In this thesis I have argued that, since Twitter and Facebook do not have the same possibilities, people can express minority views on these platforms in different ways. The possibilities I considered were networks of friends on each platform, visibility and identifiability. I argued that on Twitter people might feel freer to express minority views because Twitter networks tend to be made up of strangers or acquaintances of weak importance; therefore, people would not fear losing these relationships if they expressed divergent opinions to the same extent they would on Facebook, where people interact with real-life friends or stronger bonds, such as family members. Say no to plagiarism. Get a tailor-made essay on "Why Violent Video Games Shouldn't Be Banned"? Get an original essay. Similarly, I have argued that posts on Facebook are more visible because the platform automatically notifies people's friend networks about a post, whereas Twitter's notification system is less prominent. As a result, people would be more afraid to express a different point of view on Facebook where real friends could see it than on Twitter among strangers and weak acquaintances. Finally, I argued that Facebook requires more visibility of people's identities by requiring a real name and encouraging the inclusion of more details about a person in one's profile, whereas Twitter allows fake names and does not provide space for many details about a person. This decreased identifiability on Twitter, I argued, would make people freer to express minority views because they wouldn't worry about offending true friends because they can distance their Twitter profile from their true selves. My findings support this thesis. I found that people were more willing to express controversial opinions that they believed were in the minority on Twitter than on Facebook. I also found that users tend to use opinion-avoidance strategies more often on Facebook than on Twitter when the majority of their network disagrees with them. This offers important new insights into how affordances work on various social media platforms. These findings suggest that even if people do not consciously think about differences in affordances, they influence how they express their opinions on these platforms. Norman (1990) argues that “affordances provide strong clues to how things work” (p. 9). ). Similarly, Gibson (1982) stated that affordances could limit or encourage certain actions. My findings strongly support this thesis. My results showed that when the same people were asked whether they would express a controversial minority point of view on Twitter and on Facebook, they reported being more constrained on Facebook. This suggests that it is the opportunities of these two platforms that may lead to differences, not the fact that different people may be attracted to using Twitter rather than using Facebook. Therefore, this suggests that differences in people's behaviors in discussing opinions are not just a function of the platform they choose to use. Rather, this study provided the first evidence that it is the affordances of each platform that can lead people to change their behaviors when interacting on one versus the other. This begins to answer a question left unanswered by previous research, which had also found that people differed in how they communicated on Twitter versus Facebook. However, the study did not clarify whether different people areattracted to Twitter or Facebook and this is why opinion discussion behaviors may differ or why the same people express opinions differently. Furthermore, my findings showed that it is possible that people are more interested in agreeing on a controversial issue with certain people rather than with the rest of their network on social media sites. For example, most networks of people may agree with them on a controversial topic, but it is their key. If strong ties disagree with them, they may not want to express their opinion on that network since the their visibility and identifiability are high and they may not want to upset their key connections. In other words, the opportunities of friend networking, visibility, and identifiability work together to create a climate on Twitter in which people may be freer to express controversial minority opinions because they feel less visible and identifiable and are less concerned to offend one's weak. bond relationships or strangers, so they feel encouraged. Considering the above picture, the main theoretical contribution of this work is to suggest that a favorable opinion climate alone is not sufficient to explain people's willingness to express opinions on a controversial issue on social media platforms. Depending on each platform's offerings, people may not want to express their opinions if they feel more identifiable, more visible, or if they interact primarily with friends and family with strong ties. The concept of homophily is helpful in understanding my findings. Some studies suggest that people tend to talk about controversial issues with their homogeneous networks or close ties (Marsden, 1987), and users are likely to have a homogeneous network on Facebook (Lönnqvist & Itkonen, 2015). On the other hand, the results of this thesis suggest that people tend to avoid discussing controversial issues on homophilic Facebook, and are more likely to discuss these topics on more heterogeneous Twitter. Another explanation for these findings could be that high visibility and identifiability could increase people's awareness of diversity of opinion within users' networks on Facebook. Therefore, although users are more likely to have homophilous networks on Facebook, increased awareness of different opinions could reduce their perceived homophily and increase ambivalence towards a controversial discussion on Facebook. Therefore, this increased awareness could explain why users censor themselves and prefer not to talk about controversial issues on Facebook. This study also supported the idea that visibility and identifiability can influence users' opinion-expressing behaviors on social media sites. The result showed that people tend to express their opinion on Twitter where visibility and identifiability are low compared to Facebook where these two affordances are high. Gaver (1996) hypothesized that affordances can influence social interaction between people. The findings suggest that while users on Twitter tend to express their opinions on a controversial issue, they tend to use avoidance strategies on Facebook and avoid interacting with their network on a controversial issue. Therefore, it can be concluded that social media affordances not only influence individual actions, but can also impact social interactions on social media websites. The findings also suggest that users may feel less responsible for their actions when their visibility and identifiability are low. People do.