In order to tackle the above stated research question, for this kind of preliminary study, it was decided to apply the method of thematic analysis [
32], defining major themes that occur in the articles about COT. This kind of analysis does not give results on opinions expressed in the text, the structure of arguments, nor does it give any kind of indication on the effect of or intention behind the texts that have been analyzed. These can only be hypothesized on, by for instance, contextualizing the results with the troves on news value theory [
33‐
35].
3.2 Corpus Selection & Data Cleaning
The period chosen to select the corpus was from 01.01.2019 to 30.06.2020 (in total 18 months). This ensured that the event of May 2019 of 12.000 visitors from China visiting Switzerland and the outbreak of Covid-19 were covered. In order to find content covering COT, the keywords “China” and “Tourism” (in German language) were used in Factiva. In total, 457 articles were found published by the eleven sources, once duplicates had been removed. Next, the first round of data cleaning followed. Only if COT constituted at least 50% of the concerned article, it would be considered as a dominant topic and the article would be kept within the sample. Articles in which tourism was, for example, only mentioned as an industry suffering because of Covid-19 and China was not mentioned in relation with tourism were excluded. In the end, the database amounted to 60 articles. Interestingly, one newspaper (Sonntags-Blick) did not seem to have published any article where COT was considered a dominant topic.
For comparative reasons it was decided to also search the number of articles only mentioning the keyword “China” (in German language). One can see that the combination of “China” and “Tourism” takes up a rather small amount of mentions in comparison to articles just mentioning “China”. “China” and “Tourism” are mentioned in 3,4% (n = 457) of the sample only mentioning “China” (n = 13.320). Moreover, after the data cleaning only 13,1% (n = 60) out of the articles mentioning “China” and “Tourism” (n = 457) made it into the final sample.
Furthermore, the frequency of the mentions of the words “China” and “Tourism” over the selected time frame of 18 months (in this case with duplicates, n = 487) was considered. Interestingly, one can observe three peaks where the words “China” and “Tourism” were mostly mentioned: the first in May 2019 (in total 42 mentions), the second in July 2019 (in total 32 mentions) and the last in February 2020 (in total 69 mentions). It is striking that May 2019 was the month when the 12.000 visitors from China visited Switzerland, whereas February 2020 was two months after Covid-19 had been officially announced by the World Health Organization, and when daily cases in the Western Pacific reached their first peak [
37].
In a second step, the articles were traced back to the Facebook pages of the concerned media outlets. This could be done by manually researching the articles on the different Facebook pages of the media outlets. Facebook was selected as the first platform to collect UGC because it is a frequently used social media platform in Switzerland. In December 2019, 3,5 million users logged into Facebook in Switzerland [
38]. Considering that the number of people living in Switzerland in 2019 was 8,606 million [
39], 3,5 million Facebook users would amount to roughly 41% of the population. When checking the Facebook page of the media organizations in the sample (on 05.09.2020), the number of followers ranged from 555,633 (
20 Minuten) to 8,477 (
NZZ am Sonntag). Notably, not all of the word-for-word versions of the printed articles could traced back online, yet, the main topics of the online articles and the print articles were the same. As an example, a post about the article on the case of 12.000 tourists by
20 Minuten (print) published on May 14, 2019 could not be found online on the Facebook page of the organization. However, around the same time, two posts directing to online articles about the same topic were published.
Facebook posts by news media organizations are published online and publicly accessible for any entity with a Facebook account. Nonetheless, for privacy and legal reasons, usernames or profile descriptions of the “commentators” were not included in the data collection and are hence not part of the analysis. Furthermore, in this study, user comments were never replicated word for word, but their content was paraphrased. For the purpose of this study, only two posts about articles, which were posted in their “original” print form, from the two “highest” peaks of frequency of mentions were selected to be presented further. This was done to put emphasis on the qualitative methodological approach of the study and underline its exploratory nature. Furthermore, both articles deal with subjects related to the city of Lucerne, which was also frequently mentioned in the media discourse surrounding the visit of 12’000 Chinese tourists visiting Switzerland in May 2019. Issues such as the discrepancies between the publication of print and online versions of articles (as described above) are still to be solved and currently represent significant issues in terms of the generalization of results.
The first post was about an article that was published on May 24, 2019 by the Neue Zürcher Zeitung concerning “the case of 12’000 visitors to Lucerne” published after the event occurred (14.05.2020). It was concurrently published on the newspaper’s Facebook page (254.585 followers by 05.09.2020) on May 24, 2019, and received 71 reactions (likes, dislikes etc.), 73 comments and 12 shares (until 05.09.2020). For the analysis, only the first level of user generated comments was considered, while the comments to the comments were excluded. In total, 32 comments were analyzed.
The second post was about an article that was published on February 8, 2020 by the Luzerner Zeitung with a headline indicating that Chinese tourists were not present on one of the city’s main attractions (a square). Concurrently, a post about the article was published on their Facebook page (31.783 followers by 05.09.2020); it received 524 reactions, 223 comments and 51 shares. In total, all 52 first-level comments were analyzed.
3.3 Thematic Analysis
Thematic analysis can be considered a basic method of qualitative text analysis and is used in this context as an exploratory and descriptive approach [
32]. As a category-based method the text was analyzed according to different themes, which are presented to the reader [
32]. Moreover, according to Kuckartz [32, p.70] thematic analysis is a “method that reduces content”. For this study, after the data was cleaned, the text was thoroughly read. Afterwards an inductive theme building approach followed. The first number of categories derived from this stage is rather rough and manageable. In a next step, the main categories are divided into sub-themes. In this study, this process is shown by the in-depth description of the analysis of the two articles chosen to exemplify the research. Moreover, a theme was coded as major if it was described in at least one third of the article. In order to collect, highlight, code and analyze the different articles, the qualitative data analysis software nVivo was used. Figure 1 (see below) shows the sequential process applied, which was based on Kuckartz [32, p.70].
This process was first applied to the 60 selected news media articles. In a second step, the user-generated comments of two articles on Facebook were coded using the same process. The description of the results is presented hereafter.