Journal of Business Research 110 (2020) 474–483

Contents lists available at ScienceDirect

Journal of Business Research

journal homepage: www.elsevier.com/locate/jbusres

A consumer socialization approach to understanding advertising avoidance T

on social media
⁎,1
Sydney Chinchanachokchai1, Federico de Gregorio
Department of Marketing, College of Business Administration, University of Akron, Akron, OH 44325-4804, USA

A R T I C LE I N FO A B S T R A C T

Keywords: The past five years have seen a rapid growth of advertising on social media platforms (SMPs). The current study
Advertising avoidance adopts the consumer socialization framework to investigate predictors of advertising avoidance on SMPs
Ad avoidance (Facebook, Twitter, and Instagram) via an online survey of 693 U.S. adults. Results show that the effects of SMP
Social media usage, susceptibility to social media influence, and susceptibility to peer influence on SMP ad avoidance are all
Consumer socialization
mediated by attitude toward social media advertising in general. Greater SMP usage and higher susceptibility to
Survey
social media influence are positively related with SMP advertising attitudes, while greater peer influence sus-
ceptibility is negatively related. The data also show no differences by demographics for avoidance or attitudes.

1. Introduction deliberate physical actions to avoid commercial contact. This includes

actions such as leaving a social media page that has too many ads,
In mid-2019 approximately 90 percent of the American population scrolling past ads quickly without processing them, closing pop-up/roll-
were considered regular Internet users (Anderson, Perrin, Jiang, & over ads, and installing ad blockers. When cognitively avoiding, audi-
Kumar, 2019). Social media platforms (SMPs henceforth and used as an ences mentally ignore promotional messages or (un)consciously not
interchangeable reference to “social media”) have captured a sig- look at locations they know to commonly contain ads on SMPs. Affec-
nificant share of the public’s Internet activities. Facebook is utilized by tive avoidance in Cho and Cheon (2004) conceptualization does not
roughly 70 percent of all U.S. Internet users, while Instagram and comprise a specific set of actions, but rather a negative emotional or-
Twitter account for respectively 35 and 24 percent (Smith & Anderson, ientation towards online advertising that magnifies/reinforces the ef-
2018). Concomitant with the popularity of SMPs among Internet users fects of the cognitive and behavioral avoidance forms. This includes
is their adoption by businesses as a means of advertising to current/ orientations such as general dislike of SMP ads that get in the way of
potential customers. Facebook’s revenue in the third quarter of 2019 consumers’ content enjoyment/interaction and feeling that SMPs would
was approximately $17.4 billion from advertising on its platform and it be better online environments without commercial messages. However,
had 4 million active advertisers (Statista, , 2019a). Meanwhile, In- as with behavioral avoidance methods, it is very challenging for cog-
stagram is predicted to generate $6.8 billion and Twitter over $1.62 nitive and affective avoidance forms to prevent the successful influence
billion from ad revenue by the end of 2020 (Statista, 2019b & c). Such of all promotional messages on SMPs given the sheer number and
high revenue represents significant ad volume. However, while such varying types.
revenues sound impressive, consumers are increasingly overwhelmed Although the majority of SMPs are rife with paid advertising and/or
by the amount of online marketing they encounter and, like they do in other forms of promotional content due to reliance on revenues from
offline media, utilize numerous means to avoid such messages. marketers to provide consumers access to their services for free, thus far
Grounding their work in both Speck and Elliott (1997) research on very few studies of ad avoidance on SMPs exist, which indicates a
general advertising avoidance and Vakratsas and Ambler (1999) review significant gap in knowledge. Moreover, studies of the antecedents of
of advertising response, Cho and Cheon (2004) empirically showed that ad avoidance in general largely focus on various perceptions/attitudes
online advertising avoidance is related but distinct. Specifically, In- toward advertising as a practice, such as irritation, beliefs about ad-
ternet ad avoidance takes three forms – behavioral, cognitive, and af- vertising’s role in society, predicted goal impediment, and skepticism
fective (Cho & Cheon, 2004). Behavioral avoidance involves any (e.g., Cho & Cheon, 2004; Kelly, Kerr, & Drennan, 2010; Shin & Lin,

⁎
Corresponding author.
E-mail addresses: schinchana@uakron.edu (S. Chinchanachokchai), degrego@uakron.edu (F. de Gregorio).
1
Authorship is listed in alphabetical order by last name. Both authors contributed equally to the research.

https://doi.org/10.1016/j.jbusres.2020.01.062
Received 1 February 2019; Received in revised form 27 January 2020; Accepted 29 January 2020
0148-2963/ © 2020 Elsevier Inc. All rights reserved.
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

2016). However, what shapes those perceptions/beliefs/attitudes about specific context. Thus, the framework is able to incorporate changes
advertising in the first place? In the current exploratory study, we over time in the forms and nature of these agents. Indeed, recent re-
likewise measure attitude toward (SMP) advertising as a predictor of ad search has demonstrated the applicability of the CS framework and its
avoidance, but propose that social forces are key in helping shape such social learning mechanisms to the social media context of this study
advertising perceptions. Thus, we take a broader societal perspective to (e.g., Mishra, Maheswarappa, Maity, & Samu, 2018; Wang et al., 2012).
understand the precursors of such attitudes. Specifically, we adopt
consumer socialization (CS; Ward, 1974) as the conceptual framework 2.1.2. Outcomes – attitudes towards and avoidance of SMP advertising
and investigate the impact of two key socialization agents (peers and While no existing research we are aware of utilizes the CS frame-
media) on attitudes toward SMP advertising, which in turn mediates the work to study ad avoidance, CS-based studies in general have confirmed
effects on SMP ad avoidance. Given that SMPs are highly popular online a positive attitude-behavior relationship (Bush, Smith, & Martin, 1999;
venues and the inherently social nature of the interaction on SMPs, de Gregorio & Sung, 2010). The core of this relationship is based on the
peers and media are an important source of socialization and thus the premise that a positive orientation toward a focal idea/object (adver-
appropriateness of the CS framework (Wang, Yu, & Wei, 2012). Thus, in tising) results in greater likelihood of committing related positive ac-
this study we: (1) reinforce previous research showing that perceptions tions (purchase, clicking on an ad) and less likelihood of committing
of advertising are the key precursor of ad avoidance, while being the negative actions (avoidance) (Muehling & McCann, 1993). This atti-
first to investigate the socially-based antecedents of these perceptions of tude-behavior relationship has been found in diverse marketing con-
(SMP) advertising, (2) demonstrate that different socialization agents texts such as product placements in films (de Gregorio & Sung, 2010),
(peers and media) have differing effects on SMP advertising attitudes, social media usage in general (Wang et al., 2012), and physical retail
which mediate SMP ad avoidance, and (3) provide theoretical im- shopping (Lueg, Ponder, Beatty, & Capella, 2006).
plications to expand academic understanding of how CS contributes to Separately, avoidance researchers have also demonstrated that at-
ad avoidance, as well as managerial suggestions to help SMPs make titudes toward and beliefs about advertising are key predictors of, and
their offerings more attractive to advertisers and marketers to better inversely related with, avoidance behavior. Speck and Elliott (1997)
implement SMP campaigns. revealed that stronger positive beliefs about advertising (being useful,
interesting, believable) were linked with reduced ad avoidance across
2. Literature review four non-digital media formats, while stronger negative beliefs (an-
noying, excessive, waste of time) resulted in increased avoidance. This
2.1. Consumer socialization pattern of results has been found in virtually all subsequent studies of
ad avoidance predictors, including those examining the behavior in
Derived from the broader socialization construct explaining how online contexts. Cho and Cheon (2004) showed that perceived goal
peoples’ cognitions, attitudes, and related behaviors develop in society, impediment, ad clutter, and lack of utility were positively related with
Ward (1974, p. 2) foundational conceptualization of CS described it as Web ad avoidance. A recent replication confirmed those findings hold
the collective “processes by which young people acquire skills, up in today’s online environment (Seyedghorban, Tahernejad, &
knowledge, and attitudes relevant to their functioning as consumers in Matanda, 2016), as well as in the context of avoiding location-based
the marketplace”. CS provides a framework for analyzing the influences advertising (Shin & Lin, 2016), MySpace and Facebook ads (Kelly et al.,
on/sources of how people learn to enact their roles as consumers in 2010), and social media ads in general (Dao, Le, Cheng, & Chen, 2014).
society. While the bulk of the CS literature has investigated children’s While in the current study attitude toward advertising serves as an
and adolescents’ socialization outcomes/antecedents, and childhood is Outcome component, we consider it essentially a mediating variable.
a key formative period in developing consumption-related behaviors, We propose that two socialization agents (peer communication and
the process continues during the adult life cycle (Moschis, 1987) as media; see sub-sections 2.1.3 and 2.1.4 for further discussion) impact
adults modify existing behaviors and adapt to new/changing consumer advertising avoidance indirectly via attitude toward advertising. As a
roles. While four components (Antecedents, Socialization Processes, corollary:
Socialization Agents, and Outcomes) are central to CS (Moschis, 1987;
Moschis & Churchill, 1978), the framework is flexible regarding the H1: Attitude toward SMP advertising will be inversely related with
specific variables that can be investigated within each of these com- SMP advertising avoidance.
ponents. For example, family members could be incorporated as part of
the Socialization Agents component depending on appropriateness of fit 2.1.3. Socialization agent – peers
with the outcome(s) being assessed. According to the CS framework, interaction with peer groups is a
fundamental human trait, arising from basic psychological, social, and
2.1.1. Overview of CS foundations and processes practical needs (Moschis, 1987; Ward, 1974). Via consumption-related
Although not a theory in itself, the theoretical grounding of CS is communications, peers serve as significant transmitters of attitudinal
found in its Socialization Processes component. One of the most com- and behavioral standards by which one’s own beliefs and actions may
monly utilized theories to explain the CS process is social learning. be gauged or adjusted (Bush et al., 1999; Moschis & Churchill, 1978).
Social learning theory places emphasis on the external, environmental Peers not only serve as a base of comparison, but also provide a means
sources of learning/socialization, with peers and media (Socialization of learning how to respond to new consumption-related stimuli in the
Agents) being two of the most impactful (Moschis, 1987). These ex- environment. Peer groups offer a real-world way of assessing to what
ternal sources explicitly and implicitly transmit norms of attitudes, extent one’s orientations and actions are “appropriate” as compared
behaviors, and responses from which learning/socialization occurs. with the norm among groups that one belongs to (e.g., the reaction
There are three social learning mechanisms by which Socialization among one’s peers to sharing a brand’s SMP ad with them or otherwise
Agents influence CS Outcomes – modeling, reinforcement, and social indicating approval of such ads). Thus, susceptibility to peer influence
interaction (Ward, 1974; Moschis & Churchill, 1978). These mechan- comprises both willingness to conform with the opinions of one’s peers
isms are complementary, operating together to varying degrees de- as well as openness to receiving/acting on the information received
pending on the specific context. In turn, these mechanisms influence the from them. The strength of the influence peers have as reference groups
Outcomes, which can be actual behaviors or cognitions such as atti- (i.e. the willingness to conform to peers) on consumer’s perceptions and
tudes, values, or beliefs. While the CS framework was developed before purchase behaviors is well documented (e.g., Bearden & Etzel, 1982;
the advent of SMPs, its conceptualizations of socialization agents such Bearden, Netemeyer, & Teel, 1989). Such conformity in response to
as peers and media were broad in nature and not restricted to any peer opinions, requests, and behaviors has also been shown to operate

475
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

in online contexts, such as willingness to participate in and share sur- content on SMPs, greater usage of SMPs likely results in greater overall
veys when requested via e-mail (De Bruyn & Lilien, 2008), influencing exposure to advertisements and other promotional communications. In
others on SMPs (Trusov, Bodapati, & Bucklin, 2010), and attitudes to- the same vein as mere exposure, the more one is exposed to SMP ads
wards and willingness to share incorrect information on Facebook (particularly in a pre-attentive manner with little/no intent to do so) via
(Colliander, 2019). greater SMP usage, the more familiar such ads become and thus result
Such susceptibility to the opinions of close others by its nature has in more positive attitudes toward them as a whole (Janiszewski, 1993;
been considered a type of conformity, theorized to be commonly a trait- Zajonc, 1968). This positive attitude towards SMP ads in turn likely
level orientation (Boush, Friestad, & Rose, 1994; Mangleburg & Bristol, decreases the likelihood of avoiding such promotional messages. Si-
1998). By extension, consumers who are higher on peer susceptibility milarly, cultivation posits the more media content one is exposed to, the
would also be more accepting broadly of (i.e., more likely to conform more one’s perceptions of and beliefs about the world broadly fall in
with) persuasive communications from other entities, including mar- line with the themes and norms of that content. While the original
keters and their messages. Indeed, susceptibility to peer communication conceptualization of cultivation was in the context of television con-
about consumption has been found to be negatively related with tent, emerging extensions have demonstrated its applicability to social
skepticism toward advertising (Boush et al., 1994; Mangleburg & media (Intravia, Wolff, Paez, & Gibbs., 2017; Tsay-Vogel, Shanahan, &
Bristol, 1998; Martin, Wentzel, & Tomczak, 2008). In addition, studies Signorielli, 2018). Thus, in the CS context of this research, the more one
have demonstrated a positive relationship between peer influence and uses SMPs, the more one is likely to be exposed to ads, thereby in-
consumers’ attitudes about non-interactive advertising (Bush et al., directly reinforcing the notion that ads are a necessity if SMPs are to
1999; de Gregorio & Sung, 2010) as well as social media marketing, remain free for use.
including brand-originating tweets on Twitter (Kwon, Kim, Sung, & CS researchers have predominantly operationalized assessment of
Yoo, 2014), SMP ads (Taylor, Lewin, & Strutton, 2011), and willingness the socializing effects of media via measuring the amount of time spent
to purchase from companies they have interacted with on SMPs (Wang using specific forms of media (e.g., Bush et al., 1999; Kwon et al.,
et al., 2012). Thus, greater susceptibility to peer influence about con- 2014). However, in line with Bearden et al. (1989) and Mascarenhas
sumption should lead to a more positive attitude towards SMP adver- and Higby (1993) we posit that extent of usage should not inherently be
tising and thus be negatively related to SMP ad avoidance. As one en- conflated with susceptibility to being persuaded by the (promotional)
gages in greater consumption-focused communications with one’s peer content found in a particular medium. This is particularly a con-
groups, one is more open to being influenced by those peers. In turn, sideration for SMPs because the main reasons for their usage are for
one will be more positively disposed towards consumption-related ob- social interaction or entertainment and not seeking marketing content
jects such as brands and thus more willing to process brand messages (Valentine, 2018). Greater social media usage may not necessarily have
from marketers (i.e., one will avoid ads less). Thus: a similar magnitude of effect on ad attitudes and avoidance as sus-
ceptibility to its influence. Thus, we assess susceptibility to social media
H2a: Susceptibility to peer influence about consumption will be influence separately from extent of SMP usage. As in the case of peer
positively related with attitude toward SMP advertising. influence, consumers vary as to their openness and susceptibility to
H2b: The relationship between susceptibility to peer influence about being influenced (in)directly media content (Mascarenhas & Higby,
consumption and SMP advertising avoidance will be mediated by 1993; Ward, 1974). This susceptibility is posited to be a form of general
attitude toward SMP advertising. tendency toward conformity (Boush et al., 1994; Mangleburg & Bristol,
1998). Thus, if people are more susceptible to aligning their con-
2.1.4. Socialization agent – media sumption beliefs based on marketing content on social media, they
As with one’s peer groups, the CS framework views media as a would also be more positively oriented to the content from businesses
significant source of consumption-related attitudes, beliefs, and beha- that are part of the SMP ecosystem. Based on this conceptualization, the
vioral norms. A key difference between media and peer groups as so- greater the usage of SMPs (and thus greater exposure to their attendant
cialization agents is that the former offers a mediated, indirect means of advertisements and other promotional forms), and the more susceptible
gauging one’s behaviors and attitudes in relation to social norms to social media influence consumers are, the more likely they are to be
(Moschis, 1987). For example, if one is regularly exposed to content positively disposed towards SMP ads and thus less likely to avoid such
such as documentaries or scripted shows discussing the dark side of ads.
digital technologies, it may influence one’s decision about whether to
share a company’s social media ads. There are initial indications that H3a: SMP usage will be positively related with attitude toward SMP
media exposure/usage influences advertising attitude-related out- advertising.
comes, though the evidence is mixed. Amount of media usage has been H3b: The relationship between SMP usage and SMP advertising
found to be positively related with attitudes toward advertising in avoidance will
general, as well as specific forms of advertising such as product place- be mediated by attitude toward SMP advertising.
ment (Bush et al., 1999; de Gregorio & Sung, 2010), but also have been H4a: Susceptibility to social media influence will be positively re-
found to have no impact (Smith & Moschis, 1984). With regard to In- lated with attitude toward SMP advertising.
ternet marketing, the evidence is limited to only a single study, which H4b: The relationship between susceptibility to social media influ-
found no relationship between Twitter usage and attitudes toward ence and SMP advertising avoidance will be mediated by attitude
brand-originating communications on the platform (Kwon et al., 2014). toward SMP advertising.
Based on the limited number of studies looking at this issue, the
preliminary data seem to indicate a weak or negligible relationship Based on the above discussions, Fig. 1 models the current study’s
between media exposure and attitude toward advertising, both in framework.
general and on specific media platforms. This is somewhat under-
standable. In a social media context, ads generally appear as part of the 3. Method
regular design of a platform rather than being focal points of content or
discussion (like they might be as part of a TV show scene). However, we 3.1. Sample and data collection
propose that norms of SMP ad attitudes and avoidance would be
transmitted and reinforced in a manner generally akin to the processes A survey was administered to an online sample from Amazon’s
underlying mere exposure (Zajonc, 1968) and cultivation (Gerbner, Mechanical Turk (MTurk) service. Research using MTurk participants
1970). As it is not possible to completely avoid all forms of promotional has become more common in marketing, advertising, and general

476
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

Susceptibility to Peer
Influence
H2

H3 H1
Attitude toward SMP SMP Advertising
SMP Usage
Advertising Avoidance

H4
Susceptibility to
Social Media
Influence

Fig. 1. Conceptual model.

business research due to availability, speed of data collection, and ease Table 1
of access, and has appeared in top outlets (e.g., Langan & Kumar, 2019; Demographic profile of the sample.
Mazodier, Henderson, & Beck, 2018). While no data source is without Frequency Percentage
flaws/challenges, the quality of data resulting from MTurk samples has
been reviewed and empirically investigated numerous times and has Gender
received strong support across numerous domains (Coppock, 2019; Male 340 49.4%
Female 353 50.9%
Goodman & Paolacci, 2017; Kees, Berry, Burton, & Sheehan, 2017). For Total 693 100%
example, MTurk data has been found to outperform both panel data
Age group
procured from professional marketing research companies such as
< 25 117 16.9%
Qualtrics (Kees et al., 2017) and nationally drawn samples (Coppock, 25–35 314 45.3%
2019) across various measures of data quality. Moreover, Huff and 36–45 147 21.2%
Tingley (2015) compared the characteristics of MTurk participants with 46–55 68 9.8%
those from the nationally representative samples used by the yearly Over 55 47 6.8%
Total 693 100%
Cooperative Congressional Election Survey and found MTurk samples
to be quite similar in representativeness and quality. Education Level
High school graduate 145 20.9%
Using MTurk’s Premium Qualifications respondent-specification Associate degree 136 19.6%
options, only U.S.-based respondents who had both a Twitter and a Bachelor’s degree 306 44.2%
Facebook account were invited to participate in the study (having an Master’s degree 90 13%
Instagram account was not available as a specification option). The Doctorate 13 1.9%
Other 3 0.4%
remuneration per respondent was $1.00 plus the premium qualification
Total 693 100%
cost ($0.20). We followed Kees et al. (2017) best practices for online
data collection by checking the IP addresses and location parameters of Household Income Level
Less than $30,000 246 35.5%
the participants, as well as including multiple attention check measures $30,000–$59,999 269 38.8%
in different parts of the survey. All respondents were located in the U.S. $60,000–$89,999 118 17%
A total of 693 participants completed the survey after removing five $90,000–$99,999 24 3.5%
whose answers either exhibited an extreme pattern of results or lack of $100,000–$149,999 27 3.9%
More than $150,000 9 1.3%
attention. Respondents ranged from 18 to 82 years old, 50.9% were
Total 693 100%
female, and spent 40.03 min a day on SMPs (Facebook, Twitter, and
Ethnicity
Instagram combined mean). Additional demographic characteristics of
African American 40 5.8%
the respondents are shown in Table 1. Asian 109 15.7%
Caucasian 479 69.1%
Hispanic or Latino 35 5.1%
3.2. Independent variables Middle Eastern 5 0.7%
Native American 7 1%
Construct measurement items were adapted from existing scales, Other/Mixed Race 18 2.6%
Total 693 100%
with minor adaptations for the social media context. Susceptibility to
peer influence about consumption was assessed using a seven-point
scale adapted from Mangleburg and Bristol (1998). During the analysis,
combined across all platforms to calculate the overall social media
one item (“If I don't have a lot of experience with a product, I often ask
susceptibility. Similar to the susceptibility to peer influence measure,
my friends about it.”) was dropped due to low standardized factor
the adapted social media susceptibility scale assessed the impact of both
loading. To measure susceptibility to social media influence, we
normative and informative influences. SMP usage was measured by
adapted Mascarenhas and Higby (1993) susceptibility to media scale
requesting respondents estimate how many minutes per day they spend
that addresses traditional forms of advertising (e.g., TV, radio). Our
on the three SMPs separately. The minutes for each of the three plat-
adapted scale focused solely on SMPs. The items measure the degree to
forms were then averaged.
which consumers are susceptible to the promotional content that exists
within the social media ecosystem on a five-point Likert-type scale
(1 = strongly disagree, 5 = strongly agree). Each item in the scale was 3.3. Dependent variables
further adapted into three variations, specifically applied to each plat-
form (Facebook/Twitter/Instagram) in the current study separately and Attitude toward SMP advertising was assessed using a scale

477
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

Table 2
Item statistics and measurement model results.
Loading CR AVE Cronbach’s alpha

Susceptibility to Peer Influence

Normative:
When buying products, I usually buy the ones that I think my friends will approve of. 0.91 0.91 0.62 0.89
I like to know what products and brands make a good impression on my friends. 0.91
It is important that my friends like the products and brands I buy. 0.91

Informative:
I often ask my friends to help me choose the best product. 0.86
I often get information about a product from friends before I buy. 0.83
To make sure that I buy the right product or brand, I often look at what my friends are buying and using. 0.85

Susceptibility to Social Media Influence

Normative:
I buy only those products/brands that are advertised on (Facebook/Twitter/Instagram). 0.86 0.94 0.75 0.94
(Facebook/Twitter/Instagram) ads often determine my brand loyalties. 0.87
I continue buying the same brands as long as my favorite (Facebook/Twitter/Instagram) stars endorse them. 0.83

Informative:
I always consult (Facebook/Twitter/Instagram) to determine the best buys. 0.93
I always look at the ads on (Facebook/Twitter/Instagram) before I buy. 0.92

Attitude toward SMP Advertising

Overall, I consider (Facebook/Twitter/Instagram) advertising a good thing 0.92 0.96 0.85 0.96
Overall, I like (Facebook/Twitter/Instagram) advertising 0.96
I consider (Facebook/Twitter/Instagram) advertising very essential 0.84
I would describe my overall attitude toward (Facebook/Twitter/Instagram) advertising very favorably 0.97

SMP Advertising Avoidance

Cognitive ad avoidance
I intentionally ignore any ads on the (Facebook/Twitter/Instagram). 0.80 0.95 0.66 0.95
I intentionally don't put my eyes on any ad on (Facebook/Twitter/Instagram). 0.80
I intentionally don't pay attention to (Facebook/Twitter/Instagram). 0.87
I intentionally don't click on any ads on (Facebook/Twitter/Instagram), even if the ads draw my attention. 0.78

Affective ad avoidance
I hate seeing ads on (Facebook/Twitter/Instagram) 0.90
It would be better if there were no ads on (Facebook/Twitter/Instagram) 0.84

Behavioral ad avoidance
I scroll down past ads on (Facebook/Twitter/Instagram) 0.82
I hide/close the ads to avoid them on (Facebook/Twitter/Instagram) 0.76
I do any action to avoid ads on (Facebook/Twitter/Instagram) 0.77

modified from Boateng and Okoe (2015). The respondents rated their concluded that self-report measures are likely to provide a conservative
agreement with four statements regarding each platform separately on estimate of such behaviors, thus implying that by extension there is
a five-point Likert-type scale (1 = strongly disagree, 5 = strongly more actual ad avoidance on SMPs than indicated by self-report.
agree). The average across all platforms was then calculated to de-
termine the overall attitude toward SMP advertising. SMP advertising
4. Results
avoidance was assessed using a 10-item, seven-point Likert-type scale
adapted from Cho and Cheon (2004). The scale measured cognitive,
4.1. Indices of the measurement model
affective, and behavioral ad avoidance on each platform separately.
During analysis, one item (“I install ad-blocking software/apps to re-
The reliability and validity of the latent variables in the proposed
move ads on (Facebook/Twitter/Instagram).”) was dropped due to low
model was assessed using confirmatory factor analysis in AMOS 25. The
standardized factor loading. There were no significant differences
results revealed a satisfactory fit (Chi-square = 635.28, df = 236, non-
across all three platforms. The overall SMP ad avoidance was then
normed fit index [NFI] = 0.97, confirmatory fit index [CFI] = 0.98,
computed by averaging the results of the three platforms. Table 2
incremental fit index [IFI] = 0.98, root mean square error of approx-
contains the final pool of items and their reliability values.
imation [RMSEA] = 0.049). According to Hu and Bentler (1999), the
We note that all of our IV and DV measures are self-reported in
overall fit was acceptable and the relevant factor loadings were sub-
nature. There are no existing studies we are aware of that compare self-
stantial and highly significant. The results indicate that all measures
report vs non-self-report measures of CS, SMP ad attitude, or SMP
achieved acceptable levels of reliability based on composite reliability
avoidance, and there exist only a very limited handful doing such a
scores exceeding 0.70 (Fornell & Larcker, 1981). Similarly, Cronbach’s
comparison as applied to any kind of ad avoidance. Ferguson (1994)
alpha coefficients of the latent constructs ranged from 0.89 to 0.96.
compared participants’ responses to a self-report questionnaire on their
Discriminant validity was assessed by comparing the average variance
TV channel flipping behavior with their observed behaviors (TV
extracted (AVE) values with their associated pair of correlations. The
channel flipping behavior was not solely limited to, but included the
correlation coefficients among latent variables were less than the
behavior of, ad skipping). The results showed correlations between self-
square roots of AVEs, which indicates adequate discriminant validity
reported responses and actual TV channel flipping behavior, suggesting
(Chin, 1998). In addition, the AVE for each construct exceed the 0.50
that self-report scales provide acceptable estimates of ad avoidance. In a
threshold, varying from 0.62 to 0.82. The standardized factor loadings
review of research on remote control usage behaviors (including skip-
were all greater than 0.6. Thus, the scale items provided a good re-
ping of TV ads via changing channels), Bellamy and Walker (1996)
presentation of the constructs (see Table 3 for the correlation coefficient

478
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

Table 3
Correlation coefficient matrix.
Peer susceptibility SMP usage Social media susceptibility Attitude toward SMP advertising SMP advertising avoidance Mean (SD)

Peer susceptibility (0.787) 3.62 (1.45)

SMP usage 0.178 (0.710) 40.03 (35.69)
Social media susceptibility 0.565 0.289 (0.866) 2.10 (1.08)
Attitude toward SMP advertising 0.343 0.297 0.686 (0.922) 2.76 (1.09)
SMP advertising avoidance −0.171 −0.191 −0.450 −0.686 (0.812) 4.93 (1.47)

Note: Square root of average variance extracted (AVE) is shown on the diagonal of the matrix.

matrix). H2b hypothesized the mediation effect of attitude toward SMP adver-
Although Fuller, Simmering, Atinc, Atinc, and Babin (2016) sug- tising. Therefore, H2b is supported. With regards to SMP usage, the
gested that common method variance (CMV) does not represent a ser- direct effect on SMP ad avoidance was found to be insignificant
ious threat to the validity of survey research, they recommended that a (β = 0.02, p > .1). However, the indirect effect test was significant
priori steps be taken to minimize risks of bias. In the current research, (β = −0.067, p < .01) showing a full mediation effect of attitude
several techniques were used to lessen CMV effects, including pro- toward SMP advertising (thus H3b is supported). Lastly, when ex-
tecting respondent anonymity, asking the respondents to answer as amining the mediating role of attitude toward SMP advertising on the
honestly as possible, and carefully wording and scaling the items relationship between susceptibility to social media influence and SMP
(Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Fuller et al. (2016) ad avoidance, we found the predicted relationship. The direct effect of
concluded that the relationships between variables in single source data susceptibility to social media influence on SMP ad avoidance was in-
will be biased only when the levels of CMV are high. In this research, significant (β = −0.014, p > .1), whereas the indirect effect test was
we performed a Harman’s single-factor test (Podsakoff & Organ, 1986). significant (β = −0.603, p < .001), suggesting a full mediation effect
The results show that the highest covariance explained by one factor is of attitude toward SMP advertising (thus H4b is supported). See Table 4
47.16%. This is lower than the 50% threshold suggested by Podsakoff for a summary of mediation effects of attitude toward SMP ads.
et al. (2003). The model fit remains similar after the inclusion of a
common latent factor (model without common latent factor: χ2/ 4.4. Additional analysis - demographics
df = 2.69, model with common latent factor: χ2/df = 2.71) (Podsakoff
et al., 2003). These results suggest that common method bias is not an The CS framework addresses demographic characteristics as part of
issue for concern. its Antecedents component. However, the framework does not provide
a theory-oriented means of explaining/predicting how each possible
4.2. Structural model testing and hypotheses testing antecedent influences CS outcomes nor which specific ones might do so,
as this would vary greatly depending on topic/outcome being studied.
The fit indices of the structural model are as follow: Chi- Thus, the sample’s demographic characteristics were not included as
square = 664.18, df = 256 (χ2/df = 2.59), NFI = 0.96, RFI = 0.96, part of our model or predictions. Notwithstanding this, we were in-
IFI = 0.98, TLI = 0.97, CFI = 0.98, and RMSEA = 0.048. In com- terested in the extent to which SMP advertising attitude and avoidance
parison with values suggested by Hu and Bentler (1999), the overall fit are influenced by demographic characteristics. Thus, regression ana-
of the proposed model is acceptable. lyses were conducted with all demographic variables as independent
Attitude toward SMP advertising (Asoc) negatively affects SMP ad- variables. Results of the analyses revealed none of the demographic
vertising avoidance (β = −0.772, p < .001, thus H1 is supported). characteristics to be related with attitude toward SMP advertising or
Furthermore, susceptibility to peer influence about consumption was SMP advertising avoidance, neither at an overall level nor platform by
negatively related with attitude toward SMP advertising (β = −0.132, platform. Multi-group analyses showed no significant differences
p < .001. This result is statistically significant but is in the opposite among the demographic groups.
direction predicted, thus H2a is not supported. For SMP usage, re-
spondents reported that they spent an average of 62.33 min per day on 5. Discussion
Facebook, 24.64 min on Instagram, and 28.65 min on Twitter. The total
average time respondents spent daily on all platforms combined was 5.1. Theoretical/Conceptual contributions
115.44 min and the average time per platforms was 40.03 min. SMP
usage and susceptibility to social media influence also show significant Although they have measured advertising avoidance in different
positive effects on attitude toward SMP advertising [SMP usage, ways while using varied scales, prior studies have largely focused on
β = 0.087, p < .01, thus H3a is supported; Social media susceptibility, one key predictor - consumers’ perceptions of advertising in general. If
β = 0.781, p < .001, thus H4a is supported]. See Table 4 for the one has a positive attitude toward advertising in general, one will
summary of hypotheses and results. generally put less effort into avoiding instances of it (actual ads). At a
base level, the current study supports this established finding in the
4.3. Mediating effects of attitude toward SMP ads specific context of SMP advertising. However, we are among the first to
investigate the antecedents of the advertising attitude predictor. We do
Mediation was tested using Mathieu and Taylor (2006) re- so by situating ad avoidance and attitudes/perceptions toward adver-
commended method of using 2000 bias corrected bootstrapping sam- tising within a broader social(ization) context by adopting the CS fra-
ples at 95 BC Confidence Level in AMOS while applying the Shrout and mework. At a general level, the results demonstrate the viability of CS
Bolger (2002) approach for assessing the presence of mediation effects. as a means of understanding the socially driven antecedents of adver-
We found that attitude toward SMP advertising mediates the relation- tising attitudes and by extension the resulting ad avoidance behaviors.
ship between susceptibility to peer influence and SMP ad avoidance. More specifically, the results reveal three predictors of both SMP ad-
The direct effect of susceptibility to peer influence about consumption vertising attitudes and avoidance – susceptibility to peer influence
on SMP ad avoidance was found to be significant (β = 0.088, p < .05). about consumption, SMP usage, and susceptibility to social media in-
The indirect effect test was also significant (β = 0.102, p < .01), fluence.
showing a partial mediation effect of attitude toward SMP advertising. Perhaps the most unexpected result of the current study is with

479
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

Table 4
Structural equation modeling results for hypotheses.
Hypothesized effect Standardized coefficient Standard error p-value Conclusion

H1: Attitude toward SMP advertising → SMP ad avoidance −0.772 0.046 < 0.001 Supported
H2a: Peer susceptibility → Attitude toward SMP advertising −0.132 0.041 < 0.001 Not supported1
H3a: SMP usage → Attitude toward SMP advertising 0.087 0.03 < 0.01 Supported
H4a: Social media susceptibility → Attitude toward SMP advertising 0.781 0.033 < 0.001 Supported

Mediation hypotheses (Attitude toward SMP advertising, Asoc) Direct effects Indirect effects (Boot Boot 95% CI Conclusion
SE)
H2b: Asoc mediates the relationship between peer susceptibility and SMP 0.088* 0.102** [0.048, 0.157] Partial mediation
advertising avoidance
H3b: Asoc mediates the relationship between SMP usage and SMP advertising 0.02 −0.067** [−0.107, −0.029] Full mediation
avoidance
H4b: Asoc mediates the relationship between social media susceptibility and SMP −0.014 −0.603*** [−0.689, −0.532] Full mediation
advertising avoidance

*p < .05; **p < .01; ***p < .001

1
H2a is statistically significant, but in the opposite direction than predicted

regards to H2a. Based on the CS framework and its related mechanisms, about consumption and perceptions of SMP ads indicates that the
we proposed that the more susceptible consumers are to the influence of conformity to norms implied by the susceptibility construct may be
their peers, they will in turn have more positive predispositions towards more domain/context specific than otherwise thought. While certainly
SMP advertising. Indeed, prior studies support this logic of suscept- some people are generally more conformist by nature regardless of the
ibility to peer communications about consumption as indicating a re- context, in the case of SMP ad attitudes and avoidance, being more
ceptiveness to consumption-related messages such as advertisements susceptible/conforming to peers with regards to consumption-related
(Bush et al., 1999; de Gregorio & Sung, 2010; Kwon et al., 2014; Taylor communications does not mean that one will automatically be more
et al., 2011). However, H2a is statistically significant in the opposite easily persuaded by such messages from other sources such as mar-
direction predicted. If one is more susceptible to being influenced by keters. Indeed, our result showing that social media susceptibility is
one’s peer groups in terms of consumption-related behaviors, our result negatively related with SMP ad perception further indicates that sus-
shows one will actually be more negatively disposed towards SMP ad- ceptibility-linked conformity can vary by the domain (or in this case
vertising messages and in turn more likely to avoid such messages. social force) under consideration.
A possible explanation for this finding comes from the word-of- In addition to demonstrating the influence of peers, the study also
mouth (WOM) literature as consumption-related peer communications validates the key socialization agent of the media as an influence on
are a form of such communication. One of the earliest formal studies of SMP advertising attitudes and avoidance. As noted earlier, prior results
WOM found that consumption-related discussion among peers was the regarding the impact of media usage on advertising outcomes are in-
most influential source of product information and purchase decisions, consistent. The few studies looking at media usage and advertising at-
more so than advertisements (Katz & Lazarsfeld, 1955). In the inter- titude also have shown mixed results, with some finding a positive
vening time, industry and academic studies have demonstrated that association of usage (Bush et al., 1999; de Gregorio & Sung, 2010) and
WOM is (still perceived to be) more credible, trustworthy, and preferred others finding no relation at all (Kwon et al., 2014; Smith & Moschis,
than marketer-driven messaging (e.g., Bickart & Schindler, 2001; 1984). Our results support the influence of media as a socialization
Murray, 1991). Thus, rather than indicating a receptiveness to corpo- agent on both SMP ad attitude and ad avoidance. In contrast to peer
rate messaging, (susceptibility to) peers in a sense replaces and to a influence, greater usage of SMPs and having a higher susceptibility to
certain extent counters such communications. Peers are essentially a being influenced by social media content both predict attitudes towards
competing, rather than complementary, and more trusted source of SMP advertising and less SMP ad avoidance. Attitudes also mediate the
information and decision-making than companies and brands. These relationships between both amount of SMP usage and susceptibility to
results are interesting in light of the fact that mainstream SMPs all social media content and SMP ad avoidance. Thus, if consumers use
provide some form of user-visible, peer-related information regarding SMPs to a greater extent, and are more open to being influenced by the
posted content (e.g., showing that X number of one’s connections/ content they consume and interact with, they are likely to be more
friends have liked or shared a marketer-originating post containing an positively disposed towards company-originating messages on SMPs
ad). Our results imply that the mere presence of such peer information and less likely to avoid such messages.
about sponsored content is not enough to counter the negative per- In our lead up to H3a/b, we noted that CS researchers have largely
ception of marketer messages such as ads among those who are high in operationalized the assessment of media as a socialization agent by
susceptibility to peer influence. Such consumers trust their peers sig- measuring the total amount of time participants spend using a parti-
nificantly more in terms of consumption decisions than they do mar- cular medium. In the current study, we similarly measured (social)
keters. Thus, they regard such peer-related statistics about marketing media usage. However, in line with researchers such as Bearden et al.
content as simply another tactic marketers are using to persuade them. (1989) and Mascarenhas and Higby (1993) who propose that suscept-
A further implication of H2a’s result relates with the nature of the ibility in response to a stimulus is not synonymous with stimulus
conformity implied by being susceptible to peer discussions. Although duration, we additionally incorporated a measure of susceptibility to
not investigated extensively, being susceptible to the opinions of close social media influence. While our results show that both social media
others is considered a form of conformity and has been posited as usage and susceptibility are positively related with SMP ad attitudes,
generally being a trait as opposed to a contextual phenomenon that the strength of the relationships differs. As shown in Tables 3 and 4,
varies much by the situation or nature of the close other (Boush et al., social media susceptibility had a notably stronger impact on SMP ad
1994; Mangleburg & Bristol, 1998). This further implies that such attitude than SMP usage. These results provide some preliminary in-
susceptibility extends beyond the peers context to being more accepting dication that the usage of any particular medium should not be con-
of (more likely to conform with) persuasive communications from other flated as being the same thing as being susceptible to the messages on
entities, such as marketer-originating messages. Our finding of a ne- that medium. While usage and susceptibility may have effects in the
gative relationship between susceptibility to peer communications same direction, they seem to have differing strengths of impact. Thus,

480
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

we recommend that CS researchers incorporate dual measures of both confirm, it is possible a positive feedback loop might occur. If SMPs
amount and influence susceptibility when assessing impact of media. allow marketers more/different options (such as extent of SMP usage
and susceptibility to social media) by which to target consumers, this
5.2. Managerial implications would result in presumably more accurately targeted ads. In turn, this
should increase the effectiveness of such ads. As SMP ad effectiveness
For SMPs and marketers using SMPs for promotional purposes, the increases, those who have higher susceptibility to social media (i.e.,
results of this research provide interesting implications. Most SMPs rely they are open to and use SMPs to help in consumption choices) would
on revenue from advertising rather than charging consumers for access, become further susceptible since the ads are now more accurately ad-
and thus are constantly seeking to improve the attractiveness of their dressing their needs and thus their perception of SMPs as aids for
platforms to marketers. Our results indicate that SMPs should provide consumer decisions would increase.
clients with means of selecting and targeting audiences based not only Regarding the impact of peers, in contrast to SMP usage levels and
on the typical demographic and psychographic criteria, but also pro- susceptibility to social media influence, marketers should take caution
pensity to be receptive to SMP advertising. in using peer susceptibility as an additional basis by which to find those
Our results suggest that focusing on consumers who (1) spend more with a greater openness to company-originating messages. Upon first
time on SMPs and/or (2) are more susceptible to social media influence glance, and based on our original H2a, it may seem beneficial to find
regarding their consumption habits would assist marketers in attaining individuals who demonstrate a pattern of responding to (e.g., liking,
greater ROI from SMP advertising. Regarding SMP usage, marketers sharing) brand-related postings of peers or requesting product re-
and SMPs should not, however, interpret our result of greater usage commendations among their social connections. These actions would
being related with SMP ad attitude as an indicator of an actively and seem to indicate an overall positive disposition to brands and, by ex-
inherently positive disposition towards ads (and, by extension, less ad tension, receiving messages about said brands from companies.
avoidance). Most people use SMPs primarily to engage in activities that However, our results show that such individuals are actually less likely
are social in nature (e.g., interacting, keeping up to date on life events to have positive attitudes towards SMP advertising and would avoid
and milestones) and less so to find product information (Valentine, such ads to a greater extent. What is plausibly occurring is that peers
2018). As we discuss in Section 2.1.4, the basis for the positive SMP are serving as a competing, alternate source of consumption and pro-
usage-ad attitude relation is due more to processes similar to mere duct decisions versus advertisements. Thus, marketers could still pursue
exposure and cultivation. The more people are exposed to the eco- users who actively engage with their online connections regarding
system of marketing that exists on SMPs via greater usage, the more the products and purchases, but they should minimize the use of paid ads.
presence of such marketing messages is normalized and, in turn, the These consumers are not necessarily anti-consumption or anti-brands
more there is a broadly positive acceptance regarding their presence. (given that they, by definition, discuss consumption-related matters
Most SMPs certainly track usage by analyzing average duration of use frequently), but have negative perceptions of the overt promotion of
of their service and number of visits/logins in a defined period, as in- products via SMP advertising. A common reason for such negative at-
dicated by the plentiful public data regarding time spent on specific titudes is the perception that ad content holds little relevance to the
SMPs. Allowing marketers to target based on these and other usage receiver and is an interruption to the activities people are engaged in
level indicators would be a relatively easy additional option SMPs could (e.g., watching an online video, reading a blog post) (e.g. Cho & Cheon,
provide marketers to assist in their campaign effectiveness. Further- 2004; Kelly et al., 2010). Thus, marketers could alternatively commu-
more, by adding such a targeting mechanism, SMPs could charge higher nicate with consumers on a more direct basis with highly targeted,
rates to marketers for ads that used it as it would be an extra choice that individually customized, messages such as direct tweets or personalized
currently goes beyond the standard demographics and interests options. e-mails. Such direct outreach carefully conducted based on accurate
However, unlike in the case of SMP usage level, social media sus- analysis of personalized data may well circumvent negative attitudes
ceptibility implies a general receptivity to using SMPs for the purposes toward ads being irrelevant and spam-filled. In addition, such direct
of at least facilitating consumption decisions. As shown in Table 2, the communication methods as e-mails and tweets are less susceptible to
susceptibility scale specifically focuses on the use of SMPs for con- the more overt and easy to implement behavioral avoidance methods
sumption-oriented purposes such as determining best buys and speci- such as ad blockers.
fically seeking out what is being advertised or promoted by companies An additional implication of our result regarding peer commu-
people might follow. While the main use of SMPs is for primarily social nications about consumption relates to the means by which social in-
activities and entertaining, almost a third of users also use them to help formation is incorporated into marketers’ promotional posts. In Section
make consumption decisions (Valentine, 2018). However, short of di- 5.1, we proposed that a plausible reason people who exhibit greater
rectly asking its users to complete the susceptibility to social media peer communication feel more negatively towards SMP ads and in turn
influence scale (which may not be particularly welcomed by users), avoid such ads to a greater extent is because, for such consumers, peers
assessing susceptibility would necessitate indirect measurement via a essentially supplant marketers as sources of information about pur-
deeper dive into an SMPs massive trove of user data. This susceptibility chasing. We also noted that SMPs including social statistics as part of
could be inferred via data regarding the overall volume of clicks, shares, advertisers’ posts (e.g. X number of people have liked an ad) is not
purchases, and other actions individuals take in response to marketer- enough to override peoples’ negative perception of SMP ads. However,
originating content such as ads, sponsored posts, and interactive games. that is not to say that such social information is not important for
The greater the number of such actions, the more likely an individual is marketers. Indeed, our results reinforce that peers are highly important
open to “official” messaging from a specific brand and thus less likely to (Valentine, 2018). Thus, we recommend that SMPs provide more op-
avoid ads from said brand. tions for marketers in terms of how they may include social statistics
Combined with the now-standard demographics and user interests regarding their posted promotional content. These options should be
targeting criteria that many large SMPs provide, using the more indirect focused around making it easier in terms of (1) prominence for con-
predictors of SMP usage and susceptibility to social media would allow sumers to notice such social information and (2) connecting it to online
SMPs to provide added value and generate more revenue. In addition, peers. Currently, information such as number of likes and shares are
because susceptibility to social media would require more effort to often small in size relative to the posted content. One simple suggestion
ascertain by SMPs (given that it cannot be as easily collected as de- would be to increase the size of such social data relative to the mar-
mographic information), SMPs could charge marketers a premium to keters’ post/ad size, thus making it more attention-capturing. In addi-
access this targeting criteria as well as for any ads that used such data. tion, such information is typically underneath marketer posts. We
Furthermore, although speculative and requiring research to (dis) suggest moving them above such posts – given that people generally

481
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

read content from top to bottom, consumers would encounter this data our knowledge, there exist only two published academic field studies of
immediately prior to seeing marketer content, thereby providing them a online ad avoidance - neither are specifically in the context of SMPs and
pre-cue to its popularity. both focus solely on ad-blocker technology usage (Shiller, Waldfogel, &
In addition to changes in the prominence of how social statistics Ryan, 2018; Söllner & Dost, 2019). However, the benefits of such sec-
about marketer posts are provided, we suggest SMPs make it easier for ondary/big data approaches would further enhance understanding of
marketers to assist consumers in connecting said statistics to their peers. the ways by which consumers’ socialization factors and processes im-
Certainly, SMPs do provide marketers the means of displaying social pact their ad avoidance. In addition, such studies would also allow the
data in terms of both quantity (e.g., number of shares, likes/dislikes, exploration of other important issues that have not yet been in-
comments, views) and quality/valence (e.g., visual representations of vestigated in the context of avoidance, such as the extent to which
people’s reactions, such as love, dislike, surprise), this data is often not consumers who avoid ads purchase products and how the accuracy/
easily or explicitly connected to specific users. Thus, we suggest that appropriateness of ad targeting quality impacts avoidance.
SMPs add options to make such connections easier without consumers Finally, social media is modifying/expanding the conceptualization
having to take extra steps such as clicking on/rolling over information. of who is considered a “peer”, as well as providing varied means of
For example, in addition to showing users how many people have ex- interacting with them. Thus, studies assessing differences in the extent
pressed approval of a piece of content, SMPs could allow adding specific to which face-to-face video interactions vs. written-only, real-time vs.
information that X number of one’s friends/connections (perhaps even lagged time (e.g. live chatting on a SMP platform vs delayed messa-
including their names) have liked or otherwise reacted to said content. ging), and weak tie vs. strong tie peer communications impact SMP ad
While data regarding one’s direct connections/friends on SMPs would avoidance would add greatly to understanding in this area.
be ideal, such data would not be available if none of one’s connections
has been exposed or interacted with the content in question. In such References
cases, the data would not have to strictly be limited to direct connec-
tions but could also be slightly broader in encompassing the notion of a Anderson, M., Perrin, A., Jiang, J., & Kumar, M. (2019). 10% of Americans don’t use the
peer. For example, if one’s direct connections have not responded to a internet. Who are they? accessed July 23, 2019 at: https://www.pewresearch.org/
fact-tank/2019/04/22/some-americans-dont-use-the-internet-who-are-they/.
movie trailer, information about how similar others have reacted would Bearden, W. O., & Etzel, M. J. (1982). Reference group influence on product and brand
also help in making psychological connections between a user and the purchase decisions. Journal of Consumer Research, 9(2), 183–194.
content (e.g. displaying that X number of people who live in one’s town Bearden, W. O., Netemeyer, R. G., & Teel, J. E. (1989). Measurement of consumer sus-
ceptibility to interpersonal influence. Journal of Consumer Research, 15(4), 473–481.
or zipcode have shared the trailer or that Y people who also enjoy Bellamy, R. V., Jr, & Walker, J. R. (1996). Television and the remote control: Grazing on a
science fiction have liked the trailer). vast wasteland. New York, NY: Guilford Press.
Bickart, B., & Schindler, R. M. (2001). Internet forums as influential sources of consumer
information. Journal of Interactive Marketing, 15(3), 31–40.
5.3. Limitations and future research directions Boateng, H., & Okoe, A. F. (2015). Consumers’ attitude towards social media advertising
and their behavioural response: The moderating role of corporate reputation. Journal
Our study expands knowledge about SMP ad avoidance by de- of Research in Interactive Marketing, 9(4), 299–312.
Boush, D. M., Friestad, M., & Rose, G. M. (1994). Adolescent skepticism toward TV ad-
monstrating that susceptibility to consumption-related peer influence,
vertising and knowledge of advertiser tactics. Journal of Consumer Research, 21(1),
extent of SMP usage, and susceptibility to social media content are 165–175.
predictors of attitude toward SMP advertising. It also verifies the key Bush, A. J., Smith, R., & Martin, C. (1999). The influence of consumer socialization
mediating effect of SMP attitudes on the relationship between these variables on attitude toward advertising: A comparison of African-Americans and
Caucasians. Journal of Advertising, 28(3), 13–24.
predictors and SMP ad avoidance, such that more positive attitudes lead Chin, W. W. (1998). The partial least squares approach to structural equation modeling.
to less avoidance. However, a limitation of this work is that it does not In G. A. Marcoulides (Ed.). Modern methods for business research (pp. 295–336). New
provide an in-depth means of assessing the underlying “why” of the York: Lawrence Erlbaum Associates.
Cho, C. H., & Cheon, H. J. (2004). Why do people avoid advertising on the Internet?
results. There have been few analyses of how exactly socialization Journal of Advertising, 33(4), 89–97.
processes operate in the CS framework to influence its outcomes and Colliander, J. (2019). “This is fake news”: Investigating the role of conformity to other
the extent to which each process contributes to the overall socialization, users’ views when commenting on and spreading disinformation in social media.
Computers in Human Behavior, 97, 202–215.
in this context specifically in reference to avoiding SMP advertising. Coppock, A. (2019). Generalizing from survey experiments conducted on Mechanical
Future researchers in the area are encouraged to delve into the un- Turk: A replication approach. Political Science Research and Methods, 7(3), 613–628.
derlying mechanism, thoughts, feelings, and factors that drive such Dao, W. V. T., Le, A. N. H., Cheng, J. M. S., & Chen, D. C. (2014). Social media advertising
value: The case of transitional economies in Southeast Asia. International Journal of
socialization processes in regard to SMP advertising perceptions, SMP Advertising, 33(2), 271–294.
usage, and SMP ad avoidance. The data-rich in-depth qualitative De Bruyn, A., & Lilien, G. L. (2008). A multi-stage model of word-of-mouth influence
methods such as those used by Kelly, Kerr, and Drennan (2010, 2018) through viral marketing. International Journal of Research in Marketing, 25(3),
151–163.
and Knittel, Beurer, and Berndt (2016) would result in deeper insight
de Gregorio, F., & Sung, Y. (2010). Understanding attitudes toward and behaviors in
into the internal(ized) processes by, and reasons for, which people response to product placement. Journal of Advertising, 39(1), 83–96.
avoid SMP ads. Such approaches would expand understanding of not Ferguson, D. A. (1994). Measurement of mundane TV behaviors: Remote control device
just the more easily observable (and measurable) behavioral avoidance flipping frequency. Journal of Broadcasting and Electronic Media, 38(1), 35–47.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with un-
methods (e.g. installing ad-blockers, clicking to close ads), but also the observable variables and measurement error. Journal of Marketing Research, 18(1),
less trackable ones such as cognitively choosing to ignore or reflexively 39–50.
disbelieve any ads encountered. Fuller, C. M., Simmering, M. J., Atinc, G., Atinc, Y., & Babin, B. J. (2016). Common
methods variance detection in business research. Journal of Business Research, 69(8),
The current study is exploratory in nature, designed to determine 3192–3198.
the extent to which CS can be utilized as a framework to understand the Gerbner, G. (1970). Cultural indicators: The case of violence in television drama. Annals
antecedents of SMP ad perceptions and by extension ad avoidance on of the American Academy of Political and Social Science, 388(1), 69–81.
Goodman, J. K., & Paolacci, G. (2017). Crowdsourcing consumer research. Journal of
SMPs. A fruitful path for future research to build on our work is field Consumer Research, 44(1), 196–210.
study approaches that, for example, conduct field experiments with Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indices in co-variance structure
consent from an SMP or acquire and analyze avoidance and CS-related analysis: Conventional criteria versus new alternatives. Structural Equation Modeling,
6(1), 1–55.
data from SMPs or third-party tracking companies. Key variables such Huff, C., & Tingley, D. (2015). “Who are these people?” Evaluating the demographic
as ad avoidance could be assessed by ad-blocker presence as well as characteristics and political preferences of MTurk survey respondents. Research &
mouse movement tracking, and susceptibility to peer influence mea- Politics, 2(3), 1–12.
Intravia, J., Wolff, K. T., Paez, R., & Gibbs, B. R. (2017). Investigating the relationship
sured by text/sentiment analysis performed on users’ product/brand
between social media consumption and fear of crime: A partial analysis of mostly
conversations or posts. Access to such data is notably challenging. To

482
S. Chinchanachokchai and F. de Gregorio Journal of Business Research 110 (2020) 474–483

young adults. Computers in Human Behavior, 77, 158–168. Investigating the relationship between user perceptions and advertising avoidance.
Janiszewski, C. (1993). Preattentive mere exposure effects. Journal of Consumer Research, Computers in Human Behavior, 63, 444–452.
20(3), 376–392. Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental stu-
Katz, E., & Lazarsfeld, P. A. F. (1955). Personal influence: The part played by people in the dies: New procedures and recommendations. Psychological Methods, 7(4), 422–445.
flow of mass communication. Glencoe. Illinois: Free Press. Smith, A., & Anderson, M. (2018). Social media use in 2018, accessed February 11, 2019
Kees, J., Berry, C., Burton, S., & Sheehan, K. (2017). An analysis of data quality: at https://www.pewresearch.org/internet/wp-content/uploads/sites/9/2018/02/PI_
Professional panels, student subject pools, and Amazon's Mechanical Turk. Journal of 2018.03.01_Social-Media_FINAL.pdf.
Advertising, 46(1), 141–155. Smith, R. B., & Moschis, G. P. (1984). Consumer socialization of the elderly: An ex-
Kelly, L., Kerr, G., & Drennan, J. (2010). Avoidance of advertising in social networking ploratory study. Advances in Consumer Research, 11(1), 548–552.
sites: The teenage perspective. Journal of Interactive Advertising, 10(2), 16–27. Söllner, J., & Dost, F. (2019). Exploring the selective use of ad blockers and testing banner
Kelly, L., Kerr, G., & Drennan, J. (2018). Triggers of engagement and avoidance: Applying appeals to reduce ad blocking. Journal of Advertising, 48(3), 302–312.
approach-avoid theory. Journal of Marketing Communications. Speck, P. S., & Elliott, M. T. (1997). Predictors of advertising avoidance in print and
Knittel, Z., Beurer, K., & Berndt, A. (2016). Brand avoidance among Generation Y con- broadcast media. Journal of Advertising, 26(3), 61–76.
sumers. Qualitative Market Research, 19(1), 27–43. Statista (2019a). Facebook's global revenue as of 3rd quarter 2019, by segment, accessed
Kwon, E. S., Kim, E., Sung, Y., & Yoo, C. Y. (2014). Brand followers: Consumer motivation December 1, 2019 at: https://www.statista.com/statistics/277963/facebooks-
and attitude towards brand communications on Twitter. International Journal of quarterly-global-revenue-by-segment.
Advertising, 33(4), 657–680. Statista (2019b). Projected revenue of Instagram from 1st quarter 2017 to 4th quarter
Langan, R., & Kumar, A. (2019). Time versus money: The role of perceived effort in 2020, accessed December 1, 2019 at: https://www.statista.com/statistics/448157/
consumers' evaluation of corporate giving. Journal of Business Research, 99, 295–305. instagram-worldwide-mobile-internet-advertising-revenue.
Lueg, J. E., Ponder, N., Beatty, S. E., & Capella, M. L. (2006). Teenagers’ use of alternative Statista (2019c). Twitter's advertising revenue in the United States from 2017 to 2021,
shopping channels: A consumer socialization perspective. Journal of Retailing, 82(2), accessed July 23, 2019 at: https://www.statista.com/statistics/232384/forecast-of-
137–153. twitters-advertising-revenue.
Mangleburg, T. F., & Bristol, T. (1998). Socialization and adolescents' skepticism toward Taylor, D. G., Lewin, J. E., & Strutton, D. (2011). Friends, fans, and followers: Do ads
advertising. Journal of Consumer Research, 27(3), 11–21. work on social networks? Journal of Advertising Research, 51(1), 258–275.
Martin, B. A. S., Wentzel, D., & Tomczak, T. (2008). Effects of susceptibility to normative Trusov, M., Bodapati, A. V., & Bucklin, R. E. (2010). Determining influential users in
influence and type of testimonial on attitudes toward print advertising. Journal of Internet social networks. Journal of Marketing Research, 47(4), 643–658.
Advertising, 37(1), 29–43. Tsay-Vogel, M., Shanahan, J., & Signorielli, N. (2018). Social media cultivating percep-
Mascarenhas, O. A. J., & Higby, M. A. (1993). Peer, parent, and media influences in teen tions of privacy: A 5-year analysis of privacy attitudes and self-disclosure behaviors
apparel shopping. Journal of the Academy of Marketing Science, 21(1), 53–58. among Facebook users. New Media & Society, 20(1), 141–161.
Mathieu, J. E., & Taylor, S. R. (2006). Clarifying conditions and decision points for Valentine, O. (2018). Top 10 reasons for using social media. accessed July 22, 2019 at:
mediational type inferences in organizational behavior. Journal of Organizational https://blog.globalwebindex.com/chart-of-the-day/social-media/.
Behavior, 27(8), 1031–1056. Vakratsas, D., & Ambler, T. (1999). How advertising works: What do we really know?
Mazodier, M., Henderson, C. M., & Beck, J. T. (2018). The long reach of sponsorship: How Journal of Marketing, 63(1), 26–43.
fan isolation and identification jointly shape sponsorship performance. Journal of Wang, X., Yu, C., & Wei, Y. (2012). Social media peer communication and impacts on
Marketing, 82(6), 28–48. purchase intentions: A consumer socialization framework. Journal of Interactive
Mishra, A., Maheswarappa, S. S., Maity, M., & Samu, S. (2018). Adolescent's eWOM in- Marketing, 26(4), 198–208.
tentions: An investigation into the roles of peers, the Internet and gender. Journal of Ward, S. (1974). Consumer socialization. Journal of Consumer Research, 1(2), 1–14.
Business Research, 86, 394–405. Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social
Moschis, G. P. (1987). Consumer socialization. Lexington, MA: Lexington Books. Psychology, 9(2), 1–27.
Moschis, G. P., & Churchill, G. A., Jr. (1978). Consumer socialization: A theoretical and
empirical analysis. Journal of Marketing Research, 15(4), 599–609. Sydney Chinchanachokchai (Ph.D., University of Illinois at Urbana-Champaign) is an
Muehling, D. D., & McCann, M. (1993). Attitude toward the ad: A review. Journal of Assistant Professor in the Department of Marketing of the College of Business
Current Issues and Research in Advertising, 15(2), 25–58. Administration at the University of Akron. Her primary research interests are in the areas
Murray, K. B. (1991). A test of services marketing theory: Consumer information acqui- of multitasking, mind wandering, media avoidance, attention, and cross-cultural con-
sition activities. Journal of Marketing, 55(1), 10–25. sumer behavior. Her research has appeared in journals such as the Journal of Business
Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research: Problems Research, Computers in Human Behavior, Journal of Marketing Management, and Social
and prospects. Journal of Management, 12(4), 69–82. Marketing Quarterly, among others.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method
biases in behavioral research: A critical review of the literature and recommended
remedies. Journal of Applied Psychology, 88(5), 879–903. Federico de Gregorio (Ph.D., University of Georgia) is an Associate Professor in the
Seyedghorban, Z., Tahernejad, H., & Matanda, M. J. (2016). Reinquiry into advertising Department of Marketing of the College of Business Administration at the University of
avoidance on the Internet: A conceptual replication and extension. Journal of Akron. His research focuses on advertising phenomena, digital marketing, and product
Advertising, 45(1), 120–129. placement. His research has appeared in journals such as the Journal of Advertising,
Journal of Advertising Research, International Journal of Advertising, and Journal of Consumer
Shiller, B., Waldfogel, J., & Ryan, J. (2018). The effect of ad blocking on website traffic
and quality. RAND Journal of Economics, 49(1), 43–63. Marketing, among others.
Shin, W., & Lin, T. T. C. (2016). Who avoids location-based advertising and why?

483