Social phenomena have been studied extensively in small scales by social scientists. With the increasing popularity of Web 2.0 and online social networks/media, a large amount of data on social phenomena have become available. In this dissertation we study online social phenomena such as social influence in social networks in various contexts. This dissertation has two major components: 1. Identifying and characterizing online social phenomena 2. Leveraging online social phenomena for economic and commercial purposes. We begin the dissertation by developing multi-level revenue sharing schemes for viral marketing on social networks. Viral marketing leverages social influence among users of the social network. For our proposed models, we develop results on the computational complexity, individual rationality, and potential reach of employing the Shapley value as a revenue sharing scheme. Our results indicate that under the multi-level tree-based propagation model, the Shapley value is a promising scheme for revenue sharing, whereas under other models there are computational or incentive compatibility issues that remain open. We continue with another application of social influence: social advertising. Social advertising is a new paradigm that is utilized by online social networks. Social advertising is based in the premise that social influence can be leveraged to place ads more efficiently. The goal of our work is to understand how social ads can affect click-through rates in social networks. We propose a formal model for social ads in the context of display advertising. In our model, ads are shown to users one after the other. The probability of a user clicking an ad depends on the users who have clicked this ad so far. This information is presented to users as a social cue, thus the click probability is a function of this cue. We introduce the social display optimization problem: suppose an advertiser has a contract with a publisher for showing some number (say B) impressions of an ad. What strategy should the publisher use to show these ads so as to maximize the expected number of clicks? We show hardness results for this problem and in light of the general hardness results, we develop heuristic algorithms and compare them to natural baseline ones. We then study distributed content curation on the Web. In recent years readers have turned to the social web to consume content. In other words, they rely on their social network to curate content for them as opposed to the more traditional way of relying on news editors for this purpose -- this is an implicit consequence of social influence as well. We study how efficient this is for users with limited budgets of attention. We model distributed content curation as a reader-publisher game and show various results. Our results imply that in the complete information setting, when publishers maximize their utility selfishly, distributed content curation reaches an equilibrium which is efficient, that is, the social welfare is a constant factor of that under an optimal centralized curation. Next, we initiate the study of an exchange market problem without money that is a natural generalization of the well-studied kidney exchange problem. From the practical point of view, the problem is motivated by barter websites on the Internet, e.g., swap.com, and u-exchange.com. In this problem, the users of the social network wish to exchange items with each other. A mechanism specifies for each user a set of items that she gives away, and a set of items that she receives. Consider a set of agents where each agent has some items to offer, and wishes to receive some items from other agents. Each agent would like to receive as many items as possible from the items that she wishes, that is, her utility is equal to the number of items that she receives and wishes. However, she will have a large dis-utility if she gives away more items than what she receives, because she considers such a trade to be unfair. To ensure voluntary participation (also known as individual rationality), we require the mechanism to avoid this. We consider different variants of this problem: with and without a constraint on the length of the exchange cycles and show different results including their truthfulness and individual rationality. In the other main component of this thesis, we study and characterize two other social phenomena: 1. friends vs. the crowd and 2. altruism vs. reciprocity in social networks. More specifically, we study how a social network user's actions are influenced by her friends vs. the crowd's opinion. For example, in social rating websites where both ratings from friends and average ratings from everyone is available, we study how similar one's ratings are to each other. In the next part, we aim to analyze the motivations behind users' actions on online social media over an extended period of time. We look specifically at users' likes, comments and favorite markings on their friends' posts and photos. Most theories of why people exhibit prosocial behavior isolate two distinct motivations: Altruism and reciprocity. In our work, we focus on identifying the underlying motivations behind users' prosocial giving on social media. In particular, our goal is to identify if the motivation is altruism or reciprocity. For that purpose, we study two datasets of sequence of users' actions on social media: a dataset of wall posts by users of Facebook.com, and another dataset of favorite markings by users of Flickr.com. We study the sequence of users' actions in these datasets and provide several observations on patterns related to their prosocial giving behavior.

This research sought to explore media framing theory, first introduced by Erving Goffman, which asserts that the media portray certain items in a way that affects awareness, salience and tone of those items. There has long been debate about media framing especially as it pertains to the framing of violent events. Mass shootings are of particular interest because of the graphic and often senseless nature of the crime. This study looked particularly at the 2011 Tucson, Arizona mass shooting. A content analysis of articles between February 7, 2010, and November 8, 2011, was conducted to explore media framing of gun control after the Tucson, Arizona mass shooting involving U. S. Representative Gabrielle Giffords. Results showed an equal amount of articles with a negative or neutral tone, with the tone shifting to being more frequently neutral after the shooting. The topic shifted toward legislation, adding further support to media framing theory. These findings have impacts for media and public relations.

The present study analyzes the way in which Toni Morrison, an established author, was covered by U.S. newspapers in the year before and year following her selection for Opraha's Book Club. The content analysis method was used in the research, and the results were used to test 6 hypotheses and 6 research questions. The results indicated that there was a significant increase in the total number of mentions of Toni Morrison in the year after her inclusion in Opraha's Book Club. The overwhelming trend that was seen in all variable comparisons was that while there were obviously more mentions of Toni Morrison post-Book Club, there were also increases and changes in the tone, page placement, and story placement of the mentions of Toni Morrison .

Karst landscapes are vulnerable to human influence, especially agricultural practices. The interconnectedness between surface activities and subsurface environments make karst landscapes particularly susceptible to soil erosion and water contamination. The likelihood of these two phenomena happening increases when agricultural intensification, irrigation, or fertilizer application occurs. This situation arises frequently in Vietnam, where 18% of the country is karst terrain and 60% of the population depends on agriculture for their livelihoods (Farming First 2009). In order to mitigate the negative consequences of agriculture on karst landscapes, effective implementation of policy to regulate human activities and increased communication of these policies to appropriate communities is needed. This study occurred in Phong Nha- Kẻ Bàng National Park, Vietnam, a UNESCO World Heritage site dominated by karst landscapes, extensive agricultural communities, and minimal regulation efforts specific to karst terrains. Interviews, observation, and GPS analysis were used to analyze the effectiveness of policy communication and karst protection in PN-KB. The research revealed that karst protection policy in the region is minimally communicated and, when communicated, often delivered in an ineffective manner to the wrong individuals. Despite the known harm agriculture causes to karst landscapes, intensification, irrigation, and the use of fertilizers still occurs frequently and is often supported by government officials in PN-KB. Policy and karst landscape information is concentrated among park officials and rarely presented in an informal setting, leaving those in most frequent contact with the karst landscape—the farmers—without any information about the vulnerability of karst terrain to agricultural activities and the subsequent consequences to human health. Through analyzing the interactions between farmers and management officials in Phong Nha-Kẻ Bàng National Park, general conclusions on communicating policies to protect karst terrain in agricultural regions can be drawn. The communication of karst science and the implementation of policy to protect karst landscapes must be presented both formally to governing officials and local representatives, as well as through informal networks to general citizens. Through these means of communication, protection for karst landscapes and their inherent natural resources can successfully be implemented.

This thesis examines the television reporters' verbal depictions of two rescue events, the rescue of Jessica McClure in 1987 and the rescue of three whales at Pt. Barrow, Alaska in 1988, in order to discover what rhetorical techniques were used to appeal to the public interest. Analog criticism, metaphorical analysis and pentadic analysis were used to discover the dominant language reporters chose. Three main conclusions were drawn from this analysis: (1) use of the dramatistic pentad showed how reporters focused public attention away from the purpose,(2) verbal and visual depictions cannot be separated when studying television news stories and (3) two different styles of depiction were used to portray two very similar events.

Working from a communication theory paradigm and from previous literature, the purpose of this study was to empirically examine significant receiver correlates which predict adoption behavior of Sigma Nu Fraternity on three campuses. Drawing from past research, the present study utilized a unique combination of social and communication variables as predictors of fraternity membership. A stepwise multiple discriminant analysis using five factors derived from the thirteen independent variables yielded a highly significant three factor discriminant function (p < .001) which explained 46.65 percent of the common variance in adoption/non-adoption behavior of Sigma Nu. The significant factors were labeled peer group behavior, parental economic support, and social attitudes. Based on intuitive interpretation features of the data, the researcher suspects that adopters are more socially oriented, have a wider variety o: friends, are less dependent on parental economic support, and exhibit higher scores on measures such as self-esteem, liberalism, conservatism, and the need for identification with groups.

Because of the increasing importance of local government in America's cities, it is worthwhile to note what factors influence the election of our local leaders. This study sought to isolate the communication factors and rhetorical strategies which influenced the election of Mayor Richard Fulton in Nashville's 1975 Mayoral Race. Data was gathered from the written and video media, a private interview with the candidate, and campaign materials. Research also provided the candidate's previous political background. Various factors in Nashville's 1975 Mayoral Race combined to produce an insightful episode in rhetorical and nonrhetorical communication campaign strategies. During the initial stages of the study, emphasis was placed on the rhetorical strategies of candidate Richard Fulton, and to a lesser degree, those of his opponent, Earl Hawkins. As the study progressed, it became increasingly evident that although rhetorical strategies were influential in Fulton's campaign victory, the major reason for his political success could not be totally attributed to his campaign speaking. After acknowledging that fact, a search throughout the available data began to determine what factors were responsible for his ultimate campaign success. It was found that the raising of the candidate's ethos was the most influential campaign factor. Evidence throughout the available data documented over and over again the conclusion that Richard Fulton built an attractive image appealing to Nashville's voters. Fulton's rhetorical content and past experiences served to connunicate this "winning image." It was revealed that: The national and local political situation was favorable to the image which Fulton projected, Fulton's ethos was very high, Fulton's image was the chief factor in this successful campaign, and Fulton's rhetorical strategies were in accord with the majority of Nashville's voters' attitudes. Because specific categories have not been formed by the communication community pertaining to the political candidate's image, this study has been one of discovery and new insight in the area of local politics. A proper rhetorical analysis could not simply proffer an evaluation of traditional aspects of speaking, such as ethos, logos, and pathos. Truly these categories enter the campaign on an important level, but with the increasing role of the media, even in local campaigns, new categories need to be developed and explored. It is the hope of the writer that this particular study will encourage greater interest and future investigations into the local political workings in our cities, particularly those in the South. Why are the Americans of each city choosing their particular leaders? What role is public rhetoric playing in the local choice? How is the media affecting those choices on the local level? Is the American public being "sold" a false image by local politicians due to local advertising, and if so, how might greater public speaking help to dispell these misleading images? These are just a few questions which may stir future interest in pursuing the rhetorical study of our nation's many, local political campaigns and their place in the future development of the contemporary South.

What is the importance of political speechmaking? Do state governors discuss presidential priorities? This study addresses these questions by analyzing the contents of annual State of the State addresses given by governors from 2012 to 2014 during the presidency of Barack Obama. A descriptive paper provides evidence that governors primarily discuss employment and economic issues in their addresses, are discussing greater number of policy issues than in previous decades, and are delivering their address before the presidential State of the Union message. Examining health care and immigration policy in separate empirical papers, I theorize that contextual factors, including legislative partisanship, public approval, and presidential influence may affect the extent to which policies supported by the Obama administration are rhetorically referenced by governors. Empirical analyses found limited support for the influence of divided government, but demonstrated significant evidence for the importance of including state-centric factors, including annual employment rate and proximity to Mexico, as well as temporal effects, into future analyses of gubernatorial rhetoric.

Building on research that has focused on understanding how peers contribute to students' engagement, this dissertation explores the extent to which peer group influences on students' engagement may add to and be contextualized by qualities of the relationships they maintain with their teachers and their parents. To focus on how each of these adult contexts work in concert with peer groups to jointly contribute to changes in students' engagement, the two studies used data on 366 sixth graders which were collected at two time points during their first year of middle school: Peer groups were identified using socio-cognitive mapping; students reported on teacher and parent involvement; and teachers reported on each students' engagement. In both studies, models of cumulative and contextualized joint effects were examined. Consistent with models of cumulative effects, peer group engagement, parent involvement, and teacher involvement each uniquely predicted changes in students' engagement. Consistent with contextualized models suggesting differential susceptibility, peer group engagement was a more pronounced predictor of changes in engagement for students who experienced relatively low involvement from teachers. Similarly, peer group influences on changes in students' engagement were stronger for students who experienced relatively low involvement from their parents. In both cases, these peer effects were positive or negative depending on the engagement versus disaffection of each student's peer group. Both studies also used person-centered analyses to reveal cumulative and contextualized effects. Most engaged were students who experienced support from either both teachers and peers, or both parents and peers; the lowest levels of engagement were found among those students who affiliated with disaffected peers who also experienced either their teachers or parents as relatively uninvolved. Both high teacher and high parent involvement partially protected students from the motivational costs of affiliating with disaffected peers. Similarly, belonging to engaged peer groups partially buffered students' engagement from the ill effects of low teacher and parent involvement. These findings suggest that, although peer groups and teachers and parents are each important individually, a complete understanding of their contributions to students' engagement requires the examination of their joint effects.

This content analysis of mainstream and alternative news narratives interprets the use of the crisis media frame, and describes the relationship between local policy initiatives, media discourse and public opinion on tent cities, organized by people experiencing homelessness in Portland, Oregon. Framing homelessness and housing as a crisis intensified the public debate, attested by an increase in mainstream media reports on tent cities, and by controversial policy changes that addressed the individually-experienced traumatic impacts of the City's anti-camping ordinance, as well as the systemic lack of affordable housing and emergency shelter. Media discourse related to city-sanctioned tent cities blurs the lines between Shanto Iyengar's episodic and thematic media frames because of the simultaneous acknowledgement of individual and systemic circumstances. The crisis frame is a discursive mechanism in the production of knowledge on homelessness and housing, and is considered as an integral characteristic of Henri Lefebvre's conceptual model of socio-spatial production, which describes the interdependency between discourse, practice and meaning in the material and symbolic production of space.