CHAPTER-7

RESEARCH FINDINGS,
IMPLICATIONS AND CONCLUSION
 CHAPTER 7
RESEARCH FINDINGS, IMPLICATIONS AN

7.1 Introduction

P
resent chapter is the concluding chapter of the thesis; accordingly, chapter
summarizes the research efforts so far, and suggests future research directions
along with implications and limitations of the research. First of all, chapter provides an
overview of the research approach undertaken by the researcher. After this, chapter presents
summary of the research results to provide a quick glance of the main findings, in reference
to the objectives of the study, to the reader. Next section elaborates on theoretical, practical
and social implications of the study. This section is followed by the section enumerating the
possible limitations of the study. Concluding section of the chapter gives suggestions about
future research possibilities.

7.2 Overview of Research Approach

In accordance with the research problem and research questions, relevant literature
related to social media, consumer behavior, credibility, engagement and reference group
influences were reviewed to define the concepts and constructs and to frame six research
objectives (and their sub-objectives) and corresponding hypotheses. Based on existing
theoretical and empirical studies, multi-item rating scales were developed for measuring
research variables. Survey instrument so developed was refined based on feedback of
experts from industry and academia as well as on the basis of results of pilot study. Data
collected from users of social media residing in NCT of Delhi, using an online
questionnaire, resulted in 530 relevant responses. Various measurement scales used in

196
 Research Findings, Implications and

questionnaire were checked for internal consistency taking help of cronbach‘s alpha. Value
of coefficient alpha indicated the reliability of the measurement scale of each construct.

7.3 Research Findings

This section pertains to summary of research results. First, important findings related
to demographic details of the sample have been given followed by findings related to social
media usage profile. After this, major findings related to research objectives have been
given.
Analysis of demographic profile of the sample revealed that social media platforms
are majorly used by youth for making purchase related decisions with mean age of the
respondents being 26.66 years (SD 9.11 years) and with around 78 % of the respondents
being less than 30 years in age. This corroborates well with other researchers in India and
abroad (Li, 2007; Lenhart et al., 2010; Vizisense, 2010; Chui et al., 2012; Ortiz-Ospina,
2019; Chaffey, 2019; Pragati, 2019; Kemp, 2020; Diwanji, 2020a; SannamS4, (n. d.)).
Finding, that around 22 % of respondents in the age group of 30 years and above were
using social media platforms for making purchase related decisions, seems to be in
agreement with Chaffey (2019). As per Chaffey (2019), social media is attaining maturity
and slowly generational divide may diminish but as India is largely a young country
(Pragati, 2019), as of now that seems to be little far ahead in India. More number of males
was using social media in comparison to females, a fact again corroborated by Kemp
(2020) and SannamS4 (n. d.) that there is digital gender divide prevalent in India. Only
40.6 % women were found to be using social media in comparison to 59.4 % of men and
that too in a state like Delhi evidences gender divide. It was found that unmarried
population (72.8%) is more active on social media. Education and social media usage seems
to be related as 94.5% respondents were graduates and above. Students were found to be the
heaviest users followed by private jobholders, again corroborating the youth dominance of
these mediums. Majority respondents (69.2%) were having monthly household income up
to Rs. 02 Lakhs. Average Internet usage per day for sample respondents was 4.28 hours (SD
3.6 hours). 68.3% of respondents were using internet for up to 4 hours in a day. Only 11.1
% liked to use internet for more than 8 hours in day. It was also revealed that social media
was used by respondents in all districts of Delhi state. However, East Delhi topped the list
with 12.1 % users coming from the district.

19
 Research Findings, Implications and

Major insights from social media usage profile of the respondents using social media
for purchase related decisions in Delhi state revealed that respondents using social media for
purchase related decisions in Delhi state spend average 2.74 hours per day on social media
(SD 2.59 hours), which is higher than the national average of 2.42 hours (Pragati, 2019;
Kemp, 2020). 81.2% of the respondents were using social media for more than 2 years and
thus may be considered mature and proficient at handling various social media platforms.
87% of the respondents were engaged on social media platforms up to o4 hours per day.
Only 3.4% used social media for more than 08 hours a day. Analysis further elucidated that
most preferred mode to access social media in Delhi state was mobile (70.2%) followed by
laptop (56%). This is also in alignment with early researches (Mohsin, 2019; Kemp, 2020).
Tab (15.1%) was the least preferred mode for assessing social media. Home (57.4%) was
the most preferred location for accessing social media. Though, people do not mind
accessing social media from anywhere (52.8%), office (13%) was least preferred. Delhiites‘
most preferred time to access social media or social media prime time of Delhiites is 20:00
to 24:00 (49.5%) followed by 16:00 to 20:00 (44.2%). This corroborated well with preferred
location, as people prefer to access social media from home and during 16:00 to 24:00 time
slot majority people are back from their offices and may browse social media platforms
sitting in the cozy and safe environments of their homes.
First objective of the study was ―To identify the factors of social media influencing
consumer Decision-making process‖. EFA conducted on collected data using principal
component analysis with varimax orthogonal rotation and using Kaiser‘s criterion resulted
in five factors viz. Perceived Ease of Information Exchange, Perceived Purchase Relevant
Benefits, Perceived Risk, Perceived Hedonism and Perceived Fluency. Total variance
explained was 63.276 %. Out of which, first factor explained 22.951%, second factor
explained 14.805%, third factor explained 11.104%, fourth factor explained 9.436% and
fifth factor explained 4.980% of total variance. Reliabilities of the extracted five factors
were .847, .820, .838, .858, .794 respectively. Confirmatory factor analysis established the
dimensionality and validity of the latent factors. Selected AMOS output revealed that all
unstandardized estimates were reasonable and statistically significant with all standard
errors being appropriate and all critical ratios being greater than ± 1.96. All path coefficients
were significant and standardized regression weights ranged between .60 and .91.
Goodness-of-fit indices indicate that hypothesized five factor measurement model is an
adequate fit to sample data (χ2/DF = 1.335, GFI=.958, AGFI= .947, CFI= .985, RMSEA =

19
 Research Findings, Implications and

.025 and ECVI = .706 (least compared to other two models)). Factors were found to be
having content, convergent and discriminant validity.
Second objective of the study was ―To assess the credibility of information
available at selected social media platforms in influencing consumer decision-making
process‖. One- way ANOVA adjusted using Welch resulted in F (4, 1314.392) = 33.853,
p<.05, which was significant and thus it was concluded that there was a significant
difference in mean values of credibility of information available at selected five social
media platforms in influencing consumer decision- making process. Pairwise comparison
evidenced credibility of information available at Twitter (Mean 3.29, SD .98) was highest
and significantly more credible in influencing consumer decision-making process than
almost all other platforms. Information available at Reviewer sites and Blogs was also found
out to be more credible compared to other platforms though not significant in all one-to-one
comparisons. Credibility of information available at Facebook was least (Mean 2.73, SD
.78) and significantly less credible than all other social media platforms.
Third objective was ―To assess credibility dimensions viz. Currency,
Coverage, Accuracy, Authority and Objectivity of information available at selected social
media platforms in influencing consumer decision-making process‖. To achieve third
objective five sub-objectives were framed, each sub-objective was to test a different
dimension of information credibility. For first sub-objective, one-way ANOVA adjusted
using Welch resulted in F (4, 1319.559) = 29.723, p<.05, which was significant meaning
thereby that there was a significant difference in mean values of Currency of information
available at selected social media platforms in influencing consumer decision- making
process. Pairwise comparison revealed that information at Twitter was most current (Mean
3.26, SD 1.37) followed by Reviewer sites and Blogs. Information at Twitter was
significantly more current than information at Facebook and YouTube, but in case of
Reviewer sites and Blogs, differences were not significant. Facebook information is
considered significantly least current (Mean 2.52, SD1.13) by the Delhi consumes for
purchase related decision- making.
For second sub-objective, one-way ANOVA adjusted using Welch resulted in F(4,
1320.754) = 11.346, p<.05 which meant that there was a significant difference in mean
values of Coverage of information available at selected social media platforms in
influencing consumer decision- making process. Pairwise comparison revealed that
information available at Reviewer sites (Mean 3.33, SD 1.31) was most comprehensive

19
 Research Findings, Implications and

followed by Blogs (Mean 3.23, SD 1.27) and Twitter (Mean 3.19, SD 1.28), but it was not
significantly more detailed than Twitter, YouTube and Blogs. Facebook information was
considered significantly least detailed (Mean 2.88, SD 1.09). This finding seems to be along
expected lines as the format of Reviewer sites and Blogs makes it plausible to provide
detailed content compared to Twitter. However, it was surprising that customers found
YouTube to be less detailed than Twitter, though the difference was not significant. One
possible reason for this can be sharing of web links via Twitter, which when clicked take
customers to detailed web content.
Result for third sub-objective came out to be a significant difference in mean values
of Accuracy of information available at selected social media platforms in influencing
consumer decision- making process with Welch adjusted ANOVA F(4, 1321.910) = 23.109,
p<.05. Post hoc paired comparison using Games-Howell‘s test indicated information at
Twitter to be significantly more accurate (Mean 3.23, SD 1.36) than all other platforms
except Reviewer sites. Information at Facebook turned out to be significantly least accurate
(Mean 2.56, SD 1.24) of all.
For fourth sub-objective, Value of Welch‘s F(4, 1320.260) = 14.980, p<.05 was
significant meaning thereby there was a significant difference in mean values of Authority
of information available at selected social media platforms in influencing consumer
decision- making process. Information available at Twitter was considered having
significantly more authority authoritative (Mean 3.39, SD 1.20) than all other platforms,
followed by information at Reviewers sites and Blogs, but the authority of information at
Twitter is not significantly difference than that at Reviewer sites. Information available at
Facebook was found to be having significantly least authority (Mean 2.90, SD 1.02).
Result of fifth sub-objective was that there is a significant difference in mean values
of Objectivity of information available at selected social media platforms in influencing
consumer decision- making process with value of Welch‘s F(4, 1320.615) = 20.584, p<.05.
Once again, information available at Twitter was found to be significantly more objective
(Mean 3.35, SD 1.26) than most of the platforms and information available at Facebook was
considered to be significantly least objective (Mean 2.78, SD 1.06).
Possible reason for information at Twitter being most accurate, most current, most
authority and most objectivity can be the use of Twitter by most of print and TV media to
break news to their subscribers (Mayfield, 2008; Kaplan & Haenlein, 2011). As media
news comes through gatekeepers, it has higher currency, accuracy, authority and objectivity

20
 Research Findings, Implications and

(Metzger, 2007) and perhaps same attributes have been attributed to purchase decision-
making related information too in accordance with congruity theory of Osgood &
Tannenbaum (1955). In addition, the format of microblogging site makes it possible to
upload latest information faster, hence the higher currency.
Statistical analysis of fourth objective using One way ANOVA resulted in Welch‘s
F(4, 1321.644) = 170.436, p<.05 meaning thereby that there was a significant difference
between mean scores of engagement for different stages of Social Media Engagement for
consumer decision-making. Delhi social media users were found to be primarily engaged in
Watching (Mean 2.29, SD 1.05), followed by Sharing (Mean 2.86, SD 1.09), Commenting
(Mean 3.31, SD 1.10), Producing (Mean 3.65, SD 1.21) and Curating (Mean 3.86, SD 1.17)
and these engagements were significantly different as per Pairwise comparisons (refer Table
6.2).
Results of the fifth aim of the study indicated that out of five Social Media
Engagement Stages viz. Watching, Sharing, Commenting, Producing and Curating,
Watching is significantly most effective in generating influence on consumer decision-
making with value of Welch‘s F(4, 1322.362) = 70.315, p<.05. Watching stage (Mean 2.50,
SD 1.15) of social media engagement was having between ―Moderately high‖ and ―Neither
High Nor Low‖ influence on Delhi consumer decision-making. Sharing was next most
effective stage (Mean 3.07, SD 1.18) and generated ―Neither High Nor Low‖
influence. Influence of Commenting (Mean 3.27, SD 1.20), Curating (Mean 3.47, SD 1.23)
and Producing (Mean 3.27, SD 1.18) lied between ―Neither High Nor Low‖ and
―Moderately Low‖ in a decreasing order. Out of these, influence of commenting was
significantly better than influence of Producing stage. No significant difference was noticed
between influence of Commenting and Curating and between influence of Commenting and
Producing.
Results of fourth and fifth objectives when combined translated to that whenever
Delhi consumers are confronted with purchase related decisions, they ―most of the
times‖ like to visit home pages of various products/ services, read blogs or comments or
product ratings and watch videos. Consumers ―sometimes‖ update status on social sites
or twitter, upload/share photos, videos, articles etc. and write comment on blogs or news
stories, review or rate product. They were only ―occasionally‖ engaged in writing Blogs or
views to start new discussions and/or are involved in discussions on fan pages or boards/
forums, organization of such discussions and promoting brands and Watching stage

20
 Research Findings, Implications and

influences them most followed by Sharing, Commenting, Curating and Producing.

20
 Research Findings, Implications and

To achieve the last aim of the study i.e. to assess the Informational Influence,
Normative Influence and Value-Expressive Influence of social media platforms at various
stages of consumer decision-making process, nine sub-objectives were framed. First six
sub-objectives were to analyze any significant difference between Informational Influence,
Normative Influence and Value-Expressive Influence of social media platforms at various
stages of consumer decision-making process, where as last three sub-objectives were
framed to analyze any significant difference within Informational Influence, Normative
Influence and Value-Expressive Influence of social media platforms at various stages of
consumer decision-making process.
Analysis revealed that Social media platforms exert a considerable influence over
various purchase decision stages as all mean values were between 2.61 to 3.22 (meaning
thereby that values lie on the relevance side of the continuum), which translates to that
customers find these influences sufficiently relevant at various stages of purchase decision-
making. Statistical Analysis indicated that there were significant differences in
Informational Influence, Normative Influence and Value-Expressive Influence of social
media platforms at most of the stages of consumer decision-making process (Only H 05C
was supported and H05A, H05B, H05D, H05E, H05F were not supported). Informational
influence on various purchase decision stages is highest, followed by Normative influence
and Value-Expressive Influence. At almost each stage of consumer decision-making
process, Informational influence is significantly more than Normative Influence and Value-
Expressive Influence. However, consumers did not perceive any difference in Informational
influence, Normative influence and Value-Expressive influence while evaluating purchase
alternatives. In addition, at Purchase stage, Normative Influence was significantly (Mean
3.04, SD 1.11) better than Value-Expressive Influence (Mean 3.22, SD 1.20).
Further, last three sub-objectives of sixth objective framed to assess any significant
difference within Informational Influence, Normative Influence and Value-Expressive
Influence of social media platforms at various stages of consumer decision-making process.
Results indicated that there were significant differences in mean values of Informational
influence, mean values of Normative influence and mean values of Value-Expressive
Influence of social media platforms on various stages of consumer decision-making.
Informational influence is highest at Need Recognition stage (Mean 2.61, SD 1.02) and
lowest at Post-Purchase Dissatisfaction stage (Mean 2.85, SD 1.07). Normative influence is
highest at Alternative Evaluation stage (Mean 2.64, SD 1.07) and lowest at Purchase stage

20
 Research Findings, Implications and

7.3.1 Summary of Research Results at a Glance

Table 7.1: Summary of Research Results

Sub-Objective Corresponding Null Hypothesis Result

Objective

1. To identify the factors NONE No Hypothesis Five Factors Resulted:

of social media 1. Perceived Ease of
influencing the Information Exchange
consumer decision- 2. Perceived Purchase
making process Relevant Benefits
3. Perceived Risk
4. Perceived Hedonism
5. Perceived Fluency

2. To assess the NONE H01: There is no significant difference in Not Supported

credibility of mean values of credibility of information
information available at available at selected social media
selected social media platforms in influencing consumer
platforms in influencing decision-making process
consumer decision
making process

3. To assess credibility 3A. To assess Currency of H02A: There is no significant difference Not Supported
dimensions viz. information available at selected in mean values of Currency of

20
 Research Findings, Implications and

Currency, Coverage, social media platforms in information available at selected social

Accuracy, Authority influencing consumer decision- media platforms in influencing consumer
and Objectivity of making process decision-making process.
information available at
3B. To assess Coverage of H02B: There is no significant difference Not Supported
selected social media
information available at selected in mean values of Coverage of
platforms in influencing
social media platforms in information available at selected social
consumer decision-
influencing consumer decision- media platforms in influencing consumer
making process
making process decision-making process

3C. To assess Accuracy of H02C: There is no significant difference Not Supported

information available at selected in mean values of Accuracy of
social media platforms in information available at selected social
influencing consumer decision- media platforms in influencing consumer
making process decision- making process

3D. To assess Authority of H02D: There is no significant difference Not Supported

information available at selected in mean values of Authority of
social media platforms in information available at selected social
influencing consumer decision- media platforms in influencing consumer
making process decision-making process

3E. To assess Objectivity of H02E: There is no significant difference Not Supported

information available at selected in mean values of Objectivity of
social media platforms in information available at selected social

20
 Research Findings, Implications and

influencing consumer decision- media platforms in influencing consumer

making process decision-making process

4. To identify the most NONE H03: There is no significant difference Not Supported
engaging stage of between mean scores of engagement for
Social Media different stages of Social Media
Engagement Stages for Engagement for consumer decision-
consumer decision- making
making

5. To identify the most NONE H04: There is no significant difference Not Supported
effective stage out of between mean values of influence on
Social Media consumer decision-making generated by
Engagement Stages in different stages of Social Media
generating influence on Engagement
consumer decision-
making

6. To assess the 6A. To ascertain the Informational H05A: There is no significant difference Not Supported
Informational influence, Normative influence and in mean values of Informational
Influence, Normative Value-Expressive influence of influence, Normative influence and
Influence and Value- social media platforms at need Value-Expressive influence of social
Expressive Influence of recognition stage of consumer media platforms at need recognition
social media platforms decision-making process stage of consumer decision-making

20
 Research Findings, Implications and

at various stages of process

consumer decision-
6B. To measure the Informational H05B: There is no significant difference Not Supported
making process
Influence, Normative Influence and in mean values of Informational
Value-Expressive Influence of influence, Normative influence and
social media platforms at Value-Expressive influence of social
information search stage of media platforms at information search
consumer decision-making process stage of consumer decision-making
process

6C. To assess the Informational H05C: There is no significant difference Supported

influence, Normative influence and in mean values of Informational
Value-Expressive influence of influence, Normative influence and
social media platforms in Value-Expressive influence of Social
alternative evaluation media platforms in alternative evaluation
stage of consumer decision-making
process

6D. To assess the Informational H05D: There is no significant difference Not Supported
influence, Normative influence and in mean values of Informational
Value-Expressive influence of influence, Normative influence and
social media platforms at the stage Value-Expressive influence of Social
of purchase media platforms at the stage of purchase
in consumer decision-making process

20
 Research Findings, Implications and

6E. To analyze the Informational H05E: There is no significant difference Not Supported
influence, Normative influence and in mean values of Informational
Value-Expressive influence on influence, Normative influence and
sharing of post-purchase Value-Expressive influence on sharing of
satisfaction at social media post-purchase satisfaction at social media
platforms platforms

6F. To analyze the Informational H05F: There is no significant difference Not Supported
Influence, Normative Influence and in mean values of Informational
Value-Expressive Influence on Influence, Normative Influence and
sharing of post-purchase Value-Expressive Influence on sharing
dissatisfaction at social media of post-purchase dissatisfaction at social
platforms media platforms

6G. To analyze the Informational H05G: There is no significant difference Not Supported
influence of social media platforms in mean values of Informational
on various stages (viz. Need influence of social media platforms on
recognition, Information search, various stages (viz. Need recognition,
Alternative evaluation, Purchase, Information search, Alternative
Post-purchase satisfaction, Post- evaluation, Purchase, Post-purchase
purchase dissatisfaction) of satisfaction, Post-purchase
consumer decision-making dissatisfaction) of consumer decision-
making

20
 Research Findings, Implications and

6H. To analyze the Normative H05H: There is no significant difference Not Supported
influence of social media platforms in mean values of Normative influence
on various stages (viz. Need of social media platforms on various
recognition, Information search, stages (viz. Need recognition,
Alternative evaluation, Purchase, Information search, Alternative
Post-purchase satisfaction, Post- evaluation, Purchase, Post-purchase
purchase dissatisfaction) of satisfaction, Post-purchase
consumer decision- making dissatisfaction) of consumer decision-
making

6I. To analyze the Value- H05I: There is no significant difference Not Supported
Expressive influence of social in mean values of Value-Expressive
media platforms on various stages influence of social media platforms on
(viz. Need recognition, Information various stages (viz. Need recognition,
search, Alternative evaluation, Information search, Alternative
Purchase, Post-purchase evaluation, Purchase, Post-purchase
satisfaction, Post-purchase satisfaction, Post-purchase
dissatisfaction) of consumer dissatisfaction) of consumer decision-
decision-making making

Source: Based on researcher‘s calculations from collected data

20
 Research Findings, Implications and

(Mean 3.04, SD 1.11). Possible reason for this can be that as customers generally like to
seek suggestions of other significant people in order to maintain behavioral parity with them
and for their acceptance (Assael, 2005). Value-Expressive influence is highest Alternative
Evaluation stage (Mean 2.71, SD 1.00) and lowest at Purchase stage (Mean 3.22, SD 1.20).
Table 7.1 provides a summary of the research results at a glance by putting together
research objectives, their corresponding sub-objective(s) and null hypothesis and the
outcome of the hypothesis testing. Practical and theoretical implications of the research
have been discussed in section 7.4.
7.4 Implications of the Research
No research is complete without discussing its value addition to the academic
domain and to the field of action. Contributions of research justify the entire research
exercise by isolating distinct theoretical and practical inputs emanating from research.
Accordingly, theoretical and practical implications of the present research have been
discussed in the next sub-sections.

7.4.1 Theoretical Implications

As far as theoretical contributions of the research are concerned, study has

made significant contributions to the existing body of knowledge. Literature review
evidenced that there was a paucity of literature with respect to social media in India. First,
study explored and filled academic void by isolating five factors, viz. Perceived Ease of
Information Exchange, Perceived Purchase Relevant Benefits, Perceived Risk, Perceived
Hedonism and Perceived Fluency of social media, which influence Indian consumers‘, more
specifically Delhi consumers‘, purchase decision-making. Results corroborated earlier
researches that perceived ease of use, perceived usefulness (Davis 1989; Davis et al., 1989;
Gefen & Straub, 2000; Wu & Wang, 2005; Wang, Chung, Park, McLaughlin & Fulk,
2011; Lorenzo-Romero, Constantinides, Alarco´n-del-Amo, 2011; Cha, 2011),
perceived risk (Forsythe & Shi, 2003; Swinyard & Smith, 2003; Javadi, Dolatabadi,
Nourbakhsh, Poursaeedi & Asadollahi, 2012), enjoyment (Davis et al, 1992; Atkinson
& Kydd, 1997; Venkatesh, 2000; Bagozzi & Dholakia, 2002; Dholakia, Bagozzi &
Pearo, 2004; Venkatesh et al., 2008; Ha & Stoel, 2009) and flow experience play an
important role in influencing consumer behavior in online contexts (Hoffman & Novak,
1996; Mahnke, Benlian & Hess, 2015; Pelet, Ettis & Cowart, 2017). Researchers can
take a cue from here and further refine these factors and add more factors to the list.

21
 Research Findings, Implications and

Outcome of the second objective is another value addition to academic terrain as

there was hardly any empirical research available in Indian context that assesses and
compares information credibility of social media. This study, comparing over all credibility
as well as credibility of information along different dimensions of information credibility at
five social media typologies, is perhaps first of its kind in India and may serve as basis for
future researches. Research also contributed a social media credibility measurement scale
adapted from existing literature, which can be further tested and refined. Though the
research results exhibit that user find various social media to be only fairly credible (values
lying between 2.73 and 3.29), it must be kept in mind that user‘s credibility assessment
varies with user‘s ability, knowledge, experience, motivation and availability of time
(Metzger, 2007; Li et al., 2015). Past researches have also exhibited that netizens were ill-
equipped and unprepared for credibility evaluation in online environments (Amsbary et al.,
2003; Meola, 2004; Metzger et al., 2003; Scholz-Crane, 1998) and on many occasions,
users did not report actual evaluation behavior (Flanagin et al., 2000; Scholz-Crane,
1998). Users, especially experienced ones, exaggerated their evaluation behavior and
reported to be using five criterion (Flanagin et al., 2007), while they were not using all five
dimensions for credibility evaluation simultaneously (Scholz-Crane, 1998; Flanagin et al.,
2000; Metzger, 2007; Li et al., 2015). Hence, there is a possibility of such variations and
exaggeration in credibility assessment in current study as well. Though, in respect of above
observations, this should also be noticed that most of these past studies were based on small
samples and were not using actual users. Present study is based on actual users and has a
fairly big sample comparatively. In addition, in online environment, users were found to be
assessing both source and message together (Rieh, 2002; Eysenbach et al., 2002; Fogg et
al., 2003). Though, present study was specifically based on message credibility, such mix
up may have crept in. In addition, Li & Suh (2015) reported Facebook information to be
credible, where as in current study, it was found to be least credible. This contradiction
perhaps is due to the fact Li & Suh (2015) studied only one platform, where as the research
instance focused on five platforms of social media.
Findings of fourth and fifth objective related to social media engagement level as
well as its influence on purchase decisions of Delhi consumers corroborated earlier
researches, based on engagement pyramid (Rosenblatt, 2010; Owyang, 2010; Evans,
2010; Zailskaite-jakste et al., 2012; Starck, 2016; Percolator Consultancy, 2018) and
Participation Inequality principle or 90-9-1 principle (Jake McKee and 90-9-1.com;

21
 Research Findings, Implications and

Nielsen, 2006; Zailskaite-jakste et al., 2012), in Indian context as more respondents were
at the base of the pyramid engaged in Watching and lesser were engaged in Curating
(Rosenblatt, 2010; Owyang, 2010; Evans, 2010; Zailskaite-jakste et al., 2012; Starck,
2016; Percolator Consultancy, 2018).
Results of sixth objective and its sub-objectives (pertaining to assessment of
Informational Influence, Normative Influence and Value-Expressive Influence of social
media platforms at various stages of consumer decision-making process) exhibited
Informational influence to be significantly better than Normative influence and Value-
Expressive Influence on most of the purchase decision stages barring purchase alternative
evaluation stage, where consumer did not perceive any difference in three influences. This
result is aligned with earlier researches (Burnkrant & Cousineau, 1975; Park & Lessig,
1977), which exhibited more informational influence on the respondents compared to other
two influences. In addition, it was found that on most occasions, there was no significant
difference between Normative influence and Value-Expressive influence barring at the
stage of purchase in consumer decision-making process (where Normative Influence was
significantly better than Value-Expressive Influence) and Normative influence and Value-
expressive influence exhibited same pattern of influence at various stages of consumer
decision-making. These empirical results have departed from tri-dimensional concept of
reference group influence construct (composed of informational, normative and value-
expressive influence) as proposed by Park & Lessig (1977) and lends support to empirical
researches (Brinberg & Plimpton 1986; Bearden, Netemeyer & Teel, 1989; Tudor &
Carley, 1998) taking reference group influence as two dimensional construct. Though
social media has become a indispensable part of consumer purchase journey, social
commerce is yet to catch on (Chaffey, 2019; Mohsin, 2019; Kemp, 2020), which can be a
possible reason for a significantly better Normative influence at the stage of purchase in
consumer decision-making process as family and friends accompanying the customer might
influence his/her choices. In addition, as reported by Bearden & Etzel (1982) and Childers
& Rao (1992), it is not necessary for all three influences to be present all the time and
simultaneously, same is the case in present research. Yet another theoretical contribution is
a scale measuring Social Media‘s reference group influence on consumer decision-making,
though scale needs to be further tested and refined in future researches. In addition,
documentation of reference group influence of social media on consumer decision-making
is a value addition to existing literature due to paucity of research on reference group

21
 Research Findings, Implications and

influence on consumer decision-making stages and in specific reference to social media in

Indian context (Ilavarasan, 2018).

7.4.2 Practical Implications

The study is significant for the practitioners in the field of consumer behavior in
India as it provides them better insights to understand and explore new customer touch
points made available by these computer mediated platforms, which will help them to
propose more effective marketing strategies. Study is significant to platform and content
developers also, as they get to know what features are better suited to attract more users to
their platforms
Analysis of demographic profile and social media usage profile implies that while
framing their social media strategy, marketers must keep in mind that social media users
exhibit gender disparities, generational and demographic variations and also differ greatly in
their time, place and device preferences. For example, Facebook, and Twitter are more
preferred sites by females whereas YouTube is more preferred by males (financesonline,
2020; Gibson, (n. d.)). In addition, females use Facebook to keep in touch with their friends
and post more on their Facebook wall compared to men. Significantly higher number of
women access social media on mobile phone compared to men. Millennials are biggest
users of social media on mobile phones, where as Gen X use tablets to access social media
(Mohsin, 2019; Kemp, 2020; Gibson, (n. d.)). Hence, it is advisable to marketers that
instead of focusing on the entire gamut of social media platforms, they should build social
presence only on those platforms, which are most preferred by their customers. Thus, a
balanced device strategy is needed based on use cases and contexts.
Out of the five factors, resulted from EFA and CFA, most important factor of social
media influencing Delhi consumers‘ decision-making process related to ease of information
exchange. The fact that people use social media for obtaining purchase related information
has been corroborated by researches (Kemp, 2020). In addition, it is also well established
that people like to vent irk and anger on these platforms. The fact that people find it easy
and like to get information from social media presents an excellent opportunity to marketers
to use various platforms to market and advertise their products and brands. Marketers can
also nurture influencers and brand advocates for generating positive e-word of mouth for
their brands. However, it is worrisome that consumers use social media for disseminating
product related information via product reviews and ratings and by sharing their opinions

21
 Research Findings, Implications and

about products, brands and other aspect of product service. Because one, marketers have no
control over this user generated content and two, these reviews and ratings influence the
opinions of the consumers reading the content (Duhan & Singh, 2014b; O'Connor, 2015).
Thus, marketers need to carefully comb and analyze the information being exchanged about
their brands to understand the emotional valence and intensity towards their brands. Social
listening will help in understanding what resonates with the customers, thereby helping
marketers to build on customer goodwill and to mitigate the impacts of negative information
effectively (Duhan et al., 2014b; O'Connor, 2015).
Second and third factor of social media relates to perceived purchase relevant
benefits and perceived risks of social media for making purchase related decisions. As
social commerce shares various tenets with e-commerce and m-commerce, this appears a
natural outcome that social media is instrumental in providing purchase related useful
information and saving time and physical efforts of consumers being leveraged through
Information and communication technologies as well as it encompasses various risks
associated with online commerce. Marketers‘ role here is to provide relevant, detailed and
accurate information to the customers as well as to provide more and easier means to
disseminate information. Millennials, which constitute the most lucrative market right now,
have higher propensity to share on social media than other generations, so providing more
information sharing means to them have a positive impact on them as well as it may
enhance the reach and visibility of the brand without any apparent marketing efforts.
Millennials also get more influenced by user generated content (whether by friends or
strangers) in comparison to other generations (O'Connor, 2015, Duhan & Singh, 2016).
In addition, marketers should try to ward off perceived risks by adopting means such as
easy return policies, cash on delivery option, registering only authorized dealers for
supplying products, stringent data privacy and security policies, black listing the vendors
compromising customer information and/or supplying defective products and use of newer
technologies such as Virtual reality and augmented reality to reduce psychological risk etc.
Fourth factor, Perceived Hedonism, indicates the pleasure associated with purchase
activities at and through social media. Marketers can take a cue from here and should
provide content not only enlightening but also engaging, entertaining, exciting, aesthetically
appealing and out of box content at various social media touch points to grab and hold
attention of customers and turn brands in to experiences, something most sought after by
Millennials and most apt way to target them (Meadows-klue, 2008; Manning, 2014,

21
 Research Findings, Implications and

O'Connor, 2015, Duhan et al., 2016). Fifth factor, Perceived Fluency, indicates that
marketers should ensure that social media interactions should be smooth and uninterrupted.
Provision for more user-friendly interfaces, seamless navigation and assignment of
dedicated staff to assist may ensure this. Such measures may also lure older generations to
frequent social media more, in general, and for purchase related activities, in specific.
Results of second objective and sub-objectives of third objective (assessing over all
information credibility of selected five social media platforms as well as information
credibility along five dimensions) have two fold implications- one, for the social media
platform developers and content creators and two, for the marketers. First of all, social
media platforms developers and content creators should take these results with a pinch of
salt as most of the mean values (for over all credibility and for individual dimensions) are in
the range of 3.39 to 2.52, which means that customers only ―occasionally‖ find the
information to be credible. Hence, steps should be taken to enhance the credibility of the
platforms. Facebook should be most worried as it fails on all the counts. Though steps has
been taken by Facebook, YouTube and Twitter to increase transparency such as removing
fake accounts, screening of fake information, authentication of members, improved privacy
and data protection, monitoring communications and content, enhanced security features
etc. (Sysomos Inc., 2009; Twitter Inc., 2010; MLOT, 2013; Anderson, 2017; Dell'Oca,
2018; Sullivan, 2019), but still they need to go a long way for enhancing credibility. As
people are becoming increasingly concerned about the authenticity of digital information
(Kemp, 2020), enhancing credibility is important. Enhanced credibility will lead to hiked
―virtual footfalls‖ and in turn, to generation of more ―big data‖- organized and unorganized.
This big data can be dissected for behavioral targeting (Sarkar, 2014) and segmentation of
customers (Schiffman, Kanuk & Kumar, 2010) as well as for tracking customers‘
emotional valence (positive or negative) and emotional intensity for the brands (Bellini,
D‟Ascenzo, Ghi, Spagnoli & Traversi, 2012). This also means enhanced profits for the
platform developers as data can be sold to marketers. In addition, enhanced credibility and
more users will lure more marketers to a platform to promote their brands as it will give
more exposure and visibility to their brands and will enhance footfalls in virtual spaces as
well as brick and mortar spaces.
Marketers as an initial step should include those social media platforms in their
marketing strategies, which customers find more credible. Use of more credible platforms
will lend credibility to their products/ brands and services being promoted and marketed

21
 Research Findings, Implications and

through these platforms as per Congruity theory of attitude change, which talks about
associative assertion (Osgood et al., 1955) and asserts linking of ‗netizens‘ social media
credibility experience with their credibility experiences of brands promoted through these
platforms. Social Judgment theory also asserts that new information about attitude objects is
assimilated in the light of existing information, which acts as frame of reference for the new
information (Solomon, 2010, pg. 247). Thus, prior experiences with a social media platform
becomes a frame of reference for a brand‘s social media pages and hence for the brand and
may lead to higher evaluations of the brand (Tugrul, 2014). Hence, making favorable
impressions by enhancing social media credibility is imperative due to increasing impact of
these on virtual interactions and online as well as offline consumer decision-making
(Constantinides, 2004; Tugrul, 2014).
Results of fourth and fifth objective when combined translated to that majority of
Delhi consumers were at the base of the engagement pyramid engaged in Watching
activities and vertical movement exhibited that number reduced at the higher levels and
lesser were engaged in Curating. Marketers may use this knowledge of the level of customer
engagement for framing appropriate marketing strategies catering to the needs of the
customers spread along the vertical dimension of the engagement pyramid by segmenting
customers as per their level of engagement. There is lesser number of people at the top of
the engagement pyramid, called brand advocates, yet their contributions are much higher
than that of large number at the bottom of the pyramid. But, marketers should focus on both
highly engaged as well as lesser engaged because few highly engaged are not true
representatives of large number of lesser engaged, though different strategies should be
framed for lesser-engaged vis-à-vis more engaged as also advocated by practitioners (Jake
McKee and 90-9-1.com; Nielsen, 2006). In addition, as a customer moves from lower to
higher levels of engagement, his/her brand associations, brand preference, brand
confidence, satisfaction, trust, brand loyalty, attachment, support and emotional bonding
increases so much so that ultimately customer turns in brand advocate (Zailskaite-jakste et
al., 2012). Thus, marketers must look for ways to take more customers from lower levels of
engagement to higher levels of engagement. Personalized conversations through dedicated
social media pages are an excellent way to engage customers (Pansari et al., 2017).
Organizational sponsorship of customer-initiated activities such as customer blogs and chat
forums, creating possibility of sharing stories as on Instagram, holding contests and polls
also lead to better customer engagement (Vivek et al., 2012; Ashley & Tuten, 2015;

21
 Research Findings, Implications and

Mohsin, 2019) of current as well as potential customers. Marketers can go beyond

sponsorship of customer-initiated events and can actively participate in creating brand
communities, organizing trainings, announcing participation rewards and incentives,
facilitating easy interactions, providing instantaneous responses and possibilities for co-
creation for better engagement (Van Doorn et al., 2010; Vivek et al., 2012; Barger et al.,
2016). Various other actions to be undertaken by organizations are attractive display,
providing more sharing options (such as video uploading, RSS etc.), more commenting
opportunities (such as discussion forums, blogs), identification and recognition of e-opinion
leaders. Organizations can also resort to following User-generated content and sharing best
User-generated content on brand pages (Owyang and Li, 2010; Zailskaite-jakste &
Kuvykaite, 2012; Ashley & Tuten, 2015; Barger et al., 2016).
As Watching is the most engaging as well as the most influence producing stage for
Delhi social media users, it is important for marketers to offer relevant content to customer
as it is a fundamental to engagement. Content is the king and marketers must strive to have
the content which may elicit intense emotional responses from the viewer for heightened
engagement. One size does not fit all, so while designing campaigns cultural and linguistic
specifics must be taken care of to avoid marketing faux pas. Hence, message is clear that
instead of translating a copy, social nuances need to be taken in to account. Use of
appropriate channels and modes along with culturally relevant imagery and media, use of
slangs and jargons and appropriate inter-personal communication is needed (Sedley, 2010).
Content should be current, exclusive, functional, informative, branded, entertaining,
remunerative, using experiential, interactive and vivid images (Cvijikj & Michahelles,
2013; Ashley & Tuten, 2015; Duhan and Singh, 2014b; Barger, Peltier & Schultz,
2016; Chaffey, 2019) posted at multiple platforms (Ashley & Tuten, 2015). Companies
like Dove and Coca-cola have been able to successfully engage customers majorly due to
the video content as per experts (Pansari et al., 2017). However, marketers must keep the
content confined to selected few platforms only as multiplicity of social media platforms
coupled with content abundance may put pressure on consumers‘ cognitive faculties.
Consumers, thus, consume only selective content ignoring the rest (Barger, Peltier &
Schultz, 2016). In addition, youth dominance means posted content should be thought
provoking (Meadows-klue, 2008) and suitable to taste and preferences of target customers
to ensure better engagement (Chaffey, 2019; Kemp, 2020).

21
 Research Findings, Implications and

Analysis of sixth objective and its sub-objectives revealed that social media
platforms exert a considerable influence over various stages of consumer decision-making,
though, marketers should use these results with caution. Obtained mean values indicated
sufficiently relevant influences at various stages of purchase decision-making, yet these
values could have been different as the studies have indicated that reference group
influences vary according to the members‘ attitudes towards the group (Assael, 2005). If the
beliefs, opinions and suggestions of group members are considered important (Calder &
Burnkrant, 1977), if the group is considered as a credible source of information (Miniard
& Cohen, 1983) and the rewards and punishments accorded (for accepting and not
accepting the norms of the group) are accepted by the customer (Allen, 1965) then the
propensity of groups to influence consumer choices is higher. Reference group influences
also vary according to the nature of the group (Bearden & Etzel, 1982; Childers & Rao,
1992) (Cohesive groups, Primary groups and Distinctive and exclusive groups have been
found to be exerting more influence) (Leigh & Gabel, 1992) and the type of product
(Bourne, 1957; Bearden & Etzel, 1982; Childers & Rao, 1992; Assael, 2005).
Conspicuous and exclusive, status exhibiting product purchases are more influenced by
reference groups (Bourne, 1957; Bearden & Etzel, 1982; Childers & Rao, 1992). As the
present study was based on social media as a single variable without make any
demarcations in various social media platforms with respect to importance of a platform to a
member, cohesiveness, membership, credibility, distinctiveness etc., these mean value of
influences should be used with caution by the marketers. Social media platforms differ from
each other in terms of characteristics (e.g. present research exhibited social media platform
vary in credibility) as enumerated above, accordingly the influences exerted by the platform
may also vary as explained above. Also, as the study has not been conducted with specific
reference to a particular product, marketers should make concessions for same while
utilizing the results.
Informational influence on most of the purchase decision stages was significantly
better than Normative influence and Value-Expressive Influence implying that customers
are influenced more by the informational aspect of social media and like to go to social
media to fulfill their informational needs. That people like to use social media for
satisfaction of their informational needs has been corroborated by other practitioners and
researchers (Chaffey, 2019; Mohsin, 2019; Kemp, 2020) albeit not in reference group
influence context. In addition, as indicated by the results of first objective on the research

21
 Research Findings, Implications and

instance too, customers find the information at social media useful and easy to exchange for
purchase related decisions. Hence, marketers must include social media in marketing
strategy contingent to fulfillment of customers‘ information need fulfillment expectations
(Duhan & Singh, 2019). As enumerated above in practical implications, social media
content should be current and relevant for effective decision-making. Easily available and
accessible information about products, brands and services, using virtual reality and
augmented reality tools, provided at various stages of decision-making can help in assessing
the quality of the products better and reduces search costs and customers‘ expectations and
perceptions gap (Luo, Ba & Zhang, 2012). Lesser mean values of Normative influence and
Value-expressive influence of social media may be resulting from absence of face-to-face,
frequent interactions and less stringent norms of groups on social media. Marketers should
look for opportunities to enhance the reference group influences by facilitating the
formation of strong brand communities, discussion forums, cultivating brand influencers,
use of bloggers and brand advocates strategically.

7.4.3 Social Implications

Study has social significance too as results can help marketers to find out the most
effective ways to incorporate social media into their marketing strategy leading to reduced
marketing costs, reduced prices, better products and enhanced customer experiences. In
addition, platform developers can use the results to improve the features; this once again
enhances customer experiences. This will benefit consumers as well as leads to optimum
utilization of resources. Further, consumers will be getting information in the desired format
and at the desired platform.

7.5 Limitations of the Study

Limitations of a study are the factors that may influence the findings of the research.
These constraints may arise either due to some characteristics of design or research
methodology or due to researcher. These limitations may put constraints on the
generalizations of the study, practical applications of the research or utility of the findings
(Malhotra et al., 2007; "Organizing…", n.d.; Wordvice, 2018). As no research is
without limitations; accordingly, this research is no exception. Even with the best efforts on
the part of the researcher, certain limitations do exist in research either due to researcher or
design or research methodology. Limitations of the study are as given below:-

21
 Research Findings, Implications and

First, study uses non-random sampling technique for data collection. It is a known
and established fact that random sampling techniques are more rigorous compared to non-
probability sampling techniques. However, as it was difficult and costly to get a sampling
frame of users of social media, it was decided to use non-random technique for data
collection. However, choice of random sampling technique may have resulted in better
results.
Second, data were collected using self-administered questionnaire, the link of which
was shared through e-mail and various social media platforms. Hence, there was no way to
motivate and/or to assist the respondent to give desired information. Though utmost care
was taken while designing the questionnaire to remove such obstacles, however possibility
of withholding of the desired information on the part of few respondents cannot be ruled
out. In addition, as in schedules, there were no means to crosscheck the authenticity of
responses through gestural language.
Third, as the respondents have self-selected themselves to fill out the questionnaire,
there is a possibility that only most enthusiastic or highly motivated users of social media
may have filled the questionnaire. In addition, the possibility of high users of social media
filing the form cannot be ruled out.
Fourth, there is also a possibility of self-reported bias in the responses. Researchers
have to accept whatever respondents say or fill at face value even if the data may suffer
from various potential biases such as selective memory, selective recall, attribution, socially
desirable answers and exaggeration. Current research also has this limitation, as there are
no means to independently verify the self-reported data.
Yet another limitation that may have influenced the results of the study can be
language. Users of social media are net savvy and are considered to be comfortable with
English language. In addition, utmost care was taken to keep the language of the
questionnaire simple and free of technical jargon taking help of experts as well as pilot
survey. However, as English is not the native language of Delhiites, some may have
encountered problem due to language too.
Further, any limitations inherent in any social science research such as subjectivity
of the respondents may also present in this research. In addition, the limitations inherent in
analytical techniques also add to the limitations of the study such as the generalizations of
the results of exploratory factor analysis beyond the sample should be done only when EFA
using different samples results in same factor structure. In addition, while testing

22
 Research Findings, Implications and

convergent validity, AVE for the factor Perceived Purchase Relevant Benefits was less
than 0.5 (though as Composite reliability was higher than 0.6, it was accepted with proper
citations (Fornell et al., 1981; Lam, 2012), future research may look into this issue and fix
it.
Another limitation associated with the generalization of the result relates to
geographical scope of the study. Present study is limited only to National Capital region of
Delhi. Hence, the generalization of the results may be done with caution. Further, as there
are not many research studies available directly dealing with the research problem and
objectives especially in Indian context, validity of the results cannot be established.

7.6 Suggestions for Further Research

Mayfield (2008), likened social media to a genie, which has come out of the bottle
and is not going to go back into it ever, though, it might be referred to by different names in
future. Ever present dynamism, influx of new platforms, added features to existing ones and
enhanced membership on existing as well as newer social media platforms turn Mayfield‘s
observations into a prophecy. Ever evolving and ever-changing social media scenario and
its attendant influences on masses make it an exciting field for researches. Some of the
possible research avenues for future research endeavors have been detailed below:
First, due to advent of new technologies, social media are ever evolving and ever
changing. On one hand, new and better features are being added to the existing platforms to
make them more competitive and on the other hand, older platforms are making way for
younger ones. In such a scenario, to understand and capture the changing social media
consumption habits as well as the changing uses and usage of social media, longitudinal
study may be conducted.
Second, EFA and CFA resulted in five factors of social media that influence
consumer decision-making. Total variance explained by these factors was 63.276 %. Result
leaves almost 37 % of unexplained variance and this creates a scope for researchers to
explore additional factors of social media, which may influence consumer decision-making.
Third, future research may be conducted to find out the impact of extracted factors
on purchase intention of consumers. Also, it may be researched whether the extracted
factors differ significantly with reference to demographic variables and social media usage
variables. Further, future research may also throw some light on if the informative
influence, normative influence and value-expressive influence differ significantly with

22
 Research Findings, Implications and

respect to demographic variables such as gender, age, income etc. and social media usage
variables like heavy usage and light usage.
Fourth, lot of qualitative, structured, unstructured and unadulterated data in the form
of comments, reviews, conversations, uploads is available at social media. Future research
endeavors may be centered on these big data to gather deeper consumer insights.
Fifth, future research can also be attempted using focused group interviews to gather
qualitative data and better consumer insights. Also, experimental research design may be
used to validate the current results and to throw more light on why and how consumers use
social media for decision-making.
Sixth, as user‘s credibility assessment varies with user‘s ability, knowledge,
experience, motivation and availability of time, future researchers may attempt to study
impact of these variables on information credibility and its dimensions at social media in
Indian context.
Future research can focus on other influences exerted by social media. Studies may
also focus on social media‘s reference group influence on consumer decision-making
process with reference to the purchase of a high involvement versus low involvement
product or service as well as with respect to varying characteristics of the social media
platforms to get further insights.

7.7 Chapter Summary

In the concluding chapter of the thesis, first an overview of the research approach
was explained in which summary of the research efforts put in so far have been given. Next,
summary of the main findings of the research was given. After this, theoretical and practical
implications of the research have been discussed. Further, potential limitations of the study
have been given. Chapter concludes with elaboration on future research possibilities in the
researched domain.

22