NOW YOU SEE ME
,
NOW YOU DON
’
T
–
H
OW THE VISIBILITY
OF LIKES AFFECTS CONSUMERS
M
ASTER THESISM
SC
– Communication Science – Persuasive Communication Sophia Weiss Supervisor: Dr. Eva van Reijmersdal Submission: 26. June. 2020, final version Word Count: 7030A
BSTRACTThis study examined the relationship between the visibility of likes on an influencers’ post on Instagram and intention to engage with the post and the consumer attitudes towards the post, in moderation with experience with the social media platform. Engagement with the post includes the intention to like the post and comment on the post. The consumer attitudes include attitudes towards the post, the brand, and the influencer. Results of a one factorial design (visibility of likes: yes or no) with a measured moderator (experience with the social media platform) revealed a significant interaction effect between the visibility of likes, experience with the platform, and attitudes towards the brand and the post.
I
NTRODUCTIONInstagram has risen to be one of the most popular social media apps, with the highest popularity among teens and young adults. The platform has more than 1 billion monthly users and is valued at more than $100 billion. Due to the network’s young users and increasing content, the influence the platform has on peoples’ well-being is increasingly being questioned. More specific questions about the effect of the social media platform on the mental health of its users have arisen. Instagram has been said to affect the mental health of its users through various ways such as cyberbullying or body image issues through online comparisons with peers. To counter some of these issues the CEO, Adam Mosseri stated that, towards the end of 2019, the platform will begin to hide the number of likes of a post (Visca, 2020). This change to the platform already occurred in some ‘test
countries’ such as Australia, Brazil, Canada, Ireland, Italy, Japan, and New Zealand, which began in July of 2019.
Taking away likes on the platform is said to reduce anxiety and social comparisons on the platform which could help the current cyberbullying and body image issues (Shaban, 2019). This change does not completely remove the likes, but users will only be able to see the number of likes on their own posts and not on those posted by other users. The idea behind this change to the platform is that people might feel less self-conscious if their followers cannot see the number of likes they received (Sal19, 2019). Additionally, people have speculated that this change may have been caused by other underlying reasons such as wanting people to post more content, due to the expected lowered social pressure of posting. More content on the platform would mean more time spent on the platform, providing a better channel for advertisement. The idea of removing likes on Instagram has gotten a lot of backlash from influencers on Instagram who claim that this change could lead to less engagement with the posts, which could negatively influence their social media careers and their income (Sal19, 2019). Influencers benefit from the “like momentum” which means that Instagram users might be more inclined to like a post if they can see that a lot of people including their friends liked the post, which they could join by adding to the count. This could therefore not only effect the number of likes a post receives but also, for example, the amount that an influencer could charge a brand for sponsoring their posts.
What causes people to like a post and what effects this can have has been studied in previous literature, although the results that can be found are often inconclusive or even contradictive (Schreiner, Fischer & Riedl, 2019). On the one hand, ' liking' has previously been found to be the most common form of engagement and is seen as less valuable than other forms of engagement on social media. Liking can be characterized by the low effort, meaning that the
user does not have to extensively think about the action (Hellberg, 2015). On the other hand, it has been shown that “liking an image where the 'like' would not necessarily be noticed was found to be pointless”, indicating that an important thought process takes place before 'liking' an image (Hellberg, 2015). There are many factors that have been found to influence the number of likes a post on social media receives. Specifically, factors such as humor and emotion were shown to increase engagement with a post (Lee, Hosanagar & Nair, 2018). Cvijikj and Michahelles (2013) also found the posting time and the media type to affect engagement with the post. The effect of the number of previous likes and comments on a post, on the other hand, is still uncertain. Therefore, The First aim of this study is to understand the effect of visibility of likes on engagement with the post.
Such an increase in engagement with social media content, especially brand content, was found to have a positive effect on brand equity (Coursaris, Osch & Balogh, 2016) and is therefore very important for brands and their survival. Comments and likes can be seen as an indication of not only the number of likes a post receives but also the attitudes towards the post and especially towards the person or the brand represented in that post. Previous findings have already found significant relationships between the number of likes on a post and consumers' attitudes towards the post and the brand (Phua & Ahn, 2016). Phua and Ahn (2016), found that the higher the number of likes on a post, the more likely consumers were to have a more positive brand attitude. Social media users have been found to view their SNSs contacts as an important source of product information (Chu & Kim, 2011), supporting the idea that the number of likes on a post can be a factor determining attitudes towards a post and maybe even the person or the product in the post. Therefore, the second aim is to understand the effect of visibility of likes on the consumer attitudes towards the post and the brand and influencer in the post.
In addition to the visibility of the likes, the present study will also include the prior experience with the social media platform, in this case, Instagram. The variable of prior experience will be included, reflecting the knowledge and fluency of the participants’ use of the platform. Prior experience will be included as a moderator and has previously been found to be a significant influence on social media (Visca, 2020), specifically of engagement intention (Lee & Ma, 2012). Additionally, studies have found that engaging with posts, especially in the form of liking a post can be influenced by prior experience with the platform (Paek, Hove, Jung & Cole, 2013). It is therefore expected that differing experience with the platform causes people to engage with the platform and the posts on it differently. Experience with Instagram could lead to people valuing likes and comments differently, which can affect their engagement with the post. Additionally, it is expected that if the experience with Instagram is high the change to the platform (taking away the visibility of likes) will be noticed more drastically by the participant, which could lead to a change in the way people engage with the posts. Thus, the third aim is to enhance our knowledge of whether prior experience with the social media platform moderates the effect of the visibility of likes on consumer responses, including engagement intentions and attitudes.
This study will add insights to the moderating role of experience with the platform and how this condition plays a role in the effect of visibility of likes on engagement with the post and the resulting consumer attitudes. Thus, adding knowledge to advertising practices in the field of social media. Additionally, this knowledge can be helpful for society, especially when acknowledging the young group of easily influenced consumers on Instagram and the livelihood of influencers on Instagram, which depend on the consumers engagement and attitudes. Studying the visibility of likes is especially important at this point in time since Instagram is changing these features currently, making this a very immediate concern. Only a single study was found that addresses the
visibility of likes on Instagram and the current change to the platform which was published this spring (15.05.2020) by Visca (2020). The study focused on different variables and questioned whether “levels of trait narcissism and peer attachment anxiety in adolescence predict satisfaction with the decision to hide publicly viewable likes on Instagram and authenticity in Instagram use since the implementation of this decision”. Additionally, similar to the present study, the study by Visca (2020) included frequency of use as a variable with significant results of its influence, supporting the inclusion of the same variable in this study
T
HEORETICALF
RAMEWORKVisibility of likes
The effects of likes on social media have been widely studied, although the concept of visibility of likes as an independent variable as presented in this study, has only been studied in scientific literature once before. The article that addresses the visibility of likes by Visca (2020) is a scholarly master thesis and was not published by a scientific journal. Thus, the hypotheses and therefore also the theoretical framework will be based on articles that summarize what other characteristics of social media posts can influence the engagement with the post. The number of likes a post receives on Instagram is a measure of engagement with the photo. It quantifies the number of individual users who pressed the “like button” for the photo. A "like" is a strong social signal on Instagram that demonstrates the extent to which users liked the photo (Phua & Ahn, 2016). “Liking” has often been characterized as being a low-cost activity (Surma, 2016), even though various significant findings have been made on the effects this so-called low-cost activity has. Seemingly small factors of a post have been found to increase or decrease the number of likes a post receives dramatically, such as the gender and the age of the person (Hong, Chen & Li, 2017) or even the
simple existence of a face in the photo (Bakhshi, Shamma & Gilbert, 2014). This leads to the question of whether also the visibility of the number of likes on a post can affect the consumers engagement with the post.
Effect of visibility of likes on engagement
Engagement with the post in this study is operationalized as the number of likes and comments on the post. Comments are seen as a measure of explicit action on the content, which indicates the extent to which users publicly conversed about the photo, and hence it can be considered as a measure of discussion (Phua & Ahn, 2016). The act of liking and commenting precedes purchasing (Barnes, 2015) and is therefore very important in gaining market share in the social commerce space.
Reciprocity theory might help to shed a light on why likes on a post, specifically a high number of likes on a post might increase the consumers' intention to like it. Reciprocity theory suggests that people award kind actions and punish unkind ones. The theory entails that people perceive an action as kind or unkind by evaluating the outcomes of action and the actions’ underlying intentions. (Falk, Armin & Fischbacher, Urs, 2001). This theory can be linked to the concept of “likes for likes” which has been mentioned in scientific literature and is often seen on various social media sites (Dumas, Maxwell-Smith, Davis, & Giulietti, 2017). This behavior, called “likes for likes” means that an Instagram user likes another person’s photo primarily to increase the likelihood that the other person will reciprocate by liking one of the user’s photos in the future. This concept can be linked to reciprocity theory since people reciprocate an action because they perceive it as kind, therefore they want to “give back” and do so with a "like". Giving likes on social media was found to be perceived as a form of gift-giving and can be seen as people's
intentions to help achieve desired impressions (Schlenker & Britt, 1999; Hong, Chen, & Li, 2017). Similarly, Hong, Chen, and Li (2017) found that likes can be perceived as a shortcut for a person to express affective responses to a post, which can be considered as an acquisitive strategy to gain social approval. Therefore, the idea of linking likes to reciprocity theory is not new. Significant findings have already been made, proving that an increase in the number of reciprocity messages a person broadcasts, increases the reciprocity reactions from their audience (Surma, 2016). If the count of likes on a post gets removed it could therefore possibly influence the way that people interact on social media, specifically what actions are perceived as kind and how these can be reciprocated. The evaluation of the action of giving a like might change due to its visibility and could, therefore, influence how this action is reciprocated.
Another explanation for the effects of visibility of likes can be found in the theory of reference groups which postulates that individuals adopt the values expressed by the majority of the members of their reference group (Hyman, 1942). Reference group theory could explain the need of people on social media to engage with a post simply because they can see that their friends or other users (reference group) have engaged with it. A reference group is a group to which an individual or another group is compared, such groups, can be used as a standard for evaluating themselves and their behavior as a reference group. This leads to the belief that such reference groups could have an important impact on the way people engage with content on social media. This can help in explaining the above mentioned “like momentum”, where users might be more inclined to like a post if they can see that a lot of people including their friends previously liked the post, providing the post with the so-called momentum which rapidly increases likes. Both, the reference group theory and the reciprocity theory indicate that the visibility of previous likes on an Instagram post could lead to an increase in engagement with the post, due to wanting to a)
give back a kind gesture and b) be part of the group of people that liked the post. Therefore, it is expected that the visibility of likes will increase the amount of engagement with the presented Instagram post, meaning more likes and more comments.
H1: The visibility of likes on an Instagram post has a positive effect on consumer engagement with the post, that is it will lead to a higher intention to a) like the post and b) comment on it.
Effects of visibility if likes on consumer attitudes
In addition to likes and engagement, consumer attitudes towards the post, and especially towards the person or the brand represented in that post are very important to the brand and influencers active on the platform. Not only has it been suggested that likes can be perceived as a shortcut for a person to express their affective responses to a post (Hong, Chen, and Li, 2017), but previous findings have already found significant relationships between the number of likes on a post and consumer attitudes towards the post and the brand (Phua & Ahn, 2016). Phua and Ahn (2016), found that the higher the number of likes on a post, the more likely consumers were to have a more positive brand attitude. It was suggested that the lack of other social cues on social media leads consumers to use system-generated, numerical aggregates of sociometric and perceptual popularity to form an initial impression of the brand. Overall likes and comments on a post can be considered as such a sociometric measure of popularity (Phua & Ahn, 2016). Previous studies on a similar topic already found that the simple number of followers an influencer has can affect consumer's attitudes towards the person, which De Veirman, Cauberghe, and Hudders (2017) explained by perceptions of popularity. Their findings specifically state that a higher number of followers lead to an increased positive attitude towards the influencer, leading to the belief that the number of
likes on a post could have a similar effect. Additionally, since a positive influence of the number of likes on brand attitudes has been found (Phua & Ahn, 2016) and the post includes not only a brand but also a person, namely the influencer a similar effect is expected for attitudes towards the entire post and attitudes towards the influencer. Thus, the present study will include attitudes towards a) the post, b) the influencer, and c) the brand as dependent variables and will be included as such in the hypotheses.
H2: The visibility of likes on an Instagram post has a positive effect on consumer attitudes toward a) the post, b) the influencer, and c) the brand.
Experience with the Platform
Additionally, experience with the platform will be included as a moderator. As mentioned above the relationship between the visibility of likes and the dependent variables (engagement with the post and attitudes towards the post) will be tested with the moderation of experience with the platform. Prior experience with the social media platform has been found to increase engagement intention, specifically when looking at likes (Lee & Ma, 2012; Paek, Hove, Jung & Cole, 2013). Previous literature, mostly focusing on Facebook, found a significant link between experience with the platform and engagement with posts and even attitudes towards the post and the brand. Phua & Ahn (2016) found the intensity of Facebook use to mediate the relationship between the independent variable “overall likes” and the dependent variables: brand trust, greater brand involvement, and purchase intention. Increased platform use, increases the chance of the consumers having positive opinions about the brand advertised in the post. This can be explained by users who use the platform more being able to better interpret the number of likes and generally
pay better attention to these being displayed. This means that a regular Instagram user will be more likely to notice the number of likes on the post and can compare this number to other posts viewed in the past. This mental comparison could, therefore, lead to a quicker interpretation of the number of likes as being high or low. Thus, a high score for experience with the platform may facilitate the interpretation of the number of likes shown in the stimulus material. Therefore, the previous findings suggest that brands on social media will have more positive feedback and positive results on attitudes towards the brand if their following consists of more frequent users (Phua & Ahn, 2016; Visca, 2020)
Similarly, Lee & Ma (2012) found that social media sharing experience was a significant factor influencing intentions to engage in sharing content on social media. It was suggested that this increase in engagement could be caused by increased confidence in using the platform. This means that the people who use the platform more frequently have acquired the capability or have the confidence to execute such actions (engaging with content). The familiarity with the platform makes these actions habitual and ritualized (Lee, Goh, Razikin, & Chua, 2009; O’Brien, 2010; Diddi & LaRose, 2006), leading to more and more frequent engagement by heavy users. Additionally, the prior experience can help to establish trustworthy relationships between individuals and media use, and further encourage sharing behaviors (Hsu et al., 2007). Therefore, the following hypothesis is formulated:
H3: the positive effect of visibility of likes on a) engagement and b) consumer attitudes toward the post, influencer, and brand, is stronger for people with high experience than for people with low experience.
Figure 1. Conceptual model
M
ETHODOLOGYSample and Design
The participants were sampled through a convenience sample, meaning the sample includes mostly personal social media network connections since people could only be contacted online. The sample consisted of 122 participants with a mean age of 30 years ranging from ages 15 to 92 (M = 30.68, SD = 14.31, 59% female). The highest achieved educational level of the participants' ranged from a high school degree (15.6%) to a Ph.D. degree (3.3%). The majority of the participants completed a bachelor’s degree as their highest education (53.3%). The participants in the sample were composed of 16 different nationalities with a majority of the sample (40.2%) currently residing in the Netherlands. The experiment conducted for this study consisted of a one factorial design (visibility of likes: yes or no) with a measured moderator (experience with the
Visibility of Likes Attitudes towards Brand Engagement with Post Attitudes towards Post Attitudes towards Influencer + + Experience with Platform
social media platform). 53.3% of the participants were in the condition that did not see any likes, compared to 46.7% that did see the likes. The different pictures were also distributed relatively equally among participants with percentages ranging from 30% to 35% for each of the three pictures. Since every picture was represented once for each of the two conditions, this makes for 6 different stimuli which were also relatively equally represented in the data ranging and 12% to 19% of participants per specific condition.
Procedure
The Questionnaire firstly included a letter explaining the following content and a form to ensure the consent of the participants. The first block of questions addressed the participants' demographics, including their age, gender, education, and country of residence. Next, the participants were asked about their experience with the platform Instagram before being shown the randomly assigned stimulus material. The participants were asked to view the Instagram post as if it were in their personal feed and move on to the questions about the dependent variables. This included firstly the participants preferred engagement with the post, and their attitudes towards the post, the brand, and the influencer. Lastly, questions about the participants' familiarity with the brand and the influencer were included to make sure that these do not act as confounding variables. Additionally, a manipulation check was included at the end of the questionnaire, asking the participants if they noticed the overall number of likes for the post presented to them.
Materials
The stimulus material consisted of a post by an influencer, Megan Roup, in collaboration with a brand, meaning that a branded product was visible in the picture, posted by the influencer. The
three pictures chosen to be presented in the questionnaire were all posted by the same influencer. The influencer, that was chosen (Megan Roup) currently has 214.000 followers which is relatively moderate, to decrease the chances of the influencer being well known among the sample so that the attitudes towards the post can be analyzed in relation to the visibility of likes and do not come from previous knowledge about the influencer.
The first picture included as stimulus material presents the influencer, Megan Roup, sitting outdoors in sports clothing, presenting a beverage by the brand “Urbanremedy” which is linked in the caption below and advertised through the comment on the picture as a healthy snack and beverage company. The second post is a picture of the same woman indoors, facing the camera, presenting a face product (Futurist Hydra Rescue Moisturizing Makeup) by “Esteelauder”. The product and the partnership are further explained in the caption of the picture. The last post used for the stimulus material is a picture of the same influencer, facing the camera, indoors in front of a Christmas tree with multiple products by “Naturopathica”. The products in the pictures were subtly placed, and do not make up the center of attention of the pictures. This means that the influencer is still the focus of attention for the posts. The pictures only include the location under the influencers' name and only mention the partnership with the brand in the caption of the picture. Six different stimulus materials were used in the questionnaire, 3 pictures for each of the two conditions. The two conditions include the same three posts, with the only difference being the visibility of the likes, meaning that the pictures in the two conditions are equal. This was operationalized by including the same pictures as stimulus materials with a simple difference in the layout by hiding the number of likes in one condition. The manipulation was done by taking a screenshot of the same post in the Netherlands were likes are currently still visible and in New Zealand, were likes are not visible on Instagram posts anymore.
Measures
Dependent variables:
Engagement with the post was divided into liking and commenting the post, which was measured with one item each, asking the participants how likely he or she would be to like(M = 2.26; SD = 1.58), or comment (M = 1.30; SD =.90) on the stimulus material presented to them, on a scale ranging from 1 (extremely unlikely) to 7 (extremely likely). Consumer attitudes include the participants' attitudes towards the post, the influencer, and the brand. The three dependent variables were measured with the same 5 semantic differential scales (ranging from 1 to 7) (‘unappealing/ appealing’, ‘unpleasant/pleasant’, ‘boring/interesting’, ‘negative/positive’, ‘bad/good’). The five items were averaged to create a scale for attitudes towards the brand (Cronbach’s Alpha = .84; M = 3.83; SD = 5.20; Bruner, 1998), A scale for attitudes towards the post (Cronbach’s Alpha = .84; M = 3.93; SD = 5.85; Bruner, 1998), and lastly, a scale for attitudes towards the influencer (Cronbach’s Alpha = .86; M = 4.24; SD = 5.91; Bruner, 1998),
Moderator:
The participants' experience with the social media platform was measured with 4 items, a combination of measurements from 2 previous studies (Lee, & Ma, 2012; Paek, Hove, Jung & Cole, 2013), which were adapted to fit the hypothesis in this study. The four items measure the duration and frequency of use on a 7-point-Likert scale. For the two questions about the duration, the 7 items range from 1 (none) to 7 (3 hours or more), and for the two questions regarding frequency the 7 items range from 1 (never) to 7 (a few times a day) : (Cronbach’s Alpha = 78; M = 4.56; SD = 4.44).
Control variables:
The control variables included in this study were age, gender, level of education, familiarity with the influencer, and familiarity with the brand. For the level of education, participants could choose their highest level of education from 7 categories ranging from “No schooling completed” to “Ph.D.” (M = 4.71, SD= 1.38). Familiarity with the influencer was measured with two separate dichotomous variables asking the participant if he or she a) knew the influencer (Megan Roup) before this study and b) if he or she already follows the participant on Instagram. None of the participants answered that they did know or follow the influencer and therefore this variable was not included as a covariate in any analysis. Familiarity with the brand was also measured with two dichotomous variables, asking the participants if they a) are familiar with the brand and b) ever used the brand. In this case, 8 participants knew the brand shown to them in the post (7.2%) and 6 participants used the brand before this study (5.4%).
Additionally, a manipulation check was included to see whether the participants in the different conditions actually remember seeing the likes on the post presented to them. The question to test the manipulation asked the participant: “Did you notice the number of overall “likes” on the pictures?”, which the participant could answer with “Yes” or “No”. The test shows that 92% of the people in the condition where the likes were not visible correctly remembered that they did not see the likes. But also, that in the condition where the likes were visible, 74% of the participants say they did not notice the number of likes. Only 26.5% did recall it.
R
ESULTS:
A randomization test was conducted with an ANOVA to check whether the respondents were divided equally between conditions with respect to age and education. Age (F (1,120) = 0.16, p = 0.69) and education (F (1,120) = 0.23, p = 0.63) did not significantly differ between conditions. For the randomization test other (dichotomous) control variables such as gender and familiarity with the brand, a chi-square test was conducted. Age (chi2 (1, N = 122) = 0.02, p = .52) did not significantly vary between conditions and will thus not be included as a covariate. Familiarity with the brand was the only variable that significantly differed between conditions chi2 (1, N = 111) = 6.81, p = .008. Additionally, another ANOVA was conducted to analyze the relationship between familiarity with the brand and the moderator and the dependent variables. Familiarity with the brand is only significantly related to attitude towards the brand and will be included as a covariate in any further model, including attitude towards the brand. Furthermore, for the manipulation check a chi-square test was conducted. As mentioned above only 26.5% of the participants in the condition with likes recall seeing them, chi2(1, N = 111) = 6.87, p = .009 which indicates people
did not pay much attention to the number of likes.
Lastly, before moving on to hypothesis testing the relationship between age and experience with the platform was analyzed with help of the Pearson’s correlation coefficient, which indicated that the two variables are indeed significantly correlated r (120) = -.28, p = .002. This was done due to the users of the platform Instagram consisting of a relatively young population. Thus, the belief arose that young people are more likely to have a higher experience with the platform compared to older participants. The significant negative correlation between age and experience proves this, meaning that the older the participant, the less likely he or she is to have a high score for experience with the platform. Therefore, also age will be included as a covariate in any further analysis including experience with the platform.
Hypothesis 1 testing
To test the first hypothesis, questioning if the visibility of likes affects the intention of the participant to like or comment on the picture, a MAOVA was conducted. The analysis included visibility of likes as the independent variable, likes, and comments as the dependent variables. The results show no significant multivariate effect of the visibility of likes on the intention to like or comment on the post: Wilk's Λ = 1.0, F (2, 118) = .07, p = .94 see Table 1 for means. These results show that H1 is not supported by the data and therefore we cannot conclude that the visibility of likes affects the participants’ intention to like or comment on the Instagram post.
Table 1. Mean values for dependent variables depending on visibility of likes Visibility of likes
no yes
Intention to like the post 2.22a (1.58) 2.32a (1.59) Intention to comment on the post 1.27a (.80) 1.26a (.88)
Attitude toward the post 3.73a (1.10) 4.18b (1.22) Attitude towards the brand 3.66a (1.12) 4.02b (.90) Attitude towards the influencer 4.10a (1.22) 4.42a (1.12)
Note: Means are portrayed with standard deviations between parentheses. a, b means with different
superscripts in the same row differ significantly from each other at p < .05.
For testing the second hypothesis another MANOVA was conducted. The second hypothesis proposed an effect of the visibility of likes on consumer attitudes, which include attitudes towards the post, brand, and influencer. Additionally, familiarity with the brand was included as a covariate. The MANOVA shows a marginally significant multivariate effect of the visibility of the likes, F (3, 106) = 2.10, p = .11; Wilk's Λ = .95 and for the effect of familiarity with the brand F (3, 106) = 3.35, p = .022; Wilk's Λ = .91. The univariate tests show that that there is a significant effect of the visibility of likes for attitude toward the post (F (1) = 4.50, p = .037) and attitudes towards the brand (F (1) = 6.10, p = .016). The effects on attitudes towards the influencer are only marginally significant F (1) = 3.14, p = .079.
Hypothesis 3 testing
To test the third hypothesis the PROCESS macro (Hayes, 2013, model 1) was used. The model that is tested in this hypothesis includes the visibility of likes as the independent variable and all consumer attitudes and engagement as dependent variables, experience with the platform as a moderator, and age as a covariate. Additionally, the analysis that includes attitudes towards the brand as a dependent variable includes familiarity with the brand as a covariate. This, calls for 5 separate regression analyses using PROCESS, one for each dependent variable. The results of the PROCESS analyses showed no significant moderation effects of experience with the platform for likes F (4,117) = 2.0, b = -.10, t = -.37, p =.712, R² = 0.25, comments, F (4,117) = 0.39, b = -0.04, t = -0.24, p = .809, R² = 0.01, attitudes towards the post, F (4,106) = 2.40, b = 0.37, t = 1.79, p = .076, R² = 0.08, and attitudes towards the influencer, F (4,106) = 1.40, b = 0.31, t = 1.43, p = .156, R² = 0.05.
When including attitudes towards the brand as the dependent variable, the interaction effect of visibility and experience with the platform shows significant results: b = 0.50, t (103) = 2.73, p = .008. The PROCESS macro also shows the Slopes for visibility predicting attitudes towards the brand at each level of experience with the platform: For a low level of experience b = 0.04, t (103) = 0.14, p = .090 there is no significant relationship between the visibility of likes and attitude towards the brand. For an average level of experience b = 0.53, t (103) = 2.65, p = .009 there is a significant relationship between the visibility of likes and attitude towards the brand. The same significant effect was found for a high level of experience with the platform b = 0.95, t (103) = 3.68, p < .001. This was expected, since the participants with more experience, as mentioned above, can better interpret the number of likes they see if they do. This means that if a participant is well acquainted with the platform, he or she can make better sense of the likes and can compare that number to other posts on the platform. The Johnson-Neyman significance regions show us that a participant has to score a minimum score of 4.49 on the scale experience with the platform for the direct effect of visibility on attitude towards the brand to be significant, t (103) = 1.98, p = .05, b = 0.4. As experience with the platform increases, the relationship between the visibility of likes and attitudes towards the brand increases with the highest experience with the platform, b = 1.27, t (103) = 3.68, p < .01. Thus, only for people who score 4.49 or higher on experience, there is a positive effect of visibility on attitudes towards the brand.
Table 2. Johnson-Neyman output – Conditional effect of Visibility at different levels of Experience with the Platform
Experience Effect t p
4.50 .40 1.98 .0500
6.25 1.27 3.68 .0004
Note: the values in this table represent the lowest and the highest scores for experience with the
platform, in addition to the score at which the effect of experience with the platform becomes significant.
C
ONCLUSION ANDD
ISCUSSIONThis study examined the effect of the visibility of likes on Instagram on engagement with the post and consumer attitudes. the data collected in the survey does not support the hypothesis, that the visibility of likes increases the participants' engagement with the post through likes and comments. A possible reason why no significant results were found for the first hypothesis is that only a small amount of people in the condition where likes were visible recall it. This assumption would entail that the manipulation in the survey failed or that participants only subconsciously notice the number of likes and that this number does not affect their engagement with the post.
The visibility of likes was proven to significantly influence attitudes towards the brand and attitudes towards the post. Attitudes towards the influencer were only marginally significant. Whether likes are visible on a post or not can influence what the consumer thinks about the brand shown in the picture and how the consumer thinks about the entire post, this influence is positive, meaning that if likes are visible these attitudes will positively increase.This effect was found even though the influencer was the main focus of the picture and not the brand. Most interesting is that as mentioned above, only a small number of people in the condition with likes recall the number of likes or that they were present at all. Still, significant effects on consumer attitudes were found,
the number of likes on a post. The fact that significant findings were found for the effect of the visibility of likes on the consumer attitudes but not for the dependent variables likes and comments is interesting since it was hypothesized that likes are seen as a sort of short-cut to express affective responses to a post (Hong, Chen & Li, 2017). A possible reason for these results could be the layout of the present study. This means that the environment in which the study was done, did not reflect real life, meaning that the software on which the survey was presented was not Instagram itself but simply a survey platform (Qualtrics), with screenshots of the actual platform. This entails that the functions and features were not the same as they would have been on the actual platform. This includes the option of actually liking a post instead of the participant indicating how likely he or she would be to like this post if they were to see it on the actual platform. Choosing a high score for intention to like the post in the survey does not result in the influencer seeing the like in real life, meaning that the effect felt towards the post could not have been transmitted to the influencer or brand. A different study design, allowing the participants to access the actual platform and not just seeing screenshots of posts could lead to different results.
Additionally, even though the reference group theory did not seem to explain the way people engage with the post, by liking and commenting, the same concept could have affected the consumer attitudes. It is possible that by simply seeing the number of likes, knowing that a certain amount of people already liked the content of the post, including the brand and the influencer, this also increases personal attitudes towards the post. Furthermore, this study shows that experience with the platform is a condition for the effects of visibility of likes on brand attitude. The experience with the platform affected brand attitude as expected, namely that the effect was only visible for a moderate and high level of experience, meaning that the more experience the participant has, the more likely their brand attitude is to be positively influenced by the visibility
of likes. Since Age was negatively correlated to experience with the platform it can be concluded that, at least in this sample, younger peoples’ brand attitudes are more likely to be influenced by the visibility of likes. The significant results found for brand attitudes in moderation (or mediation: Phua & Ahn, 2016) with experience with the platform were expected and previously supported by research to be significant for this dependent variable: “Increased platform use, increases the chance of the consumers having positive opinions about the brand advertised in the post” Phua & Ahn (2016). Prior research gave a similar explanation for this, that users with high experience pay better attention to changed details on the platform and can interpret these better.
Theoretical and Practical implications
From a scientific perspective, these findings expand the knowledge in the field of digital marketing communications. It is important to note that these findings were found for Instagram specifically and it is important to further broaden the reach of the conclusion made in this study by focusing on exploring the effect of the visibility of likes on other social media platforms. The novelty of this topic and current change to the platform of Instagram increases the societal relevance and suggests the urgency of further research in this field. From a managerial perspective, these findings allow brands and social media influencers to understand the effects of hiding the likes on Instagram better when entering discussions on why this should or should not be implemented. On one hand, attitudes towards the influencer in the posts were only marginally affected by the visibility of likes and therefore, their fear of getting fewer likes and perhaps less money for their posts is not directly supported. On the other hand, this study does show significant negative effects on brand attitude when likes were hidden, especially for users with high experience with the platform. Since influencers usually get paid by brands to promote their products in their posts, this could indeed
L
IMITATIONSThe stimulus material included pictures, where products in the pictures were subtly placed, and do not make up the center of attention of the pictures. Hence, not finding significant results for the attitudes towards the influencer but finding them for the brand is interesting and raises further questions. A possible reason might be use of the same influencer for all the conditions, whereas different brands were represented in the stimulus materials. It may be beneficial for further research to include a larger variety of influencers, including a range of ages and different genders and ethnicities.As mentioned above the platform which was used for the questionnaire might have been an issue, leading up the presented results. The fact that not all the same interaction possibilities were provided compared to the platform in question, Instagram, definitely lowers the internal validity of the study. I suggest trying to alter this by testing the same variables on the actual platform. This would require access to the actual platform and the data. Instagram itself is doing this kind of testing with different variables and models at the moment by removing the likes only in certain countries as a test. Another limitation of this study lies within the measurements. Experience with the platform is measured with two different concepts, each with two separate questions, duration of use, and frequency of use. These two concepts were put on the same scale which might further reduce the internal validity of the study since this has not been done before and has not been proven to be a valid way of combining the two concepts. Lastly, the relatively small convenience sample, compared to the number and diversity of users on the actual platform leads to a relatively low external validity of the study and further research with larger, more diverse samples with the present rarely studied variable, visibility of likes, is suggested.
R
EFERENCES:
Bakhshi, S., Shamma, D. A., & Gilbert, E. (2014). Faces engage us/ Photos with faces attract more likes and comments on instagram. In Proceedings of the SIGCHI conference on human factors in computing systems. Association of Computing Machinery, pp. 965-974.
Barnes, N. G. (2015). EWOM drives social commerce: A survey of millennials in US and abroad. Journal of Marketing Development and Competitiveness, 9(2), 36.
Chu, S. C., & Kim, Y. (2011). Determinants of consumer engagement in electronic word-of-mouth (eWOM) in social networking sites. International journal of Advertising, 30(1), 47-75. Coursaris, C., Osch, W. and Balogh, B., 2016. Do Facebook Likes Lead to Shares or Sales? Exploring the Empirical Links between Social Media Content, Brand Equity, Purchase Intention, and Engagement. 2016 49th Hawaii International Conference on System Sciences (HICSS), pp.3546-3555.
Cvijikj, I. P., & Michahelles, F. (2013). Online engagement factors on Facebook brand pages. Social Network Analysis and Mining, 3(4), 843-861.
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(5), 798-828.
Dumas, T. M., Maxwell-Smith, M., Davis, J. P., & Giulietti, P. A. (2017). Lying or longing for likes? Narcissism, peer belonging, loneliness and normative versus deceptive like-seeking on Instagram in emerging adulthood. Computers in Human Behavior, 71, 1-10.
Falk, A. and Fischbacher, U., 2006. A theory of reciprocity. Games and Economic Behavior, 54(2).
Fang, Y. H. (2014). Beyond the credibility of electronic word of mouth: Exploring eWOM adoption on social networking sites from affective and curiosity perspectives. International
Journal of Electronic Commerce, 18(3), 67-102.
Hellberg, M., 2015. Visual Brand Communication On Instagram: A Study On Consumer Engagement. Master Thesis. Hanken School of Economics.
Hong, C., Chen, Z. and Li, C., 2017. “Liking” and being “liked”: How are personality traits and demographics associated with giving and receiving “likes” on Facebook?. Computers in Human
Behavior, 68, pp.292-299.
Hsu, M. H., Ju, T. L., Yen, C. H., & Chang, C. M. (2007). Knowledge sharing behavior in virtual communities: The relationship between trust, self-efficacy, and outcome expectations.
International Journal of Human Computer Studies, 65(2), 153–169.
Hyman, H. H. (1942). The psychology of status. Archives of Psychology (Columbia University),
269, 94.
Lee, D., Hosanagar, K., & Nair, H. S. (2018). Advertising content and consumer engagement on social media: Evidence from Facebook. Management Science, 64(11), 5105-5131.
Lee, C. S., Goh, D. H., Razikin, K., & Chua, A. Y. (2009). Tagging, sharing and the influence of personal experience. Journal of Digital Information, 10(1), 1-15.
Lee, C. S., & Ma, L. (2012). News sharing in social media: The effect of gratifications and prior experience. Computers in Human Behavior, 28(2), 331-339.
Paek, H. J., Hove, T., Jung, Y., & Cole, R. T. (2013). Engagement across three social media platforms: An exploratory study of a cause-related PR campaign. Public Relations Review, 39(5), 526-533.
Phua, J., & Ahn, S. J. (2016). Explicating the ‘like’ on Facebook brand pages: The effect of intensity of Facebook use, number of overall ‘likes’, and number of friends' ‘likes’ on consumers' brand outcomes. Journal of Marketing Communications, 22(5), 544-559.
Sal19. (2019, December 09). Facebook has a theory that hiding 'likes' will increase post volume, and Instagram is testing that theory. Retrieved from
https://www.cnbc.com/2019/12/06/instagram-hiding-likes-could-increase-post-volume.html Schlenker, B. R., & Britt, T. W. (1999). Beneficial impression management: Strategically controlling information to help friends. Journal of Personality and Social Psychology, 76(4), 559.
Schreiner, M., Fischer, T., & Riedl, R. (2019). Impact of content characteristics and emotion on behavioral engagement in social media: literature review and research agenda. Electronic
Shaban, H., 2019. Here's why Instagram is going to hide your 'likes'. Washington Post. Surma, J., 2016. Social exchange in online social networks. The reciprocity phenomenon on Facebook. Computer Communications, 73, pp.342-346.
Visca, M., 2020. Narcissism And Attachment Anxiety Predicting The Impact Of Hiding
Instagram “Likes” In Canada. Undergraduate Honours Thesis. Huron University College.
A
PPENDIXVisibility of Likes on Instagram
Start of Block: Introduction LetterQ32 Dear Participant ,With this letter, I would like to invite you to participate in a research study to be conducted under the auspices of the Graduate School of Communication, a part of the University of Amsterdam. The title of the study for which I am requesting your cooperation is ‘Visibility of likes on Instagram – how does this change affect engagement?’. In the online survey, a series of Instagram Posts will be displayed. Each time, you will be asked to choose how and if you would like to engage with the presented post (add a like or a comment). In addition, several questions will be asked about your Social media use, specifically related to Instagram, and your attitudes towards the presented material. Only students over the age of 18 may
participate in this study. The goal of this research is to generate insight into engagement behavior on social media. The study will take about 10 minutes. As this research is being carried out under the responsibility of the ASCoR, University of Amsterdam, we can guarantee that:
1.Your anonymity will be safeguarded, and that your personal information will not be passed on to third parties under any conditions, unless you first give your express permission for this. 2.You can refuse to participate in the research or cut short your participation without having to give a reason for doing so. You also have up to 7 days after participating to withdraw your permission to allow your answers or data to be used in the research.
3. Participating in the research will not entail your being subjected to any appreciable risk or discomfort, the researchers will not deliberately mislead you, and you will not be exposed to any
For more information about the research and the invitation to participate, you are welcome to contact the project leader Sophia Weiss at any time (weisssophia@yahoo.com).Should you have any complaints or comments about the course of the research and the procedures it involves as a consequence of your participation in this research, you can contact the designated member of the Ethics Committee representing ASCoR, at the following address: ASCoR Secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020‐525 3680; ascor‐secr‐fmg@uva.nl.Any complaints or comments will be treated in the strictest confidence. We hope that we have provided you with sufficient information. We would like to take this opportunity to thank you in advance for your assistance with this research, which we greatly appreciate.
Kind regards, Sophia Weiss
End of Block: Introduction Letter
Start of Block: Participation consent
Q30
I hereby declare that I have been informed in a clear manner about the nature and method of the research, as described in the email invitation for this study.I agree, fully and voluntarily, to participate in this research study. With this, I retain the right to withdraw my consent, without having to give a reason for doing so. I am aware that I may halt my participation in the
experiment at any time.If my research results are used in scientific publications or are made public in another way, this will be done such a way that my anonymity is completely safeguarded. My personal data will not be passed on to third parties without my express
permission.If I wish to receive more information about the research, either now or in future, I can contact Sophia Weiss (weisssophia@yahoo.com). Should I have any complaints about this research, I can contact the designated member of the Ethics Committee representing the ASCoR,
at the following address: ASCoR secretariat, Ethics Committee, University of Amsterdam,
Postbus 15793, 1001 NG Amsterdam; 020‐ 525 3680; ascor‐secr‐fmg@uva.nl.
o
I understand the text presented above, and I agree to participate in the research study. (4)o
I do not agree and will not participate (5)End of Block: Participation consent
Start of Block: General Participnt Information Q1 How old are you? (in years)
________________________________________________________________
Q2 What gender do you identify with?
o
Male (1)o
Female (2)o
Other (3)Q3 What is the highest degree or level of school you have completed? if currently enrolled click the education you are following right now
o
No schooling completed (1)o
High school graduate, diploma or the equivalent (2)o
HBO/Higher Professional Education (3)o
MBO/ Secondary vocational education (4)o
Bachelors degree (WO/University) (5)o
Masters Degree (6)o
PHD (7)Q4 In which country are you currently living?
________________________________________________________________ End of Block: General Participnt Information
Start of Block: Experience with the Platform
Q5 On average, how much time do you spend on social media platforms per day?
o
none (1)o
5 minutes (2)o
10 minutes (3)o
30 minutes (4)o
1 hour (5)o
2 hours (6)o
3 hours or more (7)Q6 On average, how frequently do you visit Instagram?”
o
never (1)o
less than once a month (2)o
once a month (3)o
a few times a month (4)o
a few times a week (5)o
once a day (6)o
a few times a day (7)Q7 On average, how much time do you spend on Instagram per day?”
o
none (1)o
5 minutes (2)o
10 minutes (3)o
30 minutes (4)o
1 hour (5)o
2 hours (6)o
3 hours or more (7)Q8 On average, how often do you post on Instagram?
o
never (1)o
less than once a month (2)o
once a month (3)o
a few times a month (4)o
a few times a week (5)o
once a day (6)o
a few times a day (7)End of Block: Experience with the Platform Start of Block: Stimulus introduction
Q27 In the next page you will see an Instagram post of an influencer. Please have a look at this post like you would normally, as if you would when scrolling through your Instagram feed. End of Block: Stimulus introduction
Start of Block: Stimulus material with likes 2 Q34
End of Block: Stimulus material with likes 2 Start of Block: Stimulus material with likes 3 Q35
End of Block: Stimulus material with likes 3 Start of Block: Stimulus material without likes 1 Q36
Start of Block: Stimulus material without likes 2 Q33
Q26
Start of Block: Engagement with post
Q9 How likely are you to click the "like" button for this post?
o
Extremely unlikely (1)o
Moderately unlikely (2)o
Slightly unlikely (3)o
Neither likely nor unlikely (4)o
Slightly likely (5)o
Moderately likely (6)o
Extremely likely (7)Q10 How likely are you to comment on the post?
o
Extremely unlikely (1)o
Moderately unlikely (2)o
Slightly unlikely (3)o
Neither likely nor unlikely (4)o
Slightly likely (5)o
Moderately likely (6)o
Extremely likely (7)Q11 If you were to comment on the picture, what would your comment be?
o
I would not comment (1)o
My comment would be: (2) ________________________________________________ End of Block: Engagement with postStart of Block: Consumer Attitudes
Q28 The next questions are about your opinions about the post. This includes your opinions about the post itself, the brand featured in the post and the influencer.
Q12 In my opinion, the Post is:
1 (1) 2 (2) 3 (3) 4 (4) 5 (5) 6 (6) 7 (7) Unappealing
o
o
o
o
o
o
o
Appealing Unpleasanto
o
o
o
o
o
o
Pleasant Boringo
o
o
o
o
o
o
Interesting Negativeo
o
o
o
o
o
o
Positive Bado
o
o
o
o
o
o
GoodQ13 In my opinion, the Brand shown in the post is: 1 (1) 2 (2) 3 (3) 4 (4) 5 (5) 6 (6) 7 (7) Unappealing
o
o
o
o
o
o
o
Appealing Unpleasanto
o
o
o
o
o
o
Pleasant Boringo
o
o
o
o
o
o
Interesting Negativeo
o
o
o
o
o
o
Positive Bado
o
o
o
o
o
o
GoodQ14 In my opinion, the Influencer shown in the post is:
1 (1) 2 (2) 3 (3) 4 (4) 5 (5) 6 (6) 7 (7) Unappealing
o
o
o
o
o
o
o
Appealing Unpleasanto
o
o
o
o
o
o
Pleasant Boringo
o
o
o
o
o
o
Interesting Negativeo
o
o
o
o
o
o
Positive Bado
o
o
o
o
o
o
GoodEnd of Block: Consumer Attitudes
Q15 The post you saw was from an influencer called Magan Roup. did you know Megan Roup before taking part in this study?
o
Yes (1)o
No (3)Q16 Do you follow the influencer (Megan Roup) on Instagram?
o
Yes (1)o
No (2)Q17 a. Are you familiar with the brand that is being promoted in the picture?
o
Yes (1)o
No (2)Q18 Have you ever used the brand that is being promoted in the post?
o
Yes (1)o
No (2)End of Block: Familiarity with Influencer and brand Start of Block: Manipulation check
Q19 Did you notice the number of overall “likes” on the pictures?
o
Yes (1)Q20 About how many likes did the picture presented to you receive?
