Words wield immense power. They help to shape realities, tell stories, and encompass deeper values and intentions on behalf of their users. Buzzwords are imprecise, trendy – and often-frustrating – words that are encountered in daily life. They frame problems, evoke emotional responses, and signal moral values. In this dissertation, I study buzzword use within the field of environmental conservation to better untangle the inherent tension they have long produced: do buzzwords help or hurt collective conservation efforts? Using a mixed methods approach, this dissertation provides descriptive and causal empirical evidence on many of the untested assumptions regarding the behavior, use, and impacts of buzzwords on conservation decision making. First, through a series of expert interviews with conservation professionals, I develop an empirically informed definition and understanding of buzzwords that builds upon the scholarly literature. It identifies eight defining characteristics, elaborates on the nuances of their use, life cycle, and context dependence, and sets forth a series of testable hypotheses on the relationship between buzzwords, trust, and perceptions. Second, I take this empirically informed understanding and employ a large-scale text analysis to interrogate the mainstream conservation discourse. I produce a list of buzzwords used across institutions (e.g., academia, NGOs) in the past five years and link them to predominant conservation frames, comparing the ways in which different institutions relate to and discuss conservation concepts. This analysis validates many long-held paradigms and ubiquitous buzzwords found in conservation such as sustainability and biodiversity, while identifying a more recently emerging framing of inclusive conservation. Third, I experimentally test a set of hypotheses on the effects that buzzwords have on decision making, as moderated through trust. This study finds evidence of a greenwashing effect, whereby buzzwords may produce marginal benefits to less trustworthy organizations through increases in credibility and group identity alignment, but do not outweigh the benefits of being trustworthy in the first place. In the face of many current global challenges requiring cooperation and collective action – such as climate change and environmental degradation – it is imperative to better understand the ways in which communication and framing (including buzzwords) influence decision making.

Exabytes of data are created online every day. This deluge of data is no more apparent than it is on social media. Naturally, finding ways to leverage this unprecedented source of human information is an active area of research. Social media platforms have become laboratories for conducting experiments about people at scales thought unimaginable only a few years ago.

Researchers and practitioners use social media to extract actionable patterns such as where aid should be distributed in a crisis. However, the validity of these patterns relies on having a representative dataset. As this dissertation shows, the data collected from social media is seldom representative of the activity of the site itself, and less so of human activity. This means that the results of many studies are limited by the quality of data they collect.

The finding that social media data is biased inspires the main challenge addressed by this thesis. I introduce three sets of methodologies to correct for bias. First, I design methods to deal with data collection bias. I offer a methodology which can find bias within a social media dataset. This methodology works by comparing the collected data with other sources to find bias in a stream. The dissertation also outlines a data collection strategy which minimizes the amount of bias that will appear in a given dataset. It introduces a crawling strategy which mitigates the amount of bias in the resulting dataset. Second, I introduce a methodology to identify bots and shills within a social media dataset. This directly addresses the concern that the users of a social media site are not representative. Applying these methodologies allows the population under study on a social media site to better match that of the real world. Finally, the dissertation discusses perceptual biases, explains how they affect analysis, and introduces computational approaches to mitigate them.

The results of the dissertation allow for the discovery and removal of different levels of bias within a social media dataset. This has important implications for social media mining, namely that the behavioral patterns and insights extracted from social media will be more representative of the populations under study.