초록 열기/닫기 버튼
이 연구는 급속한 미디어 환경 변화 속에서 개인의 소셜미디어 이용과 프라이버시 염려(privacy concern)에 대한 실증적 분석을 통해 역설적 프라이버시 현상이 어떻게 변화했는지 탐색하고자 한다. 정보통신정책연구원(KISDI)의 <한국미디어패널> 자료(2015∼2019)를 토대로 인구사회학적 변인에 따라 소셜미디어 이용과 프라이버시 염려가 어떻게 변화했으며, 소셜미디어 이용이 프라이버시 염려에 어떤 영향을 미치는지 분석하였다. 연구결과, 소셜미디어 이용에 영향을 미치는 가장 핵심적인 요인은 연령 요인으로 나타났으나, 프라이버시 염려에 대해서는 인구사회학적 요인에 따라 일관된 결과를 보이지는 않았다. 프라이버시 염려가 소셜미디어 이용에 미치는 영향이 일관되게 나타나지 않은 것으로 보아 프라이버시 역설 현상이 온전히 이론적 수준에서 받아들여지는 한계가 있다는 점을 밝혔다. 이 연구는 2차 데이터를 활용한 연구의 한계에도 불구하고, 최근 사회적 이슈로 제기되고 있는 프라이버시 염려에 대해 동태적 시각에서 실증적으로 분석했다는 점에서 의미를 지니고 있다.
1. INTRODUCTION At present, new economic values are expected to be created through convergence services in all industries, particularly by strengthening the data, network, and artificial intelligence (AI) ecosystem in the digital environment. In particular, various services based on big data are receiving particular attention in the field of media. The reason is that the business model providing personalized and customized services by analyzing and restructuring vast amounts of personal information through AI algorithms has scalability in various industries. However, given the wide range of personal information collection on the Internet, including social media and various services, the problem of privacy invasion is being raised as a social issue. Recently, there have been cases of privacy violations, such as collecting social media conversations of general service users and using them for AI chatbot. With the advent of various social media-based consumption and distribution platforms, delivery services and online classes have become common. These developments have also brought changes to the lifestyle of individual users. News consumption, online shopping, and video content consumption through social media are rapidly spreading as well. Moreover, there are various changes in the perception of risk for privacy infringement. This study focused on these changes in the media environment and user behavior and attempted to explore changes in social media use and personal privacy concerns. 2. LITERATURE REVIEW This study used media dependency theory proposed by Ball-Rokeach and DeFleur (1976) and the privacy paradox concept as bases in attempting to examine the changing patterns of people's privacy concerns in terms of social media usage experience. In relation to media use, media dependence theory has been discussed at various levels, such as individuals and individuals and individuals and society. This theory particularly provides a rich theoretical foundation at the individual level for the relationship among people through social media. The tendency of dependence is gaining strength in terms of the media environment of modern society as a new paradigm based on high-speed Internet. A clear evidence is that various Internet services have become available in recent years owing to the rapid spread of mobile smartphones. In terms of devices, smartphones are already recognized as an essential medium in modern society (KISDI, 2020). Moreover, social network platform-based Internet services play a dominant role in our society. However, nearly all users are likely to be in a situation where they have to provide their personal information to use convenient personalized services in social media platforms. In this process, tension develops between the utility of providing personal information and the risk of personal information exposure and misuse. This paradoxical phenomenon is called "privacy paradox" (Sutanto, Palme, Tan, & Phang, 2013; Belanger & Crossler, 2011; Potzsch, 2009; Norberg, Horne & Horne, 2007). Privacy paradox can be understood as a type of dissonance. That is, even if individuals' awareness of the risk of privacy invasion increases, they will still use social media services frequently. Furthermore, the higher the use of social media, the greater the privacy concerns. Barth et al. (2019) conducted a study on privacy paradox and found that paradoxical behaviors on privacy concerns are still observed. They argued that users are concerned with potential risk factors for invasion of their privacy on the Internet (mobile environment), but they are not willing to invest the time, effort, or money needed to protect them. However, other previous studies on privacy paradox in the online space have tended to appear differently, depending on the nature of Internet services. For example, the higher the use of a relatively closed private online platform centered on personal networks, the lower the privacy concerns. By contrast, the higher the use of open platforms, such as online clubs and cafes, the greater the privacy concerns (Ryu & Koh, 2021). A study on the privacy concerns of the young generation, who are the most active social media users, have yielded contradictory results. Several inferences are possible as to why previous studies on the paradoxical phenomenon of privacy concerns have not provided consistent results. First, in situations where social media properties and services diversify, such outcomes can be viewed as a result of limiting individual privacy concerns to specific social media services or caused by limitations of the analysis samples (e.g., small samples). Given that privacy concerns perceived by users may appear differently, depending on the relational attributes of specific social media or the degree of openness of information (Seo & Son, 2011), the paradoxical phenomenon of privacy concerns does not show consistent results. Second, most of the previous studies related to the privacy paradox have tended to focus on the dichotomous discussion between attitude and behavior toward privacy (Kokolakis, 2015). That is, epistemological and behavioral dimensions of privacy concerns are closely related and may appear differently, depending on the passage of time and changes in media environmental factors. Therefore, this study formulated the following research questions to explore the privacy paradox from a dynamic approach regarding the relationship between general behavioral factors of social media use and privacy concerns. <RQ 1> Are there differences in personal social media use and privacy concerns according to demographic variables over the past five years? <RQ 2> How does personal social media use affect privacy concerns and how has such an effect changed over the past five years? 3. METHOD To empirically explore the paradoxical privacy concern phenomenon, this study used the 2015-2019 <Korea Media Panel> data of KISDI as bases in analyzing how media use and privacy concerns had been changed by demographic factors and how the effect of media use on privacy concerns had been altered during the recent five years. <Korea Media Panel> data are nationally approved since 2010 for measuring annual cross-media usage behaviors for various media contents, targeting approximately 10,000 people aged 6 and over in approximately 5,000 households in 17 cities and provinces nationwide. This study used five-year data measured and collected from 2015 to 2019 among media panel data. The final analysis of this study was the response data involving 5,666 people in 2015, 5,751 people in 2016, 5,852 people in 2017, 5,965 people in 2018, and 6,056 people in 2019. Although there are some differences by year, the proportion of women (55.7%-56.9%) was slightly higher than that of men (44.0% to 44.3%). Those aged 19-29 years account for 10.9%-12.7%; 30s, 9.0%-14.3%; 40s, 21.4%-25.6%; 50s, 21.0%-22.6%; 60s, 14.9%-15.5%; and 70s and above, 13.3%-18.8%. 4. RESULTS For the first research question, personal privacy concerns were examined using demographic variables. First, privacy concerns according to gender factors showed a statistically significant difference from 2015 to 2019(F2015=3.84, p<.05; F2016=5.29, p<.05; F2017=12.06, p<.01; F2018=13.99, p<.001; F2019=25.91, p<.001). However, note that men were more concerned with privacy than women in 2015 and 2016, but women were more concerned with privacy than men from 2017 to 2019. Second, privacy concerns by age have changed with a statistically significant difference over the past five years. However, privacy concerns of specific age groups showed substantial difference depending on the period. Third, for privacy concerns according to individual education levels, the lower the education level, the lower the privacy concerns in 2015 and 2016. After the 2017 survey, the higher the education level, the lower the privacy concerns (80.662≤Fs≤441.20, p<.001). Fourth, privacy concerns according to individual income levels also showed statistically significant differences between groups (22.06≤Fs≤112.84, p<.001), but there was no consistent behavioral change. To explore the second research question, this study conducted a hierarchical regression analysis using gender, age, education level, and personal income as main demographic factors affecting individual privacy concerns as main variables in the first stage and social media use in the second stage. For hierarchical regression analysis, multicollinearity performance with tolerance and variance inflation factor VIF) applied (1.17≤VIF≤2.33) and Durbin-Watson d applied autocorrelation assumption test were conducted, confirming that all results were not violated (1.14≤Durbin-Watson’s ds≤1.32). As a result of analyzing the effect of social media use on privacy concerns by year after controlling for gender, age, education level, and personal income variables (Model 2), social media use increased to privacy concerns only in 2016 and 2018 (model 2). The result was found to have a significant effect (β2016=-.050, p<.001; β2018=-.028, p<.05). The paradoxical phenomenon that privacy concerns decrease as social media use increases was confirmed. The analysis results for <Research Question 1> and <Research Question 2> indicated that the use of social media and concerns with privacy show an increasing trend compared with the previous year. As a result of regression analysis in 2016, the paradoxical phenomenon of lower privacy concerns as social media use increased was confirmed in the same year. However, whether or not the paradoxical phenomenon is common is difficult to conclude. Note that the 2017 analysis showed that social media use and privacy concerns do not have any effect. In 2017, the use of social media decreased compared with that in the previous year (4.41→4.36), and concerns with privacy increased (3.17→4.27). In the 2018 analysis result, a paradoxical phenomenon was confirmed again, in which social media use increased (4.36→4.75) compared with that in the previous year, while privacy concerns decreased (4.27→3.98). Regression analysis in 2018 showed similar results. However, the 2019 regression analysis indicated that social media use and privacy concerns do not have a significant effect. Even when compared with the results in 2018, social media use slightly increased and privacy concerns increased. That is, there was no paradoxical phenomenon in 2019. 5. CONCLUSION AND DISCUSSION In conclusion, a temporary privacy paradox was identified between social media use and privacy concerns, but this result was limited to year-specific data. This result requires careful interpretation because it is the result of data analysis for each independent year. However, when viewed from 2015 to 2019, it appears reasonable to interpret the paradoxical phenomenon of privacy concerns as being difficult to see as a general phenomenon or trend, given that social media use and privacy concerns are on the rise. This study indicates that the term “privacy paradox” required further empirical research though the elaborated analysis model. Despite the limitation imposed by the usage of secondary data, this study has some implications for exploring the causal relation between social media use and privacy concerns and suggesting the application of the dynamic privacy paradox approach.
키워드열기/닫기 버튼
Privacy paradox, Social media use, Privacy concern, Media dependency, Media panel
