Alice E Marwick and Danah Boyd (pg.1052, 2014) research into social media and privacy sought to question and understand the societal assumption that “teenagers don’t care about privacy”. Through conducting semi-interviews as part of a wider ethnographic experiment, Marwick and Boyd were to a great extent successful in combining a variety of research methods to not only understand how young people conceptualize privacy but constructing their ‘networked privacy theory’ that challenged mainstream privacy theories. However, the objectivity and validity through utilising an ethnography and the ethical questions surrounding monetary compensation raise concerns over the findings. Nevertheless, Marwick and Boyd’s ethnography and semi-interviews allowed for a well-rounded research article.
Through the utilization of an inductive style of analysis, the ethnography allowed for Marwick and Boyd to gain detailed and insightful knowledge about privacy within social media. This allowed for the construction of the researchers own privacy model that challenged cultural assumptions of privacy. To understand the behaviours of how teenagers managed their privacy in networked publics, Marwick and Boyd (pg. 1055, 2014) conducted two waves of semi-interviews with 166 teenagers. The first wave focused on “general technology practices” from 2006 to 09 while the second wave in 2011 emphasized privacy (Marwick and Boyd, pg. 1055, 2014).
The nature of an ethnographic study allowed these semi-interviews to be situated within the real world. This resulted in personal and authentic findings, were Marwick and Boyd were able to identify themes and concepts that were prevalent throughout all the semi-interviews. Scholar Munz (pg. 455, 2017) states that ethnographies allow researchers to “learn more about members of a community from the members themselves, in their own words and in a natural setting’. This is directly related to the interpretivist paradigm, where researchers understand how individuals create meaning within the social world (Deacon, pg. 6, 1999). Through using an ethnography as the primary research method Boyd and Marwick (pg. 1055, 2014) were able to observe these teenagers in a “broader context”. For example, Boyd and Marwick attended sports games, restaurants, and religious services to not only question but also observe these teenagers in a real-life setting (Boyd and Marwick, pg. 1055, 2014). This resulted in more authentic responses, as these teenagers were able to express their values and beliefs about privacy in a comfortable setting. As Boyd and Marwick (pg. 1055, 2014) state, the use of an ethnography allowed them to understand the participants intimately in how the concept of privacy affects their lives rather than “interrogating the accuracy of their statements”. For example, “Carmen” discussed how she managed privacy through encoding messages on Facebook through pop-culture references (Marwick and Boyd, pg. 1058, 2014). This allowed Boyd and Marwick to engage in a nuanced discussion within their network privacy theory about encoding and privacy-protecting strategies teenagers conduct. Hence, the use of an ethnography resulted in authentic findings that aided in the researcher’s validity.
However, the lack of participants who discussed privacy, coupled with the years of difference between the two semi-interviews could limit the validity and quality of the ethnographic and the findings. In terms of the distribution of participants in the two waves, only 60 people in 5 states were asked about privacy compared to the 110 people across 17 states in the first interviews (Marwick and Boyd, pg. 1055, 2014). This illustrates the vast difference in not only the number of participants but the number of states where the interview was conducted. As the second interview was explicitly discussing privacy, it would have been more efficient to garner more participants within the second wave and across more states. This would allow for greater utilization of the intimate nature of the ethnography and thus, a greater insight into privacy. In addition, the two to five year time gap between both surveys could pose as a potential negative for the validity of their findings and research. Even though the time difference was most likely due to the time it takes to complete ethnographic style research, the last ten years have seen a dramatic change within technology. Hence, the time gap between the two interviews would see teenagers use different online platforms or even new privacy strategies.
The open nature of semi-interviews aided the successfulness of the ethnography as it allowed the freedom for the participants and the researchers to elaborate on specific ideas and questions. Leech (pg. 665, 2002) states that successful semi-structured interview questions “tries to enter into the world of the respondent”. Marwick and Boyd were able to effectively question participants in this manner, with questions inquiring about the way young people openly engage with technology and privacy. Even though the first interview was not specifically about privacy, conducting a semi-interview about how teenagers use technology would allow for Marwick and Boyd to grasp a better understanding of the participant’s world and social media usage. This was further solidified through follow up questions allowed by the semi-interviewing style, which inquired deeper into concepts and themes that were then later analysed. For example, when interviewing “Hunter”, Marwick and Boyd (pg. 1057, 2014) asked relevant follow up questions regarding how “one should know what is appropriate for commentary” online. These open-ended and inquisitive questions allowed for a more in-depth analysis on “context-collapse”, a concept that was integrated into their theory.
However, it is important to acknowledge the bias that could have been presented within the semi-interviews. Baxter and Babbie (pg. 90, 2004) state, “research can never be objective because researchers are human and therefore necessarily subjective”. When asking questions, the researchers could have asked follow-up questions that would have subjected the participants to their social and cultural worldviews. As Machin (pg. 4, 2002) explains, researchers are “from a very different social world”, meaning that they have “little knowledge” of how participants world views and lives. In addition, as the individuals understand that they are being interviewed, these young participants might be subconsciously inclined to answer a certain way. This has the potential for bias within the findings.
The utilisation of a diverse group of participants from different cultural, economic, and societal settings, enhanced the validity of the ethnography and its findings. Through collaborating with a community organisation, Marwick and Boyd were able to conceptualize privacy through a variety of different voices. Even though this has the potential for participation selection bias as community organisations select individuals who best represent the community rather than an individual from a lower socio-economic area, the researcher’s insistence on diverse individuals and collaboration limits this bias significantly. As Marwick and Boyd (pg. 1055, 2014) state, their participants were not a “generalised sample…reflect(ing) a variety of experiences and backgrounds”. For example, participants ranged from 13 to 19, with the average age being 16 (Marwick and Boyd, pg. 1055, 2014). This is highly valuable as age is a major factor in the usage of social media, thus changing the way privacy issues are managed. In addition, a variety of socio-economic, ethnic and racial backgrounds were selected such as 39 African Americans, 13 individuals of Asian descent and individuals from different economic status (Marwick and Boyd, pg. 1005, 2014). Hence, through selecting participants that were from different backgrounds, the validity of the findings and theory increased.
Through the semi-interviews, ethnography and data collection, Marwick and Boyd were able to utilize the findings and themes to construct their successful networked privacy theory. Throughout the semi-interviews, Marwick and Boyd(pg.1055, 2014) coded data of “emerging patterns” and themes related to privacy such as “privacy-protecting tactics”, “power of trust” and “determining context”. As a result of their inductive method of research, the theory could be changed and developed as the interviews continued. This is a strong positive of the research method as the theory was flexible to new information. This data successfully led Marwick and Boyd (pg. 1063, 2014) to the conclusion that navigating privacy within the new media landscape highlight the “underlying interactional dynamics of privacy practices”, thus deconstructing previous privacy model and leading to new research in the field.
However, as a result of all participants being from the United States of America, the researchers’ findings and theory embodies a westernised framework. This suggests that the validity of the ethnography and thus the theory may not apply to other nations or societies. Machin (pg. 1, 2002) expresses that within ethnographies, the researcher must understand that “peoples behaviours as being determined largely by culture”. Even though participants were chosen as a result of their diverse backgrounds, having the same nationality could see similar cultural and social similarities concerning privacy issues. This means that the validity and replicability of this ethnography and networked privacy theory are limited to western societies cultures. Boyd and Marwick (pg. 1053, 2014) acknowledge this issue to some degree, stating that privacy is a “social construct that reflects the values and norms of individuals within cultures”. However, interviewing international participants, even though highly challenging to achieve, would have allowed for a more nuanced understanding of privacy.
The researchers’ monetary compensation for participants is an ethical issue that has the potential for bias and invalid responses. Marwick and Boyd (pg. 1055, 2014) state that participants were paid between $30 and $40 for their time. Generally, compensating participants in research can have negative effects on the validity of the results. Participants may be more inclined to provide responses that are agreeable with the Marwick and Boyd’s research questions, thus having the potential to show bias in the results. This is echoed by Bentley and Thacker (pg. 293, 2004), stating that there is a major “potential for payments to diminish the integrity of a study’s findings”.
However, through analysing the participant’s responses, it is more likely that monetary compensation had a positive effect as it allowed these teenagers to be more open to sharing their experience. Bentley and Thacker (pg. 293, 2004) state found that monetary compensation allowed for “positive effects on respondents willingness to participate in research”. Such bias would have been limited to an extent within this semi-interview as the researchers took screenshots of their social media profile, and could legitimize the stories. In addition, the purpose behind using an ethnography and semi-interview was to allow participants to feel comfortable in their environment and for the researchers to understand rather than critique their views. Hence, this ethical issue would have been significantly limited.
Marwick and Boyd’s ethnography were successful in creating a theory that discussed how privacy has changed as a result of social media. However, issues with the participants chosen, its significance within other cultures and monetary gain are issues that might affect the findings.