The Influence of Social Media Influencer Marketing on Consumer Purchase Intention Among Young Adults

Sample Master’s (MSc) Dissertation — Marketing

Written by an AHC subject expert in Marketing. Distinction / First-class standard.

> Important note on this sample. This dissertation is an original teaching example produced by Assignment Help Center to show how a Master’s-level empirical marketing dissertation is structured, argued and referenced. The theoretical discussion and literature are grounded in real, published scholarship. All primary data reported in Chapter 4 — the sample of 312 respondents, the descriptive statistics, reliability coefficients, correlations and regression results — are ILLUSTRATIVE. They were constructed for demonstration purposes and were not collected from real participants. They are clearly labelled as illustrative throughout. Students should use this document to understand form, flow and academic convention, not as a source of empirical evidence to cite.

Abstract

Social media influencer marketing has become a central pillar of contemporary digital marketing strategy, with brands reallocating substantial portions of their communication budgets away from traditional celebrity endorsement and towards content creators who command engaged followings on platforms such as Instagram, TikTok and YouTube. Young adults, who have grown up with these platforms and who spend a disproportionate share of their media time within them, are the demographic most exposed to, and most courted by, influencer marketing. Yet the mechanisms through which influencer marketing translates into a genuine intention to purchase remain incompletely understood, particularly the relative contribution of the perceived credibility of the influencer and the affective, quasi-social bond — the parasocial relationship — that followers form with them.

This dissertation investigates the influence of social media influencer marketing on consumer purchase intention among young adults, and examines the roles of source credibility and parasocial relationship as explanatory constructs. Adopting a positivist philosophy and a deductive, quantitative approach, the study proposes a conceptual framework linking three dimensions of source credibility (trustworthiness, expertise and attractiveness), parasocial relationship strength and perceived advertising value to purchase intention. An online cross-sectional survey design was operationalised using established, validated measurement scales adapted from Ohanian (1990), Lou and Yuan (2019) and related work.

Using an illustrative dataset of 312 young-adult respondents constructed to demonstrate the analytical procedure, the study reports descriptive statistics, internal-consistency reliability (Cronbach’s alpha), bivariate correlations and multiple linear regression. In this illustrative analysis, trustworthiness, parasocial relationship and expertise emerge as the strongest positive predictors of purchase intention, while attractiveness exerts a weaker, though still positive, effect. The findings — again, illustrative — are interpreted through the lens of source-credibility theory, parasocial interaction theory and the Theory of Planned Behaviour, and are used to derive practical recommendations for marketers concerning influencer selection, authenticity, disclosure and relationship cultivation. The dissertation closes by acknowledging the limitations inherent in a single cross-sectional design and by proposing avenues for future longitudinal and experimental research.

Keywords: influencer marketing; source credibility; parasocial relationship; purchase intention; young adults; social media.

Chapter 1: Introduction

1.1 Background to the study

Over the last decade the architecture of persuasion in consumer markets has shifted decisively. Where brands once relied on mass-media advertising and the borrowed lustre of celebrity endorsement, they increasingly place their trust — and their budgets — in social media influencers: individuals who have accumulated sizeable and engaged audiences by producing content on platforms such as Instagram, TikTok, YouTube and, more recently, a range of short-form video services. Influencers are often described as ordinary people who have become “micro-celebrities” through the consistent cultivation of a personal brand, and their commercial appeal rests on a perception, whether accurate or not, that they are more relatable, more authentic and more trustworthy than either traditional celebrities or corporate advertising (Freberg et al., 2011; Abidin, 2016).

The scale of the phenomenon is considerable. Influencer marketing has grown from a peripheral experiment into a mainstream channel that absorbs a meaningful share of global digital advertising expenditure, and industry commentary consistently reports year-on-year growth in the value of the sector. For marketers, the attraction is straightforward: influencers offer targeted access to specific communities, higher engagement rates than many paid formats, and a form of endorsement that is woven into content the audience has chosen to consume rather than interrupted by. For consumers — and especially for young adults — influencers have become a routine source of product discovery, evaluation and social proof.

Young adults occupy a special position in this landscape. As so-called digital natives, they have grown up with social media as an ambient part of everyday life, they spend a disproportionate share of their leisure and information-seeking time on the very platforms where influencers operate, and their consumption identities are formed substantially through peer and quasi-peer reference groups online (Djafarova and Rushworth, 2017). They are, in short, both the most exposed to influencer marketing and the demographic whose purchase behaviour brands are most eager to shape. Understanding how influencer marketing moves this group from passive exposure to active purchase intention is therefore of both academic and practical significance.

1.2 Problem statement

Despite the ubiquity and commercial importance of influencer marketing, the mechanisms by which it shapes consumer purchase intention are not yet fully understood. Practitioners often treat reach — the raw number of followers — as the primary criterion for selecting influencers, on the intuitive assumption that a larger audience yields greater persuasive effect. Academic evidence, however, complicates this assumption: a very large follower count can, under some conditions, reduce perceived likeability and authenticity (De Veirman, Cauberghe and Hudders, 2017), and the persuasive power of an influencer appears to depend more on qualitative perceptions — how credible the influencer seems, and how connected the follower feels to them — than on audience size alone.

Two constructs recur in the literature as candidate explanations for influencer effectiveness, yet they are seldom examined together in a single, integrated model applied specifically to young adults. The first is source credibility, the classical communication-theory notion that a message’s persuasiveness depends on the audience’s perception of the communicator’s trustworthiness, expertise and attractiveness (Hovland, Janis and Kelley, 1953; Ohanian, 1990). The second is the parasocial relationship — the one-sided but emotionally real bond that audience members form with media personae they follow over time (Horton and Wohl, 1956). Both are theoretically plausible drivers of purchase intention, but their relative contribution, and whether they operate independently or in concert, remains an open empirical question. This dissertation addresses that gap.

1.3 Research aim

The aim of this dissertation is to investigate the influence of social media influencer marketing on consumer purchase intention among young adults, with particular attention to the roles played by source credibility and parasocial relationship.

1.4 Research objectives

To achieve this aim, the study pursues four objectives:

1. To critically review the theoretical and empirical literature on source credibility, parasocial relationships, influencer marketing and consumer purchase intention. 2. To develop a conceptual framework and testable hypotheses linking the dimensions of source credibility, parasocial relationship and perceived advertising value to purchase intention. 3. To operationalise and test that framework using a quantitative survey design (reported here with an illustrative dataset). 4. To derive evidence-informed recommendations for marketers and to identify limitations and directions for future research.

1.5 Research questions

• RQ1: To what extent do the three dimensions of source credibility — trustworthiness, expertise and attractiveness — influence young adults’ purchase intention?
• RQ2: To what extent does the strength of the parasocial relationship between a young adult and an influencer influence purchase intention?
• RQ3: How do source credibility and parasocial relationship compare in their relative contribution to explaining purchase intention?

1.6 Significance of the study

Theoretically, the study contributes to the marketing communications literature by integrating source-credibility theory and parasocial interaction theory within a single model of purchase intention, and by testing that model on the demographic most relevant to influencer marketing. Practically, the findings — illustrative though the present dataset is — are intended to help marketers move beyond crude follower-count heuristics towards a more discriminating basis for influencer selection and relationship management. Methodologically, the dissertation demonstrates a transparent, replicable procedure for measuring these constructs using validated scales, which practitioners and future researchers can adapt.

1.7 Scope and delimitations

The study is delimited to young adults, operationally defined as individuals aged approximately 18 to 30, who follow at least one social media influencer. It concentrates on image- and video-led platforms where influencer marketing is most concentrated. It examines purchase intention rather than actual purchase behaviour; while intention is a well-established antecedent of behaviour (Ajzen, 1991), the two are not identical, and this distinction is revisited in the limitations. The study is cross-sectional and correlational, and does not claim to establish causation.

1.8 Structure of the dissertation

Chapter 2 reviews the literature and develops the conceptual framework and hypotheses. Chapter 3 sets out the methodology. Chapter 4 reports the (illustrative) findings. Chapter 5 discusses those findings in relation to theory and prior evidence. Chapter 6 concludes, offers recommendations, acknowledges limitations and proposes future research.

Chapter 2: Literature Review

2.1 Introduction

This chapter reviews the scholarship relevant to the study’s aim. It begins by establishing the conceptual terrain of influencer marketing and situating it in relation to traditional celebrity endorsement. It then examines the two theoretical pillars of the study — source credibility and parasocial relationships — before turning to the dependent construct, purchase intention, and its grounding in the Theory of Reasoned Action and the Theory of Planned Behaviour. The chapter synthesises these strands into a conceptual framework and a set of testable hypotheses.

2.2 Influencer marketing: definition and distinctiveness

Social media influencers have been defined as a new type of independent third-party endorser who shapes audience attitudes through blogs, tweets and the use of other social media (Freberg et al., 2011). Unlike traditional celebrities, whose fame typically originates outside social media — in film, sport or music — influencers usually build their prominence natively on the platforms themselves, cultivating an audience through the sustained production of content within a particular niche such as beauty, fitness, gaming, travel or personal finance (Abidin, 2016). This origin story matters: because influencers appear to have risen through relatability and authenticity rather than through the machinery of celebrity, their audiences frequently perceive them as more accessible and more trustworthy than conventional stars (Schouten, Janssen and Verspaget, 2020).

A useful conceptual distinction has emerged between macro-influencers (large followings, broad reach, often approaching traditional-celebrity status) and micro-influencers (smaller but highly engaged niche audiences). Empirical work suggests the relationship between follower count and persuasive effectiveness is not linear: De Veirman, Cauberghe and Hudders (2017) found that a very high number of followers can enhance perceptions of an influencer’s popularity but may simultaneously undermine perceived likeability, particularly when the influencer follows few accounts in return, signalling a lack of relatability. This finding is central to the present study’s motivation, because it implies that the drivers of influencer effectiveness are qualitative and psychological rather than simply a function of audience scale.

Influencer marketing also differs from traditional advertising in its integration into organic content. Sponsored posts are frequently embedded within the ordinary stream of an influencer’s output, which can blur the line between editorial and commercial content. This blurring raises questions of disclosure and authenticity that recur throughout the literature and to which the discussion returns (Evans et al., 2017; Boerman, 2020).

A further distinguishing feature is the mechanism of opinion leadership. Casaló, Flavián and Ibáñez-Sánchez (2020) argue that influencers function as contemporary opinion leaders, occupying a position within their followers’ reference groups that lends their recommendations social weight. Ki and Kim (2019) develop a complementary account in which influence operates partly through consumers’ desire to mimic an admired other, so that following a recommendation becomes a means of approximating an aspirational identity. Jin, Muqaddam and Ryu (2019) add a comparative dimension, finding that Instagram influencers can generate stronger source trust and social presence than traditional celebrities, precisely because their self-presentation reads as more ordinary and attainable. Together these strands reinforce the study’s premise that influence is a psychological and social phenomenon rooted in perceived character and connection, not a mechanical function of audience size.

2.2.1 Young adults as the focal audience

The choice of young adults as the study’s population is not incidental but theoretically motivated. Young adults are frequently characterised as digital natives whose consumer socialisation has taken place substantially within social media environments, and for whom influencers occupy a role once reserved for peers, older siblings and, in an earlier media era, television personalities (Djafarova and Rushworth, 2017). Two features make this group especially responsive to influencer marketing. First, their identities and consumption preferences remain comparatively fluid and are actively negotiated through reference groups, of which influencer communities are now a prominent example. Second, their media diet is concentrated on precisely the visual and video platforms where influencer marketing is densest, maximising both exposure and the accumulation of the repeated contact from which parasocial bonds are built. Djafarova and Rushworth (2017), studying young female Instagram users, found that the perceived credibility of online personalities materially shaped purchase decisions, and that non-traditional, relatable figures were often more persuasive than conventional celebrities. This makes young adults not merely a convenient sample but the most theoretically appropriate population in which to observe the phenomena of interest.

2.3 Source credibility theory

The theoretical roots of source credibility lie in the persuasion research of the mid-twentieth century. Hovland, Janis and Kelley (1953) demonstrated that the same message produces different levels of attitude change depending on the perceived credibility of its source, and identified expertise and trustworthiness as the twin components of credibility. A source perceived as both knowledgeable and honest is more persuasive than one perceived as lacking in either quality.

Ohanian (1990) extended and operationalised this tradition for the endorsement context, developing and validating a tri-dimensional source-credibility scale comprising trustworthiness, expertise and attractiveness. Trustworthiness refers to the audience’s confidence in the source’s integrity and in the honesty of the claims made; expertise refers to the perceived knowledge, skill or qualification of the source relative to the subject; and attractiveness refers not merely to physical appeal but to a broader sense of likeability, familiarity and similarity. Ohanian’s scale has become one of the most widely used instruments in endorsement research and provides the measurement foundation for the credibility constructs in this dissertation.

The applicability of source-credibility theory to the influencer context has been established by a growing body of work. Lou and Yuan (2019) developed and tested a model of influencer marketing in which the trustworthiness, expertise and attractiveness of the influencer, alongside the informational and entertainment value of their content, shaped followers’ trust in branded posts and, in turn, their brand awareness and purchase intention. Their study found that trustworthiness and the informative value of content were especially important. Similarly, Sokolova and Kefi (2020) examined YouTube and Instagram influencers and found that source credibility contributed to purchase intention, though they argued that parasocial interaction played a role at least as important — a finding of direct relevance here.

A recurring theme is that the three dimensions of credibility do not contribute equally. Trustworthiness and expertise are frequently reported as stronger predictors of persuasive outcomes than attractiveness, particularly for products where functional performance or informational accuracy matters (Erdogan, 1999; Lou and Yuan, 2019). Attractiveness, by contrast, may be more influential for hedonic or image-related products, and its effect may operate partly through the affective route rather than the cognitive one. This nuance informs the hypothesis development below.

The endorsement literature also offers a complementary theoretical lens in McCracken’s (1989) meaning-transfer model, which holds that endorsers are effective not simply because they are credible but because they carry culturally constituted meanings — lifestyle, status, values — which are transferred first to the product and then, through purchase and use, to the consumer. Viewed through this lens, source credibility explains why an endorsement is believed, while meaning transfer explains why it is desired. Influencers, who construct elaborate and legible personal brands, are unusually rich carriers of such meaning, which helps to reconcile the co-existence of a cognitive credibility route and a more aspirational, identity-based route to influence (see also Ki and Kim, 2019). While the present study operationalises the credibility route directly, it recognises meaning transfer as part of the theoretical backdrop against which the results are interpreted.

2.4 Parasocial relationships

The concept of the parasocial relationship originates with Horton and Wohl (1956), who observed that audiences of mass media develop an “intimacy at a distance” with media personae — a sense of friendship, knowledge and connection that is genuinely felt by the audience member even though it is entirely one-sided and unreciprocated. Originally applied to television presenters and radio hosts, the concept has proven remarkably durable and has been revitalised by social media, whose interactive affordances — comments, direct messages, live streams, behind-the-scenes stories — create an unprecedented illusion of reciprocity and closeness (Chung and Cho, 2017).

Parasocial relationships are theoretically distinct from source credibility. Credibility is fundamentally a cognitive judgement about the reliability of a message source; the parasocial relationship is an affective bond, an emotional attachment that accrues over repeated exposure and that carries feelings of loyalty, identification and trust that are relational rather than purely evaluative. Social media influencers are especially well positioned to foster parasocial bonds because they routinely share personal, apparently unguarded glimpses of their lives, address their audiences directly and in the second person, and respond, however selectively, to audience communication. This cultivates a perception of friendship and authenticity that can be powerfully persuasive (Sokolova and Kefi, 2020).

Empirically, parasocial relationship strength has been linked to a range of consumer outcomes. Chung and Cho (2017) found that parasocial interaction with celebrities on social media enhanced trust in the celebrity and, through it, brand credibility and purchase intention. Sokolova and Kefi (2020) reported that, for certain audiences, parasocial interaction was a stronger driver of purchase intention than source credibility itself, because followers were inclined to buy what a “friend” recommended somewhat independently of a cold assessment of that friend’s expertise. Hwang and Zhang (2018) similarly found that parasocial relationships influenced followers’ intention to follow influencer recommendations and to spread positive word of mouth. These findings motivate the inclusion of parasocial relationship as a distinct predictor alongside the credibility dimensions.

2.5 Consumer purchase intention

Purchase intention is generally defined as the subjective probability or willingness of a consumer to buy a particular product or brand, and it occupies a well-established place in consumer-behaviour theory as a proximal antecedent of actual purchase behaviour. Its theoretical grounding lies principally in the Theory of Reasoned Action (Fishbein and Ajzen, 1975) and its successor, the Theory of Planned Behaviour (Ajzen, 1991). In these frameworks, behaviour is predicted by behavioural intention, which is itself shaped by the individual’s attitude towards the behaviour, subjective norms (perceived social pressure) and, in the planned-behaviour extension, perceived behavioural control. Intention is thus the pivotal mediating construct between attitudes and action.

Influencer marketing can be understood as operating on several of these antecedents simultaneously. A credible, well-liked influencer may improve the consumer’s attitude towards the endorsed product by supplying favourable information and positive affect; the influencer, and the community around them, may function as a reference group that shapes subjective norms; and the sense of connection generated by a parasocial bond may reinforce both. Because purchase intention integrates these influences into a single, measurable disposition to act, and because it is a strong (if imperfect) predictor of behaviour, it is adopted here as the dependent variable, consistent with the majority of the influencer-marketing literature (Lou and Yuan, 2019; Sokolova and Kefi, 2020).

It is important to acknowledge the intention–behaviour gap: measured intention does not translate perfectly into purchase, and the size of the gap varies with context, product type and the interval between measurement and opportunity to act (Sheeran and Webb, 2016). This study measures intention, not behaviour, and this delimitation is treated explicitly in the limitations.

2.6 Perceived advertising value and content characteristics

Beyond the person of the influencer, the characteristics of the content itself contribute to persuasive outcomes. Ducoffe’s (1996) model of advertising value, developed for web advertising and subsequently applied to social media, holds that consumers evaluate advertising according to its informativeness, entertainment and (negatively) its intrusiveness or irritation, and that this overall perceived value shapes attitudes towards the advertising and the brand. Lou and Yuan (2019) incorporated the informative and entertaining value of influencer content into their model and found both to matter. Perceived advertising value is therefore included in the present framework as a supplementary predictor, capturing the content-level, as opposed to person-level, drivers of intention.

2.7 Authenticity, disclosure and the limits of persuasion

A tension runs through the literature between the commercial logic of influencer marketing and the authenticity on which its effectiveness depends. Followers value influencers precisely because they seem authentic and independent; overt commercialisation, or the perception that an influencer will endorse anything for payment, can erode trust and, with it, persuasive power (Audrezet, de Kerviler and Guidry Moulard, 2020). Sponsorship disclosure — increasingly required by advertising regulators — introduces a further complication: disclosure activates persuasion knowledge, potentially prompting more sceptical processing of the message (Evans et al., 2017; Boerman, 2020). The evidence is mixed, with some studies finding that disclosure reduces persuasive effect while others find that transparent disclosure can, by signalling honesty, actually reinforce trustworthiness. This ambivalence underscores why trustworthiness, rather than mere exposure, is theorised here as a central driver.

2.8 Conceptual framework and hypotheses

Synthesising the foregoing, the study proposes a conceptual framework in which purchase intention among young adults is predicted by five constructs: the three dimensions of source credibility (trustworthiness, expertise and attractiveness), the strength of the parasocial relationship, and perceived advertising value. The framework treats these as parallel predictors of purchase intention, consistent with the multi-factor models of Lou and Yuan (2019) and Sokolova and Kefi (2020).

The framework can be represented conceptually as follows:

> Trustworthiness → Purchase Intention > Expertise → Purchase Intention > Attractiveness → Purchase Intention > Parasocial Relationship → Purchase Intention > Perceived Advertising Value → Purchase Intention

From this framework, the following hypotheses are derived:

• H1: Perceived trustworthiness of the influencer is positively associated with young adults’ purchase intention.
• H2: Perceived expertise of the influencer is positively associated with young adults’ purchase intention.
• H3: Perceived attractiveness of the influencer is positively associated with young adults’ purchase intention.
• H4: The strength of the parasocial relationship with the influencer is positively associated with young adults’ purchase intention.
• H5: Perceived advertising value of the influencer’s content is positively associated with young adults’ purchase intention.

Consistent with prior evidence that trustworthiness and relational closeness tend to dominate attractiveness in driving intention (Lou and Yuan, 2019; Sokolova and Kefi, 2020), it is further anticipated that H1 and H4 will exhibit the strongest effects and H3 the weakest, though these expectations are treated as directional rather than as formal hypotheses.

2.9 Chapter summary

This chapter has established that influencer marketing is theoretically distinct from traditional endorsement, that its effectiveness is better explained by qualitative perceptions than by audience scale, and that source credibility and parasocial relationships are the two most compelling explanatory constructs, complemented by perceived advertising value at the content level. Purchase intention, grounded in the Theory of Planned Behaviour, serves as the integrating dependent variable. The next chapter sets out how the resulting framework is operationalised and tested.

Chapter 3: Methodology

3.1 Introduction

This chapter explains and justifies the methodological choices made to address the research aim and objectives. It is organised using the “research onion” heuristic (Saunders, Lewis and Thornhill, 2019), moving from research philosophy through approach, strategy, choices and time horizon to the concrete procedures of sampling, measurement, data collection and analysis, and concluding with a full treatment of ethics, reliability and validity.

3.2 Research philosophy

The study adopts a positivist philosophy. Positivism holds that social phenomena can be studied through observable, measurable regularities in a manner analogous to the natural sciences, that the researcher is independent of the object of study, and that knowledge is advanced by testing theory-derived hypotheses against empirical data (Saunders, Lewis and Thornhill, 2019; Bryman and Bell, 2015). This philosophy is appropriate here because the study seeks to test a pre-specified theoretical model, to measure clearly defined constructs using validated scales, and to establish statistical relationships that can, in principle, be generalised. An interpretivist stance, by contrast, would be better suited to exploring the subjective meanings young adults attach to influencer relationships, but would not permit the hypothesis testing that the research objectives require.

3.3 Research approach

Consistent with positivism, the study takes a deductive approach: theory and hypotheses are established first (Chapter 2), and data are then collected to test them, rather than theory being built inductively from data. Deduction is the natural counterpart to a study that begins with an established theoretical framework and seeks confirmation or disconfirmation of specific predicted relationships (Bryman and Bell, 2015).

3.4 Methodological choice and strategy

A quantitative, mono-method design is used, employing a self-administered survey as the research strategy. Surveys are efficient for collecting standardised data from large samples, are well suited to measuring attitudes and intentions, and permit the statistical analysis of relationships between variables (Saunders, Lewis and Thornhill, 2019). A survey is the dominant strategy in the influencer-marketing literature the study builds upon (Lou and Yuan, 2019; Sokolova and Kefi, 2020), which also facilitates comparison of results.

3.5 Time horizon

The study is cross-sectional: data are collected at a single point in time, providing a snapshot of relationships among the variables. A cross-sectional design is appropriate given the practical constraints of a Master’s dissertation and the correlational nature of the research questions, though its limitations for causal inference are acknowledged in Chapter 6.

3.6 Population and sampling

The target population comprises young adults, defined as individuals aged approximately 18 to 30 who use social media and follow at least one influencer. Because a complete sampling frame of such individuals does not exist, a non-probability approach was adopted, combining purposive and convenience sampling, supplemented by snowball referral. A screening question at the start of the questionnaire ensured that only respondents who both fell within the age band and reported following at least one influencer proceeded, thereby enforcing the purposive criteria.

For a multiple regression with five predictors, methodological guidance recommends a minimum sample well in excess of 100; a common rule of thumb suggests at least 50 + 8k cases per predictor, giving a minimum of roughly 90, while more conservative guidance recommends 10–15 cases per predictor (Field, 2018; Hair et al., 2019). A target of approximately 300 responses was therefore set to provide ample statistical power and to accommodate incomplete responses. In this illustrative demonstration, a dataset of 312 valid responses is used (see Chapter 4, and the note on illustrative data below).

The use of non-probability sampling limits the statistical generalisability of the findings to the wider population of young adults, a limitation discussed in Chapter 6; it is nonetheless standard and defensible in this research context given the absence of a sampling frame and the exploratory-confirmatory purpose of the study.

3.7 Measurement instrument

Data were to be collected through a structured, self-completion online questionnaire comprising three sections: (i) screening and demographic questions; (ii) a battery of attitudinal items measuring the study constructs; and (iii) a short debrief. All construct items used established, previously validated scales, adapted only in wording to fit the influencer context, in order to maximise content and construct validity. Each attitudinal item was measured on a five-point Likert scale (1 = strongly disagree to 5 = strongly agree).

The constructs and their scale sources are summarised below:

Respondents were asked to answer the construct items with reference to a specific influencer of their own choosing whom they currently follow, a technique that grounds the abstract items in a concrete referent and is common in the literature.

3.8 Pilot study and instrument refinement

Prior to full deployment, the questionnaire was to be pilot-tested with a small convenience sample (n ≈ 20) drawn from the target population, to check item clarity, timing and the absence of ambiguity, and to obtain a preliminary reading of scale reliability. Minor wording adjustments would then be made before the instrument was finalised. Pilot data are excluded from the main analysis.

3.9 Data collection procedure

The finalised questionnaire would be hosted on an online survey platform and distributed via social media channels and university networks, with a participant information statement and consent presented on the landing page before any question was shown. Data collection would run until the target sample was reached. Responses failing the screening criteria, or showing patterns indicative of inattentive responding (for example, straight-lining or implausibly short completion times), would be removed during data cleaning.

3.10 Data analysis strategy

Analysis would be conducted using standard statistical software (for example, IBM SPSS). The analytical sequence comprises: (i) data screening and cleaning; (ii) descriptive statistics to characterise the sample and the distributions of the constructs; (iii) internal-consistency reliability analysis using Cronbach’s alpha, with the conventional threshold of 0.70 for acceptable reliability (Nunnally, 1978); (iv) bivariate correlation (Pearson’s r) to examine the strength and direction of associations among the constructs and to provide a preliminary check on the hypotheses; and (v) multiple linear regression, with the five predictors entered simultaneously to predict purchase intention, thereby testing H1–H5 and establishing the relative contribution of each predictor. Regression assumptions — linearity, normality of residuals, homoscedasticity and the absence of problematic multicollinearity (assessed via the variance inflation factor) — would be checked before interpretation.

3.11 Reliability and validity

Reliability — the consistency of measurement — is addressed through the use of multi-item, previously validated scales and confirmed empirically through Cronbach’s alpha. Construct validity is supported by the adoption of established scales with demonstrated validity in prior research. Content validity is strengthened by grounding every item in the theoretical constructs defined in Chapter 2 and by the pilot review. Common method bias, a risk in single-source, self-report survey designs, would be mitigated procedurally by assuring anonymity, mixing item order and using clear, concise wording to reduce evaluation apprehension and demand characteristics (Podsakoff et al., 2003). External validity is constrained by the non-probability sample, as noted.

3.12 Ethical considerations

The study was designed to conform to the ethical principles governing research with human participants and would proceed only following institutional ethics approval. The key safeguards are: informed consent, obtained via a clear information statement and an explicit opt-in before any data were collected; voluntary participation and the right to withdraw at any point without penalty; anonymity and confidentiality, with no personally identifying information collected and data stored securely in compliance with data-protection law (including the UK GDPR); no harm, the subject matter being low-risk and non-sensitive; and transparency, with a debrief explaining the study’s purpose. Because participants would be adults aged 18 and over, no additional safeguards for minors were required.

3.13 Note on the illustrative dataset

For the purposes of this teaching sample, the analysis reported in Chapter 4 is performed on an illustrative dataset constructed to demonstrate the described analytical procedure. No human participants were recruited, and the numbers presented — sample composition, means, standard deviations, alpha coefficients, correlations and regression coefficients — do not represent findings from real respondents. They are internally consistent and plausible, and they are used solely to show what a completed analysis and its interpretation look like. This is stated again at the head of Chapter 4 and reiterated wherever results are reported.

3.14 Chapter summary

This chapter has justified a positivist, deductive, quantitative, cross-sectional survey design; specified a purposive and convenience sample of young adults with a target of around 300 responses; detailed a measurement instrument built from validated scales; and set out a reliability, correlation and regression analysis strategy, together with a full account of ethics, reliability and validity. The following chapter reports the illustrative results.

Chapter 4: Findings

4.1 Introduction and reiterated note on illustrative data

This chapter presents the results of the analysis. The reader is reminded, before any figure is reported, that all data in this chapter are ILLUSTRATIVE. The 312 respondents, and every descriptive statistic, reliability coefficient, correlation and regression result below, were constructed for the purpose of demonstrating the analytical procedure described in Chapter 3. They are not empirical findings from real participants and must not be cited as such. The values were designed to be internally coherent and broadly consistent with patterns reported in the published literature, so that the interpretation in Chapter 5 can proceed realistically.

4.2 Sample profile (illustrative)

After screening and data cleaning, the illustrative analysis proceeds with 312 valid responses. The composition of this illustrative sample is summarised below.

Table 4.1 (Illustrative): Sample profile, n = 312.

The illustrative sample skews modestly female and is concentrated in the 22–25 age band, with Instagram and TikTok the dominant platforms — a distribution that is plausible for the young-adult influencer-following population. Notably, in this illustrative sample almost two-thirds report having previously purchased at least one product on the basis of an influencer recommendation, underscoring the practical relevance of the research question.

4.3 Descriptive statistics (illustrative)

Table 4.2 reports the illustrative means and standard deviations for each construct, computed as the mean of its constituent items on the five-point scale.

Table 4.2 (Illustrative): Construct means and standard deviations, n = 312.

All illustrative construct means sit above the scale midpoint of 3.0, indicating generally favourable perceptions across the board. Attractiveness records the highest mean (3.91) and perceived advertising value the lowest (3.49). Purchase intention has a mean of 3.61, suggesting a moderately positive disposition to purchase on influencer recommendation among this illustrative sample.

4.4 Reliability analysis (illustrative)

Internal-consistency reliability was assessed using Cronbach’s alpha; the conventional acceptability threshold is 0.70 (Nunnally, 1978). Table 4.3 reports the illustrative coefficients.

Table 4.3 (Illustrative): Reliability coefficients, n = 312.

In this illustrative analysis, every scale exceeds the 0.70 threshold comfortably, with all alphas at or above 0.80, indicating good to excellent internal consistency and supporting the aggregation of items into composite construct scores for the subsequent analysis.

4.5 Correlation analysis (illustrative)

Table 4.4 presents the illustrative Pearson correlation matrix among the six constructs. All correlations are statistically significant at p < 0.01 in this illustrative dataset.

Table 4.4 (Illustrative): Pearson correlation matrix, n = 312. All correlations significant at p < 0.01.

In this illustrative matrix, purchase intention correlates most strongly with trustworthiness (r = 0.64) and parasocial relationship (r = 0.62), followed by expertise (r = 0.56), perceived advertising value (r = 0.53) and, most weakly, attractiveness (r = 0.41). The inter-predictor correlations are moderate (the highest being 0.61 between trustworthiness and expertise), which is consistent with distinct but related constructs and does not immediately suggest problematic multicollinearity — a point confirmed by the VIF values reported below. The pattern provides preliminary, illustrative support for H1–H5, with the relative ordering already hinting that trustworthiness and parasocial relationship are the dominant correlates.

4.6 Regression analysis (illustrative)

To test the hypotheses jointly and to establish the relative contribution of each predictor, a multiple linear regression was conducted with the five constructs entered simultaneously and purchase intention as the dependent variable. Assumption checks (illustrative) indicated approximately normally distributed residuals, homoscedasticity, and variance inflation factors between 1.4 and 2.1 — well below the common threshold of concern (VIF > 5), indicating that multicollinearity is not a material problem.

Table 4.5 (Illustrative): Multiple regression predicting purchase intention. Model R² = 0.54, adjusted R² = 0.53, F(5, 306) = 71.9, p < 0.001. n = 312.

In this illustrative model, the five predictors together explain 54% of the variance in purchase intention (R² = 0.54), and the model is statistically significant overall. Examining the individual coefficients:

• Trustworthiness is the strongest predictor (β = 0.31, p < 0.001), supporting H1.
• Parasocial relationship is the second strongest (β = 0.27, p < 0.001), supporting H4.
• Expertise is a significant, moderate predictor (β = 0.18, p = 0.002), supporting H2.
• Perceived advertising value is a significant, smaller predictor (β = 0.14, p = 0.014), supporting H5.
• Attractiveness does not reach conventional significance in the multivariate model (β = 0.09, p = 0.076); its bivariate association with purchase intention (r = 0.41) appears to be substantially accounted for by its overlap with the other predictors. H3 is therefore only weakly and non-significantly supported in this illustrative analysis.

4.7 Summary of hypothesis tests (illustrative)

Table 4.6 (Illustrative): Summary of hypothesis testing.

4.8 Chapter summary

The illustrative analysis indicates that four of the five hypothesised predictors — trustworthiness, parasocial relationship, expertise and perceived advertising value — are positively and significantly associated with young adults’ purchase intention, with trustworthiness and parasocial relationship dominant, while attractiveness does not contribute significantly once the other constructs are controlled. These illustrative patterns are interpreted against theory and prior evidence in the next chapter. The reader is reminded a final time that all figures above are illustrative and not real empirical findings.

Chapter 5: Discussion

5.1 Introduction

This chapter interprets the illustrative findings of Chapter 4 in relation to the research questions, the conceptual framework and the wider literature. Throughout, it should be understood that the interpretations rest on illustrative data and are offered to demonstrate how findings of this shape would be discussed; they are not empirical claims about real consumers.

5.2 The primacy of trustworthiness (RQ1)

The illustrative finding that trustworthiness is the single strongest predictor of purchase intention (β = 0.31) aligns closely with the theoretical and empirical literature. Source-credibility theory, from Hovland, Janis and Kelley (1953) onwards, positions trustworthiness as a foundational component of persuasion, and Lou and Yuan (2019) reported trustworthiness among the most important influencer attributes for downstream marketing outcomes. The interpretation is intuitive in the influencer context specifically: the entire value proposition of the influencer relative to a traditional advertisement rests on a perception of honesty and independence. A follower who believes an influencer would not recommend a product they did not genuinely rate is far more likely to convert that recommendation into intention.

This result also speaks to the authenticity-and-disclosure tension identified in Section 2.7. If trustworthiness is the primary engine of purchase intention, then anything that erodes it — over-commercialisation, undisclosed sponsorship later exposed, or a mismatch between an endorsement and the influencer’s established persona — directly attacks the mechanism on which influencer marketing depends. The illustrative dominance of trustworthiness thus reframes disclosure not merely as a regulatory burden but as a potential trust-building asset, consistent with the argument that transparent disclosure can, by signalling honesty, reinforce rather than undermine credibility (Boerman, 2020).

5.3 Expertise and the informational route

The illustrative finding that expertise is a significant but more moderate predictor (β = 0.18) is likewise consistent with the literature. Expertise supplies the cognitive, informational basis for persuasion — the sense that the influencer knows what they are talking about — and matters most for products where functional performance and accurate information are salient. That it ranks below trustworthiness in the illustrative model is congruent with the recurring finding that audiences weigh honesty above competence when the two are separable, perhaps because expertise without trustworthiness (a knowledgeable but self-interested endorser) is discounted, whereas trustworthiness lends weight even to modest expertise. The moderate role of expertise also converges with the perceived-advertising-value result: informative content (Ducoffe, 1996; Lou and Yuan, 2019) and perceived expertise both operate through the same broadly cognitive, evaluative route, and each contributed significantly, if modestly, in the illustrative model.

5.4 The weak role of attractiveness

Perhaps the most theoretically interesting illustrative result is that attractiveness, although positively correlated with purchase intention at the bivariate level (r = 0.41), fails to retain significance in the multivariate model (β = 0.09, p = 0.076). This pattern is consistent with a body of endorsement research suggesting that attractiveness is the weakest of Ohanian’s (1990) three dimensions for driving purchase-related outcomes, particularly outside of image- and appearance-led product categories (Erdogan, 1999). The most plausible interpretation of the illustrative result is that attractiveness’s apparent bivariate influence is largely a proxy for, and is absorbed by, the trustworthiness, parasocial and expertise constructs with which it correlates: attractive influencers may also tend to be liked, felt close to and trusted, and once those overlapping perceptions are statistically controlled, attractiveness adds little independent explanatory power. This does not render attractiveness irrelevant — it may still function as an initial attention-getting and audience-building attribute — but it cautions marketers against treating surface appeal as a sufficient basis for endorsement effectiveness.

5.5 The affective power of parasocial relationships (RQ2)

The illustrative finding that parasocial relationship is the second strongest predictor of purchase intention (β = 0.27) — nearly matching trustworthiness — is central to the study’s contribution and echoes Sokolova and Kefi (2020), who found parasocial interaction to rival or exceed source credibility as a driver of purchase intention. The result supports the theoretical proposition that influence operates through an affective, relational route as well as a cognitive, evaluative one. Followers who feel a genuine, if one-sided, bond with an influencer — who feel they “know” them, look forward to their content and regard them as a kind of friend — appear disposed to act on their recommendations in a way that is partly independent of a cold assessment of expertise. This is the modern, social-media-amplified realisation of Horton and Wohl’s (1956) “intimacy at a distance”, supercharged by the interactive affordances of contemporary platforms.

The near-parity of trustworthiness and parasocial relationship in the illustrative model is theoretically significant because the two constructs, while correlated (r = 0.58 in the illustrative matrix), capture different things: one a cognitive judgement of reliability, the other an affective attachment. Their joint prominence suggests that effective influencer marketing engages consumers on both registers simultaneously — persuading the head that the source is honest and knowledgeable, and the heart that the source is a trusted companion.

5.6 Relative contribution and the integrated model (RQ3)

Addressing RQ3, the illustrative results indicate a clear ordering of predictors: trustworthiness > parasocial relationship > expertise > perceived advertising value > attractiveness. The model’s illustrative explanatory power (R² = 0.54) is substantial for a cross-sectional attitudinal study and comparable to that reported in analogous published models, lending face validity to the demonstration. The substantive implication is that the person of the influencer — specifically their perceived honesty and the relational bond they cultivate — matters more than either their demonstrated expertise, the qualities of any individual piece of content, or their physical appeal. This finding, if it held in real data, would carry a pointed message for an industry that still frequently selects influencers on the basis of follower count and aesthetic fit: reach and good looks are poor proxies for the trust and connection that actually move purchase intention.

5.6.1 Situating the results within the intention–behaviour framework

It is worth restating that the dependent variable is intention, not observed purchase, and interpreting the illustrative results accordingly. Within the Theory of Planned Behaviour (Ajzen, 1991), the constructs examined here can be read as operating primarily on the attitudinal and normative antecedents of intention. Trustworthiness and expertise plausibly shape the consumer’s attitude towards purchasing the endorsed product by supplying credible, favourable evaluative information; the parasocial relationship, and the wider community of co-followers, plausibly shape subjective norms by making purchase feel socially endorsed within a valued reference group; and perceived advertising value contributes to attitude through the affective and informational quality of the content itself. That the illustrative model accounts for a little over half the variance in intention is consistent with the expectation that these attitudinal and normative routes are important but not exhaustive determinants — perceived behavioural control, for instance, captured here only implicitly, and situational factors such as price and availability will also bear on whether intention is realised. This framing tempers any temptation to read the illustrative effect sizes as direct predictors of sales; they are, more precisely, predictors of a disposition that is itself a strong but imperfect antecedent of behaviour (Sheeran and Webb, 2016).

5.7 Theoretical implications

The study’s integration of source-credibility theory and parasocial interaction theory within a single model of purchase intention, tested (illustratively) on young adults, contributes to marketing communications scholarship in three ways. First, it demonstrates that the two theoretical traditions are complementary rather than competing: both contribute significant, independent variance. Second, it reinforces the disaggregation of source credibility into its component dimensions, showing (illustratively) that they behave very differently — trustworthiness and expertise persist as predictors while attractiveness does not. Third, by situating purchase intention within the Theory of Planned Behaviour (Ajzen, 1991), it connects the specific machinery of influencer persuasion to a general theory of intention formation, suggesting that influencers act on attitudes and subjective norms simultaneously.

5.8 Practical implications

For practitioners, the illustrative findings suggest a reorientation of influencer strategy around trust and relationship rather than reach and appearance. Marketers should prioritise influencers with a demonstrated record of authentic, honest engagement over those with merely large or attractive followings; should protect that trust by insisting on genuine product fit and transparent disclosure; and should value the depth and longevity of an influencer’s relationship with their audience — the raw material of parasocial bonds — as a strategic asset. These implications are developed into concrete recommendations in Chapter 6.

The illustrative pattern also has a bearing on the perennial macro-versus-micro-influencer debate. If trustworthiness and parasocial connection are the operative drivers, and if very large followings can dilute perceived relatability (De Veirman, Cauberghe and Hudders, 2017), then micro-influencers — whose smaller, niche audiences often sustain denser and more reciprocal relationships — may deliver disproportionate persuasive value per follower, notwithstanding their limited reach. This does not imply that macro-influencers lack a role; their reach remains useful for awareness objectives higher in the funnel. Rather, it suggests that objectives should be matched to influencer type: macro-influencers for breadth of exposure, micro-influencers for the depth of trust and connection that the illustrative results identify as the true engine of purchase intention. This nuance is consistent with a growing practitioner consensus and provides a theoretically grounded rationale for the tiered-portfolio approaches many brands now adopt.

5.9 Chapter summary

Interpreted against theory and prior evidence, the illustrative findings tell a coherent story: young adults’ purchase intention is driven principally by how much they trust an influencer and how connected they feel to them, secondarily by the influencer’s expertise and the value of their content, and scarcely at all, once these are accounted for, by attractiveness. The next chapter draws conclusions, offers recommendations and sets out the study’s limitations.

Chapter 6: Conclusion and Recommendations

6.1 Conclusion

This dissertation set out to investigate the influence of social media influencer marketing on consumer purchase intention among young adults, with particular attention to source credibility and parasocial relationships. Through a critical review of the literature it established that influencer effectiveness is better explained by qualitative, psychological perceptions than by raw audience scale, and it developed an integrated conceptual framework in which three dimensions of source credibility, parasocial relationship strength and perceived advertising value predict purchase intention. Adopting a positivist, deductive, quantitative survey design, and using an explicitly illustrative dataset to demonstrate the analytical procedure, the study found — illustratively — that trustworthiness and parasocial relationship are the dominant drivers of purchase intention, that expertise and perceived advertising value play significant supporting roles, and that attractiveness contributes little once the other constructs are controlled.

Returning to the research questions: RQ1 is answered by the finding that trustworthiness and expertise, but not attractiveness, significantly drive purchase intention; RQ2 by the finding that parasocial relationship strength is a powerful, near-primary driver; and RQ3 by the clear predictor ordering that places trustworthiness and parasocial relationship at the top. The overarching conclusion — offered as a demonstration of what such a study would conclude — is that influencer marketing works on young adults principally by being trusted and by feeling personal, and that industry heuristics centred on follower count and physical appeal are poorly aligned with the mechanisms that actually generate purchase intention.

6.2 Recommendations for marketers

1. Select for trust, not just reach. Because trustworthiness is the strongest driver of purchase intention, influencer selection should weight demonstrated authenticity and audience trust above follower count. Auditing an influencer’s track record of honest, consistent endorsement is more predictive than headline reach. 2. Protect authenticity and disclose transparently. Since trust is the central mechanism, brands should insist on genuine product fit and clear sponsorship disclosure. Handled well, disclosure signals honesty and can reinforce, rather than diminish, credibility. 3. Cultivate and value parasocial relationships. The strength of the follower–influencer bond is nearly as important as trust. Brands should favour longer-term partnerships with influencers who have deep, engaged relationships with their audiences over one-off placements, allowing the relational bond to work in the brand’s favour. 4. Match expertise to product category. Where products are functional or information-intensive, the influencer’s demonstrable expertise matters; brands should ensure a credible knowledge fit between influencer and product. 5. Invest in content value. Because perceived advertising value contributes independently, sponsored content should be informative and entertaining in its own right, not merely a placement, so that it earns rather than interrupts attention. 6. Do not over-rely on attractiveness. Surface appeal may build an audience but does not, by itself, convert intention. It should be treated as a hygiene factor rather than a primary selection criterion.

6.3 Limitations

Several limitations must be acknowledged, over and above the foundational point that the reported data are illustrative rather than empirical. First, the design is cross-sectional and correlational; it cannot establish causation, and reverse or reciprocal relationships (for example, that intention shapes perceptions rather than only the reverse) cannot be ruled out. Second, the study measures purchase intention, not behaviour, and the intention–behaviour gap means the findings may overstate real purchasing (Sheeran and Webb, 2016). Third, the non-probability sample of young adults limits statistical generalisability. Fourth, reliance on single-source self-report exposes the design to common-method and social-desirability biases, mitigated procedurally but not eliminated. Fifth, the study treats “influencer” and “purchase intention” generically, without disaggregating by product category (hedonic versus utilitarian) or influencer tier (micro versus macro), each of which the literature suggests may moderate the relationships. Finally, the model, while integrative, is not exhaustive: constructs such as product involvement, brand familiarity, persuasion knowledge and subjective norms were not modelled.

6.4 Recommendations for future research

Future work should, first, adopt longitudinal or experimental designs to move beyond correlation towards causal inference, for example by manipulating disclosure or influencer type. Second, researchers should link intention to actual behaviour, ideally using behavioural or purchase-record data, to quantify the intention–behaviour gap in this context. Third, studies should test the model across product categories and influencer tiers to identify moderators, and across cultures, since the salience of trust and relationship may vary. Fourth, the model could be extended to incorporate persuasion knowledge, product involvement and subjective norms, and analysed with structural equation modelling to test mediation — for instance whether parasocial relationship mediates the effect of content characteristics on intention. Finally, given the rapid evolution of platforms, comparative work across emerging formats (short-form video, live commerce) would keep the evidence base current.

6.5 Concluding remark

Influencer marketing has reshaped how young adults discover and evaluate what they buy. This dissertation has argued, and illustratively demonstrated, that its power rests not on the size of an audience or the polish of an image but on something older and more human: being trusted, and feeling like a friend. For marketers, the lesson is that the most valuable currency in the influencer economy is credibility and connection — and both are far easier to lose than to buy.

References

Abidin, C. (2016) ‘Visibility labour: engaging with influencers’ fashion brands and #OOTD advertorial campaigns on Instagram’, Media International Australia, 161(1), pp. 86–100.

Ajzen, I. (1991) ‘The theory of planned behavior’, Organizational Behavior and Human Decision Processes, 50(2), pp. 179–211.

Audrezet, A., de Kerviler, G. and Guidry Moulard, J. (2020) ‘Authenticity under threat: how social media influencers maintain authentic self-presentation’, Journal of Business Research, 117, pp. 557–569.

Boerman, S.C. (2020) ‘The effects of the standardized Instagram disclosure for micro- and meso-influencers’, Computers in Human Behavior, 103, pp. 199–207.

Bryman, A. and Bell, E. (2015) Business Research Methods. 4th edn. Oxford: Oxford University Press.

Casaló, L.V., Flavián, C. and Ibáñez-Sánchez, S. (2020) ‘Influencers on Instagram: antecedents and consequences of opinion leadership’, Journal of Business Research, 117, pp. 510–519.

Chung, S. and Cho, H. (2017) ‘Fostering parasocial relationships with celebrities on social media: implications for celebrity endorsement’, Psychology & Marketing, 34(4), pp. 481–495.

De Veirman, M., Cauberghe, V. and Hudders, L. (2017) ‘Marketing through Instagram influencers: the impact of number of followers and product divergence on brand attitude’, International Journal of Advertising, 36(5), pp. 798–828.

Djafarova, E. and Rushworth, C. (2017) ‘Exploring the credibility of online celebrities’ Instagram profiles in influencing the purchase decisions of young female users’, Computers in Human Behavior, 68, pp. 1–7.

Ducoffe, R.H. (1996) ‘Advertising value and advertising on the web’, Journal of Advertising Research, 36(5), pp. 21–35.

Erdogan, B.Z. (1999) ‘Celebrity endorsement: a literature review’, Journal of Marketing Management, 15(4), pp. 291–314.

Evans, N.J., Phua, J., Lim, J. and Jun, H. (2017) ‘Disclosing Instagram influencer advertising: the effects of disclosure language on advertising recognition, attitudes, and behavioral intent’, Journal of Interactive Advertising, 17(2), pp. 138–149.

Field, A. (2018) Discovering Statistics Using IBM SPSS Statistics. 5th edn. London: Sage.

Fishbein, M. and Ajzen, I. (1975) Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley.

Freberg, K., Graham, K., McGaughey, K. and Freberg, L.A. (2011) ‘Who are the social media influencers? A study of public perceptions of personality’, Public Relations Review, 37(1), pp. 90–92.

Hair, J.F., Black, W.C., Babin, B.J. and Anderson, R.E. (2019) Multivariate Data Analysis. 8th edn. Andover: Cengage.

Horton, D. and Wohl, R.R. (1956) ‘Mass communication and para-social interaction: observations on intimacy at a distance’, Psychiatry, 19(3), pp. 215–229.

Hovland, C.I., Janis, I.L. and Kelley, H.H. (1953) Communication and Persuasion: Psychological Studies of Opinion Change. New Haven, CT: Yale University Press.

Hwang, K. and Zhang, Q. (2018) ‘Influence of parasocial relationship between digital celebrities and their followers on followers’ purchase and electronic word-of-mouth intentions, and persuasion knowledge’, Computers in Human Behavior, 87, pp. 155–173.

Jin, S.V., Muqaddam, A. and Ryu, E. (2019) ‘Instafamous and social media influencer marketing’, Marketing Intelligence & Planning, 37(5), pp. 567–579.

Ki, C. and Kim, Y. (2019) ‘The mechanism by which social media influencers persuade consumers: the role of consumers’ desire to mimic’, Psychology & Marketing, 36(10), pp. 905–922.

Lou, C. and Yuan, S. (2019) ‘Influencer marketing: how message value and credibility affect consumer trust of branded content on social media’, Journal of Interactive Advertising, 19(1), pp. 58–73.

McCracken, G. (1989) ‘Who is the celebrity endorser? Cultural foundations of the endorsement process’, Journal of Consumer Research, 16(3), pp. 310–321.

Nunnally, J.C. (1978) Psychometric Theory. 2nd edn. New York: McGraw-Hill.

Ohanian, R. (1990) ‘Construction and validation of a scale to measure celebrity endorsers’ perceived expertise, trustworthiness, and attractiveness’, Journal of Advertising, 19(3), pp. 39–52.

Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y. and 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), pp. 879–903.

Rubin, A.M., Perse, E.M. and Powell, R.A. (1985) ‘Loneliness, parasocial interaction, and local television news viewing’, Human Communication Research, 12(2), pp. 155–180.

Saunders, M., Lewis, P. and Thornhill, A. (2019) Research Methods for Business Students. 8th edn. Harlow: Pearson.

Schouten, A.P., Janssen, L. and Verspaget, M. (2020) ‘Celebrity vs. influencer endorsements in advertising: the role of identification, credibility, and product-endorser fit’, International Journal of Advertising, 39(2), pp. 258–281.

Sheeran, P. and Webb, T.L. (2016) ‘The intention–behavior gap’, Social and Personality Psychology Compass, 10(9), pp. 503–518.

Sokolova, K. and Kefi, H. (2020) ‘Instagram and YouTube bloggers promote it, why should I buy? How credibility and parasocial interaction influence purchase intentions’, Journal of Retailing and Consumer Services, 53, 101742.