Abstract
This study examines the extent to which the leaders of business schools engage with Twitter to reach diverse audiences, the possible links between Twitter usage and the ranking of the Dean’s respective business college, and the linguistic/stylistic approaches adopted. We employed sentiment analysis to examine the linguistic approaches among the various tweets from the Dean’s account. The findings of the study suggest speaking at stakeholders from a public microblog may not be the most effective way to connect with them. Notwithstanding, biological and cognitive constraints limit the economy of attention and relationships in an online world.
Deans, particularly business school Deans, face the unenviable task of having to communicate across a broad spectrum of audiences with a variety of different interests and concerns (Gioia & Corley, 2002; Kambil & Budnik, 2013). These audiences range from Generation Z high school students making decisions as to where to go to college to Baby Boomer alumni and corporate leaders who are making crucial decisions as to where to invest and/or donate their financial resources (Hawawini, 2005). To complicate matters further, the midrange of the spectrum consists of audiences composed of Millennial undergraduate and graduate students, younger alumni groups, and young faculty members as well as Generation X alumni and midlevel to senior-level faculty groups (Bennis & O’Toole, 2005; DeShields, Kara, & Kaynak, 2005). Each group looks to the Dean as a leader and judges his or her messages from that perspective. And, of course, Deans want and need to connect with these multiple stakeholders (Kambil & Budnik, 2013).
Until a decade ago, the challenge of communicating across these multiple dimensions was ameliorated by the limited number of channels to reach these diverse audiences. Newsletters, annual reports, and, more recently, e-mails were the venues of choice for reaching these different audiences (Davies & Thomas, 2009). But the simplicity of those days is long past. Today, a plethora of new channels ranging from Facebook, LinkedIn, and Snapchat to Slack, Disqus, and Twitter provide a litany of ways that Deans can and are expected to communicate to their diverse and ever growing followers.
Study Overview
At the outset, let us state that this is an exploratory study that will contain more speculation than it will academic certitude. Hair, Wolfinbarger, Money, Samouel, and Page (2015) advocate for the importance and value of such studies when research is in its discovery phase. To this end, Blanche, Durrheim, and Painter (2006) suggest that exploratory studies should be designed as open and flexible investigations which seek to look for new, preliminary insights into phenomena. We have designed our analysis to follow in that tradition.
The specific purpose of this exploratory study is threefold. First, we want to examine the extent to which the leaders of business schools engage with a particular social media platform, Twitter, to reach diverse audiences. We chose Twitter both because of its ubiquitous coverage—a reported 320 million active monthly users at the end of 2015, with 1 billion unique monthly visitors to the site (Twitter, 2016)—and because of its already well-established history as a venue for academic study (Zimmer & Proferes, 2014). Second, once we made a determination of the leaders’ Twitter Engagement, we wanted to explore for possible links between the Twitter usage from their office and the ranking of the Dean’s respective business college. We realized at the outset that any such findings are speculative, yet we thought the possibility of the existence of such links worthy of exploration. Finally, we wanted to explore the linguistic/stylistic approaches that business schools use on Twitter. More specifically, we wanted to examine the leader/language approaches that business schools use within this channel and sought to determine the manner in which they do so. We used the linguistic approach of Sentiment Analysis, which we will define in more detail shortly, to examine the linguistic approaches among the various tweets from the Dean’s account.
Background
Recent academic research has advanced the study of social media in general and Twitter in particular toward a perspective of a being a leadership tool. Numerous scholars (e.g., Honey & Herring, 2009; Krishnamurthy, Gill, & Arlitt, 2008; Zhao & Rosson, 2009) see social media and especially Twitter as a part of a leader’s conversational repertoire. Others see social media and Twitter as a means to convey positional power (Smith & Place, 2013), credibility and perceptions (Schmierbach & Oeldorf-Hirsch, 2012), and to create stakeholder engagement (Rybalko & Seltzer, 2010). Building on these ideas, Kawasaki (2015) views social media overall as a means of conversational—or, as he labels it “evangelical”—career advancement. Specifically, Kawasaki advocates using Twitter and other tools of social media as a way to gain visibility for one’s self, one’s products, and one’s company.
A particularly unique feature of Twitter that has contributed to its rapid adoption rate involves its power of immediacy. Twitter is the “go to” platform when a person needs to have a message disseminated and read immediately in a public forum (Glaser, 2016). This feature of Twitter has lent itself to a number of particularly relevant applications. Twitter has thus become a communication staple in crisis situations, ranging from the Japanese tsunami (Acar & Muraki, 2011) to the Boston Marathon bombing (Buntain, Golbeck, Liu, & LaFree, 2016). Additionally, Twitter has come to play a particularly vital role in ongoing political conversations ranging from its present usage by U.S. political candidates to its assistance in helping overthrow governments during the Arab spring (Harlow & Johnson, 2011; Hermida, Lewis, & Zamith, 2014; Howard & Hussain, 2013). In short, Twitter’s role as a leadership force cannot be overlooked.
Twitter’s pragmatic influence on the domain of microblogging also needs to be acknowledged. Stated simply, Twitter may be at the forefront of a reconfiguration of the way microblogs are seen and used. Historically, microblogs have provided a forum for two-way communication (Venters, Green, & Lopez, 2012). Lovejoy, Waters, and Saxton (2012), however, analyzed more than 4,500 tweets from the nation’s largest nonprofits and found the platform being used more than 80% of the time as a one-way communication channel. Hence, the channel seems to be transitioning from a forum to establish dialogs or share alternative views to one that trumpets organizational causes and specific promotions. In a related, but similar vein, L. V. Porter and Sallot (2005) found public relations practitioners used online engagement as a means of enhancing personal organizational power and status. Each of these studies implies some contradiction to Twitter’s stated mission, “To give everyone the power to create and share ideas and information instantly, without barriers” (Twitter, 2016).
Paralleling this shift in microblogging is another shift that we choose to call unbridled optimism. Stated simply, this shift means that as social networks continue to proliferate and grow they begin to be viewed as invulnerable. Rapid increase in adoption equates to increase in growth, which correlates to increased value and usefulness (Choudary, 2014; Stewart, 2012). This ideology is not unique, finding consistency with Metcalfe’s Law which suggests that as a network scales, the expectation is that the communication network becomes more useful and valuable for users because of positive network externality or positive network effects (Gilder, 1993; Metcalfe, 1996). Taken from the perspective of a business school Dean, as we shall see shortly, the perception develops that these microblogs are “here to stay” and each respective business school must have a dominating presence on them.
It is important to add, however, that an opposing view also exists. This view, stated theoretically, suggests that network effects may also work in the reverse direction. In other words, imitating Yogi Berra’s famous dictum “Nobody goes there anymore. It’s too crowded,” the overriding success of microblogs such as Twitter may lead to users abandoning the network in large numbers and leaving it defunct (Choudary, 2014). Social media platforms like Myspace and Piczo bear testimony to this tendency. So, as enrollment on Twitter continues to increase, looming concerns about sustainability and reduced benefit from excessive growth may emerge from certain key stakeholders (Harvard Business School, 2015). Again, taken from the perspective of business school Deans, reverse network effects may compromise the efficacy of the platform as an effective leadership communication tool.
Method
Study Design
As previously mentioned, our analysis of Twitter usage by business school Deans focuses on three specific areas: Twitter Engagement, Twitter/Business School Rankings, and Tweet Sentiment Analysis. The first area, Engagement, is examined within a descriptive framework. Deans’ tweet usage is initially characterized through metrics designed primarily by M. C. Porter, Anderson, and Nhotsavang (2015), with descriptive statistics used to interpret the basic features of the data. Deans’ use of Twitter is thus computed on a 5-point scale ranging from inactive to very active based on the total number of tweets made during the 5-year time period analyzed. Then, this measure, along with Retweets, Replies, and Follower numbers, are analyzed using descriptive statistics. Various scholars (e.g., Aichner & Jacob, 2015; Hoffman & Fodor, 2010; Peters, Chen, Kaplan, Ognibeni, & Pauwels, 2013) view Retweets, Replies, and Follower numbers as important usage descriptors to gather.
The second part of this study compares the Dean’s level of engagement with the ranking of the Dean’s respective graduate business school. Rankings from the U.S. News & World Report were used to identify the placement of the graduate business schools. We chose this listing and this graduate school perspective because of the multiple measures used to rank the colleges along with the prominence and influence of this source (Peters, 2007; Sauder & Fine, 2008). Stated simply, a broad array of university Presidents and Deans aspire strategically to getting the best possible ranking in the U.S. News & World Report. An improving position in this ranking helps both for recruiting future students and signify prominence and organizational effectiveness (Dowdall, 2009). Correlation methods were employed to measure the relationship between the respective rankings and engagement levels.
The third and final section of the study compares the content of the tweets from the perspective of Sentiment Analysis. This linguistic approach uses text analysis and research on natural language processing to reveal descriptors of a writer or a speaker’s attitude (Nasukawa & Yi, 2003). While this approach has already been used on Twitter to predict the outcomes of political elections (Tumasjan, Sprenger, Sandner, & Welpe, 2010), we will use it from the perspectives of a leader’s projected persona as well as his or her affect.
At the outset, it is important to reemphasize that the focus for the first and third parts of this study is largely descriptive. As an exploratory study, we primarily want to see how Twitter is being used at the @Deans addresses. Numerous scholars (see Ang & Van Dyne, 2015; Liu, Parelius, & Singh, 1999; Ostle, 1963) support the validity of this research approach. Likewise, because of this exploratory/descriptive focus, we have refrained from ranking Twitter usage and thereby avoid implying that one approach is better than or more effective than another.
Sample and Sampling Technique
The study set out to explore the Twitter communication patterns of a specific population of interest: Deans of 14 business colleges within the Southeastern Conference (SEC). The decision to limit the sample to tweets of Deans within this prescribed academic domain was essential to the exploratory nature of the study. This smaller, more manageable data set enabled a deeper, more granular examination of overall usage and patterns of engagement.
The purposive sample consisted of the most recent 3,200 tweets from individuals holding a “@Dean” twitter profile within the respective business colleges. All 14 SEC business colleges have such accounts.
The intention, at the outset, was to study tweets from the perspective of positions, not individuals. In other words, the study focuses on tweets sent from these @Dean’s accounts. We acknowledge at the outset that more than one person may have been sending tweets from that account. But, again, our intention from the outset was not to study tweets sent by an individual. Instead, we wanted to study tweets that came from particular accounts, that is, Dean’s—at specific business schools within the SEC. By focusing solely on the “@Dean” address, we collected tweets that held the imprimatur of the Dean and were therefore ultimately associated with that position as well as that of the particular college and university that the Dean represents. Whether the actual creator of the tweet holds the position of Dean, Dean’s assistant, or student worker is immaterial to this study. From our perspective, the tweet represents the office of the Dean.
For informational purposes, we did a background analysis of the people holding the position of Dean during the timeframe studied. This analysis showed that the person holding the Dean’s position changed in 7 of the 14 schools during the time frame studied (from January 1, 2011 to March 9, 2016). A listing of position changes that occurred during the study period is available on request. The listing is not provided here in order to protect the anonymity of the individual schools studied.
Data Collection
The Twitter GET statuses/user timeline Application Programming Interface (API) was used to retrieve data. Requests to the APIs contain parameters that include, but are not limited to, hashtags, keywords, and Twitter user IDs. Twitter APIs can be accessed only via authenticated requests, and access is limited to a specific number of requests within a time window identified as the rate limit. For this reason, our data sources consist of the most recent 3,200 tweets from each Dean’s account. Veletsianos (2012) investigated higher education scholars’ participation and practices on Twitter by examining only 100 tweets from each individual and found that sample size quickly approaching data saturation. We conducted the analysis after exporting the tweets to a Microsoft Excel document.
Data Analysis
Overall Engagement Statistics
Tweets were aggregated and analyzed to evaluate overall engagement on Twitter. Descriptive statistics were used to summarize the data into the following categories: Tweets, Replies, Retweets, and Followers. To assess each Dean’s relative level of engagement, a 5-point scale ranging from 1 (inactive) to 5 (very active) was employed. The assessment scale was derived from metrics designed by M. C. Porter et al. (2015), and incorporated the descriptive statistics as contributing assessment measures.
Sentiment Analysis
To help us understand the overall sentiment or impressions conveyed in the respective @Dean’s twitter updates, and to make replicable and valid inferences, LIWC2015 (Linguistic Inquiry and Word Count), a text analysis software program was employed. This program, which is used extensively in psychology and linguistics studies, was developed to measure emotional, cognitive, and structural components of text samples. This program uses a psychometrically validated internal dictionary to conduct the analysis. Specifically, LIWC determines the rate at which elements such as emotions (positive vs. negative), cognitions (a present vs. a future or past orientation) exist in the text (Tumasjan et al., 2010), and expresses this as a percentage of the total words used. LIWC has been rigorously tested for content and construct validity; intercoder reliability has been found to range from 86% to 100%, depending on the dimension being assessed (Francis & Pennebaker, 1992; Pennebaker & Graybeal, 2001).
Results and Preliminary Findings
Part I: Engagement
Table 1 shows the results of our tweet collection from January 1, 2011 until March 9, 2016. The Total Tweets category displays the total count of tweets posted on each Twitter account. A tweet is defined as any message posted to Twitter which may contain photos, videos, links, and up to 140 characters of text. As a point of clarification, while the Twitter API only provides access to the most recent 3,200 tweets, the individual Twitter accounts identify the total number of tweets initiated by the user(s) from the inception of the account. This accounts for the variance between the total number of tweets and the 3,200 limit. The Retweets category displays the percentage of Retweets posted on each Dean’s Twitter account. A Retweet is defined as a reposting of someone else’s tweet. The Replies category displays the percentage of replies posted on each @Dean’s Twitter account. A reply is defined as a response to another user’s tweet that begins with the @username of the person being replied to. The Followers category displays the total count of users following each @Dean’s Twitter account.
Overall Engagement.
1 = inactive; 5 = very active.
Preliminary Findings
Engagement levels vary among the different @Deans accounts. According to M. C. Porter et al.’s (2015) scale, four @Deans accounts (29%) qualify as being very active on Twitter. Meanwhile, seven other accounts (50%) are relatively inactive by comparison; the remaining three (21%) engage moderately.
Overall, Deans tweet relatively moderately (M = 2.93, SD = 1.49). Interestingly, when the results are compared with Twitter Engagement data for chief executive officers (CEOs) of the Top 10 corporations in the Fortune 500 list, some similarities exist. Only four of these CEOs engage on Twitter, and activity is relatively weak overall. Surprisingly, the average tweet count for Deans from the 14 SEC schools far outweighs the average tweet count for the CEOs of the Top 10 corporations in the Fortune 500 list (3,000 vs. 200, respectively). This latter finding is particularly interesting since large corporations have established dedicated social media groups to monitor and respond to events related to the firm. These responses arrive under the CEO’s imprimatur.
Part II: Engagement/Ranking Comparison
Pearson product-moment correlation coefficients were computed to assess the relationship between Deans’ level of engagement and change in ranking of their respective college’s graduate business school. The change in ranking represented the difference between the current year’s rank and the prior year’s rank. An improvement in rank was shown as a positive number, while a decline in rank was shown as a negative number. The correlation coefficient was computed for each year from 2012 to 2016 because rankings are based on data from the previous year. The results of the analysis are shown in Table 2.
Engagement/Ranking Comparison: Correlation Coefficients and p Values (N = 14).
Preliminary Findings
For all three variables, a positive association is observed for 2012 and 2013, and a negative association is observed for 2014, 2015, and 2016. Stronger associations manifest in 2012 and 2014, and generally weak associations manifest in the remaining years. In only three instances, however, statistically significant relationships are observed:
Strong positive relationship between engagement level and change in rank for 2012 (R = .675, p = .02)
Moderate to strong negative association between engagement level and change in rank for 2014 (R = −.558, p = .05)
Moderate to strong positive association between total tweets and change in rank for 2012 (R = .598, p = .05)
Part III: Sentiment Analysis
The Sentiment Analysis consists of two distinct sections. First, we will analyze the content of the @Deans tweets from a perspective we have chosen to call the Perceived Leader Communication Trait (PLCT). An explanation of how we derived the content of the PLCT follows at the beginning of that section. Second, we will take a granular linguistic look at three specific descriptors contained with the @Deans tweets. These descriptors are Function Words, Affect, and Time Orientation.
Perceived Leader Communication Trait
A significant portion of a Dean’s job involves communication (Martin, 1993). And as we noted earlier, today’s Dean has the challenge of having to communicate with multiple audiences through a variety of different channels. Yet a consistent theme that runs through every communication channel is the need to convey a message in a leader-like manner. Stated more academically, and consistent with Goffman’s (1959) observation that forms the foundation for Impression Management theory, leaders and organizations must establish and nurture an image that is consistent with the perception they want to convey to their stakeholders. Additionally, the sentiment that Dean’s express through linguistic and textual preferences shapes an audience’s perception both of the communicator and of his or her affiliates (McCroskey & Richmond, 1976). Or stated more simply, the language Deans use to tweet may determine the perceptions others have of their effectiveness as leaders.
To explore these concepts in more detail, we used LIWC to examine four specific communication traits within the content of the respective @Dean’s tweets. The four categories are Analytical Thinking, Clout, Authenticity, and Emotional Tone. We collectively labeled these four characteristics as a PLCT. Table 3 shows the results (expressed this as a percentage of the total words used) generated by LIWC for the four summary categories.
Perceived Leader Communication Traits (PLCT) a .
Range: 1 = extremely low; 50 = average; 100 = extremely high.
Preliminary Findings
Analytical thinking is the backbone of modern business education (Snyder & Snyder, 2008). Yet those familiar with Twitter argue that the venue lends itself far more toward impulsive messages (Patriarche, Bilandzic, Jensen, & Jurišić, 2013). Little evidence exists arguing that this channel serves as the primary mode for conveying a formal, logical ordered thinking style.
Yet despite the aforementioned arguments, the 14 @Dean accounts display a remarkably consistent and high level of analytical thinking. If impression management is a conscious factor in the @Dean’s tweets, then the people composing messages for these accounts consistently want to portray the Deans as measured, rational leaders.
Clout, which can also be classified as power, involves portraying messages with a sense of certainty and confidence. Power and leadership are closely tied to both of these traits (Pfeffer, 2013). Additionally, power is a central feature of business discourse (Bargiela-Chiappini, Nickerson, & Planken, 2013) and a means through which leaders demonstrate authority and control (Fiedler & Chemers, 1974; Pfeffer & Drummond, 2010).
Although not as pronounced as the analytical category, the @Deans ranking in clout was still more than 25 points above the national average. Additionally, only two @Deans accounts scored below 70 and none went below 61. It seems apparent, then, that the tweets from the @Deans accounts project authority, assurance, and a high level of subject matter expertise.
Similar to power, authenticity has become an extremely relevant topic in the conversation about leadership communication (Ilies, Morgeson, & Nahrgang, 2005). The broad consensus is that authenticity, or at least the appearance of authenticity, gives a leader a competitive edge in terms of leadership persona as well as interpersonal connections (Cuddy, Kohut, & Neffinger, 2013; Ibarra, 2015; Liedtka, 2008: Morgan, 2008).
The aforementioned importance and relevance of authenticity seems contradicted on the behavioral level in the @Deans accounts. The comparatively low scores recorded within these tweets indicate a measured, guarded, perhaps even a distant approach to tweeting. This approach certainly contradicts the observations of various scholars (e.g., Darics, 2015; Edelson, Kim, Scott, & Szendrey, 2015; Girginova, 2015), each of whom examine tweets from a more creative and innovative perspective. It seems apparent that the tweets from @Dean’s accounts are managing language in a way that seems consistent with the perceived professionalism of the office.
The final PLCT category, Emotional Tone, detects whether the linguistic tendencies are positive and upbeat, neutral, or filled with anxiety, sadness, or anger (Pennebaker, Boyd, Jordan, & Blackburn, 2015). A highly positive rating in Emotional Tone aligns closely with business communication’s long held tenets of positive psychology (Williams & Griffin, 1966), attitude (O’Rourke, 2010), and win-win scenario (Hynes, 2015). With one exception, all of the @Deans accounts score above the midpoint on emotional tone.
In summary, the PLCT dimension demonstrates an almost surprising similarity in the @Deans accounts. From the perspective of these findings, one can say that a typical tweet from these accounts is formal, logical, and guarded. Additionally, these tweets attempt to engender perceptions of authority and a positive emotional tone. At the same time, however, the low authenticity score calls into question the legitimacy of each of these perceptions. Interestingly, the personas being projected in the @Deans accounts reinforce Goffman’s (1959) observation that the pressure of idealized conduct affects the presentation of self in everyday life. He argues that interactions are actually “performances” shaped by the environment and the audience to project certain character traits and to provide society with “impressions” that are consistent with the intended goals of the actor. The overall Twitter data summary suggests, from a leadership perspective, that these @Dean’s accounts may be functioning as such actors.
Descriptor Analysis
LIWC allows a deeper, microscopic look at a variety of linguistic constructs. For purposes of this exploratory review, we chose to explicate three dimensions that are particularly apt to tweets from the @Dean domain: function words, affect, and time orientation.
Function Words (Personal Pronouns)
Table 4 displays function words as a percentage of the total words captured from each Twitter account.
Function Words.
Sensitivity to pronoun usage has been a stalwart of business communication since its inception (Saunders, 1936). Initially, this interest led to an awareness of the ways that pronoun usage added or reduced social distance, with the overriding emphasis being in favor of using the second person pronoun to reduce social distance (R. Brown & Gilman, 1960). Subsequently, a number of studies began to call into question the unbridled advocacy of “second” person pronouns (Brockman & Belanger, 1993; Campbell, Riley, & Parker, 1990; Shelby & Reinsch, 2003), which eventually led to reframing the discussion into writer-reader relationships in business prose (Ewald & Vann, 2003; Jameson, 2004a, 2004b) along with aforementioned power relationships measured by position (Shelby & Reinsch, 2003; Rogers & Lee-Wong, 2003; Rogers, Ho, Thomas, Wong, & Cheng, 2004).
Numerous scholars (e.g., Fitzsimons & Kay, 2004; Gortner & Pennebaker, 2003; Mandel & Vassallo, 1999) returned the argument to its “pronoun roots” by suggesting that the use of “we” is a marker of affiliation that reduces social distance, while the use of “she and he” can increase social distance. Meanwhile, Thayer, Evans, McBride, Queen, and Spyridakis (2010) shifted the pronoun perspective to online content and saw the first person pronoun, “I,” shifting toward informality in electronic venue. Finally, building on the core of almost all of these arguments, Kacewicz, Pennebaker, Davis, Jeon, and Graesser (2013) extended the discussion into a frame of self- versus other-orientation of low- and high-status group members. Thus, drawing on previous research which found that high-status individuals are likely to be more collectively oriented and externally focused, Kacewicz et al. (2013) posit that high-status individuals are partial to the use of “we” rather than “I.”
Table 4 provides interesting fodder for these discussions. Although variance exists between different @Dean accounts, “we” is used 40% more often than “I.” Additionally, “we” is used 20% more often than “you.”
The limited usage of “she,” “he,” and “they” may have to do either with an attempt to avoid social distance or, perhaps more likely, the current social risk associated with these words and their gender designations (Tamburin, 2015).
The heavy use of numbers results from attempts to project certitude—similar to the aforementioned findings in analytical emphasis—along with efforts to document the specific accomplishments of the respective schools.
Overall, Table 4 seems to demonstrate that the tweets from @Deans accounts attempt to steer clear from self-orientation while shooting straight at documentation. The content emphasis is thus “us,” the school, not “me” the Dean. The proof is in the numbers.
Affect
Table 5 displays affect-related words as a percentage of the total words captured from each @Dean Twitter account.
Affect.
No matter whether it is framed as prosocial organizational behaviors (Bagozzi & Moore, 1994; Brief & Motowidlo, 1986; George & Bettenhausen, 1990; Kelley & Hoffman, 1997; N. Malhotra & Ackfeldt, 2016), emotional intelligence (Edelson et al., 2015; Holt, 2015; Sigmar, Hynes, & Hill, 2012; Wells & Dennis, 2016), organizational citizenship behavior (L. A. Brown, 2015; Carter, McFadden-Wade, & Wells, 2016; Karimikia, Singh, & Olesen, 2015), or a positive attitude (Campbell, White, & Johnson, 2003; Menning, Wilkinson, & Clarke, 1976; Shelby & Reinsch, 2003), there is a broad consensus that positive and negative emotions have different affects on message recipients but an overall agreement that in the vast majority of situations positive emotions have a more favorable affect on individuals and their organizations than do negative emotions. Developing these observations, Elliot and Thrash (2002) found that positive emotions promote resource building and collective involvement in goals. Earlier, Fredrickson (2001) posited that positive emotions help prepare an organization for future challenges and threats. Finally, Lyubomirsky, King, and Diener (2005) found that when an individual’s communication demonstrates positive emotions, positive outcomes emerge.
Not surprisingly, and consistent with the aforementioned literature, the affect findings of @Deans accounts show a 10 to 1 ratio of words conveying positive emotions over those conveying negative emotions. It seems readily apparent that these accounts are being managed to project a positive image.
Time Orientation
Table 6 displays the time orientation of the respective @Dean accounts:
Time Orientation.
Time orientation is an indicator of the author’s psychological focus on a particular period. LIWC categorizes the time focus as past, present, or future. If an individual usually writes about past experiences, his or her time orientation is considered “past.” Similarly, based on his or her focus on the present events or the events that would occur in the future, time orientation is considered “present” or “future.” Kouzes and Posner (2012) suggest that leaders tend to lean toward the future in much of their communication. The idea behind this premise is that leaders are forward looking and are therefore often describing greater opportunities that are yet to come.
The data from the @Deans accounts contradict the future orientation suggested by Kouzes and Posner (2012). These accounts focus heavy attention on present events, with almost a 6:1 ratio of present to past emphasis and a 4.5:1 ratio of present to future emphasis. The preference for future events over those that occurred in the past is 1.33 to 1.
This change in temporal emphasis might be an indication of the immediate concerns of Deans over other leaders; or, more likely, we believe, this change might be indicative of the emphasis that Twitter engenders, a presence on the here and now, not the past or the future. Hence, this finding provides additional support for aforementioned aspect of Twitter’s “power of immediacy.” From a leadership perspective, this emphasis on the present might provide an interesting insight as to when Twitter functions best for a leader’s agenda.
Findings
The subtitle to this study was chosen intentionally: Are Deans Doing It? The most accurate answers to that question based on the findings of this study may be “sort of” or “conditionally so.” In essence, this study finds that SEC business schools engage relatively moderately on Twitter through the @Deans accounts.
The question the title did not ask but meant to imply should be asked is as follows: How Are Deans Doing It? Even a cursory look at this study’s linguistic findings provides an interesting answer to that question. In essence, each @Dean’s account uses essentially the same Twitter approach to its audience. Stated in modern parlance, the @Deans tweets are “nondiverse.” Each is heavily analytical, conveys extreme power, and is basically artificial or inauthentic. Additionally, the tweets emphasize “we” more than “you,” and prefer both of these descriptors to “I.” And, of course, each @Dean’s account talks positively about the present. It would be inappropriate to label these observations as a formula for success; but we certainly can substantiate that all @Dean’s accounts are “doing it” in the same way. One should not miss the irony of using this similar nondiverse approach in a venue that was initially touted as an innovative way to communicate.
The third and final question, which is directly implied in the title, is the following: Are Deans Leading by Tweeting? The answers to this question are complex but highly interesting. Scientific inquiry into the role of Twitter Engagement on organizational performance outcomes, and as a way to create competitive advantage, has gained increased importance and relevance in recent years. For example, Bollen, Mao, and Zeng (2010) found that Twitter feeds were correlated to the value of the Dow Jones Industrial Average. Zhang, Fuehres, and Gloor (2011) analyzed tweets as a predictor of stock market indices such as Dow Jones, NASDAQ, and S&P 500, and found both positive and negative associations. In her study of Twitter as a driver of stock price of the most valuable companies by market capitalization, Jubbega (2011) found significant positive relationships between the number of brand sentiment tweets and stock prices. Her analysis showed a correlation between 100 brand sentiment tweets about a company and the subsequent stock performance of companies such as Coca-Cola, Toyota, Microsoft, and Disney. These findings led Jubbega to theorize that an increase in brand awareness, due to a larger number of brand sentiment tweets, likely led to more demand for the respective companies’ products, ultimately leading to higher stock prices. The precedent set by these studies provide the basis for our inquiry into the role of Twitter Engagement in building competitive advantage at higher education institutions, using performance metrics such as ranking.
Analysis of the impact of Deans’ engagement yields a paradoxical association different from that found by Jubbega. In fact, this analysis may suggest a recursive relationship. The earlier years in the period of investigation show positive associations with rank change (greater engagement is associated with positive rank change), while more recent years exhibit a negative association (greater engagement is associated with negative rank change). The latter finding contradicts conventional paradigms with regard to communication networks. As mentioned earlier, Metcalfe’s Law posits that as a network scales, the expected outcome is a communication network that is more useful and valuable for users because of positive network effects (Gilder, 1993; Metcalfe, 1996). This study’s findings evidently refute this law.
Again, as previously mentioned, Choudary (2014) invokes the notion of “reverse network effects” to provide an explanation for the above-mentioned dichotomy. In essence, Choudary observes that as a network scales it becomes more and more useful to users. Then, however, the network hits a tipping point and moves in the opposite direction. It thus begins to become less useful to its users. Choudary notes, however, that the crossing of the tipping point does not mean the number of users begin to diminish. Instead, the size of the network continues to increase but, ironically, the addition of each new user continues to decrease the value of the network.
It is interesting to speculate that the @Dean’s Twitter accounts may have ostensibly reached this tipping point. Hence, the value creation in terms of helping the respective college rankings plateaued in 2012 and progressively decreased thereafter. It is important to note that more detailed and broader study needs to be conducted to further substantiate these observations. However, this preliminary study indicates this effort might prove most interesting.
Finally, from a leadership perspective, this study provides a different and rather unique look at the way that leaders adopt or seem to feel a need to adopt innovative communication tools. C. K. Malhotra and Malhotra (2015) talk about how usage of new technologies revert to more established, conventional ways as the usage transitions from early to midstages of development. In many ways, this study provides interesting support for their observation. Stated simply, the @Dean’s accounts seem to be using a comparatively new form of technology, Twitter, in as conventional a manner as the technology permits. Basically, the schools adopt the technology halfheartedly, use it initially to their benefit, then, having established the tradition, the schools continue to use it to what might become their detriment. And, of course, all the time, they are using the channel in as conventional a way as possible linguistically.
If Deans and other leaders really do want to connect with their stakeholders, then this study along with others suggests that speaking at them from a public microblog may not be the most effective way to do that. Biological and cognitive constraints limit the economy of attention and the relationships in an online world (Gonçalves, Perra, & Vespignani, 2011). Hence, Deans and other leaders who want to cultivate and nurture relationships may need to find other, more personal ways to form “real-world” interactions (Choudary, 2014). Twitter, we may come to discover, will simply evolve into an online message board where Deans, like politicians, rant for attention.
Footnotes
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.