Comparative dimensions of COVID-19 visual health literacy: social media news imagery in Germany and China

The number of images we collected varied across the platforms (Fig. 1, part a). Table 2 details the specific number of images coded for each social media outlet in this study. The data includes descriptive and suggestive figures and interpretations. Due to the limited number of observations, it is not possible to provide proper statistical tests. However, the identification of patterns from the populations of posts with images underscores the importance of conducting further research on the use of visual messaging and its impact on society’s response to a healthcare emergency.

Image composition (by media outlet)

Figures 2 and 3 show descriptive graphs of the total images (Fig. 2) and the topic composition of such images (Fig. 3) for these eight media outlets under consideration: ‘@CCTV’ (China/English), ‘CCTVNEWS’ (China/Mandarin), ‘@ChinaDaily’ (China/English), ‘China Daily’ (China/Mandarin), ‘@dwnews’ (Germany/English), ‘@SPIEGEL_English’ (Germany/English), ‘@SZ’ (Germany/German), and ‘@tagesschau’ (Germany/German). ‘@ChinaDaily’ on Twitter/X is the news media platform that used the largest number of images, more than 4000, while ‘@SPIEGEL_English’ was on the other extreme, with just 11 samples found.

Bar plot visualisations of all image-led posts related to COVID-19 and public health measures according to coding criteria set out in Table 1.

Bar plot visualisations of all image-led posts related to COVID-19 and public health measures according to coding criteria set out in Table 1.

The topic composition of those images also shows some differences. The proportion of Context images is larger in CCTVNews than in any other outlet and the proportion of images of People seems larger in outlets based in Germany.

Cultural, geographical and linguistic dynamics

The English-language outlets exhibited the highest image count in China, with a total of 4239 images for 2022. Conversely, the outlets posting content in Mandarin showcased only a total of 1634 images. German-language outlets, with their communication on COVID-19, public health measures, and, in preparation for the approaching endemic phase at the end of 2022, posted a total of 2833 images. German outlets communicating in English in Germany showed the lowest count of images, with just 1183.

Figures 4 and 5 present a side-by-side comparison of the total images published in media outlets in English in China or Germany, in Mandarin and German. Figure 5 shows the topic composition of such images where we can observe the proportion of images on People, Context, Activities, Objects and Settings per language and location. The graphs suggest a stronger relevance of Context images for outlets in Mandarin or those located in China and a stronger relevance of images with People in German outlets.

Bar plot visualisations of all image-led posts related to COVID-19 and public health measures according to criteria set out in Table 2.

Bar plot visualisations of all image-led posts related to COVID-19 and public health measures according to criteria set out in Table 2.

Image composition (by language and location)

Table 3 presents both the total number of images per topic (Panels A and C) and the proportion of images for each language and the location of the outlets over the total number of images (Panel B) and restricted to images of people (Panel D).

Visual composition patterns

To provide a comprehensive multi-layer analysis that aligns with our key analytical foci, Figs. 6 and 7 present synchronised insights into the distribution patterns of image content, both overall and by sub-category. Specifically, Fig. 6 highlights the representation of different types of individuals—politicians, executive authorities, civilians, and healthcare workers—within the broader category of People as outlined in Table 1. By comparing these imagery ratios across language and cultural contexts (e.g., China/Mandarin, China/English, Germany/German, Germany/English), we can assess which contexts place more emphasis on particular groups.

Dataset fingerprint. Graph showing the relative frequency distributions of the four kinds of ‘PEOPLE’ in each national-linguistic dataset layered over each other: politicians, healthcare workers, civilians, and executive authorities such as police or military. Based on the analysis of all image-led posts related to COVID-19 and public health measures according to the criteria set out in Table 1.

Dataset fingerprint. Graph showing the relative frequency distribution of all analytical study criteria in each national-linguistic dataset layered over each other for a three-country-two-language comparison of visual health literacy. Based on the analysis of all image-led posts related to COVID-19 and public health measures according to the criteria set out in Table 1.

Figure 6 shows that Mandarin-language outlets in China tend to feature a higher proportion of politicians in their imagery compared to German or English-language outlets in both China and Germany. On the other hand, outlets in Germany, both in German and English, as well as English-language outlets in China, seem to prioritise images of healthcare workers and civilians. Notably, German-language outlets in Germany appear to give relatively more weight to politicians compared to their English-language counterparts.

Based on the numbers in Table 3, we observe some key patterns in the use of images featuring politicians, civilians, and healthcare workers. Both Germany/English and China/English outlets feature civilians more frequently than their German or Mandarin counterparts. Specifically, the difference between China/English and China/Mandarin is 22%, meaning civilian images are used 22% more often in China/English outlets. In Germany, this difference is smaller, with a 10% higher usage of civilian images in Germany/English outlets compared to Germany/German outlets. This contrast is more pronounced in China than in Germany.

Similarly, images of healthcare workers are used more in China/English outlets than in China/Mandarin outlets, with a 19% difference. Healthcare worker images appear in 37% of China/English communications, compared to only 18% in China/Mandarin. In Germany, the difference between English and German outlets is much smaller, just 5%, further emphasising the larger contrast in image use patterns between English and Mandarin outlets in China.

This pattern is also evident in the use of politician images. In China/Mandarin outlets, politicians appear in 46% of communications, compared to just 6% in China/English, marking a 40% difference. This is more than triple the difference in the use of politician images between Germany/German and Germany/English outlets, where the gap is only 12%. In both countries, however, outlets in the official language (Mandarin in China and German in Germany) prioritise images of politicians more than their English counterparts.

Figure 7 provides further insights into the relative emphasis placed on different types of image elements across the studied outlets. China/Mandarin outlets focus heavily on contextualising the overall pandemic environment, whereas China/English and Germany/German outlets give more prominence to images depicting objects and settings.

In Germany, German outlets maintain a balanced approach, with relatively equal attention to Settings, Objects, People, and Activities. However, China/English outlets prioritise Objects and Settings, suggesting different visual strategies in these cultural contexts. These patterns seem to highlight the distinct approaches to visual communication and health literacy employed by social news media outlets in China and Germany, both within the nation and internationally.

The distribution of image types across country-language pairs shows some consistency, particularly when comparing the overall categories of images. However, two notable differences emerge in Fig. 7 and Table 3. Germany/English communications utilise images of people more frequently (39% of all images) than any other country/language pair, while China/Mandarin communications place a heavier emphasis on Context images, accounting for 49% of all images.

These differences suggest varying communication strategies across the outlets. Mandarin-language outlets in China appear to prioritise Context images and Politicians, while China/English outlets focus more on Healthcare Workers and Civilians. This contrast is more pronounced than the differences between Germany/German and Germany/English outlets, where the patterns are similar but less distinct. Both German and Mandarin outlets tend to feature more images of politicians compared to their English-language counterparts in both countries.

A better understanding of these differences may be crucial in shaping more effective public health communication strategies. While preliminary and limited in scope and scale, these findings suggest that the language and cultural context in which an outlet operates influence the types of images used. By analysing these variations further, we may be able to gain more detailed and reliable insights into how visual communication strategies affect public health responses and inform the development of future communication policies.