, Richmond, & Daly’s (1975) Homophily Scale
and Current Homophily Scale
McCroskey, Richmond, &
Daly
Behaves like me
Like me
From social class similar to mine
Economic situation similar to
mine
Status like mine
Morals like mine
Looks similar to me
Same size as I am
Appearance like mine
Resembles me
Similar to me

McLaughlin

71

Table 10 (cont’d)
Thinks like me
Treats people like I do
Background similar to mine
Sexual attitudes like mine
Shares my values

Intelligence
Relationship Style
Family Situation
Attitude Towards Gay Rights
Values
Sense of Humour
Personality
Attitude Towards World Issues
Attitude Towards Drinking
Attitude Towards Gaming
Life Goals
Taste in Books
Taste in Movies
Family Values
Attitude Towards Education
Way of Handling Money
Knowledge of Technology
Taste in Friends
Political
Attitude Towards Children
Attitude Towards Sports
Religion

Much of the current homophily research has focused on demographic characteristics such
as gender, age, and race (e.g., Farmer & Farmer, 1996; Roth, 2004a; 2004b; Saiki & Delong,
2006; Thelwall, 2008; Turban, Dougherty, & Lee, 2002; Zenger & Lawrence, 1989), and
dismissed the importance of similarity regarding attitudes, core values, and personality. However,
in both the current and previous scales, demographic characteristics have been eliminated and
these other psychographic characteristics have been demonstrated to be relevant. Lazarsfeld and
Merton (1954) suggested that attitude homophily may be the result of demographic homophily –
specifically, they suggested that the tendency for people to spend time with individuals with
similar attitudes to their own was present because they chose to spend time with demographically
72

similar individuals, and these demographically similar individuals would therefore have
experienced similar upbringings, resulting in attitude similarity. However, the current research
supports previous research (e.g., Allgeier & Byrne, 1973; Byrne, 1961a; Duck, 1975; Eyal &
Rubin, 2003; Singh, 1973; Steele & McGlynn, 1979; Tan & Singh, 1995) that demographic
similarity is more likely to be the product of a preference for homophily regarding personality,
attitudes, and values – the reverse of what was originally proposed (Lazarsfeld & Merton, 1954).
This study and previous studies (e.g., Allgeier & Byrne, 1973; Byrne, 1961a; Duck, 1975;
Eyal & Rubin, 2003; Singh, 1973; Steele & McGlynn, 1979; Tan & Singh, 1995) suggest that
personality, attitude, and activity homophily facilitate and encourage relationship formation.
However, there is a substantial literature suggesting that demographic homophily is present in
interactions and relationship formation as well (e.g., Farmer & Farmer, 1996; Roth, 2004a;
2004b; Saiki & DeLong, 2006; Thelwall, 2008; Turban, Dougherty, & Lee, 2002; Zenger &
Lawrence, 1989), characteristics that were not supported in the current scale. Hoffner and
Buchanan (2009) found that individuals of the same gender were perceived to have more similar
attitudes to the participant than individuals of the opposite gender, suggesting that demographic
homophily in relationships may be due to a tendency for people with similar demographic
profiles to have things in common with each other aside from their demographic profiles. This
could be why people tend to spend time with demographically similar others but do not admit to
a preference for similarity on these characteristics – it is merely a result of a preference for
similarity on other attributes.
Alternatively, participants may have felt uncomfortable stating a preference for
homophily on these characteristics, thereby influencing the scale’s construction. Laddering
interviews conducted under anonymous or confidential conditions with interviewers who have

73

no previous (or expected future) interaction with the interviewees may be able to better
understand this tendency for demographic homophily. Furthermore, techniques that use
projection (e.g., ‘What do you think Cathy would consider…?’) could be utilized to understand
this preference. However, the current work and abundant other works indicate that attitudes,
personality, and activity homophily are consistently related to attractiveness (e.g., Allgeier &
Byrne, 1973; Batchelor & Tesser, 1971; Byrne, 1961a; Duck, 1973; 1975; Duck & Spencer,
1972; Morry, 2004; 2005; 2007). Furthermore, the remaining scale items are the characteristics
that participants spoke most about in the interviews and justified when asked why homophily
was important to them. Finally, although there is a stigma associated with demonstrating a
preference for homophily on demographic characteristics such as ethnicity, expressing a
preference for individuals with the same gender or age is not socially unacceptable. Due to the
rejection of these items, it is more likely that demographic homophily is present in society as an
artefact of a preference for homophily on other characteristics, and not because people actually
search for demographic homophily.
The items that remained in the scale after eliminating all items that reduced the scale’s
reliability were often discussed in depth by interviewees when they were asked why they
preferred to spend time with similar others. Most commonly, they discussed the ease of meeting
people with similar interests, the way that similar interests facilitated conversation, and the way
that similar interests facilitated activity choice. Similar taste in and attitudes toward activities
accounted for 4 out of 22 items that remained in the scale. Since participants spent a great deal of
time discussing how this facilitated friendship formation and growth, it makes intuitive sense that
this made up approximately 20% of the scale’s items.

74

Previous research that has touted the superiority of perceived homophily to actual
homophily in predicting attraction and relationship formation has focused on attitudinal research
rather than activities (e.g., Montoya, Horton, & Kirchner, 2008; Morry, 2004; 2005; 2007;
Strauss, Barrick, & Connerley, 2001). However, Werner and Parmalee’s (1979) work suggests
that actual homophily may be predictive of attraction in certain, more easily observable,
characteristics. Now that a scale has been created based on those characteristics that participants
admit are important to making similarity judgments, future researchers can explore this
relationship by comparing actual homophily and perceived homophily in friendship pairs on
those traits that are considered when individuals make homophily judgments in natural settings.
This will allow researchers to better understand the individual effects of preference for
homophily on friendship formation and word-of-mouth source choices.
While respondents tended to talk about the ease of forming friendships with individuals
whose interests were similar to their own first, most of them later discussed the fact that having
similar opinions made them feel more at ease when conversing with others. A similar sense of
humour, political beliefs, religious beliefs, set of values, and personality all led to participants
feeling more able to be themselves without fear of being judged. When probed further,
participants stated that it was a basic human need to feel accepted, and to feel like their opinions
were correct. By spending time with similar others, they felt like their way of living life was
validated, rather than challenged. This is in support of previous research that has demonstrated
that a high need for acceptance and approval increases homophilous tendencies (Byrne, 1961b;
Erwin, 1981; Posavac, 1971).
In shopping trips, individuals expressed a preference for homophily in their partners so
that they would enjoy the same stores and be able to give better advice regarding product choice.

75

This matches the claims above stating that homophily was important to facilitate activity choice.
Since preferences were similar, both individuals would enjoy the trip and want to do the same
things. The desire for similar taste when receiving advice may be related to a desire for approval
or acceptance (Byrne, 1961b; Erwin, 1981; Posavac, 1971), once again paralleling the findings
for shopping partners and interpersonal relationships. Many participants stated that they wanted
people with similar preferences to go shopping with them so that they could feel validated that
they were making the right choice, and not because they wanted an alternative opinion.
For highly involving situations, however, preference for similarity varied between the
interpersonal and shopping situations. For highly involving interpersonal interactions,
characterized by stress, individuals desired similarity so that they could fall back on those
similarities to smooth over difficult times – specifically, participants felt that it would facilitate
activities and discussions that would bridge gaps in the friendship. Furthermore, in personally
stressful situations (such as a birth or death), similar experiences were important because this
would allow for understanding and advice giving. For shopping partners, however, highly
involving purchases required individuals with heterophilous information. Specifically,
individuals tended to look to word-of-mouth sources who had different knowledge or experience
bases in order to lend expertise to the situation. However, a certain amount of homophily was
still desirable in order to aid in developing a true understanding of the advisee’s needs.
Study Two
This study attempted to determine how involvement and source credibility influenced
word-of-mouth advisor choices. Although the elaboration likelihood model predicted that source
cues would be utilized only when peripheral processing was undertaken (Petty, Cacioppo, &
Goldman, 1981), other research has suggested that source credibility was utilized in both

76

moderately and highly involving situations in order to form attitudes (e.g., Eisend, 2007; Petty &
Cacioppo, 1984; Wu & Schaffer, 1987). The results of this study suggest that homophily,
likeability, and trustworthiness of a source were consistently influential in attitude formation.
Specifically, no matter the level of involvement, individuals would be more likely to follow the
advice of somebody they trusted, liked, and were similar to.
Expertise, however, was a more complicated facet of source credibility. Specifically,
expertise was only influential in predicting likelihood of engaging in additional search efforts in
highly involving situations. Expertise, unlike homophily, likeability, and trustworthiness, was
limited in its relevance as a cue.
Although source credibility has historically been thought of as a peripheral cue to
information processing, individuals are becoming more and more concerned with determining
the credibility of a source – especially in the online environment (e.g., Forman, Ghose, &
Wiesenfeld, 2008; Golbeck, 2009; Hong, 2006; Heit & Rotello, 2012; Huffaker, 2010). As such,
it is likely that people are paying much more attention to the credibility of information sources
regardless of involvement or environment. However, this consistent evaluation of source
credibility tends to be based on peripheral cues such as grammar, website attractiveness, and
credentials – not the actual content (a central cue).
Trustworthiness and likeability of a source were consistently present in word-of-mouth
sources regardless of involvement level, while expertise was only present in sources advising
participants in highly involving situations. This prevalence of trustworthiness is in line with the
work of McGinnies and Ward (1980). In their study of source credibility, source expertise was
judged utilizing the central processing route, while source likeability and trustworthiness were
judged utilizing the peripheral processing route. This explains the findings of Petty and Cacioppo

77

(1984) that source cues can be processed in highly involving situations. If expertise is a central
cue while trustworthiness and likeability are peripheral cues, source credibility’s role in highly
involving situations needs to be reconsidered – especially if peripheral cues are evaluated across
situations.
This tendency to seek out credible sources on these two dimensions across involvement
levels could be due to a consistent preference for likeable and trustworthy sources, or it could be
a product of the nature of friendships – people tend to spend time with others that are likeable
and trustworthy. Future research should manipulate involvement and measure the processing
route utilized to form attitudes and the source credibility cues considered when evaluating a
source that is described by the researcher, and not a source selected by the participant from their
own social network.
The strong relationship between likeability and homophilous perceptions of the source
indicate that the scale created in study one has construct validity. Specifically, this replicates past
research which has demonstrated a strong relationship between interpersonal attraction and
homophily (e.g., Acitelli, Douvan, & Veroff, 1993; Allgeier, & Byrne, 1973; Batchelor & Tesser,
1971; Byrne, 1961a; 1961b; Fiore & Donath, 2005; Schrock, 2007; Tan & Singh, 1995; Werner
& Parmalee, 1979). Furthermore, the correlation between homophilous preferences and
homophilous ratings of the word-of-mouth source make intuitive sense, and indicate predictive
validity. Previous work that has utilized McCroskey, Richmond, and Daly’s (1975) scale has
often determined that only certain elements of homophily were significant predictors of
relationship formation (e.g., Glascock & Ruggiero, 2006; Rocca & McCroskey, 1999; Wright,
2004). Furthermore, in other studies that utilized this scale, effect sizes were small (e.g., Morris,
Gorham, Cohen, & Huffman, 1996) or null (e.g., Dohanos, 2003). This previous homophily scale

78

has been unable to find a consistent relationship between homophily and attractiveness – which,
according to the research, should not be the case (e.g., Allgeier & Byrne, 1973; Batchelor &
Tesser, 1971; Byrne, 1961a; 1965; Farmer & Farmer, 1996; Huston & Levinger, 1978; Neimeyer
& Mitchell, 1988; Posavac & Pasko, 1974; Rosenbaum, 1986; Tan & Singh, 1995). By focusing
on only those characteristics that are relevant to similarity judgments in natural settings, and by
studying new components to homophily that were left out of McCroskey, Richmond, and Daly’s
(1975) scale, the current scale will allow future researchers to find out the true importance of
homophily. If these relationships were re-analyzed utilizing the current scale, a higher correlation
and explained variance will likely be observed between perceived homophily and attractiveness.

79

CONCLUSION

The first step in this dissertation was to create a new scale to measure preference for
homophily and, by extension, perceived homophily. The resulting scale had 22 items related to
activities, personality, and values. The previously utilized scale (McCroskey, Richmond, & Daly,
1975) had been utilized in the literature a limited number of times, with mixed results. Dohanos
(2003) analyzed the perceived similarity of sorority and fraternity members. According to the
previous homophily research (e.g., Allgeier & Byrne, 1973; Byrne, 1961a; Duck, 1975; Eyal &
Rubin, 2003; Singh, 1973; Steele & McGlynn, 1979; Tan & Singh, 1995) and the predictions of
the Social Identity Model of Deindividuation Effects (SIDE Model; Lea & Spears, 1991), people
who were members of the same fraternity or sorority should have been perceived to be more
similar to the participant than members of a different fraternity or sorority. However, Dohanos
(2003) found no significant effects utilizing the McCroskey, Richmond, and Daly (1975) scale.
Furthermore, Dohanos (2003) utilized the entire scale rather than breaking the analysis into the
four dimensions of the homophily scale. Given the mixed results regarding the predictive validity
of certain dimensions of the homophily scale (McCroskey, Richmond, & Daly, 1975), Dohanos
(2003) may have experienced significant results if he had analyzed the effects of group
membership on perceived homophily’s individual dimensions, rather than utilizing the whole
scale. By utilizing the scale developed in this dissertation, Dohanos (2003) would be much more
likely to find significant effects, in support of previous research, as it is a unidimensional scale
based on characteristics that are relevant to individuals when making similarity judgments rather
than a multidimensional scale that has dimensions and items that have little bearing on similarity
judgments in natural settings.

80

Morris, Gorham, Cohen, and Huffman (1996) found that, while perceived homophily was
2

positively related to instructor ratings, the explained variance was quite low (R = 2.9%). The
current study utilized perceived homophily to predict the likeability of a source, with very high
2

explained variance (R = 64.6%). Although both of these studies indicate the same positive
relationship between likeability and perceived homophily, the current study indicated a much
stronger relationship between these two variables. By focusing on only those characteristics that
are considered when individuals study how similar a target is to them, the predictive and
construct validity of the scale is greatly increased, as demonstrated by the higher explained
variance of the current scale.
Glascock and Ruggiero (2007) utilized only two dimensions of the homophily
(McCroskey, Richmond, & Daly, 1975) scale: attitude and background homophily. They utilized
these, among other variables, to predict students’ perceptions of learning. Results suggested that
attitude homophily predicted 1.5% of the variance in student learning, while background
homophily’s explained variance was not reported. The significant, but small, effect of attitude
homophily on student perceptions of learning mimics the results of Morris, Gorham, Cohen, and
Huffman (1996). Furthermore, the significance of only one dimension out of four supports the
notion that Dohanos (2003) may have found different results if the statistical analysis had been
based on the individual dimensions rather than the scale as a whole. The small effect size,
however, is an indication, once again, that the scale is lower in construct and predictive validity
than the current study’s scale.
Rocca and McCroskey (2009) studied the effects of attitude and background homophily
on the perceived immediacy and verbal aggression of a teacher. Results indicated that perceived
homophily was positively correlated to perceived immediacy and negatively related to verbal

81

aggression. Although explained variance was not reported in this study, and thus comparison
across studies is difficult, it is noteworthy that the other two dimensions of perceived homophily
(value and appearance) were not reported.
These small and nonsignificant results indicate a scale that is low in construct and
predictive validity. Furthermore, the current scale has several questions related to values and
activities that predict likeability of a target. While the scale created by McCroskey, Richmond,
and Daly (1975) included four items related to value homophily the results were not reported in
any of the four studies cited above. This indicates that the value dimension on the previous
perceived homophily scale included items that were of little relevance to participants making
judgments regarding similarity. Furthermore, the lack of any items related to activities indicates
that an entire subset of perceived homophily was missing in the McCroskey, Richmond, and
Daly (1975) scale. This is likely due to a paucity of activity-related homophily research at the
time the scale was created, but it diminishes the validity of the scale.
While perceived homophily has not been discussed in the traditional source credibility
literature, the current study and other word-of-mouth literature (e.g., Brown & Reingen, 1987;
Bruyn & Lilien, 2004; Dellande, Gilly, & Graham, 2004; Gilly, Graham, Wolfinbarger, & Yale,
1998; Steffes & Burgee, 2009, Wang, Walther, Pingree, & Hawkins, 2008; Wright, 2000)
indicate that this is an important dimension of the source evaluation process. The current study
discussed perceived homophily with a source as a precursor to source likeability. Due to the
strong relationship between source likeability and perceived homophily it is likely that this is the
extent of the effect of perceived homophily on source credibility perceptions. As such, perceived
homophily should be discussed as a component of likeability rather than a new dimension of
source credibility until researchers better understand the effect that perceived homophily has on

82

source credibility. Future research should explore whether or not this should be a new dimension
of the source credibility literature.
Traditional models of information processing and attitude formation have suggested that
source credibility will only be relevant in low involvement conditions (e.g., Chaiken, 1980; Petty,
Cacioppo, & Goldman, 1981). The current study, however, and other research (e.g., Hugstad,
Taylor, & Bruce, 1987; Larsen & Philips, 2002; Petty & Cacioppo, 1984) suggest a more
complex relationship between involvement, source credibility, information processing, and
attitude formation. The varied effects of the three dimensions of source credibility across
involvement levels suggest that the three dimensions of source credibility, in the future, should
be analyzed separately when studying how involvement influences the relationship between
source credibility and decision-making. Much like homophily research, the different dimensions
have often been analyzed as a set rather than individually. As such, complex relationships were,
very likely, overly simplified. Furthermore, unpublished articles regarding the relationship
between source credibility and the elaboration likelihood model should be revisited to explore
this relationship. With this new information regarding the role of source cues in highly involving
situations, previous literature regarding the elaboration likelihood model (Petty, Cacioppo, &
Goldman, 1981) and source cues, among other (traditionally) peripheral cues, should be revisited
with this new perspective in mind.
Traditional studies of the elaboration likelihood model have not all studied source
credibility but they have often given source cues as an example of peripheral cues, and this has
transferred to marketing textbooks’ explanations of the elaboration likelihood model (e.g.,
Schiffman, Kanuk, & Wisenblit, 2010; Soloman, Zaichkowsky, & Polegato, 2002) . By
exploring the importance of source likeability and trustworthiness across involvement levels and

83

the importance of expertise in high involvement situations, the current study suggests a need for
a new approach to the elaboration likelihood model’s central route to persuasion, and a need for
a new explanation of the peripheral route to persuasion in marketing textbooks and scholarly
literature.
Due to these new developments in the importance of source cues across levels of
involvement, marketing textbooks and advertisers will need to modify their approach to
endorsements and sources of information. Many celebrity endorsements have been focused on
low involvement products (Everybody’s Journal, 2012), which utilize the peripheral route to
persuasion. Although celebrity endorsers have been utilized for highly involving purchases in the
past (e.g., Kate Walsh’s endorsement of Cadillac) these commercials often include very little
information and focus on peripheral routes to persuasion. Celebrities that demonstrate expertise
are often knowledgeable because of their use of the product (e.g., Obama’s use of the blackberry
during his presidential campaign) rather than any demonstrated expertise in the subject matter.
When analyzing the most effective celebrity endorsements in, traditionally highly involving,
technological products, these celebrities were either involved in the development of the product
(e.g., Dr. Dre’s endorsement of his headphones brand), they emphasized the ease of use (e.g.,
Kevin Bacon explaining how easy it is to find videos on Google TV, Ashton Kutcher’s
endorsement of the Nikon Coolpix camera), or they provided peripheral cues such as humour
(e.g., Justin Timberlake’s jokes relating to Sony’s technological products; PCMag.com). These
limited uses of celebrity endorsers suggest that, even in highly involving situations, celebrities
are utilized primarily for peripherally processed messages. Celebrities are rarely utilized for
messages that demonstrate expertise or are highly content-driven. Marketers should begin to
analyze the ways that they can increase the perceived expertise of sources for highly involving

84

products, in addition to analyzing the likeability and trustworthiness of their sources for these
products.
In addition to this new approach to traditional advertising and spokesperson selection,
word-of-mouth researchers and marketers need to modify their approach to source evaluation.
Social media and Web 2.0 have allowed marketers to analyze the actual messages conveyed via
electronic word-of-mouth. Marketers that follow the recommendations inherent in the
elaboration likelihood model (Petty, Cacioppo, & Goldman, 1981) would traditionally focus on
the content of word-of-mouth messages when they were concerned with highly involving
products, and would not take source cues into account. Due to the effects of trustworthiness and
likeability across involvement levels, however, source cues should be considered when
evaluating the impact of these messages. According to the results of this study, a highly expert
review that comes from a source that is not likeable or trustworthy will be much less impactful
than a similar review from a source that is viewed positively - a prediction that contrasts with the
traditional elaboration likelihood model.
Overall, this new relationship between involvement and source credibility cues suggests a
new way of discussing the elaboration likelihood model in textbooks and theoretical research
articles. It adds a new element to the ways that an advertiser should select spokespeople in highly
involving situations and suggests a new way of identifying influential reviews and word-ofmouth sources in the online context. This new relationship expands the way that marketers and
researchers view the effect of involvement on source selection, and has both practical and
theoretical implications that merit further exploration.

85

Limitations
This study was conducted with a homogeneous group (25-35 year-old North Americans),
which may decrease the external validity of the study. Research has suggested that homophilous
preferences change as people mature (e.g., Duck, 1975; Shrum, Cheek, & Hunter, 1988). Future
research should attempt to understand how age influences the importance of homophily, and
which characteristics are emphasized by individuals.
While perceived homophily has been demonstrated to be more important than actual
homophily (e.g., Montoya, Horton, & Kirchner, 2008; Morry, 2004; 2005; 2007; Strauss, Barrick,
&Connerley, 2001), giving credence to the need for conscious measures of homophily, future
research should attempt to understand subconscious preferences as well. The scale fails to
account for subconscious preferences that individuals are unable to articulate, and also fails to
include those items that are utilized when evaluating potential friends or advisors but are socially
undesirable. Future research should attempt to apply this to real-world situations and determine if
this scale accurately predicts friendship pairs, and whether or not other items should be included
that were not consciously admitted to by the participants.
The scale was determined to be unidimensional, but the factor analysis did indicate that
there were two factors: one factor composed of positively phrased items and one factor
composed of negatively phrased items. Although this is common in scale formulation and testing
(Schmitt & Stults, 1985), future researchers should attempt to emphasize or draw attention to the
question phrasing in order to get a more accurate picture of preference for homophily.
Although this study utilized a product that had varying levels of involvement, the nature
of the purchase (a personal item that indicates status and image to the outside world) may have
influenced the way that the participants thought about the purchase and preference for homophily.

86

External search efforts and reliance on word-of-mouth sources may be different for more
utilitarian or private product purchases. Future research should attempt to replicate the current
study with a variety of products to better understand the relationship between involvement and
the varying source credibility facets.
Future Research
The current research was focused on conscious measures and reasons for homophily, but
did not explore subconscious preferences for homophily. Future research should utilize
projective techniques and a larger number of in-depth interviews to further explore the reasons
for preference for homophily without social desirability effects.
Furthermore, the high explained variance of this scale when predicting the likeability of a
source (64.6%) suggests construct and predictive validity. However, the scale has only been
utilized in one, very specific, study with a homogeneous sample. Future research should utilize
the scale with heterogeneous participants and a broader range of topics in order to further
demonstrate the validity of the scale. Researchers should emphasize the “dissimilarity” aspect of
reverse-coded items in order to improve the reliability of the scale.
The consistent role of trustworthiness and likeability in source effectiveness and the
importance of expertise in high involvement situations should be explored with heterogeneous
samples and with a heterogeneous set of products. The current study’s focus on watches served
the purpose of testing this relationship but limited the generalizability of the results. Furthermore,
the relationship should be explored in an online context in which cues to similarity and
likeability are utilized rather than interpersonal knowledge of the source.

87

APPENDICES

88

Appendix A: Study 1 Phase 1 Interview Protocol and Consent Form
Research Participant Information and Consent Form
Consumer Preference for Homophily, Part 1
You are being asked to participate in a research study of how individuals select friends, and how
individuals select who to go to for advice about a purchase. You must be at least 18 years old to
participate in this research.
Participation in this research project is completely voluntary. You have the right to say no. You
may change your mind at any time and withdraw. You may choose not to answer specific
questions or to stop participating at any time. These interviews will be recorded to facilitate
transcription of the data. If you do not wish to be recorded, you will not be able to participate in
the study.
There are no obvious physical, legal or economic risks associated with participating in this study.
However, you will be asked questions about yourself and these questions can sometimes make
people uncomfortable. You do not have to answer any questions that you do not wish to answer.
This study is not expected to yield any immediate benefit to the individual participants.
If you have concerns or questions about this study, such as scientific issues, how to do any part
of it, or to report an injury, please contact the researchers: Dr. Keith Adler at 309
Communication Arts & Sciences, adler@msu.edu, 517-353-3266 or Caitlin McLaughlin at 309
Communication Arts & Sciences, mclau115@msu.edu, 517-775-9771.
If you have questions or concerns about your role and rights as a research participant, would like
to obtain information or offer input, or would like to register a complaint about this study, you
may contact, anonymously if you wish, the Michigan State University’s Human Research

89

Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail
at 207 Olds Hall, MSU, East Lansing, MI 48824.
You indicate your voluntary agreement to participate by completing this phone interview.

90

Interview Protocol
Think about 2 or 3 of your closest friends. How similar do you think these friends are to you?
What sorts of things do you have in common with these friends?
In what sorts of ways are you different from these friends?
Do you think it’s important that you be similar to your close friends? As you’ve added new
friends to your group, how important is it that they have something in common with you? What
sorts of things do you want these new friends to have in common with you?
Why do you feel that it’s important / unimportant?
Do you think there are times when it’s more or less important to be similar to your friends?

Now, let’s talk about purchases that you’ve made. Think about the last time you bought a bigticket item. Who did you go to for advice?
Why did you go to this person for advice?
Was it more important that this person be similar to you or different from you? Why?
How important was this similarity (or difference)? Why do you think similarity (difference) was
important?
How similar do you think this person is to you?
What sorts of things do you have in common with this person?
In what sorts of ways are you different from this person?

Now think about the person you’re most likely to go shopping with for everyday things. Why do
you go with this person?
Was it more important that this person be similar to you or different from you? Why?

91

How important was this similarity (or difference)? Why do you think similarity (difference) was
important?
How similar do you think this person is to you?
What sorts of things do you have in common with this person?
In what sorts of ways are you different from this person?

Are there any times when it is more important that the person you’re shopping with or the person
you go to for advice should be similar to you? When are those times?

92

Appendix B: Study 1 Phase 2 Questionnaire and Consent Form

Consumer Preference for Homophily, Part II

Research Participant Information and Consent Form

Consumer Preference for Homophily, Part II
You are being asked to participate in a research study of how preference for similarity influences how people choose their friends and
information sources. You will need to fill out the questionnaire that follows in order to participate in the study.
You must be at least 18 years old to participate in this research.
Participation in this research project is completely voluntary. You have the right to say no. You may change your mind at any time
and withdraw. You may choose not to answer specific questions or to stop participating at any time.
There are no obvious physical, legal, or economic risks associated with participating in this study. However, you will be asked
questions about yourself and these questions can sometimes make people uncomfortable. You do not have to answer any questions
that you do not wish to answer. In return for your participation in the study, you will receive 100 Zoom Points.

93

If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please
contact the researchers: Dr. Keith Adler at 309 Communication Arts & Sciences, adler@msu.edu, 517-353-3266 or Caitlin
McLaughlin at 309 Communication Arts & Sciences, mclau115@msu.edu, 517-775-9771.
If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input,
or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State University’s
Human Research Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail at 207 Olds Hall,
MSU, East Lansing, MI 48824.
By clicking on the button below, you indicate your voluntary agreement to participate in this online survey.
Are you between the ages of 25 and 35?



Yes



No

Please indicate your level of agreement or disagreement with the following statements:
Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Disagree

94

Agree

Strongly
agree

When I ask for advice about what I
should buy, I prefer it when the person

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

advising me is different from me
In personal relationships, I prefer people
who have a similar sense of humour to
mine
I prefer the people advising me to have
different political views than my own
I am more comfortable when my
friends are different from me in
terms of how outgoing they are
I am more likely to follow advice about
product purchases when a person
advising me is more laidback than I am
When forming personal relationships, I
like friends who are as disorganized as I
am

95

When I ask for advice about what I
should buy, I am more comfortable when
the person advising me is the same

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

gender
Having similar personalities is important
to me when forming personal
relationships
I am more comfortable asking for advice
from a person who has a similar job to
my own
I prefer it when my friends' educational
backgrounds are different from my own
In asking for advice, I prefer it when the
people advising me have similar family
backgrounds as my own (siblings,
parents, etc.)
In personal relationships, I prefer to

96

spend time with people who look
different than me
I am more likely to listen to advice if the

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

person giving it is near my age
It is important to me that my friends have
a similar sense of style as me
I am more likely to listen to advice from
a person whose financial situation is
different from my own
In personal relationships, I prefer my
friends to be just as stubborn as I am
When asking for advice, I am more
influenced by friends who share the same
attitudes toward animals that I have
I prefer it when my friends' attitudes
toward physical fitness are different from

97

my own
I am more likely to listen to advice from
a person whose attitudes toward world

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

issues are similar to my own
I am more comfortable spending time
with people who have similar attitudes
toward drinking as me
It is important to me that my friends have
different hobbies than me
I am more comfortable taking advice
from a person who is similar to me in
their intelligence levels
In personal relationships, I am more
likely to choose people whose interests
are different than my own
I prefer to spend time with friends who

98

have similar attitudes toward gaming as
me
I am more likely to spend time with
friends whose attitudes toward sports are

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

different from my own
In forming friendships, I look for people
who have different life experiences than
my own
I prefer to take advice from somebody
whose life goals and priorities are similar
to my own
In personal relationships, I look for
people whose values are different from
my own
I am more comfortable getting advice
from somebody with different religious
views than my own

99

I prefer friends to have similar taste in
books as me
I am more comfortable when my friends'

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

taste in food is different from my own
I am more comfortable when my friends
have similar taste in movies as me
I prefer it when my friends have different
taste in music as me
In forming personal relationships, I tend
to choose people whose tastes in
television shows are different from my
own
I am more likely to take advice from
somebody whose family values /
priorities are similar to my own
I prefer it when my friends have similar

100

attitudes toward education as I do
I tend to spend more time with friends
who have similar relationships styles to

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

my own
I prefer asking for advice from somebody
who handles money in a way that is
similar to me
I am more likely to ask for advice from
somebody whose attitudes towards gay
marriage are different from my own
I prefer spending time with people whose
attitudes towards having children are
different from my own
I am more likely to ask for advice from
somebody whose knowledge of world
issues is different from my own

101

I am more comfortable spending time
with people whose knowledge of

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

technology is similar to my own
I tend to ask for advice from people
whose knowledge of cars is different than
my own
In personal relationships, I am more
likely to choose people whose family
situation is similar to my own regarding
children and relationship status
I am more likely to go to somebody for
advice if they have similar taste in friends
as I do

Please answer the following questions about yourself:
What is your sex?


Male



Female

102

What is your age, today, in years?
_____________
What is your ethnicity?


Caucasian / White



African American



Native American



Asian



Pacific Islander



Hispanic / Latino



Multiracial



I don’t want to disclose

What is your employment status?


Self-employed



Works for someone else, full-time



Works for someone else, part-time



Not employed



Student
103



Retired

What is the highest degree of school you have COMPLETED?


Elementary school



Some high school, but no diploma



High school graduate or the equivalent



Some college, but no degree



Associate degree



Bachelor’s degree



Master’s degree



Professional degree



Doctorate degree

What is your PERSONAL income before taxes?


Under $10,000



$10,000 to $19,999



$20,000 to $34,999



$35,000 to $49,999

104



$50,000 to $74,999



$75,000 to $99,999



$100,000 or more

What is your HOUSEHOLD income before taxes?


Under $10,000



$10,000 to $19,999



$20,000 to $34,999



$35,000 to $49,999



$50,000 to $74,999



$75,000 to $99,999



$100,000 or more

What is your marital status?


Now married



Widowed



Divorced



Separated

105



Never married

106

Appendix C: Study 1 Phase 3 Questionnaire and Consent Form
Consumer Preference for Homophily, Part III

Research Participant Information and Consent Form

Consumer Preference for Homophily, Part III
You are being asked to participate in a research study of how preference for similarity influences how people choose their friends and
information sources. You will need to fill out the questionnaire that follows in order to participate in the study. It will take you
approximately 20 minutes. You must be at least 18 years old to participate in this research.
Participation in this research project is completely voluntary. You have the right to say no. You may change your mind at any time
and withdraw. Closing the window any time before clicking the ‘submit’ button will discontinue your participation and none of your
data will be included in the analysis. You may choose not to answer specific questions or to stop participating at any time. Data will be
stored on the zoomerang website and on the computers of the researchers, and will be password-protected at all times.
There are no obvious physical, legal, or economic risks associated with participating in this study. However, you will be asked
questions about yourself and these questions can sometimes make people uncomfortable. You do not have to answer any questions
that you do not wish to answer. This study will benefit researchers and marketers in that it will allow them to better understand how
individuals choose their advisors and will allow them to facilitate better informed decision-making. In return for your participation in

107

the study, you will receive 100 Zoom Points. In order to earn these points, you are required to complete and submit the survey. To
learn the value of these points, go to
http://join.zoompanel.com/zoompanel/index.php?menu=How+It+Works&submenu=Get+Rewards.
If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please
contact the researchers: Dr. Keith Adler at 309 Communication Arts & Sciences, adler@msu.edu, 517-353-3266 or Caitlin
McLaughlin at 1-125 Ted Rogers School of Management, Ryerson University, caitlin,.mclaughlin@ryerson.ca, 416-323-0289.
If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input,
or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State University’s
Human Research Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail at 207 Olds Hall,
MSU, East Lansing, MI 48824.
By clicking on the button below, you indicate your voluntary agreement to participate in this online survey.
Are you between the ages of 25 and 35?



Yes



No

Please indicate your level of agreement or disagreement with the following statements:

108

Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Agree

Strongly
agree

Disagree
When I ask for advice about what I
should buy, I prefer it when the person

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

advising me is different from me
In personal relationships, I prefer people
who have a similar sense of humour to
mine
I prefer the people advising me to have
different political views than my own
Having similar personalities is important
to me when forming personal
relationships
I am more likely to listen to advice from
a person whose attitudes toward world
issues are similar to my own
I am more comfortable spending time
with people who have similar attitudes

109

toward drinking as me

I am more comfortable taking advice
from a person who is similar to me in

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

their intelligence levels
I prefer to spend time with friends who
have similar attitudes toward gaming as
me
I am more likely to spend time with
friends whose attitudes toward sports are
different from my own
I prefer to take advice from somebody
whose life goals and priorities are similar
to my own
In personal relationships, I look for
people whose values are different from
my own

110

I am more comfortable getting advice
from somebody with different religious

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

views than my own
I prefer friends to have similar taste in
books as me
I am more comfortable when my friends
have similar taste in movies as me
I am more likely to take advice from
somebody whose family values /
priorities are similar to my own
I prefer it when my friends have similar
attitudes toward education as I do
I tend to spend more time with friends
who have similar relationships styles to
my own
I prefer asking for advice from somebody
who handles money in a way that is
similar to me
I am more likely to ask for advice from
somebody whose attitudes towards gay
marriage are different from my own

111

I prefer spending time with people whose
attitudes towards having children are

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

different from my own
I am more comfortable spending time
with people whose knowledge of
technology is similar to my own
In personal relationships, I am more
likely to choose people whose family
situation is similar to my own regarding
children and relationship status
I am more likely to go to somebody for
advice if they have similar taste in friends
as I do

You are planning on purchasing a new pair of jeans.
Please indicate your level of agreement or disagreement with the following items:
Strongly
Disagree

Disagree

112

Moderately
disagree

Neither
Agree
nor

Moderately
agree

Agree

Strongly
agree

Disagree

In general I have a strong interest in this

O

O

O

O

O

O

O

This product is important to me

O

O

O

O

O

O

O

This product matters a lot to me

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

product

I get bored when other people talk to me
about this product
This product is very relevant to me

If the most important purchase to you is 100, where would you rate a pair of jeans on this 100 point scale?
__________
You are planning on purchasing a new pair of shoes.
Please indicate your level of agreement or disagreement with the following items:
Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Agree

Strongly
agree

Disagree
In general I have a strong interest in this
product

O

O

O

113

O

O

O

O

This product is important to me

O

O

O

O

O

O

O

This product matters a lot to me

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

I get bored when other people talk to me
about this product
This product is very relevant to me

If the most important purchase to you is 100, where would you rate a pair of shoes on this 100 point scale?
__________

You are planning on purchasing a new pair of sunglasses.
Please indicate your level of agreement or disagreement with the following items:
Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Agree

Strongly
agree

Disagree
In general I have a strong interest in this

O

O

O

O

O

O

O

This product is important to me

O

O

O

O

O

O

O

This product matters a lot to me

O

O

O

O

O

O

O

I get bored when other people talk to me

O

O

O

O

O

O

O

product

114

about this product
This product is very relevant to me

O

O

O

O

O

O

O

If the most important purchase to you is 100, where would you rate a pair of sunglasses on this 100 point scale?
__________

You are planning on purchasing a new watch.
Please indicate your level of agreement or disagreement with the following items:
Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Agree

Strongly
agree

Disagree
In general I have a strong interest in this

O

O

O

O

O

O

O

This product is important to me

O

O

O

O

O

O

O

This product matters a lot to me

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

product

I get bored when other people talk to me
about this product
This product is very relevant to me

115

If the most important purchase to you is 100, where would you rate a watch on this 100 point scale?
__________

You are planning on purchasing a new jacket.
Please indicate your level of agreement or disagreement with the following items:
Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Agree

Strongly
agree

Disagree
In general I have a strong interest in this

O

O

O

O

O

O

O

This product is important to me

O

O

O

O

O

O

O

This product matters a lot to me

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

product

I get bored when other people talk to me
about this product
This product is very relevant to me

If the most important purchase to you is 100, where would you rate a jacket on this 100 point scale?
__________

116

You are planning on purchasing a new DVD player.
Please indicate your level of agreement or disagreement with the following items:
Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Agree

Strongly
agree

Disagree
In general I have a strong interest in this

O

O

O

O

O

O

O

This product is important to me

O

O

O

O

O

O

O

This product matters a lot to me

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

product

I get bored when other people talk to me
about this product
This product is very relevant to me

If the most important purchase to you is 100, where would you rate a DVD player on this 100 point scale?
__________

Please answer the following questions about yourself:
What is your sex?


Male
117



Female

What is your age, today, in years?
_____________
What is your ethnicity?


Caucasian / White



African American



Native American



Asian



Pacific Islander



Hispanic / Latino



Multiracial



I don’t want to disclose

What is your employment status?


Self-employed



Works for someone else, full-time



Works for someone else, part-time



Not employed
118



Student



Retired

What is the highest degree of school you have COMPLETED?


Elementary school



Some high school, but no diploma



High school graduate or the equivalent



Some college, but no degree



Associate degree



Bachelor’s degree



Master’s degree



Professional degree



Doctorate degree

What is your PERSONAL income before taxes?


Under $10,000



$10,000 to $19,999



$20,000 to $34,999

119



$35,000 to $49,999



$50,000 to $74,999



$75,000 to $99,999



$100,000 or more

What is your HOUSEHOLD income before taxes?


Under $10,000



$10,000 to $19,999



$20,000 to $34,999



$35,000 to $49,999



$50,000 to $74,999



$75,000 to $99,999



$100,000 or more

What is your marital status?


Now married



Widowed



Divorced

120



Separated



Never married

121

Appendix D: Study 2 Questionnaire and Consent Form
Consumer Preference for Homophily, Part III

Research Participant Information and Consent Form

Consumer Preference for Homophily, Part III
You are being asked to participate in a research study of how preference for similarity influences how people choose their friends and
information sources. You will need to fill out the questionnaire that follows in order to participate in the study. It will take you
approximately 20 minutes. You must be at least 18 years old to participate in this research.
Participation in this research project is completely voluntary. You have the right to say no. You may change your mind at any time
and withdraw. Closing the window any time before clicking the ‘submit’ button will discontinue your participation and none of your
data will be included in the analysis. You may choose not to answer specific questions or to stop participating at any time. Data will be
stored on the zoomerang website and on the computers of the researchers, and will be password-protected at all times.
There are no obvious physical, legal, or economic risks associated with participating in this study. However, you will be asked
questions about yourself and these questions can sometimes make people uncomfortable. You do not have to answer any questions
that you do not wish to answer. This study will benefit researchers and marketers in that it will allow them to better understand how
individuals choose their advisors and will allow them to facilitate better informed decision-making. In return for your participation in

122

the study, you will receive 100 Zoom Points. In order to earn these points, you are required to complete and submit the survey. To
learn the value of these points, go to
http://join.zoompanel.com/zoompanel/index.php?menu=How+It+Works&submenu=Get+Rewards.
If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please
contact the researchers: Dr. Keith Adler at 309 Communication Arts & Sciences, adler@msu.edu, 517-353-3266 or Caitlin
McLaughlin at 1-125 Ted Rogers School of Management, Ryerson University, mclau115@msu.edu, 514-769-6381.
If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input,
or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State University’s
Human Research Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail at 207 Olds Hall,
MSU, East Lansing, MI 48824.
By clicking on the button below, you indicate your voluntary agreement to participate in this online survey.
Are you between the ages of 25 and 35?



Yes



No

Please indicate your level of agreement or disagreement with the following statements:

123

Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Agree

Strongly
agree

Disagree
When I ask for advice about what I
should buy, I prefer it when the person

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

advising me is different from me
In personal relationships, I prefer people
who have a similar sense of humour to
mine
I prefer the people advising me to have
different political views than my own
Having similar personalities is important
to me when forming personal
relationships
I am more likely to listen to advice from
a person whose attitudes toward world
issues are similar to my own

124

I am more comfortable spending time
with people who have similar attitudes

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

toward drinking as me
I am more comfortable taking advice
from a person who is similar to me in
their intelligence levels
I prefer to spend time with friends who
have similar attitudes toward gaming as
me
I am more likely to spend time with
friends whose attitudes toward sports are
different from my own
I prefer to take advice from somebody
whose life goals and priorities are similar
to my own
In personal relationships, I look for
people whose values are different from

125

my own
I am more comfortable getting advice
from somebody with different religious

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

views than my own
I prefer friends to have similar taste in
books as me
I am more comfortable when my friends
have similar taste in movies as me
I am more likely to take advice from
somebody whose family values /
priorities are similar to my own
I prefer it when my friends have similar
attitudes toward education as I do
I tend to spend more time with friends
who have similar relationships styles to
my own

126

I prefer asking for advice from somebody
who handles money in a way that is

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

similar to me
I am more likely to ask for advice from
somebody whose attitudes towards gay
marriage are different from my own
I prefer spending time with people whose
attitudes towards having children are
different from my own
I am more comfortable spending time
with people whose knowledge of
technology is similar to my own
In personal relationships, I am more
likely to choose people whose family
situation is similar to my own regarding
children
I am more likely to go to somebody for

127

advice if they have similar taste in friends
as I do

You are planning on purchasing a new watch.
To you, this purchase is:
Unimportant

1

2

3

4

5

6

7 Important

me

1

2

3

4

5

6

7 Of concern to me

Irrelevant

1

2

3

4

5

6

7 Relevant

Of no concern to

Means nothing to
Means a lot to me

1

2

3

4

5

6

7 me

Useless

1

2

3

4

5

6

7 Useful

Valuable

1

2

3

4

5

6

7 Worthless

Trivial

1

2

3

4

5

6

7 Fundamental

Beneficial

1

2

3

4

5

6

7 Not beneficial
Doesn't matter to

Matters to me

1

2

3

4

5

6

7 me

Interesting

1

2

3

4

5

6

7 Uninteresting

Significant

1

2

3

4

5

6

7 Insignificant

Vital

1

2

3

4

5

6

7 Superfluous

Boring

1

2

3

4

5

6

7 Interesting

128

Unexciting

1

2

3

4

5

6

7 Exciting

Appealing

1

2

3

4

5

6

7 Unappealing

Mundane

1

2

3

4

5

6

7 Fascinating

Essential

1

2

3

4

5

6

7 Nonessential

Undesirable

1

2

3

4

5

6

7 Desirable

Wanted

1

2

3

4

5

6

7 Unwanted

Not needed

1

2

3

4

5

6

7 Needed

Before you purchase this watch, you’ve decided to ask somebody for advice. Who would you ask for advice?

_________________________________________
Thinking about the person you want to ask for advice about this purchase, please indicate your level of agreement or
disagreement with the following items.
Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Agree

Strongly
agree

Disagree
They are different from me
Their sense of humour is similar to my

O

O

O

O

O

O

O

O

O

O

O

O

O

O

129

own
They have different political views than

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

my own
They have a similar personality to my own
Their attitudes toward world issues are
similar to my own
Their attitudes toward drinking are similar
to my own
Their intelligence is similar to my own
Their attitudes toward gaming are similar
to my own
Their attitudes toward sports are similar to
my own
Their life goals and priorities are similar to

130

my own
Their values are different from my own
Their religious views are different from

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

my own
Their taste in books is similar to my own
Their taste in movies is similar to my own
Their family values / priorities are similar
to my own
Their attitudes toward education are
similar to my own
Their relationship style is similar to my
own
The way that they handle money is similar
to me

131

Their attitude toward gay marriage is

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

different from my own
Their attitude toward having children is
different from my own
Their knowledge of technology is similar
to my own
Their family situation is similar to my own
regarding children and relationship status
Their taste in friends is similar to my own

Thinking about the person you want to ask for advice about this purchase, please indicate your level of agreement or
disagreement with the following items.
Neither
Strongly
Disagree

Disagree

Moderately

Agree

Moderately

disagree

nor

agree

Disagree

132

Agree

Strongly
agree

This person is dependable

O

O

O

O

O

O

O

This person is honest

O

O

O

O

O

O

O

This person is reliable

O

O

O

O

O

O

O

This person is sincere

O

O

O

O

O

O

O

This person is trustworthy

O

O

O

O

O

O

O

This person is an expert with watches

O

O

O

O

O

O

O

This person is experienced with watches

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

This person is skilled with watches

O

O

O

O

O

O

O

This person is friendly

O

O

O

O

O

O

O

This person is likeable

O

O

O

O

O

O

O

This person is warm

O

O

O

O

O

O

O

This person is approachable

O

O

O

O

O

O

O

I would ask this person for advice

O

O

O

O

O

O

O

I would like this person as a co-worker

O

O

O

O

O

O

O

I would like this person as a roommate

O

O

O

O

O

O

O

I would like to be friends with this person

O

O

O

O

O

O

O

This person is physically attractive

O

O

O

O

O

O

O

This person is similar to me

O

O

O

O

O

O

O

This person is knowledgeable about
watches
This person is qualified to talk about
watches

133

This person is knowledgeable

O

O

O

O

How likely are you to follow the advice of this person after they give their recommendation?
Unlikely

1

2

3

4

5

6

7 Likely

Improbable

1

2

3

4

5

6

7 Probable

Impossible

1

2

3

4

5

6

7 Possible

Uncertain

1

2

3

4

5

6

7 Certain

How likely are you to continue searching for information after hearing this person’s advice?
Unlikely

1

2

3

4

5

6

7 Likely

Improbable

1

2

3

4

5

6

7 Probable

Impossible

1

2

3

4

5

6

7 Possible

Uncertain

1

2

3

4

5

6

7 Certain

Please answer the following questions about yourself:
What is your sex?


Male



Female

What is your age, today, in years?
_____________

134

O

O

O

What is your ethnicity?


Caucasian / White



African American



Native American



Asian



Pacific Islander



Hispanic / Latino



Multiracial



I don’t want to disclose

What is your employment status?


Self-employed



Works for someone else, full-time



Works for someone else, part-time



Not employed



Student



Retired

What is the highest degree of school you have COMPLETED?
135



Elementary school



Some high school, but no diploma



High school graduate or the equivalent



Some college, but no degree



Associate degree



Bachelor’s degree



Master’s degree



Professional degree



Doctorate degree

What is your PERSONAL income before taxes?


Under $10,000



$10,000 to $19,999



$20,000 to $34,999



$35,000 to $49,999



$50,000 to $74,999



$75,000 to $99,999

136



$100,000 or more

What is your HOUSEHOLD income before taxes?


Under $10,000



$10,000 to $19,999



$20,000 to $34,999



$35,000 to $49,999



$50,000 to $74,999



$75,000 to $99,999



$100,000 or more

What is your marital status?


Now married



Widowed



Divorced



Separated



Never married

137

Appendix E: Figure 1 - Hypothesized Path Models

138

Appendix F: Figure 2 - Distribution of Purchase Involvement in Watches

139

Appendix G: Figure 3 - Study 1 Phase 3 Path Diagram

Sense of Humour

Personality

Att. Towards World Issues

Att. Towards Drinking

Intelligence

Attitude Towards Gaming
Factor 1

Life Goals

Taste in Books

Taste in Movies

Family Values

Att. Towards Education

Relationship Style

Way of Handling Money

Knowledge of Tech.

Family Situation

Taste in Friends

Political Attitudes

Factor
2

Religious Views

Values

Attitude Towards Sports

Attitudes Toward Gays

Att. Towards Children

140

BIBLIOGRAPHY

141

BIBLIOGRAPHY

Acitelli, L. K., Douvan, E., & Veroff, J. (1993). Perceptions of conflict in the first year of
marriage: How important are similarity and understanding? Journal of Social and
Personal Relationships, 10, 5-19.
Allgeier, A. R. & Byrne, D. (1973). Attraction toward the opposite sex as a determinant of
physical proximity. Journal of Social Psychology, 90, 213-219.
Andrews, J. C. & Shimp, T. A. (1990). Effects of involvement, argument strength, and source
characteristics on central and peripheral processing of advertising. Psychology and
Marketing, 7, 195-214.
Arndt, J. (1966). Word of mouth advertising: The role of product-related conversations in the
diffusion of anew food product. Unpublished doctoral dissertation, graduate school of
business administration, Harvard University.
Arndt, J. (1967). Word of mouth advertising: A review of the literature. New York: Advertising
Research Foundation, Inc.
Arrowood, A. J. & Short, J. A. (1973). Agreement, attraction, and self-esteem. Canadian Journal
of Behavioural Science, 5, 242-252.
Atkinson, D. R., Brady,S., & Casas, J. M. (1981). Sexual preference similarity, attitude similarity,
and perceived counsellor credibility and attractiveness. Journal of Counseling Psychology,
28, 504-509.
Back, K., Festinger, L., Hymovitch, B., Kelley, H., Schacter, S., & Thibaut, J. (1950). The
methodology of studying rumour transmission. Human Relations, 3, 307-312.
Batchelor, T. R. & Tesser, A. (1971). Attitude base as a moderator of the attitude similarityattraction relationship. Journal of Personality and Social Psychology, 19, 229-236.
Beach, L. R., Mitchell, T. R., Deaton, M. D., & Prothero, J. (1978). Information relevance,
content, and source credibility in the revision of opinions. Organizational Behaviour and
Human Performance, 21, 1-16.
Bearden, W. O., Lichtenstein, D. R., & Teel, J. E. (1984). Comparison price, coupon, and brand
effects on consumer reactions to retail newspaper advertisements. Journal of
Retailing, 60, 11-34.
Bearden, W. O., Netemeyer, R. G., Teel, J. E. (1989). Measurement of consumer susceptibility to
interpersonal influence. Journal of Consumer Research, 15, 473-481.

142

Beatty, S. E. & Smith, S. M. (1987). External search effort: An investigation across several
product categories. Journal of Consumer Research, 14, 83-95.
Beck, J. (2007). The sales effect of word of mouth: A model for creative goods and estimates for
novels. Journal of Cultural Economics, 31, 5-23.
Bentler, P. M. & Newcomb, M. D. (1978). Longitudinal study of marital success and failure.
Journal of Consulting and Clinical Psychology, 46, 1053-1070.
Bhattacherie, A. & Sanford, C. (2006). Influence processes for information technology
acceptance: An elaboration likelihood model. MIS Quarterly, 30, 805-825.
Bhuiyan, T. (2010). A survey on the relationship between trust and interest similarity in online
social networks. Journal of Emerging Technologies in Web Intelligence, 2, 291-299.
Bloch, P. H., Sherrell, D. S., & Ridgway, N. M. (1986). Consumer search: An extended
framework. Journal of Consumer Research, 13, 119-126.
Britt, S. H. (1978). Psychological principles of marketing and consumer behaviour. Lexington
Books: Lexington, MA.
Brown, J. J. & Reingen, P. H. (1987). Social ties and word-of-mouth referral behaviour. Journal
of Consumer Research, 14, 350-362.
Bruyn, A. & Lilien, G. L. A multi-stage model of word of mouth through electronic referrals.
Preprint, 2004.
Byrne, B. M. (2010). Structural Equation Modeling with AMOS, (2nd ed.). New York: Taylor &
Francis Group.
Byrne, D. (1961a). Interpersonal attraction and attitude similarity. Journal of Abnormal and
Social Psychology, 62, 713-715.
Byrne, D. (1961b). Interpersonal attraction as a function of affiliation and need and attitude
similarity. Human Relations, 14, 283-289.
Byrne, D. (1965). Authoritarianism and response to attitude similarity-dissimilarity. Journal of
Social Psychology, 66, 251-256.
Byrne, D., Bond, M. H., & Diamond, M. J. (1969). Response to political candidates as a function
of attitude similarity-dissimilarity. Human Relations, 22, 251-262.
Byrne, D., Clore, G. L., & Smeaton, G. (1986). The attraction hypothesis: Do similar attitudes
affect anything? Journal of Personality and Social Psychology, 51, 1167-1170.

143

Byrne, D. & Griffitt, W. (1966). A developmental investigation of the law of attraction. Journal
of Personality and Social Psychology, 79, 155-161.
Byrne, D., Griffitt, W., & Stefaniak, D. (1967). Attraction and similarity of personality
characteristics. Journal of Personality and Social Psychology, 5, 82-90.
Byrne, D. & Nelson, D. (1965). Attraction as a liner function of proportion of positive
reinforcements. Journal of Personality and Social Psychology, 6, 659-663.
Byrne, D. & Rhamey, R. (1965). Magnitude of positive and negative reinforcements as a
determinant of attraction. Journal of Personality and Social Psychology, 2, 884-889.
Campbell, K. & Wright, K. B. (2002). On-line support groups: An investigation of relationships
among source credibility, dimensions of relational communication, and perceptions of
emotional support. Communication Research Reports, 19, 183-193.
Caplow, T. (1947). Rumours in war. Social Forces, 25, 298-302.
Carli, L. L., Ganley, R., & Pierce-Otay, A. (1991). Similarity and satisfaction in roommate
relationships. Personality and Social Psychology Bulletin, 17, 419-426.
Celsi, R. L. & Olson, J. C. (1988). The role of involvement in attention and comprehension
processes. Journal of Consumer Research, 15, 210-224.
Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source
versus message cues in persuasion. Journal of Personality and Social Psychology, 39,
752-766.
Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative
analysis. Sage Publications: Los Angeles, CA.
Cheung, M. Y., Luo, C., Sia, C. L., & Chen, H. (2009). Credibility of electronic word-of-mouth:
Informational and normative determinants of online consumer recommendations.
International Journal of Electronic Commerce, 13, 9-38.
Chiou, J.-S. & Cheng, C. (2003). Should a company have message boards on its web sites?
Journal of Interactive Marketing, 17, 50-61.
Clark, L. A. & Watson, D. (1995). Constructing validity: Basic issues in objective scale
development. Psychological Assessment, 7, 309-319.
Clore, G. L. & Baldridge, B. (1968). Interpersonal attraction: The role of agreement and topic
interest. Journal of Personality and Social Psychology, 9, 340-346.

144

Condon, J. W. & Crano, W. D. (1988). Inferred evaluation and the relation between attitude
similarity and interpersonal attraction. Journal of Personality and Social Psychology, 54,
789-797.
Crosnoe, R., Frank, K., & Mueller, A. S. (2008). Gender, body size and social relations in high
schools, 86, 1189-1216.
Dellande, S., Gilly, M. C., & Graham, J. L. (2004). Gaining compliance and losing weight: The
role of the service provider in health care services. Journal of Marketing, 68, 78-91.
Dohanos, A. D. (2003). Fraternity and sorority member perceptions of homophily, supportive
communication, and group behavior as a function of control expectancies. (Doctoral
dissertation). Retrieved from Google Scholar.
Doyle, J. M. & Kao, G. (2007). Friendship choices of multiracial adolescents: Racial homophily,
blending, or amalgamation? Social Science Research, 36, 633-653.
Duck, S. W. (1973). Personality similarity and friendship choice: Similarity of what, when?
Journal of Personality, 41, 543-558.
Duck, S. W. (1975). Personality similarity and friendship choices by adolescents. European
Journal of Social Psychology, 5, 351-365.
Duck, S. W. & Spencer, C. (1972). Personal constructs and friendship formation. Journal of
Personality and Social Psychology, 23, 40-45.
Eisend, M. (2007). Understanding two-sided persuasion: An empirical assessment of theoretical
approaches. Psychology and Marketing, 24, 615-640.
Eisenstadt, S. N. (1952). Communication processes among immigrants in Israel. Public Opinion
Quarterly, 16, 42-58.
Erwin, P. G. (1981). The role of attitudinal similarity and perceived acceptance evaluation in
interpersonal attraction. The Journal of Psychology, 109, 133-136.
Erwin, P. G. (1982). The role of attitudinal similarity and direct acceptance evaluations in
attraction. Journal of Psychology, 111, 97-100.
Everybody’s Journal (2012). Top effective celebrity endorsers. Retrieved from
http://everybodysjournal.blogspot.ca/2012/03/top-effective-celebrity-endorsers.html.
Eyal, K. & Rubin, A. M. (2003). Viewer aggression and homophily, identification, and
parasocial relationships with television characters. Journal of Broadcasting & Electronic
Media, 47, 77-98.

145

Farmer, T. W. & Farmer,E. M. Z. (1996). Social relationships of students with exceptionalities in
mainstream classrooms: Social networks and homophily. Exceptional Children, 62, 431450.
Feren, D. B., Carroll, S. J., & Olian, J. D. (1988). Effects of managerial performance and
attitudinal similarity on interpersonal attraction. Basic and Applied Social Psychology, 9,
33-44.
Festinger, L., Cartwright, D., Barber, K., Fleischl, J., Gottsdanker, J., Keysen, A., & Leavitt, G.
(1948). A study of a rumour: Its origin and its spread. Human Relations, 1, 464-486.
Fiedler, F. E., Warrington, W. G., & Blaisdell, F. J. (1952). Unconscious attitudes as correlates
of socioeconomic choice in a social group. Journal of Abnormal and Social Psychology,
47, 790-796.
Fiore, A. T. & Donath, J. S. (2005). Homophily in online dating: When do you like someone like
yourself? CHI Proceedings.
Forman, C., Ghose, A., & Wiesenfeld, B. (2008). Examining the relationship between reviews
and sales: The role of reviewer identity disclosure in electronic markets.
French, J. R. P., Jr., & Raven B. (1968). The bases of social power. In D. Cartwright & A.
Zander (Eds.), Group Dynamics (pp. 259-269). New York: Harper & Row.
Friedman, H. & Friedman, L. (1979). Endorser Effectiveness by Product Type. Journal of
Advertising Research, 19, 63-71.
Geen, R. G. & Stonner, D. (1974). Similarity, conditioned affect, and interpersonal attraction.
Journal of Psychology, 87, 111-118.
Gemser, G., Leenmders, M. M., & Wijnberg, N. M. (2008). Why some awards are more
effective signals of quality than others: A study of movie awards. Journal of Management,
34, 25-54.
Gilly, M.C., Graham, J. L., Wolfinbarger, M. F., & Yale, L. J. (1998). A dyadic study of
interpersonal information search. Academy of Marketing Science, 26, 83-100.
Glascock, J. & Ruggiero, T. E. (2006). The relationship of ethnicity and sex to professor
credibility at a culturally diverse university. Communication Education, 55, 197-207.
Goldbeck, J. (2009). Trust and nuanced profile similarity in online social networks. ACM
Transactions on the Web, 3, 12:1 – 12:33.
Goldsmith, R. E., Lafferty, B. A., & Newell, S. J. (2000). The impact of corporate credibility and
celebrity credibility on consumer reaction to advertisements and brands. Journal of
Advertising, 29, 43-54.

146

Gotlieb, J. B. & Sarel, D. (1991). Comparative advertising effectiveness: The role of
involvement and source credibility. Journal of Advertising, 20, 38-45.
Grannovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 78, 13601380.
Greenleaf, E. A. & Lehmann, D. R. (1995). Reasons for substantial delay in consumer decision
making. Journal of Consumer Research, 22, 186-199.
Griffitt, W. & Veitch, R. (1974). Preacquaintance attitude similarity and attraction revisited: Ten
days in a fall-out shelter. Sociometry, 37, 163-173.
Hanish, L. D., Martin, C. L., Fabes, R. A., Leonard, S., & Herzog, M. (2005). Exposure to
externalizing peers in early childhood: Homophily and peer contagion processes. Journal
of Abnormal Child Psychology, 33, 267-281.
Harmon, D. & Boeringer, S. B. (2004). A content analysis of Internet-accessible written
pornographic depictions. In S. N. Hesse-Biber & P. Leavy (Eds.), Approaches to
qualitative research: A reader on theory and practice (pp. 402-426). Oxford University
Press: New York.
Harrison-Walker, L. J. (2001). The measurement of word-of-mouth communication and an
investigation of service quality and customer commitment as potential antecedents.
Journal of Service Research, 4, 60-75.
Hass, R. G. & Grady, K. (1975). Temporal delay, type of forewarning, and resistance to
influence. Journal of Experimental Social Psychology, 11, 459-469.
Hawkins, D. I., Mothersbaugh, D. L., & Best, R. J. (2007). Consumer behaviour: Building
marketing strategy. McGraw-Hill Irwin: New York.
Heesacker, M. & Petty, R. E., & Cacioppo, J. T. (1983). Field dependence and attitude change:
Source credibility can alter persuasion by affecting message-relevant thinking. Journal of
Personality, 51, 653-666.
Heit, E. & Rotello, C. M. (2012). The pervasive effects of argument length on inductive
reasoning. Thinking and Reasoning, 18, 244-277.
Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word-ofmouth via consumer-opinion platforms: What motivates consumers to articulate
themselves on the Internet? Journal of Interactive Marketing, 18, 38-52.
Herr, P. M., Kardes, F. R., & Kim, J. (1991). Effects of word-of-mouth and product-attribute
information on persuasion: An accessibility-diagnosticity perspective. Journal of
Consumer Research, 17, 454-462.

147

Hoffner, C. & Buchanan, M. (2009). Young adults’ wishful identification with television
characters: The role of perceived similarity and character attributes. Media Psychology, 7,
325-351.
Hong, T. (2006). Contributing factors to the use of health-related websites. Journal of Health
Communication: International Perspectives, 11, 149-165.
Horai, J., Naccari, N., & Fatoullah, E. (1974). The effects of expertise and physical attractiveness
upon opinion agreement and liking. Sociometry, 37, 601-606.
Hovland, C., Janis, I., & Kelley, H. (1953). Communication and Persuasion. New Haven, CT:
Yale University Press.
Hovland, C. & Weiss, W. (1951). The influence of source credibility on communication
effectiveness. Public Opinion Quarterly, 15, 635-650.
Huffaker, D. (2010). Dimensions of leadership and social influence in online communities.
Human Communication Research, 36, 593-617.
Hugstad, P., Taylor, J. W., & Bruce, G. D. (1987). The effects of social class and perceived risk
on consumer information search. Journal of Services Marketing, 1, 47-52.
Huston, T. L. & Levinger, G. (1978). Interpersonal attraction and relationships. Annual Review
of Psychology, 29, 115-156.
Ibarra, H. (1992). Homophily and differential returns: Sex differences in network structure and
access in an advertising firm. Administrative Science Quarterly, 37, 422-447.
Insko, C. A., Thompson, V. D., Stroebe, W., Shaud, K. F., Pinner, B. E., & Layton, B. D. (1973).
Implied evaluation and the similarity-attraction effect. Journal of Personality and Social
Psychology, 25, 208-308.
Jacobson, D. J. (1948). The affairs of dame rumor. New York: Rinehart and Company, Inc.
Jamieson, D. W., Lydon, J. E. & Zanna, M. P. (1987). Attitude and activity preference similarity:
Differential bases of interpersonal attraction for low and high self-monitors. Journal of
Personality and Social Psychology, 53, 1052-1060.
Johnson, H. & Izzett, R. (1969). Relationship between authoritarianism and attitude change as a
function of source credibility and type of communication. Journal of Personality and
Social Psychology, 13, 317-321.
Johnson, H., Torvicia, J. & Poprick, M. (1968). Effects of source credibility on the relationship
between authoritarianism and attitude change. Journal of Personality and Social
Psychology, 9, 179-183.

148

Jones, E., Moore, J. N., Stanaland, A. J. S., & Rosalind, A. J. (1998). Salesperson race and
gender and the access and legitimacy paradigm: Does difference make a difference?
Journal of Personal Selling & Sales Management, 18, 71-88.
Jones, L. W., Sinclair, R. C., & Courtneya, K. S. (2003). The effects of source credibility and
message framing on exercise intentions, behaviours, and attitudes: An integration of the
elaboration likelihood model and prospect theory. Journal of Applied Social Psychology,
33, 179-196.
Joyner, K. & Kao, G. (2000). School racial composition and adolescent racial homophily. Social
Science Quarterly, 81, 810-825.
Kahneman, D. & Tversky, A. (1982). The psychology of preferences. Scientific American, 246,
160-173.
Kamins, M. A. (1990). An Investigation into the “Match-Up” Hypothesis in Celebrity
Advertising: When Beauty May be Only Skin Deep. Journal of Advertising, 19, 4-13.
Katona, G. & Mueller, E. (1954). A study of purchase decisions. In L. H. Clark (Ed.), Consumer
behaviour: The dynamics of consumer reaction (pp. 30-87). New York: New York
University Press.
Katz, E. & Lazarsfeld, P. F. (1955). Personal influence: The part played by people in the flow of
mass communication. Glencoe, Illinois: The Free Press.
Keller, E. & Berry, J. (2003). The Influentials. New York: The Free Press.
Kelman, H. & Hovland, C. (1953). “Reinstatement” of the communicator in delayed
measurement of opinion. Journal of Abnormal and Social Psychology, 68, 547-549.
Kempf, D. & Palan, K. M. (2006). The effects of gender and argument strength on the processing
of word-of-mouth communication. Academy of Marketing Studies Journal, 10, 1-18.
Kerlinger, F. N. & Lee, H. B. (2000). Foundations of Behavioral Research (4th ed.). Toronto, ON:
Thomson Learning.
Kiesler, C. A. & Kiesler, S. B. (1964). Role of forewarning in persuasive communications.
Journal of Abnormal and Social Psychology, 68, 547-549.
Kotler, P., Keller, K., Sivaramakrishnan, S., Cunningham, P. H. Marketing Management. 14th
Edition, Toronto: Pearson.
Lafferty, B. A. & Goldsmith, R. E. (2004). How influential are corporate credibility and endorser
attractiveness when innovators react to advertisements for a new high-technology product?
Corporate Reputation Review, 7, 24-36.

149

Larsen, D. A. & Phillips, J. I. (2002). Effect of recruiter on attraction to the firm: Implications of
the elaboration likelihood model. Journal of Business and Psychology, 16, 347-364.
Lazarsfeld, P. F., Berelson, B., & Gaudet, H. (1948). The People’s Choice. New York: Columbia
University Press.
Lazarsfeld, P. F. & Merton, R. K. (1954). Friendship as social process: A substantive and
methodological analysis. In M. Berger, T. Abel, & C. H. Page (Eds.), Freedom and
control in modern society (18-66) New York: Octagon Books, Inc.
Lea, M., & Spears R. (1992). Paralanguage and social perception in computer-mediated
communication. Journal of Organizational Computing, 2, 321–341.
Lee, J., Park, D.-H., & Han, I. (2008). The effect of negative online consumer reviews on
product attitude: An information processing view. Electronic Commerce Research and
Applications, 7, 341-352.
Levinger, G. & Breedlove, J. (1966). Interpersonal attraction and agreement: A study of marriage
partners. Journal of Personality and Social Psychology, 3, 367-372.
Levinson, R. M., McCollum, K. T., & Kutner, N. G. (1984). Gender homophily in preferences
for physicians. Sex Roles, 10, 315-325.
Lichtenstein, D. R. & Bearden, W. O. (1989). Contextual influences on perceptions of merchantsupplied reference prices. Journal of Consumer Research, 16, 55-66.
Lirtzman, S. I. & Shuv-Ami, A. (1986). Credibility of source communication on products’ safety
hazards. Psychological Reports, 58, 707-718.
Lutz-Zois, C. J., Bradley, A. C., Mihalik, J. L., & Moorman-Eavers, E. R. (2006). Perceived
similarity and relationship success among dating couples: An idiographic approach.
Journal of Social and Personal Relationships, 23, 865-880.
Maddux, J. E. & Rogers, R. W. (1980). Effects of source expertness, physical attractiveness, and
supporting arguments on persuasion: a case of brains over beauty. Journal of Personality
and Social Psychology, 39, 235-244.
Mahlotra, N. K. (1988). Self concept and product choice: An integrated perspective. Journal of
Economic Psychology, 9, 1-28.
Mano, H. & Oliver, R. L. (1993). Assessing the dimensionality and structure of the consumption
experience: Evaluation, feeling, and satisfaction. Journal of Consumer Research, 20, 451466.

150

Markham, D. (1968). The dimensions of source credibility of television newscasters. Journal of
Communication, 18, 57-64.
McCroskey, L. L., McCroskey, J. C., & Richmond, V. P. (2006). Analysis and improvement of
the measurement of interpersonal attraction and homophily. Communication Quarterly,
54, 1-31.
McCroskey, J. C., Richmond, V. P., & Daly, J. A. (1975). The development of a measure of
perceived homophily in interpersonal communication. Human Communication Research,
1, 323-332.
McGinnies, E. & Ward, C. (1980). Better liked than right: Trustworthiness and expertise as
factors in credibility. Personality and Social Psychology Bulletin, 6, 467-472.
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social
networks. Annual Review of Sociology, 27, 415-444.
Meijnders, A., Midden, C., Olofsson, A., Ohman, S., Matthes, J., Bondarenko, O., Gutteling, J.,
& Rusanen, M. (2009). The role of similarity cues in the development of trust in sources
of information about GM food. Risk Analysis, 29, 1116-1128.
Michinov, E. & Monteil, J.-M. (2002). The similarity-attraction relationship revisited:
Divergence between the affective and behavioural facets of attraction. European Journal
of Social Psychology, 32, 485-500.
Miller, G. & Baseheart, J. (1969). Source trustworthiness, opinionated statements, and response
to persuasive communication. Speech Monographs, 36, 1-7.
Mills, J. (1966). Opinion change as a function of the communicator’s desire to influence and
liking for the audience. Journal of Experimental Social Psychology, 2, 152-159.
Mitchell, H. E. & Byrne, D. (1973). The defendant’s dilemma: Effects of jurors’ attitudes and
authoritarianism on judicial decisions. Journal of Personality and Social Psychology, 25,
123-129.
Mollica, K. A., Gray, B., & Trevino, L. K. (2003). Racial homophily and its persistence in
newcomers’ social networks. Organization Science, 14, 123-136.
Mondak, J. J. (1990). Perceived legitimacy of Supreme Court decisions: Three functions of
source credibility. Political Behaviour, 12, 363-384.
Montoya, R. M., Horton, R. S., & Kirchner, J. (2008). Is actual similarity necessary for attraction?
A meta-analysis of actual and perceived similarity. Journal of Social and Personal
Relationships, 25, 889-922.

151

Morris, T. L., Gorham, J., Cohen, S. H., & Huffman, D. (1996). Fashion in the classroom:
Effects of attire on student perceptions of instructors in college classes. Communication
Education, 45, 135-148.
Morry, M. M. (2004, June). A test of the attraction-similarity hypothesis among same-sex friends.
Talk presented at the 65th Canadian Psychology Association Annual Convention, St.
John`s, NF, Canada.
Morry, M. M. (2005). Relationship satisfaction as a predictor of similarity ratings: A test of the
attraction-similarity hypothesis. Journal of Social and Personal Relationships, 24, 561584.
Morry, M. M. (2007). The attraction-similarity hypothesis among cross-sex friends: Relationship
satisfaction, perceived similarities, and self-serving perceptions. Journal of Social and
Personal Relationships, 24, 117-138.
Morry, M. M., Kito, M., & Ortiz, L. (2011). The attraction-similarity model and dating couples:
Projection, perceived similarity, and psychological benefits. Personal Relationships, 18,
125-143.
Morse, J. H. (1994). Critical issues in qualitative research methods. Sage Publications:
Thousand Oaks, CA.
Moss, M. K., Byrne, D., Baskett, G. D., & Sachs, D. H. (1975). Informational versus affective
determinants of interpersonal attraction. The Journal of Social Psychology, 95, 39-53.
Mueller, E. (1958). The desire for innovations in household goods. In L. H. Clark (Ed.),
Consumer behaviour: Research on consumer reactions (pp. 13-37). New York: Harper &
Brothers.
Murstein, B. I. & Beck, G. D. (1972). Person perception, marriage adjustment, and social
desirability. Journal of Consulting and Clinical Psychology, 39, 396-403.
Neimeyer, R. A. & Mitchell, K. A. (1988). Similarity and attraction: A longitudinal study.
Journal of Social and Personal Relationships, 5, 131-148.
Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers’
perceived expertise, trustworthiness, and attractiveness. Journal of Advertising, 19, 39-52.
Olczak, P. V. & Goldman, J. A. (1975). Self-actualization as a moderator of the relationship
between attitude similarity and attraction. The Journal of Psychology, 89, 195-202.
Palmer, D. L. & Kalin, R. (1985). Dogmatic responses to belief dissimilarity in the “bogus
stranger” paradigm. Journal of Personality and Social Psychology, 48, 171-179.

152

Park, D.-H. & Kim, S. (2008). The effects of consumer knowledge on message processing of
electronic word-of-mouth via online consumer reviews. Electronic Commerce Research
and Applications, 7, 399-410.
Patzer, G. L. (1983). Source credibility as a function of communicator physical attractiveness.
Journal of Business Research, 11, 229-241.
PCMag (2012). Top 10 celebrity tech endorsements. Retrieved from
http://www.pcmag.com/slideshow_viewer/0,3253,l=260173&a=260174&po=6,00.asp.
Peterson, W. A. & Gist, N. P. (1951). Rumour and public opinion. American Journal of
Sociology, 57, 159-167.
Petty, R. E. & Cacioppo, J. T. (1984). Source factors and the elaboration likelihood model of
persuasion. Advances in Consumer Research, 11, 668-672.
Petty, R.E., Cacioppo, J.T., & Goldman, R. (1981). Personal involvement as a determinant of
argument-based persuasion. Journal of Personality and Social Psychology, 41, 847-855.
Pornpitakpan, C. (2003). The effect of celebrity endorsers` perceived credibility on product
purchase intention: The case of Singaporeans. Journal of International Consumer
Marketing, 16, 55-74.
Posavac, E. J. (1971). Need for approval as a moderator of interpersonal attraction based on
attitude similarity. The Journal of Social Psychology, 85, 141-142.
Posavac, E. J. & Pasko, S. J. (1974). Attraction, personality similarity, and popularity of the
personality of a stimulus person. The Journal of Social Psychology, 92, 269-275.
Priem, R. L. (2007). A consumer perspective on value creation. Academy of Management Review,
32, 219-235.
Priester, J. R. & Petty, R. E. (2003). The influence of spokesperson trustworthiness on message
elaboration, attitude strength, and advertising effectiveness. Journal of Consumer
Psychology, 13, 408-421.
Reagans, R. (2005). Preferences, identity and competition: Predicting tie strength from
demographic data. Management Science, 51, 1374-1383.
Reysen, S. (2005). Construction of a new scale: The reysen likability scale. Social Behaviour and
Personality, 33, 201-208.
Rich, S. U. (1963). Shopping behaviour of department store customers. Boston: Divisioin of
research, Graduate school of business administration, Harvard University.

153

Rocca, K. A. & McCroskey, J. C. (1999). The interrelationship of student ratings of instructors’
immediacy, verbal aggressiveness, homophily, and interpersonal attraction.
Communication Education, 48, 308-316.
Rosenbaum, M. E. (1986). The repulsion hypothesis: On the nondevelopment of relationships.
Journal of Personality and Social Psychology, 51, 1156-1166.
Roth, L. M. (2004a). Bringing clients back in: Homophily preferences and inequality on Wall
Street. The Sociological Quarterly, 45, 613-635.
Roth, L. M. (2004b). The social psychology of tokenism: Status and homophily processes on
Wall Street. Sociological Perspectives, 47, 189-214.
Saiki, D. & DeLong, M. R. (2006). Professionals’ relationships with clients in the apparel
industry. Qualitative Market Research: An International Journal, 9, 266-281.
Schiffman, L. G., Kanuk, L. L., & Wisenblit, J. (2010). Consumer behaviour. Prentice Hall:
Upper Saddle River, NJ.
Schmitt, N. & Stults, D. M. (1985). Factors defined by negatively keyed items: The result of
careless respondents? Applied Psychological Measurement, 9, 367-373.
Schrock, A. (2007). A test of the homophily principle using online personal advertisements.
Masters Theses & Specialist Projects.
Scissors, L. E., Gill, A. J., Geraghty, K., & Gergle, D. (2009, April). In CMC we trust: The role
of similarity. Paper presented at the annual meeting of the Computer and Human
Interaction Conference, Boston, Massachusetts.
Senecal, S. & Nantel, J. (2004). The influence of online product recommendations on consumers’
online choices. Journal of Retailing, 80, 159-169.
Shrum, W., Cheek, N. H., & Hunter, S. M. (1988). Friendship in school: Gender and racial
homophily. Sociology of Education, 61, 227-239.
Simon, H. (1990). Invariants of human behaviour. Annual Review of Psychology, 41, 1-19.
Singer, B. D. (1983). The case for using ‘real people’ in advertising. Business Quarterly, 48, 3237.
Singh, R. (1973). Attraction as a function of similarity in attitudes and personality characteristics.
Journal of Social Psychology, 91, 87-95.
Singh, R. & Teoh, J. B. P. (1999). Attitudes and attraction: A test of two hypotheses for the
similarity-dissimilarity asymmetry. British Journal of Social Psychology, 38, 427-443.

154

Slater, M. D. & Rouner, D. (1996). Value-affirmative and value-protective processing of alcohol
education messages that include statistical evidence or anecdotes.
Snyder, R. A. & Morris, J. H. (1978). Competence as a moderator of the similarity attraction
relationship: Development and application of a new index. The Journal of Psychology, 99,
235-244.
Solomon, M. R., Zaichkowsky, J. L., & Polegato, R. (2002). Consumer Behaviour: Buying,
having and being. (2nd ed.). Toronto: Prentice Hall.
Spreng, R. A., MacKenzie, S. B., & Olshavsky, R. W. (1996). A re-examination of the
determinants of customer satisfaction. Journal of Marketing, 60, 15-32.
Steele, M. E. & McGlynn, R. P. (1979). Effects of affective value and similarity on attraction.
The Journal of Social Psychology, 108, 111-119.
Steffes, E. M. & Burgee, L. E. (2009). Social ties and online word of mouth. Internet Research,
19, 42-59.
Strauss, J. P., Barrick, M. R., & Connerley, M. L. (2001). An investigation of personality
similarity effects (relational and perceived) on peer and supervisor ratings and the role of
similarity and liking. Journal of Occupational and Organizational Psychology, 74, 637657.
Sun, T., Yun, S., Wu, G., & Kuntaraporn, M. (2006). Online word-of-mouth (or mouse): An
exploration of its antecedents and consequences. Journal of Computer-Mediated
Communmication, 11, 1104-1127.
Tan, D. T. Y. & Singh, R. (1995). Attitudes and attraction: A developmental study of the
similarity-attraction and dissimilarity-repulsion hypotheses. Personality and Social
Psychology Bulletin, 21, 975-986.
Thelwall, M. (2009). Homophily in MySpace. Journal of the American Society for Information
Science and Technology, 60, 219-231.
Torgerson, W. S. (1958). Theory and Methods of Scaling, New York: John Wiley & Sons, Inc.
Torrance, E. P., & Mason, R. (1956). Instructor effort to influence: An experimental evaluation
of six approaches. Journal of Educational Psychology, 49, 211-218.
Turban, D. B., Dougherty, T. W., & Lee, F. L. (2002). Gender, race, and perceived similarity
effects in developmental relationships: The moderating role of relationship duration.
Journal of Vocational Research, 61, 240-262.
Tversky, A. (1977). Features of similarity. Psychological Review, 84, 327-352.

155

Tversky, A. & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science,
185, 1124-1131.
Vaynerchuk, G. (2011). The Thank You Economy, New York: Harper Collins Publishers.
Walster, E. & Festinger, L. (1962). The effectiveness of “overheard” persuasive communications.
Journal of Abnormal and Social Psychology, 65, 395-402.
Wang, Z., Walther, J. B., Pingree, S., & Hawkins, R. P. (2008). Health information, credibility,
homophily, and influence via the Internet: Web sites versus discussion groups. Health
Communication, 23, 358-368.
Wangenheim, F. & Bayon, T. (2004). The effect of word of mouth on services switching:
Measurement and moderating variables. European Journal of Marketing, 38, 1173-1185.
Warnick, B. (2004). Online ethos: Source credibility in an “authorless” environment. The
American Behavioral Scientist, 48, 256-265.
Watts, W. & McGuire, W. J. (1964). Persistence of induced opinion change and retention of the
inducing message contents. Journal of Social Psychology, 99, 43-50.
Weare, C., Musso, J., & Jun, K.-N. (2009). Cross-talk: The role of homophily and elite bias in
civic associations. Social Forces, 88, 147-174.
Werner, C. & Parmalee, P. (1979). Similarity of activity preferences among friends: Those who
play together stay together. Social Psychology Quarterly, 42, 62-66.
Whittaker, J. O. & Meade, R. D. (1968). Retention of opinion change as a function of differential
source credibility. International Journal of Psychology, 3, 103-108.
Wiener, J. L. & Mowen, J. C. (1986). Source credibility: On the independent effects of trust and
expertise. Advances in Consumer Research, 13, 306-310.
Wijnberg, N. M. (2004). Innovation and organization: Value and competition in selection
systems. Organization Studies, 25, 1469-1490.
Wijnberg, N.M. & Gemser, G. (2000). Adding value to innovation: Impression and the
transformation of the selection system in visual arts. Organization Science, 11, 323-329.
Wilke, W. H. (1934). An experimental comparison of the speech, the radio, and the printed page
as propaganda devices. Rural Sociology, 15, 352-364.
Wright, K. (2000). Perceptions of on-line support providers: An examination of perceived
homophily, source credibility, communication and social support within on-line support
groups. Communication Quarterly, 48, 44-59.

156

Wright, K. B. (2004). Online relational maintenance strategies and perceptions of partners within
exclusively internet-based and primarily internet-based relationships. Communication
Studies, 55, 239-253.
Wu, C. & Schaffer, D. R. (1987). Susceptibility to persuasive appeals as a function of source
credibility and prior experience with the attitude object. Journal of Personality and Social
Psychology, 52, 677-688.
Yang, S.-C., Hung, W.-C. Sung, K., & Farn, C.-K. (2006). Investigating initial trust toward etailers from the elaboration likelihood model perspective. Psychology & Marketing, 23,
429-445.
Zaichkowsky, J. L. (1985). Measuring the involvement construct. Journal of Consumer Research,
12, 342-352.
Zenger, T. R. & Lawrence, B. S. (1989). Organizational demography: The differential effects of
age and tenure distributions on technical communication. Academy of Management
Journal, 32, 353-376.
Zhang, J. & Daugherty, T. (2009). Third-person effect and social networking: Implications for
online marketing and word-of-mouth communication. American Journal of Business, 24,
53-63.
Zoomerang (n.d.), “Survey Respondents: Profile Reference Book,” Retrieved from
http://www.zoomerang.com/resources/Panel_Profile_Book.pdf

157