Abstract
Motives for “friending,” following, or connecting with others on social network sites are often positive, but darker motives may also play an important role. A survey with a novel Following Motives Scale (FMS) demonstrates accordingly that positive, sociable motives (i.e., others providing a valued source for humor and information, others sharing a common background, as well as relationship maintenance) and inspirational motives (i.e., others providing a target for upward social comparison) can be distinguished from darker motives related to insecurity (i.e., others providing reassurance, preference for online interaction, mediated voyeurism, as well as social obligation), and even darker antisocial motives related to self-enhancement (i.e., others providing a target for downward social comparison, competition, schadenfreude, gossip, as well as “hate-following”). Results show that lower self-esteem and higher levels of need for popularity, narcissism, and dispositional schadenfreude characterize users with stronger dark side motives, whereas users with more sociable motives report more satisfaction with life, thereby providing construct validity for the novel scale. Convergent validity is demonstrated by positive relations between following motives and both time spent and following counts on different social network sites. Moreover, an embedded experiment shows that antisocial motives predicted acceptance of a Facebook friendship request from a male or female high school acquaintance who suffered a setback in the domain of appearance or status (i.e., a convenient source for self-enhancement), thereby providing additional convergent validity for the Antisocial Motives subscale.
“Thank you for making me realize how little I care about my friends on Facebook.”
One of the defining characteristics of social network sites (SNSs) is the user’s ability to choose who he or she wants to connect with, referred to as “friending” (on Facebook), “connecting” (on LinkedIn), or “following” (on Instagram and Twitter). However, studies addressing specific motives for following others on SNSs are lacking, as scholars have focused predominantly on motivations for general use and typically use measures that do not reflect specific features of SNSs (cf. Smock, Ellison, Lampe, & Wohn, 2011). For example, research from a uses and gratifications perspective (Katz, Blumler, & Gurevitch, 1974) has investigated motivations for general use of MySpace (e.g., Raacke & Bonds-Raacke, 2008), Facebook (e.g., Joinson, 2008; Papacharissi & Mendelson, 2011), Twitter (e.g., Chen, 2011), or various SNSs (e.g., Krishnan & Atkin, 2014). The number, naming, and types of motivations identified for general SNS use vary between these studies, but typically include maintaining existing relationships as well as seeking new ones, entertainment, passing time, information seeking, and self-presentation. Although such motivations may directly or indirectly influence the decision to follow others on SNSs, we argue that they do not fully reflect the wide variety of motives for the specific choice to follow others. For one thing, measures assessing motivations for general SNS use tend to be skewed to the positive and sociable, focusing on harmless enjoyment and strengthening social connections, thereby neglecting the dark side of SNSs in general (cf. Fox & Moreland, 2015), and following behavior in particular (cf. Wang, 2015). This lack of attention for darker motives in past research may partly reflect shifted norms and practices on SNSs over time. Indeed, the presence of darker following motives may help explain why current users tend to maintain relatively expansive social networks. Online social networks are distinct from offline networks in allowing interaction with, and social observation of, a greater number of weak ties (Ellison, Steinfield, & Lampe, 2011). SNSs make these connections possible, but the motives for forming and maintaining them are less clear. The goal of the present research is therefore to identify a wide variety of motives for the specific choice to follow others on SNSs rather than for general SNS use. More specifically, using an online survey with a novel Following Motives Scale (FMS) and an embedded experiment, we attempt to show (a) that positive, sociable motives for following others can be distinguished from darker motives related to social insecurity and from even darker antisocial motives related to a need for self-enhancement, (b) that these motives are related differently to personality traits, (c) that these motives are related to self-reported time spent as well as following counts on SNSs, and (d) that antisocial following motives predict the inclination to accept a hypothetical Facebook friendship request from a high school acquaintance who suffered a setback, thereby providing a convenient source for self-enhancement.
Following Motives: The Good, the Bad, and the Ugly
Our approach to the development of the FMS was to identify a wide range of possible motives from relevant literature that could explain the specific choice to follow others on SNSs rather than general SNS use, thereby allowing for a more detailed understanding of how and why users relate to others on SNSs. These motives are reviewed below and can be grouped into three larger categories expected to yield discrete factors. Specifically, we distinguish between positive, sociable motives for following others (i.e., the good), darker motives related to feelings of social insecurity (i.e., the bad), and even darker antisocial motives related to a need for self-enhancement (i.e., the ugly).
We suggest that positive, sociable motives for following others are closely related to motivations that are commonly identified for general SNS use in studies from a uses and gratifications perspective. As noted earlier, maintaining relationships has often been found to be an important motivation for general SNS use. Scholars have emphasized the social capital provided by followed others (i.e., resources and benefits accrued from network relations) as well as the importance of strengthening weak ties (e.g., Ellison, Steinfield, & Lampe, 2007). Relationship maintenance is therefore likely to be also an important motive for the specific choice to follow others on SNSs. Accordingly, early research into Facebook showed that users chose mostly to follow people who they met offline, or to engage in so-called social browsing (Lampe, Ellison, & Steinfield, 2006). A strongly related motive for following someone on SNSs is that he or she shares a common background or social identity, although this does not necessarily require a pre-established offline relationship. In line with this notion, a study by Barker (2009) demonstrated that communication with peer group members was the most important motivation for general SNS use, especially for people who place a high value on group membership. Moreover, based on the aforementioned finding that entertainment and information seeking are often identified as important motivations for general SNS use, we assume that users are motivated to follow others when they are perceived as valued social sources for humor or information. Finally, consistent with the notion that social comparison processes may play an important role on SNSs (Haferkamp & Krämer, 2011; Vogel, Rose, Roberts, & Eckles, 2014) and that people are often motivated to engage in upward social comparisons with superior others to serve self-improvement goals (Lockwood & Kunda, 1997), we suggest that users can be motivated to follow others on SNSs because they provide a source for inspirational upward social comparisons.
Several studies have alluded to darker motives for following behavior on SNSs that are related to social insecurity. Research has shown that some people feel safer, more efficacious, more confident, and more comfortable with online interpersonal interactions and relationships than with traditional face-to-face activities, a preference that has been linked to problematic Internet use (Caplan, 2002). Accordingly, we suggest that some users may be motivated to follow others on SNSs because they prefer to interact with them online. A related phenomenon is that people may want to observe others—both close friends and more distant acquaintances—from a safe distance by mediated surveillance (Ellison et al., 2011). This behavior has also been characterized as mediated voyeurism, the discrete observation of others’ personal lives (Bumgarner, 2007), and is linked to a higher number of friends on Facebook (Joinson, 2008) and retention of friends on Facebook who are considered seldom contacts (Wang, 2015). Users may thus be motivated to follow others on SNSs to enable them to engage in mediated surveillance or voyeurism. In addition, feelings of insecurity may lead people to seek contact with others who they expect will provide positive feedback on their activities on SNSs. Such reassurance seeking, which has been identified as a form of problematic Facebook use (Clerkin, Smith, & Hames, 2013), may therefore constitute another motive for following others on SNSs. Finally, social insecurity may prompt users to follow others because they do not want to hurt their feelings (i.e., out of social obligation). Indeed, some people may accept others as “friends” on SNSs regardless of closeness, because rejecting or ignoring a friendship request is perceived as risky behavior (boyd, 2008). It should be noted that pursuing motives related to social insecurity (e.g., reassurance seeking) often requires two-way, reciprocal connections (afforded by most SNSs but default on Facebook) to facilitate exchange, whereas one-way connections (as afforded by Twitter and Instagram) are sufficient for pursuing positive motives such as seeking information or inspiration from others. Available options for forming unidirectional and bidirectional links are thus a relevant affordance when considering motives across different SNSs.
Some users may possess even darker antisocial motives for following others on SNSs that are related to a need for self-enhancement. Indeed, people’s need to view themselves positively is regarded by many contemporary psychologists as a primary drive of human behavior (e.g., Baumeister, 1991). Consistent with the notion that people compare themselves with less fortunate others (i.e., make downward social comparisons) to achieve, maintain, or protect a positive self-evaluation via contrast (Wills, 1981), research demonstrates that temporary exposure to SNS user profiles of inferior others increases positive affect (Haferkamp & Krämer, 2011) and self-esteem (Vogel et al., 2014). In addition, a study by Johnson and Knobloch-Westerwick (2014) shows that more negative moods result in less selective exposure to user profiles of upward comparison targets and more exposure to those of downward comparison targets. We therefore assume that some users could be motivated to follow others because they provide an easy target for self-enhancing downward social comparison or opportunities to defeat them in competition. Taking this one step further, people may actually derive pleasure from the misfortunes of others on SNSs or experience so-called schadenfreude. Schadenfreude reactions to media content have also been linked to self-enhancement. For example, research demonstrates that people with chronically low self-esteem, and whose self-evaluation is directly threatened, are more likely to experience schadenfreude when watching a contestant failing miserably in a televised talent show (Van Dijk, Ouwerkerk, Van Koningsbruggen, & Wesseling, 2012). We therefore suggest that users may follow others on SNSs because they provide a target for schadenfreude. Another motive related to self-enhancement is the inclination to follow others so that one can gossip about them with third parties. Although people who gossip tend to be disliked (Farley, 2011), sharing information about others with third parties may elevate their power (Kurland & Pelled, 2000) and social status (Baumeister, Zhang, & Vohs, 2004). Moreover, it has been noted that negative gossip often proceeds from downward social comparison and can be understood as the gossiper’s pursuit of a positive self-view (Wert & Salovey, 2004). Finally, to establish their superiority, users may even want to connect to others who they scorn and dislike. To our knowledge, this phenomenon, referred to as “hate-following” by journalists (Golby, 2014; compare with television “hate-watching”), has not received any scientific attention. Based on the motives outlined above, the first goal of the present research is to develop an FMS to test the hypothesis below:
H1. Positive, sociable motives for following others on SNSs are distinct from motives related to social insecurity and from antisocial motives related to a need for self-enhancement.
Following Motives and Personality Traits
To provide construct validity for the novel FMS, we investigate its relations with several personality traits. We have argued that darker motives for following others are related to social insecurity and a need for self-enhancement. It is therefore reasonable to assume that these following motives are inversely related to indices of social and psychological well-being such as self-esteem and satisfaction with life. By contrast, research indicating that SNS use may increase well-being, as a result of more frequent interactions with friends (Valkenburg, Peter, & Schouten, 2006) and social capital (Ellison et al., 2007), suggests that positive, sociable motives could be associated with higher levels of well-being. In addition, as opposed to positive motives, dark side motives are likely to be related to a strong need for popularity and narcissism. The former refers to a motivation to behave in ways to appear popular (Santor, Messervey, & Kusumakar, 2000), whereas the latter is a dispositional tendency toward grandiose, albeit insecure, self-views, which are maintained by seeking external validation and self-centered behavior (Morf & Rhodewalt, 2001). Indeed, both constructs correlate positively with self-enhancing activities on SNSs, such as strategic self-presentation and profile enhancement (Carpenter, 2012; Utz, Tanis, & Vermeulen, 2012). Moreover, a similar pattern of relations can be expected for the dispositional tendency to experience schadenfreude, which shows a strong link with narcissism (Porter, Bhanwer, Woodworth, & Black, 2014). Other personality traits may have similar relations with different following motives. For example, we have argued that users may follow others on SNSs to enable them to make self-enhancing downward comparisons as well as inspirational upward comparisons, suggesting that the general tendency to make social comparisons (i.e., comparison orientation; Gibbons & Buunk, 1999) should be positively related to both antisocial and positive, sociable following motives. We include the aforementioned personality traits in our study and formulate the hypothesis below:
H2. Antisocial and insecurity motives for following others (i.e., dark side motives) will be positively related to (a) need for popularity, (b) narcissism, and (c) dispositional schadenfreude and will be negatively related to (d) self-esteem and (e) satisfaction with life.
Following Motives, Time Spent on SNSs, and Friending
Because following motives should be related to the intensity of SNS use (i.e., both time spent and following counts on SNSs; Ellison et al., 2007) to demonstrate convergent validity of the novel FMS, we formulate the hypothesis below:
H3. Following motives will be positively related to (a) self-reported time spent on SNSs and (b) self-reported following counts on SNSs.
However, it has been noted that self-report measures may not always accurately reflect actual usage and number of “friends” (Junco, 2013). For example, we have suggested that users with antisocial following motives may have a strong need for popularity. They may therefore over-report their number of “friends” to appear more popular, thereby artificially strengthening the association between antisocial motives and following counts. To provide additional convergent validity, especially for our measurement of antisocial following motives, we therefore include an online experiment in our survey. In this experiment, we present participants with a hypothetical Facebook friendship request from a high school acquaintance who suffered a setback, thereby providing an easy target for self-enhancing downward social comparison, schadenfreude, and gossip. Previous research (Haferkamp & Krämer, 2011) suggests that such self-enhancing benefits on SNSs may differ depending on the gender of the observer in combination with the gender of the comparison target (male vs. female) and the domain of comparison (status vs. appearance). Although this is not the main focus of the present research, we therefore assign participants to one of four experimental conditions in which these two factors are systematically varied:
H4. Stronger antisocial following motives will increase (a) the inclination to accept a hypothetical Facebook friendship request from a male or female high school acquaintance who suffered a setback in the domain of status or appearance, (b) the level of schadenfreude toward this acquaintance, and (c) the intention to gossip about him or her with third parties.
Method
Participants and Procedure
The study was conducted online using Internet-based survey software. The link to the survey (available in Dutch or English) was distributed among the networks of several university students, resulting in a convenient and purposive sample that contained predominantly students, a population known for their intensive use of SNSs. The survey was presented as a study on “social media use and personality” and as an incentive participants could enter a raffle to win a €100 prize by providing their email address upon completion. Given the sensitive nature of some questions, we emphasized that, in accordance with ethical standards of scientific research, participants’ answers could not be linked to their name or email address. During the data collection in May 2015, 438 participants provided consent and started the survey. Only completed surveys of SNS users were analyzed (N = 284), resulting in a sample composed of 156 males and 128 females with a mean age of 29.55 years (standard deviation [SD] = 9.66). An overwhelming majority resided in the Netherlands (81.3%), followed by Germany (10.2%), whereas a small minority resided in other nations (8.5%). The survey contained questions concerning the time spent as well as following counts on different SNSs, statements assessing different motives for following others on SNSs that formed the basis for the development of the FMS, statements measuring personality traits, as well as an embedded experiment testing reactions toward, and acceptance of a hypothetical Facebook friendship request from, a high school acquaintance who was either male or female and had suffered a setback in the domain of status or appearance, thereby providing an easy target for self-enhancement. The survey used a required-response setting. At the end, respondents were thanked for participating and presented with an opportunity to provide comments.
Measures
Unless indicated, measures consisted of statements where participants could respond on 7-point scales in terms of degree of agreement (1 = total disagreement, 7 = total agreement).
Time Spent on SNSs
Participants indicated on five 7-point scales how often (1 = never, 2 = 0–1 hr, 3 = 1–2 hr, 4 = 2–3 hr, 5 = 3–4 hr, 6 = 4–5 hr, 7 = more than 5 hr) they used Facebook (M = 3.64, SD = 1.48), Twitter (M = 1.43, SD = 0.76), LinkedIn (M = 1.79, SD = 0.80), Instagram (M = 2.06, SD = 1.24), or other social media (M = 1.78, SD = 1.15) on average per day. We also created a scale for total time spent on SNSs by summing these scales (M = 10.70, SD = 3.45).
Following Counts on SNSs
If participants indicated that they used Facebook (n = 283), Twitter (n = 94), LinkedIn (n = 198), or Instagram (n = 163), they were asked how many people they followed or befriended on each SNS using 15-point scales ranging from 1 (1–50) to 15 (more than 700) with incremental steps of 50. If participants indicated that they did not use a specific type of SNS, they were assigned the value “0” for the number of people followed on that SNS (MFacebook = 9.20, SDFacebook = 4.08; MTwitter = 1.36, SDTwitter = 2.88; MLinkedIn = 3.30, SDLinkedIn = 3.63; MInstagram = 2.71, SDInstagram = 3.74). We created also a scale for total following count by summing these scales (M = 16.56, SD = 9.91).
FMS
Based on a literature review of following motives on SNSs, as outlined in the introduction, we initially identified 14 possible motives and developed three statements reflecting each motive. Some items were based on measures from studies on preference for online interaction (Caplan, 2002), voyeurism (Nabi, Biely, Morgan, & Stitt, 2003), and relationship maintenance (Lampe et al., 2006). All 42 statements (see Table 1) were preceded by the stem “I follow some people on social media” and presented in mixed order. The items were analyzed by an exploratory factor analysis using maximum likelihood as extraction method and an oblique rotation (promax), as recommended by Costello and Osborne (2005). The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy (.90) indicated that the data were appropriate for analysis. The scree test suggested that four factors should be retained that all had an Eigenvalue greater than 1 (11.59, 5.17, 2.09, and 1.96, respectively) and together accounted for 49.54% of the variance (27.59%, 12.30%, 4.98%, and 4.67%, respectively).
Exploratory Factor Analysis of Items From the Following Motives Scale (Pattern Matrix).
I: antisocial motives; II: insecurity motives; III: sociable motives; IV: inspirational motives.
Only loadings > .25 are depicted.
Item removed from further analyses.
The resulting pattern matrix in Table 1 shows that the first factor clearly represents Antisocial Motives for following others and includes all items related to others being a target for gossip (a1–3), schadenfreude (b1–3), “hate-following” (c1–3), and downward comparison (d1–3), as well as two of the three items related to others being a target for competition (e1–2). The second factor represents Insecurity Motives for following others, and includes all items related to preference for online interaction (f1–3) and reassurance seeking (g1–3), two of the three items related to social obligation (i1–2), as well as one of the mediated surveillance items (h1), and the remaining competition item (e3) that in hindsight may reflect making social comparisons to gain self-knowledge about one’s abilities and thus can be interpreted as being related to insecurity. The third factor reflects Sociable Motives for following others, and includes all items related to others being a valued social source for humor (j1–3) and information (k1–3), others sharing a common identity or being liked (l1–3), and relationship maintenance (m1–3), as well as one of the mediated surveillance items (h3) that in hindsight may not necessarily imply that surveillance takes place without knowledge of the target and thus can be interpreted as being related to relationship maintenance. Finally, all items related to others being a source for inspirational upward social comparisons (n1–3) loaded on a separate fourth factor and will therefore be referred to as Inspirational Motives for following others. The remaining item related to social obligation (i3) did not reach the recommended minimum factor loading of .32 (see Tabachnick & Fidell, 2001), whereas the remaining item related to mediated surveillance (h2) had a relatively low loading on the primary factor combined with a relatively high crossloading. Therefore, these two items were removed from further analyses and scales were constructed with the 40 remaining items representing Antisocial Motives (14 items; M = 2.44, SD = 1.10, α = .93), Insecurity Motives (10 items; M = 2.87, SD = 1.16, α = .88), Sociable Motives (13 items; M = 5.08, SD = 0.89, α = .85), and Inspirational Motives (3 items; M = 3.81, SD = 1.44, α = .83). The resulting dark side scales for Antisocial and Insecurity Motives were strongly correlated (r = .71, p < .001), whereas the scales for Sociable and Inspirational Motives were moderately correlated (r = .33, p < .001). Other correlations ranged from .13 to .27.
Personality Traits
Participants were asked to respond to the 20 statements of Heatherton and Polivy’s (1991) State Self-Esteem Scale. Importantly, when introducing these items, we did not ask participants to respond based on how they felt right now, but rather based on how they perceived themselves and their life in general, thereby reflecting trait rather than state measures of Performance Self-Esteem (7 items; M = 5.40, SD = 0.88, α = .75), Social Self-Esteem (7 items; M = 4.53, SD = 1.09, α = .80), Appearance Self-Esteem (6 items; M = 4.90, SD = 1.09, α = .82), and Total Self-Esteem (20 items; M = 4.94, SD = 0.82, α = .88). In addition, the Satisfaction With Life Scale (Diener, Emmons, Larsen, & Griffin, 1985) was presented (five items; M = 4.87, SD = 1.16, α = .84). Moreover, we selected four items from Santor et al. (2000) to assess Need for Popularity (“At times, I’ve ignored some people in order to be more popular with others”/“I’d do almost anything to avoid being seen as a loser”/“It’s important that people think I’m popular”/“At times, I’ve changed the way I dress in order to be more popular”; M = 2.85, SD = 1.25, α = .79), presented the Single Item Narcissism Scale (M = 2.80, SD = 1.63) developed by Konrath, Meier, and Bushman (2014), and included four items to assess dispositional schadenfreude (“At times it gives me pleasure when I compare myself with others who are worse off”/“I sometimes experience joy when I learn that someone else fails”/“Every now and then I cannot resist a little smile when another person suffers a setback”/“I sometimes think there is some truth in the saying that the best joy is malicious joy”; M = 2.82, SD = 1.43, α = .86). Finally, we presented the short 6-item version of the Iowa–Netherlands Comparison Orientation Measure (M = 3.75, SD = 1.33, α = .86) from Gibbons and Buunk (1999) to assess people’s general tendency to make social comparisons.
Embedded Experiment
To test whether antisocial following motives can predict the inclination to accept a hypothetical Facebook friendship request from a male or female high school acquaintance who suffered a setback in the domain of status or appearance, as well as the level of schadenfreude and intention to gossip about him or her with third parties, participants were asked to imagine themselves in a situation described in a short vignette. In this situation, they were invited to be a friend on Facebook by a high school acquaintance (“X”) who had suffered a setback in a certain domain. In the male comparison target/status setback condition, they read that X, who had excelled in high school, was now very unsuccessful (see Figure 1). Moreover, to make the described situation more realistic, they were presented with a screenshot depicting his latest status update (“Going on vacation tomorrow!”) and a third-party comment (“No job but money for a vacation? Welfare must pay well!”), which was “liked” by three people. In the male comparison target/appearance setback condition, they read that the looks of X, who had been perceived as very attractive in high school, had deteriorated substantially (see Figure 2). As in the previously described condition, participants were presented with a screenshot depicting his latest status update (“ . . . added three photos to the album 2015”) and a comment (“What happened to that good looking boy? Maybe you should find a new personal trainer?”), which was “liked” by three people. To exclude possible confounds, all names appearing in the screenshots were blurred and all profile pictures pixelated. The two other conditions were identical with the exception that X was presented as female rather than male (yet using the same blurred images).
Screenshot of the male comparison target, status setback condition.
Screenshot of the male comparison target, appearance setback condition.
Measures Embedded Experiment
After reading the described situation, participants responded to two statements that served as manipulation checks for perceived appearance of X (“I think that X looks good at the moment”; M = 3.18, SD = 1.39) and perceived status of X (“I think that X has a high status at the moment”; M = 2.79, SD = 1.22). Moreover, two scales used in previous research (e.g., Van Dijk et al., 2012) were presented to assess schadenfreude (e.g., “Thinking about the situation, I would not be able to suppress a little smile”; five items; M = 3.05, SD = 1.60, α = .91) as well as sympathy toward X as a possible contrast measure (e.g., “I would feel sympathy for X”; three items; M = 4.41, SD = 1.52, α = .87). In addition, one statement assessed the intention to gossip about X (“I would be inclined to tell the story to other friends from high school”; M = 4.40, SD = 1.87) and one statement the tendency to accept the friendship request of X (“I would be inclined to accept the friendship request from X”; M = 4.67, SD = 1.87).
Results
Correlations of FMS With Personality Traits
Table 2 (top) shows that, as predicted, people with relatively strong dark side motives for following others on SNSs are characterized by low self-esteem, a strong need for popularity, as well as high levels of narcissism and dispositional schadenfreude. That is, both antisocial motives and insecurity motives are inversely related to performance self-esteem, social self-esteem, total self-esteem, and, albeit to lesser extent, appearance self-esteem. By contrast, sociable motives are unrelated to self-esteem, and inspirational motives show only a relatively weak, albeit significant, negative correlation with performance self-esteem. Moreover, antisocial motives and insecurity motives both show strong positive correlations with need for popularity, narcissism, as well as dispositional schadenfreude. Sociable motives show only a relatively weak, albeit significant, correlation with need for popularity, as do inspirational motives with narcissism and dispositional schadenfreude. A somewhat different pattern of relations was obtained with satisfaction with life. That is, only sociable motives for following others show a positive correlation with satisfaction with life, whereas the other motives have negative, albeit non-significant, correlations. Finally, comparison orientation was positively related to antisocial motives, insecurity motives, sociable motives, and inspirational motives.
Correlations of Personality Traits, Time Spent on SNSs, and Following Counts on SNSs with Dimensions of the Following Motives Scale (FMS).
SNS: social network site.
N = 284.
Correlations of FMS With Time Spent and Following Counts on SNSs
Table 2 (middle) shows that antisocial motives and insecurity motives both have positive correlations with time spent on Facebook and Instagram, as well as with total time spent on SNSs, suggesting that these two dark side motives for following others have a substantial contribution to SNS use. In addition, sociable motives show positive correlations with time spent on Facebook and Twitter, as well as with total time spent on SNSs, whereas inspirational motives have only a positive correlation with time spent on Instagram. Table 2 (bottom) shows that antisocial motives have a positive correlation with number of friends on Facebook, whereas insecurity motives have positive correlations with following counts on Facebook and Instagram, as well as with total following count, thereby demonstrating that dark side motives may indeed play an important role in actual following behavior. In addition, sociable motives show positive correlations with following count on Facebook and Twitter, as well as with total following count, whereas inspirational motives have only a positive correlation with following count on Instagram.
Results Embedded Experiment
The two manipulation checks of the embedded experiment were analyzed by analyses of variance (ANOVAs) with Domain (status vs appearance), Gender Participant (male vs female), and Gender Target (male vs female) as factors. Then the dependent variables were analyzed by multiple regressions on the four following motives, controlling for the main effects of the aforementioned factors (see Table 3 and 4).
Multiple Regressions of Schadenfreude and Sympathy on Domain (Status, Appearance), Gender Target (Male, Female), Gender Participant (Male, Female), and Dimensions of the Following Motives Scale (FMS).
SE: standard error.
N = 284.
Multiple Regressions of Gossip Intention and Friending on Domain (Status, Appearance), Gender Target (Male, Female), Gender Participant (Male, Female), and Dimensions of the Following Motives Scale (FMS).
SE: standard error.
N = 284.
Manipulation Checks
Results show a significant main effect of Domain on perceived status, F(1, 276) = 24.38, p < .001, ηp2 = .08. Consistent with our manipulation, in the status setback condition, the target was perceived as having less status (M = 2.44, SD = 1.20) than in the appearance setback condition (M =3 .12, SD = 1.14). No other effects were significant. Results also show a significant main effect of Domain on perceived appearance, F(1, 276) = 4.69, p = .031, ηp2 = .02. Consistent with our manipulation, in the appearance setback condition, the target was perceived as looking less good (M = 3.01, SD = 1.31) than in the status setback condition (M = 3.37, SD = 1.46). In addition, we obtained a significant interaction effect between Gender Participant and Gender Target, F(1, 276) = 6.00, p = .015, ηp2 = .02. Simple main effects show that female participants perceived a female target as looking better (M = 3.50, SD = 1.18) than a male target (M = 2.89, SD = 1.28), F(1, 276) = 6.18, p = .014, ηp2 = .02, whereas no such difference was obtained for male participants (M = 3.10, SD = 1.53 and M = 3.27, SD = 1.46, respectively), F(1, 276) = 0.80, p = .371, ηp2 = .00.
Schadenfreude
As anticipated, stronger antisocial motives increased schadenfreude felt toward the target described in the experiment. No other motives were significant predictors (see Table 3).
Sympathy
Results in Table 3 show that stronger antisocial motives decreased sympathy reactions, whereas stronger insecurity motives increased sympathy reactions. In addition, we obtained an effect of Gender Participant, indicating that female participants showed more sympathy (M = 4.85, SD = 1.44) than male participants (M = 4.06, SD = 1.49).
Gossip Intention
As expected, stronger antisocial motives increased gossip intention. Moreover, stronger sociable motives also increased gossip intention, whereas stronger inspirational motives decreased gossip intention (see Table 4). In addition, we obtained an effect of Gender Participant, indicating that female participants had a stronger gossip intention (M = 4.73, SD = 1.81) than male participants (M = 4.13, SD = 1.89).
Friending
As predicted, stronger antisocial motives increased the inclination to accept a friendship request. Moreover, results in Table 4 show that stronger sociable motives also increased friending. In addition, we obtained an effect of Domain, indicating a stronger inclination to accept a friendship request in case of an appearance setback (M = 4.84, SD = 1.85) rather than a status setback (M = 4.49, SD = 1.88), as well as an effect of Gender Participant, indicating that female participants were more inclined to friending (M = 4.91, SD = 1.81) than male participants (M = 4.46, SD = 1.90).
Discussion
To provide more insights into following behavior in general, and dark side following in particular, we investigated motives for the specific choice of following others on SNSs rather than motives for general use. We believe that such an approach uniquely contributes to a more detailed understanding of how and why users relate to others on SNSs. By developing a novel scale (FMS), we were able to demonstrate that positive sociable and inspirational motives for following others on SNSs can be distinguished from darker insecurity motives, as well as from even darker antisocial motives related to a need for self-enhancement, thereby supporting H1. However, it should be noted that the mean scores for sociable and inspirational motives were substantially higher than those for insecurity and antisocial motives, suggesting that users possess much stronger positive motives than dark side motives. This difference may reflect a “positivity offset,” a general human tendency to be positively motivated in absence of a clear threat (e.g., Cacioppo, Gardner, & Berntson, 1997). This does not imply, however, that negative motives are not important. They should become more relevant when threats to the self do arise. Indeed, although stronger positive motives may indicate that social networks largely consists of positively evaluated close friends, darker following motives may partially explain why users maintain relatively expansive social networks that also consist of many weak ties. Future research that distinguishes between motives for connecting to close others and weak ties could shed light on this issue. Alternatively, the relatively low scores on darker following motives may have been caused by participants answering in a socially desirable way. However, we have no compelling reason to believe that participants would downplay dark side motives more compared to exaggerating positive motives. Moreover, several participants remarked that they enjoyed completing the survey in an honest manner, despite the fact that some questions were “really bitchy-oriented” and “unexpectedly confronting.”
To provide construct validity for the FMS, we investigated the relations of following motives with several personality traits. Consistent with H2 and our argument that darker motives for following others are related to social insecurity and a need for self-enhancement, the results of the present research show that, as opposed to positive sociable and inspirational motives, insecurity motives and antisocial motives both have strong negative relations with self-esteem. Moreover, our findings show that higher levels of need for popularity, narcissism, and dispositional schadenfreude also characterize users with relatively strong dark side motives. By contrast, and in line with the notion that SNS use may enhance well-being as a result of more frequent interactions with friends (Valkenburg et al., 2006) and social capital (Ellison et al., 2007, 2011), only sociable following motives were positively related to satisfaction with life. Finally, all four motives were associated with the general tendency to make social comparisons (i.e., comparison orientation), thereby emphasizing the importance of social comparison processes on SNSs (cf. Haferkamp & Krämer, 2011; Johnson & Knobloch-Westerwick, 2014; Vogel et al., 2014) and reflecting the wide variety of social comparison motives that have been identified in the literature, including making downward comparisons for self-enhancement (Wills, 1981), making comparisons with similar others to obtain an accurate self-evaluation (Festinger, 1954), and engaging in inspirational upward comparisons for self-improvement (Lockwood & Kunda, 1997).
As predicted by H3, following motives were positively related to self-reported time spent as well as following counts on different SNSs, thereby demonstrating convergent validity of the novel FMS. Moreover, consistent with H4, the results of the embedded experiment show that stronger antisocial following motives increased the inclination to accept a hypothetical Facebook friendship request from a male or female high school acquaintance who suffered a setback in the domain of status or appearance, as well as the level of schadenfreude toward this acquaintance and the intention to gossip about him or her with third parties. In addition, stronger antisocial motives decreased sympathy toward the high school acquaintance. These findings provide additional convergent validity for the FMS in general and the Antisocial Motives subscale in particular. Moreover, they suggest that self-enhancing activities on SNSs are not limited to strategic self-presentation and profile enhancement (e.g., Utz et al., 2012), but may also include strategies that are dependent on connections with specific others, as they can provide convenient sources for self-enhancement activities (e.g., targets for downward comparison or gossip).
Sociable motives were also positively related to the inclination for “friending” the high school acquaintance and the intention to gossip about him or her. The former finding underlines the importance of relationship maintenance (including for weak ties, Ellison et al., 2011) in following behavior, whereas the latter finding is consistent with the notion that gossiping with third parties may not only serve a self-enhancement goal, but is also an important tool for social bonding (Dunbar, 2004). Although stronger antisocial motives increased schadenfreude and decreased sympathy toward the high school acquaintance, insecurity motives were unrelated to schadenfreude and positively related to sympathy. This suggests that, in contrast to true soldiers of misfortune on SNSs that are characterized by even darker antisocial motives, users with strong insecurity motives may be reluctant to express reactions to others’ misfortune that violate social conventions, because they perceive this as risky behavior. Given that participants were only presented with downward comparison targets, we did not expect to find any effects of inspirational motives. However, inspirational motives were inversely related to the intention to gossip about the high school acquaintance, indicating that these motives may not only be associated with a tendency to make upward comparisons, but also to avoid talking about uninspiring downward comparison targets.
When accounting for the effects of following motives, the systematic variations of the comparison domain and comparison target had relatively little effect on the dependent variables of our embedded experiment. Most relevant for the focus of the present research was the finding that friending was more likely following a setback in appearance rather than status. However, additional analyses revealed that this effect was only obtained for female participants. This may reflect the idea that appearance is a more relevant social comparison dimension for females rather than males because of gender-role socialization (Harter, 1998) or evolutionary-based competition (Buss, 1988).
A number of limitations should be highlighted. First of all, the FMS measured motives regarding social media broadly construed. However, specific platforms were considered in terms of usage and network size, and some differential patterns emerged. Future research should examine how affordances of particular social media platforms are able to satisfy the four following motives. Next, given the investigation’s focus on highlighting negative motives, the embedded experiment only employed scenarios depicting a high school acquaintance suffering a setback. However, future research could also use good fortune scenarios as a point of comparison and to further validate the FMS. Also, presenting the FMS and trait scales before the embedded experiment may have sensitized participants in their friend request response. However, asking about a diverse range of motives and traits, both positive and negative, likely mitigated any influence on responses. Future research might also improve on the single-item measure of intention to accept (or initiate) a friendship request, and also use more intensive measures of SNS activity. Moreover, including a Social Desirability Scale (e.g., Strahan & Gerbasi, 1972) could assess whether socially desirable responding influences scores on the different motives.
Conclusion
Our findings provide clear evidence that a variety of “good, bad, and ugly” motives can explain the formation and maintenance of online connections, and that these motives are linked to personality traits and online behavior in predictable ways. However, research on motives for the use of specific features of SNSs requires the development of new measures that follow formal and rigorous methods for scale development and validation. From this perspective, the present research provides only a first, albeit important, step in developing a final measure for following motives on SNSs. In future studies, we will attempt to provide additional validity for the FMS by using random samples rather than the non-random, albeit purposive, sample of the present research and by investigating relations with other individual personality traits. Another goal would be to produce a short-form version without sacrificing the reliability and validity of the subscales (for example, by including only the highest loading item of the 14 different motives we identified initially). We also encourage other researchers to build on the present research by using the FMS in their studies and by identifying other possible motives for following others. By doing so, we should be able to increase our understanding of behavior at the very core of SNSs.
Footnotes
Acknowledgements
We thank Sven Ulrich and Amber Hellmeister for their help with collecting data and creating stimulus material.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.