A preliminary search of ST in YouTube videos also indicated that these terms would pull up more relevant and popular videos than other ST search terms. Other commonly used ST terms, such as dip and chew, were not used as search terms because they also refer to nontobacco topics and resulted in many videos that were unrelated to tobacco. Two searches were performed for each term selleck chemical FTY720 (a) by relevance and (b) by view count. The search terms and methods were chosen to both mimic user behavior by using common search terms and the default search strategy (search by relevance) as well as cast a wide net to capture the most-viewed videos (search by view count). The sample was limited to the first 20 results for each search because previous research on a few Internet search engines indicates that the majority of people will only click on the first page of search results (Jansen & Spink, 2006), which is 20 videos on YouTube.
The initial sample included a total of 160 videos: 20 videos for each of the four search terms and by both search strategies. Basic information was collected from each of the videos, including title; uploader alias (username of person posting the video); and number of views, likes, and dislikes (YouTube users can rate videos by whether they ��like�� or ��dislike�� a video). Exclusion Criteria Videos that were not in English (n = 14) and videos that were not relevant (n = 26) were excluded from any analysis. Videos were considered ��not relevant�� if ST was not central to the content of the video.
Videos with only a brief mention or image of ST and videos about e-cigarettes (which are not traditionally considered ST products) were not included in the sample. Duplicate videos that appeared more than once using different search strategies and terms were also eliminated (n = 40). On two occasions, the exact same video was uploaded under different titles by two different users. The duplicate video (n = 2) was eliminated from the sample to prevent double coding, but their view counts and number of likes and dislikes were combined to accurately represent the total number of people who watched the video. A total of 78 unique videos were coded and analyzed. Coding A master��s level researcher adapted the coding scheme from a previous YouTube analysis of smoking cessation (Backinger et al., 2011).
After a preliminary viewing of ST videos, all the authors discussed and finalized categories and definitions of genres and other content variables. Videos were then coded by the master��s level researcher. When video content was difficult to classify or was ambiguous, the authors met and came to a consensus decision on the appropriate classification. Videos were first classified by whether the overall portrayal of ST was pro-ST, anti-ST, or ��sensationalized.�� Videos that promoted the use of ST or made it look enjoyable AV-951 or socially acceptable were considered pro-ST videos.