Graduate school of Communication
Will you eat #healthy too?
A study on the effects of social media influencers (micro vs. macro) on the attitude towards eating healthy and the intention to eat healthy among women (18-25 years old)
Name Jeanne Winkelman Student number 10821597
Supervisor Prof. Dr. S.J.H.M. van den Putte Master Persuasive Communication
Master’s Thesis
Date 27th of June, 2019 Word count 7.467
Abstract
The majority of the Dutch population eats too little fruit and vegetables in their daily lives, which could have severe consequences for their health. Nowadays, however, there is also a healthy food trend going around on social media platforms such as Instagram. Yet, little research has been done on the effects of influencer marketing regarding this popular trend. As young adulthood is proposed to be a time to develop unhealthy eating habits, it is important to investigate whether social media and influencers could induce healthy eating habits. The current study focuses on the possible positive consequences of the healthy food trend on social media using influencer marketing for young women (18-25 years old). It examines whether a post by a micro or a macro influencer leads to a more positive attitude towards eating healthy and a higher intention to eat healthy. Additionally, the study investigates whether the credibility and likeability of the source explain these relationships and whether the relationships are stronger when involvement with eating healthy is low. The results of the online experiment (N = 146) show that source credibility and source likeability do not explain these relationships, though involvement with eating healthy was found to strengthen the relationship between source likeability and intention, but only when involvement with eating healthy is high. Moreover, the results indicate that micro influencers generate more
credibility, a more positive attitude towards eating healthy and that macro influencers have higher likeability. Finally, implications of the findings are discussed and suggestions for future research are presented.
Introduction
“I prep meals for the entire week, using vegetables, chicken and fish. In this way, I can eat healthy all week.” These are the words of Dutch model and influencer Sanne Vloet. Every week she posts pictures on Instagram to tell her followers how she manages to eat healthy and maintain a tight body.
Young adulthood (18-25 years old) has been proposed to be a risky time period for the development of poor diet (Nelson, Story, Larson, Neumark-Sztainer, & Lytle, 2008). In the Netherlands, eight out of ten individuals eat too little vegetables and six out of ten eat too little fruit. Incorrect nutrition has the risk that blood vessels clog up due to a high cholesterol level, which can lead to heart- and brain-attacks in the long term (University of Twente, n.d.). A varied and healthy diet reduces the risk of these disorders. Moreover, the World Cancer Research Fund (2007) found that consumption of fruits and vegetables decreases the risk of developing head and neck cancer. Therefore, it is imperative to stimulate the intake of at least 250 grams of vegetables and 200 grams (two pieces) of fruit a day (Voedingscentrum, n.d.).
The healthy food movement has become an unstoppable trend. Social media is arguably the source of this new healthy food trend and thus one of the most efficient ways to market healthy food, particularly via influencer marketing (Kinski, 2018). With the
emergence of social media platforms (SMPs) and food posts, people are now exposed to an abundance of information about the eating behaviour of other people online (Hefner & Vorderer, 2017). With the ability to share information via social media, there is a possibility that social media has become an important information source about healthy food on the internet (Rutsaert et al., 2013). One of these SMPs is Instagram. It is of great importance to investigate whether SMPs like Instagram could have a beneficial effect on young adults’ attitude towards eating healthy and their intention to eat healthy to prevent the development of poor dietary habits and their severe consequences, focusing on women specifically because they engage more on social media than men (Jones, n.d.).
Instagram makes great use of influencer marketing because consumers value other people’s judgements higher than from an advertiser (De Veirman, Cauberghe, & Hudders, 2017). Research states that influencers who promote healthy food leads to 32% of their followers listening to advice for healthier food options and getting motivated to eat healthy themselves (Byrne, Kearney, & MacEvilly, 2017). This shows that influencers have the power to
influence other people with their content about healthy food choices. However, not every influencer is the same. There is a difference between micro and macro influencers. Macro influencers (100.000-300.000 followers) have much greater reach than micro influencers (1.000-20.000 followers). Therefore, macro influencers seem to generate greater source likeability. However, micro influencers are considered to have a better relationship and direct connection with their audience and to listen more to the fan base when producing content, which leads to higher source credibility (Domingues Aguiar & Van Reijmersdal, 2018). Therefore, this study investigates if these two source characteristics explain the relationship between influencers and women’s attitude towards eating healthy and intention to eat healthy.
The influence of source likeability and source credibility on women’s attitude and intention could depend upon involvement with eating healthy. Fogg (2003) stated that issue involvement with the information is a prominent factor influencing information
interpretations. A consumers’ level of involvement with eating healthy can be a processing cue that motivates the receiver to actively or passively process the perceived information by an influencer (Petty & Cacioppo, 1979; Xiao, Wang, & Chan-Olmsted, 2018). Therefore, the message of an influencer about healthy food could also be affected by the level of
involvement the receivers have with eating healthy themselves.
Attitudes and intentions towards eating healthy are investigated as dependent
variables in relation to the two-step flow theory by Maurer (2008). Maurer (2008) argues that media first provide opinion leaders, for the purpose of this research, influencers, with
information. Influencers then form their own opinions about this information and go on to communicate this to their audience. Based on this theory, it is thus expected that, when information about eating sufficient amounts of fruits and vegetables reaches the public via SMIs, it may get them to adopt the message.
As previously mentioned, young adulthood (18-25 years old) is a risky time period for the development of a poor diet, and social media and influencers have a direct influence on young people as social media content is a regular part of their psychosocial experience (McHale, Dotterer, & Kim, 2009). Therefore, it is important to investigate the influence of social media and influencers on attitudes and intentions towards eating healthy among young adults. However, the effect of Instagram influencers (micro vs. macro) on individuals’ attitude towards eating healthy and intention to eat healthy, together with the underlying processes such as credibility and likeability of the source and involvement with eating healthy, has
never been studied before. The current study addresses this research gap because it is not clear yet how social media and influencers could improve young adults’ attitudes to eat healthy and improve their intentions to do so. Moreover, the study clarifies how source
credibility, source likeability and involvement with eating healthy may play a role in this. The results can help professionals in health communication because it will be more clear what kind of influencer should be imbedded in communication on SMPs like Instagram that encourage young adults to eat sufficient amounts of fruit and vegetables in their daily lives.
Thus, the aim of the study is to identify the potential influence of influencers on Instagram on young women’s attitude towards eating healthy and their intention to eat healthy. The study will address the following research questions: (RQ1) To what extent can exposure to healthy food posts on Instagram by micro versus macro influencers influence women’s (18-25 years old) attitude towards eating healthy and intention to eat healthy? (RQ2) Is this relationship explained by source credibility and source likeability and influenced by women’s involvement with eating healthy?
Theoretical framework Instagram
Young adults between the ages of 18 to 25 years old spend more time on engaging with the internet than any other activity (Coyne, Padilla-Walker, & Howard, 2013). They have reported to use the internet as a health information source (McKinley & Wright, 2014). Especially media that facilitate interaction with other people, like SMPs, are preferred among young adults (Xenos & Foot, 2008). One of the most popular feature provided by these SMPs is photo sharing. Social media are now also frequently used by food consumers to inform others about what they eat through photos and recipes (Duggan, 2013).
Instagram is an online mobile application that provides users with the possibility to share their own photos and videos. It is one of the most popular SMPs and currently has one billion monthly active users worldwide (Statista, 2018). Instagram has the highest levels of user engagement compared to other SMPs (McCullough, 2015; Phua, Jin & Kim, 2017) and is mostly used by younger users below the age of 35 years (Statista, 2018).
SMPs such as Instagram are being used to promote health in online communities (Galica & Chou, 2014). In their research, Chou, Hunt, Beckjord, Moser and Hesse (2009) argued that health communication through social media has the broadest reach and the most
impact when it is targeted at young adults. In comparison to SMPs such as Facebook and Twitter, Instagram has a main focus on the visual aspect rather than the text. Galica and Chou (2014) argued that this visual emphasis of Instagram is an important element that evokes emotions and engagement, which could potentially enhance the impact of public health communication. Therefore, this study will focus on communication about healthy eating behaviour by influencers on Instagram.
Social media influencers: micro vs. macro
Social media influencers (SMIs) are increasingly being used on Instagram. Influencer marketing refers to business to consumers brand partnerships with SMIs that can have an advertising or a relationship focus. Influencer marketing makes it possible to reach audiences who are difficult to reach via traditional media and is mainly used to reach a young target audience (Domingues Aguiar & Van Reijmersdal, 2018).
SMIs are seen as people who have built a sizable social network of people who follow them (De Veirman et al., 2017) and they reach a mass audience on whom they have an impact (McQuarrie, Miller, & Philips, 2012). The use of SMIs is a strategy to overcome possible resistance among the audience against the targeted behaviour, by making the message more likeable and attractive (Fransen, Verlegh, Kirmani & Smit, 2015; Knowles & Linn, 2004). Therefore, influencer marketing is a less obtrusive way to reach the target audience
(Boerman, Willemsen, & Van der Aa, 2017). The strategy could thus be beneficial for young adults who feel reluctant to eat healthy food as it is promoted in an attractive way, which might lead to young adults feeling more positive about eating healthy.
Instagram provides the possibility to reach users with SMIs and pay the SMIs to endorse a product (Long, 2016). The content these SMIs produce is likely to be interpreted as highly credible and authentic rather than paid advertising because the advertisements are often a part of the daily narratives the SMIs post on their Instagram accounts (Abidin, 2016), which in turn also leads to lower resistance to the message and less avoidance (De Vries, Gensler, & Leeflang, 2012). Advertising with SMIs on Instagram also causes consumers to want to live up to the same standards that the SMIs do (Silvera & Austad, 2004). However, the effectiveness is not the same for every type of SMI (Silvera & Austad, 2004). The number of followers reflects on the network size of a SMI and serves as an indication for popularity (De Veirman et al., 2017). As previously mentioned, this study will focus on the difference
between micro and macro influencers. Micro influencers have 1.000 to 20.000 followers and macro influencers have 100.000 to 300.000 followers (Domingues Aguiar & Van
Reijmersdal, 2018). This study investigates whether a micro or a macro influencer should be used in order to positively influence attitudes towards eating healthy and intentions to eat healthy.
Attitude towards eating healthy and intention to eat healthy
As previously mentioned, the two-step flow theory maintains that opinion leaders are
provided with information by the media and form their own opinion about this. Subsequently, they communicate this opinion to their audience (Maurer, 2008). Opinion leaders, for the purpose of this study, SMIs, are more trusted than the media and have a certain status. It is thus expected that through this process SMIs influence the attitudes and intentions of their audience towards eating healthy.
Attitudes can be defined as relatively durable evaluations of objects, issues and individuals (Petty, Barden, & Wheeler, 2009). Besides attitudes towards healthy eating, intentions regarding eating healthy will also be measured. This operationalization is derived from the construct ‘purchase intentions’ which is defined as “an individual’s readiness and willingness to purchase a certain product or service” (Ajzen & Fishbein, 1980, p. 102). In this study, this definition will be slightly adjusted to the intention to adopt a certain behaviour, namely healthy eating. Based on the theory of planned behaviour by Azjen (1991), it is important to study attitudes and intentions in this area as attitudes towards a behaviour shape an individual’s behavioural intentions to perform the behaviour, which in turn is a predictor of actual behaviour.
Source credibility
To further understand the relationship between a micro or macro influencer who posts about healthy food and attitudes towards eating healthy and the intention to eat healthy of the receiver of the message, the credibility of the SMI will be studied. The credibility of the source, in this case the micro or macro influencer, is of high importance when it comes to the effectiveness of a promotional message (Gotlieb & Sarel, 1991; Ohanian, 1990). O’Keefe (1990, p. 181), defined source credibility as “judgements made by a perceiver concerning the believability of a communicator.” Source credibility is an important factor in persuasion
effectiveness as it determines the response towards advertisements and how much the message receiver believes the claims the sender makes (Amos, Holmes, & Strutton, 2008; Buda & Zhang, 2000; Gotblieb & Sarel, 1991; Wu & Wang, 2011). When a source is believed to be highly credible, it is said to result in more favourable attitudes towards the message, leading to higher purchase intentions (Pornpitakpan, 2004). Research by Pérez, del Mar García de los Salmones and Rodríguez des Bosque (2013) supported these findings as they have studied the effect of source credibility of celebrities on attitudes towards the
advertisement and found a positive influence. Moreover, two studies revealed the positive impact of source credibility on attitudes towards applying appropriate eating habits (Arora & Arora, 2004; Johnston & Warkentin, 2010). Additionally, credible sources lead to behaviour in line with the source recommendations (Pornpitakpan, 2004). These findings were also found in the context of healthy eating, where it was concluded that source credibility positively influences intentions to follow guidelines to eat healthy (Arora & Arora, 2004).
However, there seems to be a difference in credibility between micro and macro influencers. According to Cruz (2018), macro influencers often achieve a celebrity status due to their high number of followers and reach. Because there is limited research done regarding the source credibility of macro influencers, it is assumed that their credibility is comparable to those of celebrities. According to Hsu, Lin and Chiang (2013), celebrities are seen as trustworthy spokespersons. Their recommendations are effective because they offer insights into their personal lives which creates a parasocial relationship with their followers. As a result, consumers trust the message the SMI communicates. On the other hand, Djafarova and Rushworth (2017) found that celebrity endorsement on Instagram is effective, but their results also suggested that micro influencers have even more power because they are more relatable and more likely to live a normal life, compared to celebrities. Smith, Menon and Sivakumar (2005) found that people tend to rely on recommendations made by their peers, rather than other information. Similarly, Swant (2016) stated that followers often perceive micro influencers as their peers and therefore also trust their recommendations. Thus, micro influencers are more relatable and are perceived as peers which is likely to lead to higher credibility. Furthermore, the persuasion knowledge model states that consumers learn to recognize celebrity endorsements which include persuasion tactics that could be resisted (Friestad & Wright, 1994). As macro influencers are compared to celebrities, messages by macro influencers could be perceived as a sales tool rather than an information channel which leads to consumers feeling threatened (Dichter, 1966). However, when the message does not
feel like a sales tool, the consumer is more likely to accept the recommendation.
Therefore, it is expected that a micro influencer, compared to a macro influencer, will be perceived as more credible, which will lead to a more favourable attitude towards eating healthy. Furthermore, it is expected that a micro influencer, due to higher credibility, will encourage higher intentions to eat healthy (See Figure 1).
H1: Source credibility mediates the effect of SMIs on healthy eating such that an Instagram post about healthy food by a micro influencer, compared to a macro influencer, will lead to (H1a) a more favourable attitude towards eating healthy and (H1b) to a higher intention to eat healthy.
Source likeability
A second characteristic of SMIs that could explain the relationship between influencers and attitudes towards eating healthy and intentions to eat healthy is source likeability. Source likeability refers to the positive evaluations of a SMI (Cohen, 2001; Giles, 2002; Hoffner & Cantor, 1991). Due to the growing popularity of social media, consumers are likely to use a peripheral process to evaluate SMIs on social media. This is also the case on Instagram, where consumers base their judgments on peripheral cues like number of followers, which indicates higher likeability and therefore higher popularity (Chaiken & Maheswaran, 1994; De Veirman et al., 2017; Metzger & Flanagin, 2013; Petty & Cacioppo, 1986). Furthermore, research by Dholakia and Sternthal (1977) concluded that communicator likeability increases positive attitude change and that a likeable source leads to more persuasion such as positive attitude change than a non-likeable source (Chaiken, 1980). Moreover, Moyer-Gusé (2008) argued that liking a character reduces reactance, the most prominent from of resistance (Brehm, 1966). This means that liking reduces the threat to an individual’s freedom to choose his or her own attitudes and behaviours and thus enhances persuasive effects.
It is expected that macro influencers will produce greater likeability because they have a higher number of followers and therefore greater popularity than micro influencers. Accordingly, it is expected that a macro influencer will be more likeable and will lead to a more favourable attitude towards eating healthy. Furthermore, it is expected that a macro influencer, due to higher likeability, will result into higher intentions to eat healthy (see Figure 1).
H2: Source likeability mediates the effect of SMIs on healthy eating such that an Instagram post about healthy food by a macro influencer, compared to a micro influencer, will lead to (H1a) a more favourable attitude towards eating healthy and (H1b) to a higher intention to eat healthy.
Involvement with eating healthy
The previous sections have focused on the impact of SMIs on attitudes and intentions, as mediated by source likeability and source credibility. These relationships could depend upon one’s own involvement with eating healthy. Involvement is defined as the degree to which an issue is personally important (Petty & Cacioppo, 1979). According to Wilson and Sherrell (1993), involvement with an issue may moderate the impact of the source on attitudes. This effect can be explained further by the elaboration likelihood model by Petty and Cacioppo (1986). The model proposes that people with high involvement process a message via the central route with more motivation and a higher ability to process the message. People with low involvement will use the peripheral route which means that they have less motivation and a lower ability to process the message. High involved message recipients consciously consider arguments central to the message whereas low involved message recipients rely on peripheral cues like source characteristics as source credibility when they decide to accept the message or not (Andrews & Shimp, 1990). High involved message recipients also use source characteristics to help understand message arguments but strong argumentation is more important to them (Petty & Cacioppo, 1984). According to Petty and Cacioppo (1981) and Stiff (1986), high source credibility led to significantly more favourable attitudes, compared to low credibility, but only for low involvement message recipients. Similarly, Wilson and Sherrell (1993) argued that source characteristics had a significant influence on attitude, but only when recipients have low involvement. Moreover, source characteristics that are relevant for persuasion such as credibility and likeability serve as peripheral cues under low involvement, but are also seen as strong message arguments under high involvement (Petty et al., 2009). Therefore, these source characteristics may also result in attitude change under conditions of high involvement. Thus, research is somewhat divided on whether source characteristics impact attitudes more positively under low- or high involvement.
Based on most of the research on high- and low involvement, it is assumed that individuals who are low involved with eating healthy are more likely to rely on peripheral cues like the aforementioned source characteristics of the SMI. Therefore, it is expected that
the credibility of a micro influencer on Instagram will lead to more positive attitudes towards healthy eating and higher intentions to eat healthy and that this effect is stronger when the message receiver has low involvement with eating healthy rather than high involvement. The same effect is expected for the likeability of macro influencers (see Figure 1).
H3: The relationship between source credibility and (H3a) attitude towards eating healthy and (H3b) intention to eat healthy will be stronger when involvement with eating healthy is low.
H4: The relationship between source likeability and (H4a) attitude towards eating healthy and (H4b) intention to eat healthy will be stronger when involvement with eating healthy is low.
Figure 1 | Conceptual model
Method
Design and participants
The current experiment used a one (type of SMI: micro vs. macro) between-subject factorial design (post-test only). One independent variable was manipulated and the participants were randomly assigned to one of the two conditions. As such, it could be investigated whether there were differences between the two conditions. Because each participant was only measured once, it is fairly certain that no other factors influenced the results (Gravetter &
Forzano, 2012). The online experimental study was conducted from the 3rd until the 17th of May 2019 amongst 154 participants. The participants were recruited via e-mail and the link to the online experiment was distributed via social media (Instagram, Facebook and
LinkedIn). The participants who did not meet the requirements for age (18-25 years old) and gender (woman), did not have Instagram or did not complete the online survey were excluded for further analysis (n = 8). The final sample consisted of 146 women with a mean age of 22.99 years (SD = 1.66).
Procedure
The online experiment was executed via a survey that was created in Qualtrics. At the beginning of the survey, the participants had to sign an informed consent sheet before participating in the study. The participants were randomly assigned to one of the two conditions. First, the participants were exposed to the Instagram page of the fictitious SMI Chloe Smith as either a micro influencer or macro influencer. After fifteen seconds, the participants were able to click further to continue with the survey. In this way, it was
prevented that the participants paid little or no attention to the stimulus material. Thereafter, both conditions were shown four Instagram posts including pictures of a healthy dish. Again, the participants could click further after fifteen seconds. Afterwards, the participants filled out a questionnaire concerning their opinion of the healthy dishes, source likeability, source credibility, involvement with eating healthy, attitude towards eating healthy and intention to eat healthy. Next, questions for the manipulation check and three demographic questions regarding gender, age and level of education were asked. Lastly, two questions about Instagram use and current eating behaviour were asked. The questionnaire was identical in both conditions.
Stimulus material
During the online experiment, two different Instagram posts were used: photos with healthy dishes by the fictitious SMI Chloe Smith as either a micro or macro influencer (see Appendix A). Both conditions contained a screenshot of the Instagram page of Chloe Smith. In the micro condition, she had 1.356 followers and in the macro condition she had 231.000
followers. The macro condition contained a blue mark, which referred to a verified Instagram account. Thereafter, four Instagram posts with healthy dishes were shown in both conditions.
The photos were taken from Voedingscentrum (n.d.). The posts in the micro condition had between 50-250 likes (see Appendix B) whereas the posts in the macro condition had between 15.000-25.000 likes (see Appendix C).
Pre-test
Prior to the study, a pre-test was performed with the stimulus material in order to establish which out of seven dishes taken from Voedingscentrum were assessed as the most healthy (see Appendix D). Within the personal network of the researcher, a total of 24 participants were recruited. The participants had to fill out to what extent they believed the food on the photo was (1) very unhealthy to (7) very healthy on a seven-point semantic differential scale. The four photos who were rated as most healthy, (M = 6.50, SD = 0.72; M = 6.25, SD = 0.79; M = 6.25, SD = 0.94; M = 6.29, SD = 0.81) were used in the survey (see Appendices B & C).
Measures
Filler questions about Instagram posts
After the four Instagram posts, the participants were asked three questions about the healthy dishes on a five-point Likert scale ranging from (1) strongly disagree to (5) strongly agree: “I believe the dishes on the photo are tasty”, “I believe the dishes on the photos are healthy” and “I believe the dishes on the photos are easy to prepare.”
Source credibility
Source credibility was measured on a validated scale including five items derived from research by Kim, Lee and Prideaux (2014). On a five-point Likert scale ranging from (1) strongly disagree to (5) strongly agree, the participants answered to what extent they
perceived the SMI as “honest”, “reliable”, “dependable”, “sincere” and “ethical” (α = .93, M = 3.37, SD = 1.03).
Source likeability
To measure the likeability of the SMI, a validated scale developed by Reysen (2005) was used. On a five-point Likert scale ranging from (1) strongly disagree to (5) strongly agree, the participants had to fill out to what extent they perceived the SMI as “friendly”, “likeable”,
“warm" and “approachable” (α = .80, M = 3.62, SD = 0.73). Involvement with eating healthy
The items measuring involvement with eating healthy were based on eight items who formed a validated scale, as used by Bae (2008). The statement “To me, eating healthy is …” had to be answered on a five-point semantic differential scale ranging from (1) strongly disagree to (5) strongly agree with the items “unimportant/important”, “boring/interesting”, “irrelevant/ relevant”, “unexciting/exciting”, "means nothing to me/means a lot to me”, “mundane/ fascinating”, “worthless/valuable” and “not needed/needed” (α = .89, M = 3.85, SD = 0.61).
Attitude towards eating healthy
To measure attitude towards eating healthy, a validated scale by Spears and Singh (2004) was adapted. The participants answered the question “Do you find healthy eating …” on a five-point semantic differential scale from (1) strongly disagree to (5) strongly agree with the items “bad/good”, “unfavourable/favourable”, “dislikeable/likeable”, “unappealing/ appealing” and “unpleasant/pleasant” (α = .92, M = 3.87, SD = 0.88).
Intention to eat healthy
The intention to eating healthy was measured based on a validated scale by Conner, Norman and Bell (2002). Participants answered the five following statements: "I intend to eat a healthy diet in the coming days”, “I want to eat a healthy diet in the coming days”, “I try to eat a healthy diet in the coming days”, “I expect to eat healthy in the coming days”, and “In the coming days, it is very likely that I will eat healthy.” The participants answered them on a five-point Likert scale ranging from (1) strongly disagree to (5) strongly agree (α = .87, M = 4.02, SD = 0.64).
Manipulation check
In order to find out whether participants perceived the micro and macro influencer as intended, three questions were asked which the participants answered on a five-point
semantic differential scale based on research by de Veirman et al. (2017). The questions were as follows: “Chloe Smith has a … amount of followers” (1 = very small to 5 = very large), “Do you think that Chloe Smith is …” (1 = unpopular to 5 = very popular) and “Chloe Smith
is followed by … people than an average influencer” (1 = less to 5 = more) (α = .88, M = 3.42, SD = 1.08). Together, these three items measured the perceived popularity of the influencers.
Control variables
To control for confounding factors, several control variables were included. First, the
participants were asked two demographic questions regarding their age and educational level (1 = no school completed to 9 = university education). Thereafter, they were asked a question asked about their Instagram use, namely: “Do you follow influencers on Instagram who post about healthy food?” (1 = no, 2 = yes). The final question regarded the healthy eating
behaviour of the participants, and asked “Before filling out this survey, I was already eating healthy” (1 = strongly disagree to 5 = strongly agree). The question included the following footnote: “Eating healthy means eating at least 250 grams of vegetables and 200 grams (two pieces) of fruit a day (Voedingscentrum, n.d.).
Results
Randomization check
The results showed that the experimental groups did not differ with respect to age, F (144) = 1.76, p = .40, current healthy eating behaviour, r = .07, p = .40, but did on following of influencers on Instagram who post about healthy food, χ2 (1) = 10.63, p < .05 and level of
education, χ2 (2) = 6.32, p < .05. Furthermore, the correlation with the dependent variables
attitude and intention was tested, with respect to age, current healthy eating behaviour, following of influencers on Instagram who post about healthy food and level of education (see Table 1). Based on these results it was concluded that level of education and following of influencers on Instagram who post about healthy food were used in the analyses as
Table 1
Correlation between control variables and attitude and intention
* Correlation is significant at the 0.05 level (2-tailed).
Manipulation check
To check whether the participants perceived the experimental conditions as intended, an independent samples t-test was performed with the type of influencer (micro vs. macro) as independent variable and perceived popularity of the influencers as the dependent variable. A significant effect was found, t (144) = -8.27 p < .001, 95% CI [-1.51, -0.93]. The participants who were exposed to a macro influencer indicated higher popularity (M = 4.01, SD = 0.92) than the participants who were exposed to a micro influencer (M = 2.80, SD = 0.85). This indicates that the manipulation of the type of influencer was successful.
Descriptive analysis
A descriptive analysis was conducted with all the dependent variables of which an overview is presented in Table 2. Most of the respondents scored higher than the midpoint of the five-point scale for all the variables. Looking at the outcome variables, source likeability was negatively correlated with source credibility (r = -.24), attitude (r = -.13) and intention (r = -. 06) Attitude Intention Age .06 -.02
Following of influencers on Instagram -.10 -.31*
Currenthealthy eating behaviour -.08 .06
Table 2
Model correlation matrix with mean and standard deviations
* Correlation is significant at the .05 level (2-tailed). ** Correlation is significant at the .01 level (2-tailed).
Note: Correlations, means and standard deviations of outcome variables, all variables on five-point scales.
Direct effects
Before testing the hypotheses, independent samples t-tests were performed to test whether any direct effects could be observed for an Instagram post by a micro influencer or a macro influencer on any of the outcome variables (see Table 3). The analysis showed that there were significant effects for source credibility, source likeability and attitude. A micro influencer was significantly more credible (M = 4.07, SD = 0.82) than a macro influencer (M = 2.73, SD = 0.75). A macro influencer was significantly more likeable (M = 3.78, SD = 0.50) than a micro influencer (M = 3.44, SD = 0.88). Lastly, a post by a micro influencer led to
significantly more positive attitudes towards eating healthy (M = 4.23, SD = 0.57) than a post by a macro influencer (M = 3.56, SD = 0.99). Variables 1 2 3 4 5 1. Source credibility 1.00 - - - - 2. Source likeability -.24** 1.00 - - - 3. Involvement .18* .05 1.00 - - 4. Attitude .33** -.13 .31** 1.00 - 5. Intention .11 -.06 .28** .21* 1.00 Mean 3.37 3.62 3.85 3.87 4.02 SD 1.03 0.73 0.61 0.88 0.64
Table 3
Effects of micro- and macro influencer on outcome variables
Hypotheses testing
Mediation effect of source credibility
With regard to H1a, it was expected that source credibility mediates the effect of SMIs on healthy eating such that an Instagram post by a micro influencer, compared to a macro
influencer, will lead to a more favourable attitude towards eating healthy. Regression analyses were conducted in order to test the mediation effects, using the method of Baron and Kenny (1986) (see Table 4). First, it was tested whether there was a direct effect of the type of influencer on attitude (b* = -.38, p < .001). The results show that attitude significantly decreases when a macro influencer is used, instead of a micro influencer. Next, it was tested whether the type of influencer correlated with source credibility (b* = -.63, p < .001), and results show that source credibility significantly decreases when a macro influencer is used, instead of a micro influencer. Furthermore, the relationship between source credibility and attitude was tested, controlling for type of influencer. The model was significant, F (4, 141) = 6.63, p < .001. However, the effect of source credibility on attitude, controlling for the type of influencer, was not significant (b* = .14, p = .17). Given this finding, a mediation effect was excluded. Thus, source credibility was not found to mediate the relationship from the type of influencer to attitude, and H1a was rejected.
In H1b it was expected that source credibility mediates the effect of SMIs on healthy eating such that an Instagram post by a micro influencer, compared to a macro influencer, will lead to a higher intention to eat healthy. First, a regression analysis was conducted to test the relationship between the type of influencer and intention, which was not significant (b* = .04,
Micro influencer (n = 70) Macro influencer (n = 76) M SD M SD t p Source credibility 4.07 0.82 2.73 0.75 10.38 <.001 Source likeability 3.44 0.88 3.78 0.50 -2.87 <.01 Involvement 3.84 0.64 3.86 0.58 -0.17 .86 Attitude 4.23 0.57 3.56 0.99 5.08 <.001 Intention 4.09 0.59 3.95 0.69 1.28 .20
p = .65). Secondly, it was established that, as for H1a, the type of influencer significantly influenced source credibility (see Table 4). Thirdly, the relationship between source
credibility and intention, controlling for the type of influencer was tested, F (4, 141) = 4.33, p = <.05. The relationship was not significant (b* = -.09, p = .42). As only one relationship was significant, no mediation effect was found. Source credibility has not been found to mediate the relationship from type of influencer to intention, and H1b was rejected.
Table 4
Direct effects between type of influencer, source credibility, attitude and intention
Note: TOF = Type of influencer.
Note 2: LBCI is the lower bound of the confidence interval; UBCI is the upper bound of the confidence interval.
Note 3: Level of education and following of influencers on Instagram who post about healthy food are used as confounding variables.
Mediation effect of source likeability
With regard to H2, it was expected that likeability mediates the effect of SMIs on healthy eating such that an Instagram post about healthy food by a macro influencer, compared to a micro influencer, will lead to (H1a) a more favourable attitude towards eating healthy and (H1b) to a higher intention to eat healthy. To test the potential mediation effect, it was already established for the previous hypothesis that the relationship between the type of influencer and attitude was significant (see Table 5). Furthermore, the type of influencer was
significantly correlated with source likeability (b* = .21, p < .05). The results show that source likeability significantly increases when a macro influencer is used, instead of a micro influencer. The model for the relationship between source likeability and attitude, controlling for the type of influencer was significant, F (4, 141) = 6.15, p < .001. However, the
Variables 95% CI
b SE b* p LBCI UBCI
TOF -> attitude -0.67 0.14 -.38 <.001 -0.95 -0.38
TOF -> intention -0.05 0.11 -.04 .65 -0.26 0.16
TOF -> source credibility -1.30 0.14 -.63 <.001 -1.57 -1.03 Source credibility -> attitude 0.27 0.07 .32 <.001 0.14 0.41 Source credibility -> intention 0.05 0.05 .08 .36 -0.05 0.15
relationship between source likeability and attitude, controlling for the type of influencer, was not (b* = -.04, p = .61). As the last relationship is not significant, it was concluded that source likeability does not mediate the relationship between type of influencer and attitude, and H2a was rejected.
As for H2b, and similarly to H1b, the relationship between type of influencer and intention was not significant (see Table 5). Moreover, as for H2a, the type of influencer is significantly correlated with source likeability (see Table 5). The regression model for the relationship between source likeability and intention, controlling for the type of influencer, was significant, F (4, 141) = 4.25, p = < .05, but the relationship was not (b* = -.05, p = .54). Given that the last effect is not significant, a mediation effect was excluded. Source
likeability does not mediate the relationship between type of influence and intention, thereby rejecting H2b.
Table 5
Direct effects between type of influencer, source likeability, attitude and intention
Note: TOF = Type of influencer.
Note 2: LBCI is the lower bound of the confidence interval; UBCI is the upper bound of the confidence interval.
Note 3: Level of education and following of influencers on Instagram who post about healthy food are used as confounding variables.
Moderating effect of involvement with eating healthy
In the third hypothesis, it was expected that the relationship between source credibility and (H3a) attitude towards eating healthy and (H3b) intention to eat healthy will be stronger when involvement with eating healthy is low. A multiple regression with moderation was performed in order to check for the moderation effect of involvement for both attitude and
Variables 95% CI
b SE b* p LBCI UBCI
TOF -> attitude -0.67 0.14 -.38 <.001 -0.95 -0.38
TOF -> intention -0.05 0.11 -.04 .65 -0.26 0.16
TOF -> source likeability 0.30 0.12 .21 <.05 0.06 0.55 Source likeability -> attitude -0.14 0.10 -.11 .18 -0.34 0.06 Source likeability -> intention -0.05 0.07 -.06 .48 -0.19 0.09
intention, using standardized variables. The moderation of involvement was not significant for attitude (b* = -.04, p = .64), nor for intention (b* = .14, p = .09). Involvement does not moderate the relationships between source credibility and attitude (see Table 6) and intention (see Table 7). Thus, H3 was rejected.
Table 6
The effects of source credibility on attitude with involvement as a moderator
Note: R2 = .18, F (5, 140) = 6.01, p < .001.
Note 2: LBCI is the lower bound of the confidence interval; UBCI is the upper bound of the confidence interval.
Note 3: Level of education and following of influencers on Instagram who post about healthy food are used as confounding variables.
Table 7
The effects of source credibility on intention with involvement as a moderator
Note 1: R2 = .18, F (5, 140) = 6.08, p < .001.
Note 2: LBCI is the lower bound of the confidence interval; UBCI is the upper bound of the confidence interval.
Note 3: Level of education and following of influencers who post about healthy food on Instagram are used as confounding variables.
The fourth hypothesis expected that the relationship between source likeability and (H4a) attitude towards eating healthy and (H4b) intention to eat healthy will be stronger when involvement with eating healthy is low. Again, a multiple regression with moderation was
95% CI SE b* p LBCI UBCI Source credibility (X) b1 0.08 .28 <.001 0.12 0.44 Involvement (M) b2 0.08 .26 <.001 0.10 0.41 Involvement*Source credibility (XM) b3 0.08 -.04 .64 -0.20 0.12 95% CI SE b* p LBCI UBCI Source credibility (X) b1 0.08 0,04 .60 -0.12 0.20 Involvement (M) b2 0.08 .24 <.05 0.08 0.40 Involvement*Source credibility (XM) b3 0.08 .14 .09 -0.02 0.30
performed in order to check for the moderation effect of involvement, using standardized variables. Involvement did not moderate the relationship between source likeability and attitude as the effect was not significant (b* = .10, p = .21), thus H4a was rejected (see Table 8).Moreover, involvement did moderate the relationship between source likeability and intention (b* = .18, p = .03) such that when involvement increases, the effect of source likeability becomes more positive on the intention to eat healthy, and when involvement decreases, the effect of source likeability on the intention to eat healthy becomes more negative (see Table 9). Thus, the relationship between source likeability and intention is stronger when involvement with eating healthy is high, thereby rejecting H4b.
Table 8
The effects of source likeability on attitude with involvement as a moderator
Note 1: R2 = .13, F (5, 140) = 4.32, p < .05.
Note 2: LBCI is the lower bound of the confidence interval; UBCI is the upper bound of the confidence interval.
Note 3: Level of education and following of influencers who post about healthy food on Instagram are used as confounding variables.
95% CI
SE b* p LBCI UBCI
Source likeability (X) b1 0.08 -.16 .06 -0.32 0.01
Involvement (M) b2 0.08 .33 <.001 0.17 0.49
Table 9
The effects of source likeability on intention with involvement as a moderator
Note: R2 = .20, F (5, 140) = 6.76, p < .001.
Note 2: LBCI is the lower bound of the confidence interval; UBCI is the upper bound of the confidence interval.
Note 3: Level of education and following of influencers who post about healthy food on Instagram are used as confounding variables.
Conclusion and discussion
The current study examined the effects of micro vs. macro influencers on attitude towards eating healthy and intention to eat healthy, together with the mediating effects of source credibility and source likeability and the moderating effect of involvement with eating healthy. Although research suggests that micro influencers are more credible than macro influencers and macro influencers are more likeable than micro influencers, neither of these source characteristics are found to explain the effects of SMIs on attitude and intention. Moreover, the effects of the source characteristics credibility and likeability on attitude and intention were not stronger when involvement with eating healthy was low.
Upon analyzing the relationships between the type of influencer and attitudes and intentions, the results for the first hypothesis showed that source credibility did not explain these
relationships. However, a micro influencer (vs. macro influencer) did generate higher source credibility and a more positive attitude. Higher source credibility also led to a more positive attitude. The results for the second hypothesis demonstrated that neither source likeability explained the relationship between the type of influencer and attitudes and intentions. Nonetheless, a macro influencer (vs. micro influencer) led to higher source likeability. Next, the results for the third hypothesis showed no interaction effect between source credibility and involvement with eating healthy on attitude, nor on intentions. Lastly, the results for the
95% CI
SE b* p LBCI UBCI
Source likeability (X) b1 0.08 -.11 .16 -0.27 0.05
Involvement (M) b2 0.08 .27 <.001 0.12 0.42
fourth hypothesis showed no interaction effect between source likeability and involvement on attitude. Yet, there was an interaction effect between source likeability and involvement on intention, and the effect was stronger when involvement with eating healthy was higher.
Although none of the four hypotheses were confirmed, there are valuable insights to be drawn from this research. For H1 and H2, neither of the mediators could explain the relationship between the type of influencer and attitude or intention. This means that there are other potential underlying constructs that may better explain this relationship, which need to be studied in the future. However, results for H1 do show that a micro influencer is perceived as more credible, which supports the previous findings regarding the credibility of micro
influencers (Djafarova & Rushworth, 2017; Swant, 2016). Similarly, the finding that higher source credibility leads to more positive attitudes is consistent with previous research (Arora & Arora, 2004; Johnston & Warkentin, 2010; Pérez et al., 2013)and resonates with the two-step flow theory by Maurer (2008) that concludes that opinion leaders not only convey information, but also influence opinions of their audience. As no direct effect was found for source credibility on intentions, the findings contradict the research of Pornpitakpan (2004), and partly those of Arora and Arora (2004). Thus, a credible source leads to a more positive attitude towards eating healthy but does not lead to the intention to actually eat healthier. This could be explained by the fact that, similar to actual behaviour, intentions are difficult to influence. Perhaps in particular when the message is presented by a relatively unfamiliar source, which will be further discussed in the suggestions for future research.
As for H2, macro influencers are found to generate greater likeability than micro influencers, which is in line with previous studies (Chaiken & Maheswaran, 1994; De Veirman et al., 2017; Metzger & Flanagin, 2013; Petty & Cacioppo, 1986). However, source likeability did not lead to more positive attitudes or intentions, which is inconsistent with research by Chaiken (1980) and Dholakia and Sternthal (1977). A striking finding was that when a macro influencer (vs. micro influencer) was used, attitudes got significantly more negative. This could be explained by the fact that macro influencers are seen as celebrities, who are perceived as less approachable and as ‘sales tools’ who try to persuade.
For H3, the moderating effect of involvement was not significant for both attitude and intention. As for H4, the relationship between source likeability and intention became
stronger when involvement was high. This finding contradicts studies by Petty and Cacioppo (1981), Stiff (1986) and Wilson and Sherrel (1993), who stated that source characteristics are
only influential under low involvement and that likeability serves as a peripheral cue. However, research on involvement is inconclusive, as there are also studies that found that source characteristics have a greater impact under high involvement (Petty & Cacioppo, 1984; Petty et al., 2009). The current research finds support for the latter. This suggests that individuals with a high level of involvement towards eating healthy value source likeability as a source characteristic and do not merely process it as a peripheral cue. As a result, they hold stronger intentions to eat healthy themselves. A positive consequence of this finding is that intentions formed under high involvement, thus through central processing, will lead to more consistent intentions towards eating healthy, thereby potentially making intentions better predictors of actual behaviour. However, it is important to note that this effect was only found for intention and not for attitudes. This is not in line with the theory of planned
behaviour by Ajzen (1991), which means that an effect for intention does not always precede an effect for attitude.
Theoretical and practical implications
Interesting conclusions can be drawn from the results of this study that are relevant for further research and the practical area. Research on the topic of influencer marketing in relation to healthy food habits is scarce. The current study contributes to the existing literature by examining the effect of two different SMI’s on attitude towards eating healthy and intention to eat healthy. One of the main contributions of this research is the confirmation that a micro influencer is more credible and a macro influencer is more likeable. Moreover, attitudes towards eating healthy become more positive when a micro influencer is used. This effect is a remarkable finding regarding the influencer marketing trend that mainly focuses on the use of macro influencers. Macro influencers are indeed perceived as more likeable but not as more credible, and likeability does not necessarily create more positive attitudes.
From a managerial perspective, the findings could enhance the marketing strategy of brands who want to promote healthy food. With respect to influencer marketing, the findings suggests that companies who want to trigger young adults to adopt healthy eating habits should not automatically agree to a partnership with a large and popular macro influencer and instead use a micro influencer. Micro influencers have a higher chance of positively changing attitudes towards eating healthy and they charge less for their work, which is also a great benefit. Moreover, choosing a likeable influencer is only advisable when consumers are highly involved with eating healthy. When the level of involvement is unknown, it is advised
to choose a credible influencer, as this will result in more positive attitudes towards eating healthy.
Limitations and suggestions for future research
Although this study gives several interesting insights, it also has limitations. First, only one influencer was shown, who could have been recognized as it was a friend of the researcher. Using various SMIs in the future would be preferred to generalize the results.
Second, the high mean score on the question about current healthy eating behaviour proves that participants already thought of themselves as eating healthy (M = 4.04, SD = 0.94). This might have led to that the intention to eat healthy did not differ between the two conditions as on average level the participants already perceived themselves as eating healthy and were not influenced by the SMIs. For future research, it is recommended to examine participants who score lower on the perception of their current healthy eating behaviour.
Thirdly, the online experiment did not show participants their own Instagram timeline so they might not have felt a connection to the SMI and/or the content on the Instagram posts. The material might not have been as effective as seeing posts about healthy food from an influencer they actually chose to follow might have been. Therefore, it is advised for future research to study the effects of SMIs whom the participants are familiar with.
Further research is necessary in order to fully understand the relationship between different influencers and attitudes and intentions towards eating healthy. This study has already
established that different type of influencers have different effects on attitudes. Unfortunately, these findings were not explained by source credibility or source likeability and it is
somewhat unclear how involvement plays a role. Hence, future research should investigate different variables that could act as mediators that explain the relationship between the type of influencers and consumers’ attitudes and intentions, as well as different moderators that strengthen this effect. For example, it would be interesting to measure the effects of
persuasion knowledge and associated sponsorship disclosures, which are known to influence the interpretations of advertisements. It could then be investigated whether micro and macro influencers cause different levels of persuasion knowledge among consumers, and if
sponsorship disclosures have an effect that leads to differences in attitudes and intentions. It would also be optimal to study the actual behaviour of participants through a longitudinal study. In conclusion, the results highlight the potential of using micro influencers on SMPs,
as they could encourage the development of healthy eating habits of young adults in the future.
References
Abidin, C. 2016. Visibility labour: Engaging with influencers’ fashion brands and #OOTD advertorial campaigns on Instagram. Media International Australia, 161(1), 86–100. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human
Decision Processes, 50, 179-211.
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Upper Saddle River, N.J: Prentice-Hall.
Amos, C., Holmes, G., & Strutton, D. (2008). Exploring the relationship between celebrity endorser effects and advertising effectiveness: A quantitative synthesis of effect size. International Journal of Advertising, 27, 209-234.
Andrews, J. C., & Shimp, T. A. (1990). Effects of involvement, argument strength, and source characteristics on central and peripheral processing of advertising. Psychology & Marketing, 7, 195-214.
Arora, R., & Arora, A. (2004). The impact of message framing and credibility: Findings for nutritional guidelines. Services Marketing Quarterly, 26(1), 35-53.
Bae, H. S. (2008). Entertainment-education and recruitment of cornea donors: The role of emotion and issue involvement. Journal of Health Communication, 13(1), 20-36. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.
Boerman, S. C., Willemsen, L. M., & Van der Aa, E. P. (2017). “This post is sponsored” Effects of sponsorship disclosure on persuasion knowledge and electronic word of mouth in the context of Facebook. Journal of Interactive Marketing, 38, 82-92. Brehm, J. (1966). A theory of psychological reactance. New York: Academic Press. Buda, R., & Zhang, Y. (2000). Consumer product evaluation: The interactive effect of
message framing, presentation order, and source credibility. Journal of Product & Brand Management, 9, 229-242.
Byrne, E., Kearney, J., & MacEvilly, C. (2017). The role of influencer marketing and social influencers in public health. Proceedings of the Nutrition Society, 76, E103.
Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology, 39, 752-766.
Chaiken, S., & Maheswaran, D. (1994). Heuristic processing can bias systematic
processing: Effects of source credibility, argument ambiguity, and task importance on attitude judgment. Journal of Personality and Social Psychology, 66, 460-473.
Chou, W. S., Hunt, Y. M., Beckjord, E. B., Moser, R. P., & Hesse, B. W. (2009). Social media use in the United States: Implications for health communication. Journal of Medical Internet Research, 11, e48.
Cohen, J. (2001). Defining identification: A theoretical look at the identification of audiences with media characters. Mass Communication & Society, 4, 245-264.
Conner, M., Norman, P., & Bell, R. (2002). The theory of planned behavior and healthy eating. Health psychology, 21, 194-224.
Coyne, S. M., Padilla-Walker, L. M., & Howard, E. (2013). Emerging in a digital world a decade review of media use, effects, and gratifications in emerging adulthood. Emerging Adulthood, 1(2), 125-137.
Cruz, F. (2018). Influencers: Micro & macro. Retrieved from https://www.tapinfluence.com/ influencers-micro-macro/
De Veirman, M., Cauberghe, V., & Hudders, L. (2017). Marketing through Instagram influencers: the impact of number of followers and product divergence on brand attitude. International Journal of Advertising, 36, 798-828.
De Vries, L., Gensler, S., & Leeflang, P. S. H. (2012). Popularity of brand posts on brand fan pages: An investigation of the effects of social media marketing. Journal of
Interactive Marketing, 26(2), 83-91.
Dholakia, R. R., & Sternthal, B. (1977). Highly credible sources: persuasive facilitators or persuasive liabilities? Journal of Consumer Research, 3, 223-232.
Dichter, E. (1966). How word-of-mouth advertising works. Harvard Business Review, 44, 147-166.
Djafarova, E., & 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, 1-7.
Domingues Aguiar, T., & Van Reijmersdal, E. (2018). Influencer Marketing. Amsterdam: SWOCC.
Duggan, M. (2013). Photo and video sharing grow online. Retrieved from http:// www.pewinternet.org/2013/10/28/photo-and-video-sharing- grow-online/.
Fogg, B. J. (2003). Prominence-interpretation theory: Explaining how people assess
credibility online. CHI ’03 extended abstracts on human factors in computing systems (722–723). New York, NY, USA: ACM.
Fransen, M. L., Verlegh, P. W., Kirmani, A., & Smit, E. G. (2015). A typology of consumer strategies for resisting advertising, and a review of mechanisms for countering them. International Journal of Advertising, 34(1), 6-16.
Friestad, M., & Wright, P. (1994). The persuasion knowledge model: How people cope with persuasion attempts, Journal of Consumer Research, 21(1), 1-31.
Galica, K., & Chou, W. Y. S., (2014). Instagram for health: Peer-to-peer fitness motivation. Annals of Behavioral Medicine, 47, 89-89.
Giles, D. C. (2002). Parasocial interaction: A review of the literature and a model for future research. Media Psychology, 4, 279-305.
Gotlieb, J. B., & Sarel, D. (1991). Comparative advertising effectiveness: The role of involvement and source credibility. Journal of Advertising, 20(1), 38-45.
Gravetter, F. J., & Forzano, L. B. (2012). Research methods for the behavioral sciences. Belmont, CA: Wadsworth Cengage Learning.
Hefner, D., & Vorderer, P. (2017). Digital stress: Permanent connectedness and multitasking. In L. Reinecke & M. B. Oliver (Eds.), The Routledge handbook of media use and well-being: International perspectives on theory and research on positive media effects (pp. 237-249). New York, NY, US: Routledge/Taylor & Francis Group. Hoffner, C., & Cantor, J. (1991). Perceiving and responding to mass media characters. In J.
Bryant & D. Zillmann (Eds.), Responding to the screen: Reception and reaction processes (pp. 63–101). Hillsdale, NJ: Lawrence Erlbaum.
Hsu, C. H., Lin, J., & Chiang, H. S. (2013). The effects of blogger recommendations on customers’ online shopping intentions. Internet Research, 23, 69-88.
Johnston, A., & Warkentin, M. (2010). The influence of perceived source credibility on end user attitudes and intentions to comply with recommended IT actions. Journal of Organizational and End User Computing, 22(3), 1-21.
Jones, M. (n.d.). Studies show women engage more on social media than men. Retrieved from
https://www.coxblue.com/studies-show-women-engage-on-social-media-more-than-men/
perception of corporate image, corporate credibility and corporate loyalty. International Journal of Hospitality Management, 37, 131-145.
Kinski, N. (2018). Influencer marketing and the rise of healthy eating habits. Retrieved from https://upfluence.com/influencer-marketing/influencer-marketing-healthy-food-industry
Knowles, E., & Linn, J. (2004). Approach–avoidance model of persuasion: Alpha and omega strategies for change. In E. Knowles, & J. Linn (Eds.), Resistance and Persuasion (pp. 117-148). Lawrence Erlbaum Associates Publishers.
Long, J. (2016). 7 Marketing tips to help grow your brand on Instagram. Retrieved from https://www.entrepreneur.com/article/280964
Maurer, M. (2008). Two‐step flow of communication. In The International Encyclopedia of Communication, W. Donsbach (Ed.).
McCullough, A. (2015). What every marketer needs to know about Instagram. Retrieved from https://www.socialbakers.com/blog/2472-what-every-marketer-needs-to- know-about-instagram
McHale, S. M., Dotterer, A., & Kim, J. Y. (2009). An ecological perspective on the media and youth development. American Behavioral Scientist, 52, 1186–1203.
McKinley, C. J., & Wright, P. J. (2014). Informational social support and online health information seeking: Examining the association between factors contributing to healthy eating behavior. Computers in Human Behavior, 37, 107–116.
McQuarrie, E. F., Miller, J., & Phillips, B. J. (2012). The megaphone effect: Taste and audience in fashion blogging. Journal of Consumer Research, 40, 136-159. Metzger, M. J., & Flanagin, A. J. (2013). Credibility and trust of information in online
environments: The use of cognitive heuristics. Journal of Pragmatics, 59, 210–220. Moyer-Gusé, E. (2008). Toward a Theory of Entertainment Persuasion: Explaining the
Persuasive Effects of Entertainment-Education Messages. Communication Theory, 18(3), 407-425.
Nelson, M. C., Story, M., Larson, N. I., Neumark-Sztainer, D., & Lytle, L. A. (2008). Emerging adulthood and college-aged youth: An overlooked age for weight related behavior change. Obesity, 16, 2205–2211.
Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers' perceived expertise, trustworthiness, and attractiveness. Journal of advertising, 19(3),
39-52.
O'Keefe, D. J. (1990). Current communication: An advanced text series, Vol. 2. Persuasion: Theory and research. Thousand Oaks, CA, US: Sage Publications, Inc.
Pérez, A., del Mar García de los Salmones, M., & Rodríguez del Bosque, I. (2013). The effect of corporate associations on consumer behaviour. European Journal of Marketing, 47, 218-238.
Petty, R. E., Barden, J., & Wheeler, S. C. (2009). The Elaboration Likelihood Model of Persuasion: Developing health promotions for sustained behavioral change. In R. J. DiClemente, R. A. Crosby, & M. C. Kegler (Eds.), Emerging theories in health promotion practice and research, (pp. 185-214). San Fransisco, CA: Jossey- Bass. Petty, R. E., & Cacioppo, J. T. (1979). Issue involvement can increase or decrease persuasion
by enhancing message-relevant cognitive responses. Journal of Personality and Social Psychology, 37, 1915-1926.
Petty, R. E., & Cacioppo, J. T. (1981). Issue involvement as a moderator of the effects on attitude of advertising content and context. Advances in Consumer Research, 8, 20-24.
Petty, R. E., & Cacioppo, J. T. (1984). Source factors and the elaboration likelihood model of persuasion. Advances in Consumer Research, 11, 668-672.
Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer-Verlag.
Phua, J., Jin, S. V., & Kim, J. J. (2017). Gratifications of using Facebook, Twitter, Instagram, or Snapchat to follow brands: The moderating effect of social comparison, trust, tie strength, and network homophily on brand identification, brand engagement, brand commitment, and membership intention. Telematics and Informatics, 34, 412-424. Pornpitakpan, C. (2004). The persuasiveness of source credibility: A critical review of five
decades' evidence. Journal of Applied Social Psychology, 34, 243-281.
Reysen, S. (2005). Construction of a new scale: The Reysen likability scale. Social Behavior and Personality, 33(2), 201-208.
Rutsaert, P., Regan, Á., Pieniak, Z., McConnon, Á., Moss, A., Wall, P., & Verbeke, W. (2013). The use of social media in food risk and benefit communication. Trends in Food Science & Technology, 30, 84-91.
Silvera, D. H. & Austad, B. (2004). Factors predicting the effectiveness of celebrity endorsement advertisements. European Journal of marketing, 38, 1509-1526.
Smith, D., Menon, S., & Sivakumar, K. (2005). Online peer and editorial recommendations, trust and choice in virtual markets. Journal of interactive marketing, 19(3), 15-37 Spears, N., & Singh, S. N. (2004). Measuring attitude toward the brand and purchase
intentions. Journal of Current Issues & Research in Advertising, 26(2), 53–66. Statista. (2018). Instagram - statistics and facts. Retrieved from https://www.statista.com/
topics/1882/instagram/
Stiff, J. B. (1986). Cognitive processing of persuasive message cues: A meta‐ analytic review of the effects of supporting information on attitudes. Communications Monographs, 53, 75-89.
Swant, M. (2016). Twitter says users now trust influencers nearly as much as their friends. Retrieved from http://www.adweek.com/digital/twitter says%20users-now trust-influencers-nearly-much-their-friends-171367/
University of Twente (n.d.). Gevolgen ongezond eten. Retrieved from https://
www.utwente.nl/nl/hr/vg/oud/gezondheid/gezonde-leefwijze/voeding/gevolgen/ Voedingscentrum. (n.d.) Retrieved April 28th, 2019, from https://www.voedingscentrum.nl/
nl.aspx
Wilson, E. J., & Sherrell, D. L. (1993). Source effects in communication and persuasion research: A meta-analysis of effect size. Journal of the Academy of Marketing Science, 21, 101-112.
World Cancer Research Fund, & American Institute for Cancer Research. (2007). Food, nutrition, physical activity, and the prevention of cancer: a global perspective (Vol. 1). American Institute for Cancer Research. Washington DC: AICR.
Wu, P. C., & Wang, Y. C. (2011). The influences of electronic word-of-mouth message appeal and message source credibility on brand attitude. Asia Pacific Journal of Marketing and Logistics, 23, 448-472.
Xenos, M., & Foot, K. (2008). Not your father’s internet: The generation gap in online politics. In W. L. Bennett (Eds.), Civic life online: Learning how digital media can engage youth (pp. 51–70). Cambridge, MA: The MIT Press.
marketing credibility: A heuristic-systematic model. Journal of Media Business Studies, 15, 188-213.
Appendix A - Instagram profiles of micro and macro influencer
Appendix B - Instagram posts in micro influencer condition
