The produsing expert consumer: Co-constructing,resisting and accepting health-related claims on social media in response to an infotainment show about food

Produsers and expert consumers

As discussed in the Introduction, audiences are interested in and influenced by health-related media content, but are no longer passive receivers of such information. The rise of new communication technologies, such as blogs, wikis, social media and instant messaging, has changed the role of mass media as the sole providers of information, as there no longer is a distinction ‘between mass communication and interpersonal communication, and therefore between mass media and interpersonal media’ (Lüders, 2008: 683). Loosen and Schmidt (2017) similarly state that new information technologies have blurred ‘the boundaries between news producers and consumers as well as between production and consumption’ (p. 3). Consumers nowadays also produce content on online platforms, and actively engage with any media content, both user generated or mass media generated, by taking part in interpersonal follow-up communication within smaller networks (Loosen and Schmidt, 2017), which Bruns (2008) labels as produsage. A similar concept is Schmidt’s (2014) personal public, which refers to a new kind of publicness that has emerged; produsers produce content they deem relevant on a personal level (rather than on a societal level), which they communicate with audiences with which they have explicit ties (rather than with anonymous mass audiences), and in a conversational way (rather than one-way publishing). In doing so, they manage identities, relationships and information, in complex interplays with one another (Schmidt, 2014).

According to critical scholars, the omnipresence of new communication technologies and its new form of publicness have led to ‘the cult of the amateur’ (Keen, 2008), in which ‘trivial babble dominates over thoughtful knowledge of the experts’ (Schmidt, 2014: 12). This potentially results in a critical stance towards experts and mass media organizations and other institutions such as the government. More optimistically, Schmidt (2014) sees it as a potential site for inclusion and participation. Bruns (2008) also sees these opportunities when exploring the consequence of produsage for democracy, pointing to the fluid heterarchy (as opposed to traditional, strictly hierarchical organizations). However, fluid heterarchies have also changed knowledge production; classic experts are not automatically accepted as experts in digital produsage spheres, but need to re-earn and re-establish their expert position (Bruns, 2008). Consequently, ‘we are moving from the established, taxonomic, expert-driven paradigm into a new and uncharted territory’ (Bruns, 2008: 222); as experts are no longer at the heart of knowledge production, this may change knowledge and production processes in ways we cannot yet fully anticipate, and which we need to explore and research.

This concept of the produser is connected with a concept that is more specific for the consumption of health-related media content: the expert patient (Dumit, 2012), or the expert consumer (Hallin and Briggs, 2016). An expert consumer uses information provided by mass media and other sources to actively make individual and rational choices about health, apart from the supervision of their physicians. She or he is the expert about their own situation, personally constructs what is relevant and adequate knowledge, and becomes empowered because of ‘the genuine potential for making choices’ (Andreassen and Trondsen, 2010: 281), that is, to make informed, personal choices. In doing so, she or he tries to not only minimize the risk of illness, but also maximize well-being and freedom. However, the concept of the expert consumer and the idea of empowerment is also contested, for not taking into account factors such as health literacy and social conditions and determinants, which might impede making informed, empowering decisions (Greenhalgh, 2009).

In any case, similar to the produser who blurs the boundaries between mass media as an omniscient, authoritative producer of information, and the audience as a passive consumer, the expert consumer blurs the boundaries between authoritative professional biomedical experts and laypeople, between who develops and who receives knowledge. Both concepts also raise the question of new forms of expertise and to what extent these forms really change power dynamics (between mass media and audience or between health professionals and patients), and what their impact is on knowledge production and reception.

The concept of the expert consumer is also highly relevant for our analysis with respect to nutrition. Food is increasingly promoted as a means to stay healthy, in which people have a high level of autonomy in making their own choices. In Huovila and Saikkonen’s (2016) analysis of Finnish dietetic blogs, readers are encouraged to prioritize an individualistic understanding of food, bodies and knowledge, over the traditional population-based recommendations concerning nutrition. In the traditional media too, it is a salient view on health; in British and German newspaper coverage on obesity, self-control and individual responsibility are the most frequently used frames (Atanasova and Koteyko, 2017). Consequently, healthy eating is nowadays constructed as a personal, moral responsibility (Henderson et al., 2009; Huovila and Saikkonen, 2016; Sukhan, 2012). This is also reflected in and exemplified by the increased interest in functional foods (Niva, 2007), and the increase of diet shows, cooking shows and books, and non-health professionals campaigning for healthier diets, like Jamie Oliver (Gray and Leahy, 2013).

However, food choices are more complex than nutritional and health considerations, for two reasons. First, the great interest in food has led to an ever-growing supply of information, which is increasingly competing and contradictory (Aschemann-Witzel and Grunert, 2015; Huovila and Saikkonen, 2016). Second, although health might be an important, even moral, imperative, food choices are also surrounded by many other incentives, like taste, peer pressure, ethical considerations (Deliens et al., 2014), financial, religious, ethical and emotional ones, often in complex interplay with one another. Mass media take up different discourses on food and, in doing so, prioritize different incentives; in MasterChef Australia (which has also been broadcasted in Belgium, as well as produced in a local Flemish version), for instance, the dominance of considerations about taste makes nutritional discourses appear irrelevant, implicitly sending confusing and contradictory messages to the audience about which foodstuffs to limit in a normal diet (Phillipov, 2013). De Backer and Hudders (2016) similarly argue that food shows on TV, Belgian and other ones, are nowadays about entertaining and for instance emphasize the pleasurable, fun aspect of cooking, or focus on exploring uncommon ingredients or different food cultures. In Sukhan’s (2012) analysis of Canadian weight-loss shows (which are similar in format to several shows about weight and diet that were broadcasted or produced in Belgium, like Je bent wat je eet (‘you are what you eat’)), on the other hand, food is reduced to ‘an instrumental necessity that is robbed of both pleasure and positive identification of the self’ (p. 199). In sum, nowadays, ‘food choice has become an ongoing negotiation process in which consumers are faced with diverse intentions and expectations’ (Niva, 2007: 385). These intentions and expectations can be social, cultural, ethical, religious, and financial and health related, and, as a result of being so diverse, are likely to contradict each other at some point. Consequently, individuals will have to constantly negotiate with themselves, and with their environments, which ones to prioritize. Moreover, media take part and complicate this process as they produce different, contradictory messages on how to prioritize. This complexity of food choices thus is challenging for the audience, who has these many considerations to include when constructing what is adequate and relevant nutritional knowledge. Moreover, there is a possible tension between this complexity and the infotainment format, which we will further explore next.

Infotainment

Infotainment has incited long-standing debates among scholars, as this increasingly popular format in the commercialized media ecology is often seen as a form of tabloidization (the trend of increasingly bringing media content and news in popular, entertaining and supposedly more artificial formats). Criticasters are concerned that tabloidization leads to a loss of journalism’s watchdog function, and the loss of the audience’s access to important sociopolitical information (Hauttekeete, 2005). Infotainment genres have also been assumed to generate a different and worse understanding of media content in the audience, mainly in relation to comprehension and the capacity to retain and remember information later (Nguyen, 2012; Prior, 2003). Tabloidization practices, especially in political media content on television, have also been associated with media malaise, eroding trust in institutions, and political cynicism (Norris, 2000).

More optimistic scholars argue that dichotomizing entertainment and information and content and style is a simplistic rendition of the media landscape (Hartley, 1996; Hauttekeete, 2005), and that there is no uniform strong and linear trend towards tabloidization (Hauttekeete, 2005). Recent research has also shown that a different understanding of infotainment content is an effect on attitudinal changes, rather than factual knowledge (Kim and Vishak, 2008), that infotainment aspects can even have positive effects on recall, and do not significantly affect comprehension (Mujica and Bachmann, 2016).

In the case of health journalism, both scholars and practitioners have different takes on to what extent media and media content is or should be entertaining, and whether this is positive or negative. Entertainment is often listed as a news value (O’Neill and Harcup, 2009), but health journalists have different takes on what the main aim of health journalism ought to be; some see it as a form of health promotion and public health, while others see entertainment as its primary function, not the potential public health effects (Hallin and Briggs, 2016).

Among health professionals, a similar debate on the role of media in health promotion, and on how to reconcile media logics with health promotion logics, especially in the case of entertaining genres, has been going on. For a long time, health professionals considered health and entertainment as irreconcilable opposites, and health-related media coverage was believed to be health damaging (Seale, 2002). However, health promoters and educators have reached out to the media to reach a wider audience, and have shown to be willing to compromise and accept ‘the pleasure principle that drives most mass media organisations’ relationship with their audiences’ (Seale, 2002: 6), and adapt and incorporate such and other media logics. At the same time, media have reached out to biomedical professionals to produce health-related content (Hallin and Briggs, 2016), incorporating biomedical logics and increasingly coproducing health-related content in close collaboration with biomedical professionals.

Data collection and coding

We collected tweets associated with the hashtag of the show as broadcasted in 2016 through an exhaustive search. Because the free Twitter application programming interface (API) only returns a selection of tweets corresponding to a search term, we collected tweets by searching for the hashtag on the Twitter website, and then programmatically extracting tweets from the search results. A comparison with tweets returned through API calls revealed that our method retrieves a proper superset of the tweets found through the API, that is, all tweets the API returned were also in our data set (which were 1192 tweets). After data collection, we manually removed unrelated tweets that did not concern the TV show, for example, Twitter bots, from the data set, which resulted in a final set of 1181 tweets.

Facebook data were collected using the Pattern Python package (Smedt and Daelemans, 2012). We opted to collect reactions to posts made by the official Facebook page of the infotainment show in 2016, selecting all posts that related to the eight episodes of this season. We manually looked up the ID of the page of the show, and scraped all posts from the page. Following this, we extracted all comments associated with each post and all replies to comments. This resulted in a set of 4787 comments. We manually verified that the extracted set of posts and comments were exhaustive.

One of the posts on the Facebook channel consisted of 2638 comments. Of this post, only the first 160 comments were coded (to match the number of reactions of the second biggest post) to avoid that one post would distort the findings for the total sample. In addition, some comments to other posts were excluded from the analysis when foreign languages were used, or when the post featured only emojis (and no text) of which the meaning was unclear (e.g. hearts, thumbs-up and smiley faces were included; other ones with less canonical meanings were excluded). Other comments were deleted because they merely tagged another person and did not contain a reaction to the show. These alterations resulted in an analysed data set of 1736 Facebook reactions. Our final combined data set consisted of 2917 items, produced by 1054 unique Facebook users and 693 unique Twitter users, resulting in 1747 unique users in total.

The codebook for the content analysis was partly predefined but further inductively developed in a first wave of qualitative content analysis of the first 250 tweets in the data set. New categories were added in a collaborative process of conventional coding (Hsieh and Shannon, 2005). In a second wave, the finalized codebook was applied to the complete data set (N = 2917), including the first 250 tweets, resulting in a descriptive overview of the frequency of use of the different categories and variables in the codebook. Ambiguous cases were marked by coder 1 (the first author) and co-coded by coder 2 (the third author). For those cases, in a next step, the two coders decided together which category was most appropriate to ensure the reliability and validity of the coding procedure. Statistical analysis was carried out using SPSS 22. Finally, the third wave of the analysis was again qualitative, but this time summative coding, which allows for interpretation of the context associated with the use of these categories and variables, which we illustrate in the article with concrete examples of tweets and Facebook posts (Hsieh and Shannon, 2005).

The codebook consisted of three main categories: actors, topic of reaction and evaluation of the information. The actor subcategories were based on a validated codebook from previous studies on health journalism (Deprez and Van Leuven, 2017), and included media, ordinary citizens, government organizations, academic experts and so on.

The topic categories coded the topic of the reaction (RQ1). The first three categories (food/nutrition health, food/production and food/marketing) were predefined based on the ethnographic fieldwork, as they mirror the categorization used by the editors to talk about the different kinds of items they were working on. As we expected reactions on format and entertaining elements, we included these as categories, and an ‘other’ category. During the conventional coding process, we added categories for (self-)promotion (when someone advertised the show and/or told the audience to watch it), the interactive responses to questions to the audience, general evaluative comments on the show, ethics and comments on recipes.

With the evaluation categories, we aimed to capture the users’ attitude towards the information as presented in the TV show (RQ2). We predefined codes for refutation, doubt, acceptance, and as a result of the conventional coding process added suggestions, general resistance, reported intentions to change behaviour and taste. General resistance concerned reactions of users who did not refute a claim, but generally resisted looking at a claim from the particular angle the editors used.

Ethnographic data collection

As mentioned above, we want to add to the understanding of the results by contextualizing them with a few insights from another part of this study, which was linguistic ethnographic in nature. Linguistic ethnography is a specific form of ethnography based on the assumption that language and meaning on one hand and the social world on the other hand are mutually shaping (Creese, 2008; Rampton, 2007), that is, that language and a community’s social and cultural practices both shape and result from each other. Consequently, linguistic ethnographers study language to understand the context, and vice versa (Jacobs and Slembrouck, 2010; Rampton, 2007). The fieldwork took place on 31 non-consecutive observation days (March to May 2016). The data set comprises audio-recorded interviews with editors, presenters, the editor-in-chief, the on-screen experts invited by the editors and camera crew; audio-recorded meetings; audio recordings of shooting days; field notes; the broadcasts and a collection of the press materials on the show. The editor-in-chief, who acted as a gatekeeper, signed an informed consent. Because of privacy concerns, the names of the informants as well as the name of the show are fictionalized.

Ethnographic context

The show was produced by Belgium’s Dutch-language department of the public TV broadcaster, more specifically by the current affairs department. The department produces multiple daily news broadcasts, several debate and information shows and documentaries. Being a state-funded public broadcasting channel, the department has the explicit mission to provide all Dutch-speaking Belgians, regardless of age, ethnicity and other demographic factors, with high-quality programs that contain good information and stimulate public debate. In the same vein, the aim of the infotainment show was to reach a broad audience and properly inform them, in an entertaining way, about all aspects of food: health, nutrition, diet, food production and marketing, as well as production and marketing. The show’s season under scrutiny consisted of eight 45-minute episodes that were programmed in prime time; consequently clear, accessible and straightforward information was key. This became increasingly clear during the ethnography, as the editors construct their audience as (potential) expert consumers, in the web of ever-increasing supply of information on food and health, which is increasingly competing. In the interviews, the editors say that they aim to be a trustworthy source of correct and practical information, to support their audience in making adequate food choices. They are aware of the complexity of these choices. Consider the following extract from an interview with a newspaper, which was part of a larger marketing campaign to promote the show, as it was the first season of the show. In this extract, one of the presenters formulated the unique selling proposition of the show as follows:

We are not conscious enough of what we eat, as a result of the abundance of food. We used to eat to stay alive, now we eat just to fill us up, out of boredom, or because it’s comforting.

The presenter constructs food and eating as belonging to different aspects of life: as a basic physical need, as social and psychological processes and experiences and as a pastime. He consequently claims that we have neglected the most important aspect of food, the basic physical need to stay alive. The editors then argue that, to stay alive and healthy, we need to understand what food does to our bodies, and therefore consider food from a biomedical perspective. This way, we can become expert patients (Dumit, 2012), critical consumers who are actively learning and adapting to nutritional advice and knowledge, and food is constructed as a moral responsibility to stay healthy.

The editors are also aware of the complexity of nutritional information in itself and fear that, because of this double complexity, the focus on nutrition and health could easily become boring, overly didactic or preachy. They aim to avoid this in two ways. They acknowledge that the produsing expert consumer wants, in the words of Domingo et al. (2008), a conversation with, rather than a lecture from the media. Therefore, they address the audience as an active participant in the show, actively monitoring Facebook and Twitter and using them to interact with the audience, for example in the form of quiz questions.

Second, they opt for an infotainment format. This, for example, involves a weekly item with celebrities talking about their food choices and behaviour, or playful interactions between the two presenters. The editors believe an entertaining style was key to reaching a big audience, as it keeps the messages accessible. Moreover, it avoids sounding preachy, and having the produsing consumer expert resisting the information, or not watching the show at all. Consequently, although the editors construct the main aim of the programme as informing the audience about nutrition and health (and food marketing and production), they often do mention or foreground social and psychological, or more entertaining lifestyle aspects, of eating. For example, the presenters regularly discuss their guilty pleasures or weaknesses in terms of eating habits on-screen. In the final episode, in which all items are centred around barbecuing, they filmed and used shots of the crew drinking wine and going for seconds. The editors aimed to highlight that food is not all about making healthy choices, but also about enjoying the food and the company of the people you eat with.

Topics of reactions (RQ1)

To answer RQ1, whether audience members react more to information or entertainment aspects of the show, we take a look at Table 1, giving an overview of to which topics the posts and comments are linked.

OPEN IN VIEWER

Table 1.

Topic of reaction versus platform (N = 2917).

Link with programme	Twitter (n = 1181)	Facebook (n = 1736)	Total (N = 2917)
Information about food/health	258 (21.8%)	292 (16.8%)	550 (18.9%)
Information about food/production	193 (16.3%)	436 (25.1%)	629 (21.6%)
Information about food/marketing	43 (3.6%)	44 (2.5%)	87 (3.0%)
Response to (quiz) questions	106 (9.0%)	153 (8.8%)	259 (8.9%)
Formal aspects of the show	59 (5.0%)	26 (1.5%)	85 (2.9%)
Entertainment	179 (15.2%)	70 (4.0%)	249 (8.5%)
(Self)promotion	103 (8.6%)	62 (3.6%)	165 (5.7%)
Evaluation of the show as a whole	106 (9.0%)	120 (6.9%)	226 (7.7%)
Ethics	31 (2.6%)	163 (9.4%)	194 (6.7%)
Recipe	4 (0.3%)	85 (4.9%)	89 (3.1%)
Other link with programme	132 (11.2%)	316 (18.2%)	448 (15.4%)

*

The sum of all categories can be more than one hundred because some posts contained more than one link with the programme.

The most discussed topics are the factual information as put forward by the editors: most comments and tweets relate to information about health (18.9%), food production (21.6%) and marketing (3.0%). If we look into the division of reactions to the information categories, we see that these correspond with the division of the information as found in the TV show for Twitter, and that Facebook reactions correspond with the topic of the Facebook posts. Of all items in the 8 episodes, 11 related to health (28.2%), 14 to food production (35.9%), and 6 to marketing (15.4%). A similar distribution is visible in the tweets; tweets mainly relate to information about health and nutrition (21.8%) and food production (16.3%), and only to a lesser extent to marketing (3.6%). In the case of Facebook, although most of the 83 posts on the official account of the show were self-promotional or related to quiz questions (45, or 54.2%), we find a slightly different but similar tendency. Twelve posts (14.5%) presented food information related to health, 12 posts (14.5%) presented food information related to food production and 3 posts (3.6%) presented food information related to marketing. This is reflected in the distribution of reactions (N = 1653), which are mainly focused on food information in the show about production (25.7%) and health (16.9%), and to a lesser extent marketing (2.5%). Nonetheless, when taking into account that the exceptionally big Facebook post, where we only coded the first 160 reactions, discussed food information about production (more specifically about colour additives), we expect that the number of Facebook audience reactions in terms of production are in fact higher. Consequently, Facebook users react to information on production relatively more than on Twitter, and relatively more than put forward by the show’s editors.

With respect to entertainment aspects of the show, formal aspects elicited few reactions: only 2.9 per cent of comments discussed issues such as directing style, music and language; 8.7 per cent were comments discussing the celebrity presenters, the celebrities featured in the show and purely entertaining elements (e.g. footage of peeling an apple with a drill). Both these commenting practices were found relatively more on Twitter. However, for both platforms, these results suggest that the entertaining style does not greatly distract the audience from the actual information on food in the show, and that about half of the reactions are directly prompted by the actual, factual information in the show. Moreover, the reactions to the different informational subcategories (nutrition/health, food production and marketing) are distributed similarly to the show’s items and Facebook posts.

Furthermore, during the first coding round, we observed that many users, especially on Facebook (9.4%) (vs. 2.6% on Twitter), started discussions on ethical aspects of eating, mainly on the environmental aspects of eating and not eating meat. However, the show hardly ever discussed these ethical aspects; just once, a vegan celebrity briefly touched upon the ethical motivations of his choice. It thus is remarkable that, especially on Facebook, ethics discussions account for 9.4 per cent of reactions, while these were not prompted or elicited by the show itself. It confirms that food choices are complex, and that people actively bring other aspects into the picture than the ones the editors highlighted.

Evaluation of information (RQ2)

To answer RQ2, Table 2 gives an overview of the users’ evaluations of the claims on health, food production and marketing (for this analysis, we only included the 1229 comments and tweets in the informational categories). Approximately 34.3 per cent of all reactions accept the information presented in the show, about equally divided between Twitter and Facebook; 10.9 per cent of the users report they will change their lifestyle, confirming or following what they learned from the show. Yet, in most cases, audience members present themselves as doubting (4.7%), refuting (14.3%) or resisting (5.9%) the presented food information. Twitter reactions contain more refutations (17.4% vs. 12.3%), and express more doubt (4.9% vs. 4.6%) and resistance (9.4% vs. 3.5%) compared with Facebook reactions. In a quarter (25.3%) of all reactions, users make further suggestions adding to the information presented in the show; this is especially the case for Facebook (31.3%, vs. only 16.2% on Twitter).

OPEN IN VIEWER

Table 2.

Evaluation of information versus platform (n = 1229).

Evaluation	Twitter (n = 488)	Facebook (n = 741)	Total (n = 1229)
Refutation	85 (17.4%)	91 (12.3%)	176 (14.3%)
Doubt	24 (4.9%)	34 (4.6%)	58 (4.7%)
Acceptance	172 (35.2%)	249 (33.6%)	421 (34.3%)
Suggestion	79 (16.2%)	232 (31.3%)	311 (25.3%)
Resistance	46 (9.4%)	26 (3.5%)	72 (5.9%)
Intention to change behaviour	58 (11.9%)	76 (10.3%)	134 (10.9%)
Taste	24 (4.9%)	39 (5.3%)	63 (5.1%)

Interestingly, some reactions take a different perspective by focusing on the importance of taste (5.1%). For instance, an item on the use of carmine, a pigment extracted from a scale insect, elicited a lot of disgusted reactions, including many reactions of users stating they would no longer consume products containing carmine. However, some users expressed taste took precedence over considerations concerning food production or health, stating for example that they would keep consuming certain products because they are tasty:

(1) Don’t give a fuck, it tastes good so I’ll keep drinking it (Facebook)

In other cases, some users suggested the tastiness of healthy products as a secondary incentive to consume them:

(2) Light and tasty (Facebook reaction to a recipe of a healthy, low-calorie Caesar salad)

Again, this points to the fact that food choices are complex, and that the audience often considers and takes preference of other aspects, like taste, than the one the editors use as the dominant frame in their content.

To gain a better understanding of the dynamics of acceptance, refutation and other evaluations, we look at these evaluations in relation to the information type in Table 3 (considering the low number of cases in the category of information about marketing (N = 78), the numbers for this category need to be interpreted with prudence). For production, acceptance is high (43.4%) and refutation rather low (6.6%). For health, we see a different trend: acceptance is notably lower (22.5%) and refutation is high (23.0%), even slightly higher than acceptance. In addition, audience members more often suggest different angles when discussing information about health (31.7%), and as such more actively participate in constructing and debating health knowledge than knowledge on production (19.5%) and marketing (17.9%). This suggests that most audience members see themselves as expert consumers, who do not just accept the health information as presented to them, but actively add to the existing information and debates.

OPEN IN VIEWER

Table 3.

Evaluation of information versus topic of reaction.

Evaluation	Information about health (n = 521)	Information about food production (n = 595)	Information about marketing (n = 78)
Refutation	120 (23.0%)	39 (6.6%)	17 (21.8%)
Doubt	25 (4.8%)	27 (4.5%)	4 (5.1%)
Acceptance	117 (22.5%)	258 (43.4%)	35 (44.9%)
Suggestion	165 (31.7%)	116 (19.5%)	14 (17.9%)
Resistance	47 (9.0%)	20 (3.4%)	2 (2.6%)
Intention to change behaviour	24 (4.6%)	103 (17.3%)	1 (1.3%)
Taste	23 (4.4%)	32 (5.4%)	5 (6.4%)

*

The total number of reactions linking with the show in terms of information about health/food production/marketing is lower than in Table 1 because not all reactions contain an evaluation of the presented information.

In the following section, we examine some relevant categories and use examples to gain a deeper understanding of the different reactions in these categories. The refutations both concerned the information itself (4) and the science behind it (5) and how facts were presented by the editors (6):

(4) Another myth confirmed in #hashtagshow: that a daily glass of red wine supposedly is healthy. (Twitter)

(5) Lots of talking about calories again. That’s getting #oldschool when talking about food and more importantly about remaining slim. What about #glycaemicindex anyone? (Twitter)

(6) Who lets his teeth soak in coca cola for a month anyways?! #bullshit #hashtagshow #hashtagbroadcaster (Twitter)

To prove that the acidity of fizzy drinks is harmful for our teeth, the editors set up an experiment in which they soaked teeth, eggs (egg shells have a calcified structure, similar to teeth) and a rusty screw in several fluids, like orange juice and coca cola, for a month. After this month, the egg shells had disappeared and the teeth were blackened, which clearly conveyed that acid drinks are harmful for teeth. However, the audience member in (6) found the experiment too far-stretched.

The suggestions category contains explicit requests towards the editors (9); reactions in which audience members state that the information was incomplete and add new aspects (7-8); reactions which, without suggesting the information was incomplete, highlight another aspect of the topic discussed; and all further discussions among users following these suggestions (8):

(7) Yes … if they withhold the truth about the sickening wheat in bread well then it’s not worth watching this show! Not to mention the glycaemic index of bread being higher that of refined sugar like that (Facebook)

(8) User 1: And … it’s a pity you haven’t mentioned puffed rice cakes are unhealthy. They contain rat poison! A whole grain sandwich is way healthier.

User 2: puffed rice cakes contain arsenic, and there’s also arsenic in rat poison. But there’s NO rat poison in puffed rice cakes. Don’t be fooled! (Facebook)

Finally, these suggestions sometimes reframe the items of the show, highlighting other aspects of food and touching upon the complexity of food choices:

(9) Please also pay attention to the unhealthy aspect of not eating ENOUGH. 2/6 Eating Disorder day. #hashtagshow (Twitter)

Another piece of evidence of audience actively taking up the role of the expert consumer is the reactions to report intentions to change behaviour (10.9%). Interestingly, information about food production (17.4%) most often provokes such reactions, while information about health and nutrition elicits fewer responses (4.6%). Further analysis of the data shows that most (95 of 103, or 92.2%) of these behaviour change reactions are elicited by the ‘the way back’ items about food production, which shows the trajectory of a foodstuff from the plate to the unprocessed product, played backwards. In four of the eight episodes, these items contained graphic images of animals being butchered. Another episode graphically showed how a chicken laid an egg. These images often elicited reactions of disgust, especially in the case of young or photogenic animals, like rabbits, as well as many audience members expressing they never ate these animals, intended to no longer do so, or encouraged others to no longer do so:

(10) I already didn’t like it, but now never again for sure! (Facebook)

(11) Because of #hashtagshow I am seriously considering becoming a vegetarian @publicbroadcaster #fromhamburgertolamb #seriouslydisgusting (Twitter)

The audience members thus seemingly experience the information about food production as more shocking, and tend to construct its impact as bigger. Although these numbers do not provide us with any information on actual behaviour, they again illustrate the complexity of food choices, and, in this case, point to a dominance of emotional and psychological aspects of eating, over nutritional considerations.

Finally, the reactions expressing resistance (5.9%) indicate the audience members’ critical attitude towards nutritional information, as resistance is higher for nutrition/health (9%) than for food production (3.4%) and marketing (2.6%). This category contains reactions which, explicitly or through more implicit means like irony, state the information provided by the show unnecessarily complicates food choices, that resist the health frame in case of nutritional and health information, or that express a concern that everything we eat nowadays is somehow problematic.

(12) No I didn’t know. Is there something you still can eat without it containing all kinds of bad stuff? (Facebook)

(13) This show actually proves we eat crap every day. Do we really benefit from knowing that? (Facebook)

(14) What is the goal of #hashtagshow, to keep us all from eating? (Twitter)

These reactions all express a concern with the increasing considerations concerning food choices and the growing body of contradictory information. Example (13) even implies that ignorance might be better than being knowledgeable, implying that information is overwhelming rather than helpful, and does not always lead to more informed and better choices. These reactions thus construct the stream of information as disempowering, rather than empowering, and resist being constructed as expert consumers.

Some resistance reactions are humorous, some audience members proudly posited they were enjoying unhealthy food:

(15) [Watch the show] calmly? With two bags of crisps and a bottle of coke alright! #ifthatainthealthy #hashtagshow #watchit (Twitter)

Although the show aimed to positively influence food choices and health behaviour, these reactions construct watching the show as a leisure activity, meant to relax and have fun, which includes eating snacks. The higher percentage of resistance for health information thus indicates that audience members find that nutritional and health considerations are more restrictive and invasive than food production and marketing considerations.