From the preserves of the educated elite to virtually everywhere: A content analysis of Danish science news in 1999 and 2012

Abstract

In an attempt to qualify changes to science news reporting due to the impact of the Internet, we studied all science news articles published in Danish national newspapers in a November week in 1999 and 2012, respectively. We find the same amount of science coverage, about 4% of the total news production, in both years, although the tabloids produce more science news in 2012. Online science news also received high priority. Journalists in 2012 more often than in 1999 make reference to scientific journals and cite a wider range of journals. Science news in 2012 is more international and politically oriented than in 1999. Based on these findings, we suggest that science news, due partly to the emergence of online resources, is becoming more diverse and available to a wider audience. Science news is no longer for the elite but has spread to virtually everywhere in the national news system.

Keywords

online and print newspapers quantitative content analysis science news web-based information and communication technologies

1. Introduction

On 4 July 2012, CERN physicists announced the discovery of a possible Higgs boson. A few days prior to the announcement, rumours began to spread on Twitter and soon became ‘truly global’ (De Domenico et al., 2013: 4). Social media, news feeds and CERN’s own webcast made sure that the news about the Higgs boson candidate, one of the most elusive particles in the history of modern science, reached millions of users, readers and viewers across the globe. As the CERN Communication Group, slightly surprised, observed, the Higgs-like boson received rock star treatment, making its appearance in unexpected media, such as the most popular tabloid newspaper in the United Kingdom, The Sun, where the Page 3 model ‘reflected’ on the discovery, and in the tweets of celebrities such as MC Hammer (Kahle, 2012).

Although in many respects atypical, the production and circulation of news about the Higgs-like boson demonstrate the impact of the Internet on science reporting and the flow of science news. The number of science communication channels has amplified, and science news has become omnipresent. In ‘the electronic embrace of the Internet’, science reporting seems characterized by the emergence of new producers of information, new channels for distributing information and new audiences, all of which not only provide new opportunities for news media, research institutions, scientific journals and concerned citizens but also challenge traditional news values and practices (Trench, 2007).

The purpose of this study is to compare quantity, triggers, sources and demographics of science news before and after web-searching, social media and online media became pervasive in professional journalism. As Danish researchers, we conveniently focus on Denmark, a small country where the number of national newspapers is low. Even with limited resources, it is possible to collect and code all of the news items relating to science that have been published in Danish media over an extended period of more than a few days. Partly for the same reason, Denmark has been home to a unique quantitative and longitudinal study of general news. The continuing research project A News Week in Denmark coded all news items published in a given November week in 1999 and again in 2008 and 2011, aiming to trace the circulation of news in Danish media (Lund, 2013; Lund et al., 2009).

2. Trends in newspaper coverage of science

Science news in the new media ecosystem

The rise of online environments for news has been accompanied by several, contradictory changes in the way that traditional news media like newspapers process news stories. At the level of management, editorship, reporting and marketing, newspapers, throughout the 1990s and 2000s, still struggled to make best use of the web (Boczkowski, 2005; Van der Wurff and Lauf, 2005). Still today, many newspapers are undergoing financial problems, while at the same time having to produce additional round-the-clock content for their online platforms.

There are no precise accounts of how newspapers in their coverage of science respond to the challenge of producing more news on multiple platforms in an environment characterized by rapid technological development and increased competition. The emergence of numerous web-based resources for science journalists, such as online news services provided by high-ranking scientific journals, specialized science news agencies and academic institutions, surely has made it easier for science journalists to access science stories without leaving the office or even picking up the phone. In an age of tight resources, Williams and Clifford (2009) suggest, this would tempt science journalists, as other journalists, to pick the ‘low-hanging fruits’ (p. 42) in the guise of readymade news stories and consequently lead to fewer independently researched science stories in the news. Granado (2011), surveying European science journalists working for general news media and news agencies, concludes that, indeed, science journalists are becoming more dependent on the services provided by scientific journals and in their daily work spend a lot of time on the Internet. Similarly, British science journalists in interviews have conjectured that the idea that ‘the web is never full’ generally encourages online science reporters to increase news production at the expense of news quality (Williams and Clifford, 2009). Fahy and Nisbet (2011), on a more positive note, see no reason to predict uniform decline in the standards of science news, but rather develop a typology for the plurality of roles that science journalists are expected to play in the new ‘science media ecosystem’ (p. 783).

Content-wise, the advent of online news platforms so far appears to have had little impact on science news in general. Although there are only very few studies available, all of which focus on special scientific topics, they all find that online science news items tend to differ but slightly from their offline counterparts (Gerhards and Schäfer, 2010; Habel et al., 2009; Hyde, 2006; Wilson et al., 2009). To take an example, Hyde (2006) found that online news articles about genetic cloning, produced from 1996 to 1999, tended to include fewer quoted sources than printed articles in the sample; however, the difference was partly explained by the fact that the print news stories were as much as 20%–70% longer than their online stories (Hyde, 2006: 239). With respect to their language, tone and emphasis on the possible future benefits that may be associated with cloning, online and print articles were found to be strikingly similar.

Longitudinal studies of science news

Longitudinal studies provide valuable information on long-term changes in the amount and characteristics of science news (Bauer et al., 1995, 2006; Bucchi and Mazzolini, 2003; Clark and Illman, 2006; Elmer et al., 2008; Krieghbaum, 1941; Logan et al., 2000; Pellechia, 1997). Although it may be difficult to compare results across these studies, since with a few exceptions they all employ different protocols for sampling and coding articles, most conclude that the amount of science news in the media have increased since the 1980s.

The first longitudinal study of science in the news probably was Krieghbaum’s (1941) analysis of 97 American newspapers from 1939 to 1941, showing that about 5% of non-advertising space was dedicated to science reporting. Surprisingly, almost the same percentage of science news was found about 50 years later in the study performed by Bauer et al. (1995), who included 6000 articles from seven daily British newspapers from 1946 to 1986. Like Krieghbaum (1941), Bauer et al. (1995) defined science news in broad, ‘catholic’ terms and measured the amount of science news relative to the effective number of news pages exclusive of advertising space. They found a slight increase in the amount of science news starting in the early 1980s.

In Italy, where the British study was replicated by Bucchi and Mazzolini (2003), the amount of science news found in the leading Italian newspaper Il Corriere della Sera grew dramatically from about 1% in the period from 1946 to 1950 to 21% in the late 1980s, reaching 29% by the early 1990s. Bauer et al.’s (2006) comparative study of science news in the Daily Telegraph (Great Britain) and in the Rabotnichesko Delo/Duma (Bulgaria) from 1946 to 1995 (1992 for the Telegraph) concluded that the intensity of science coverage in both newspapers could be modelled by a waved cubical curve. The curve peaked in the early 1960s, reached a low by the end of the 1970s (the Telegraph) and early 1980s (the Rabotnichesko Delo) and then continued to rise throughout the 1980s and into the 1990s. In a more recent longitudinal study of science news in European newspapers, Elmer et al. (2008), looking specifically at three nationwide newspapers in Germany in two periods, 2003–2004 and 2006–2007, reported an overall increase in science reporting by 48%. Although they employed a stricter definition of science news than the British/Bulgarian and the Italian studies, they found the absolute increase in science news outside the science sections to be a staggering 136% (Elmer et al., 2008: 883).

Groves et al. (2015) are the first to conduct a longitudinal study of online science news. They sampled 50,763 text units from three major Spanish newspapers from 2002 to 2011. In absolute terms, the quantity of science news quadrupled, but in relative terms the amount of science coverage remained steady at around 7%.

On the other side of the Atlantic, similar, although more modest, growth in the absolute and relative amount of science news has been detected. Pellechia (1997), in her study of science news reporting in the New York Times, the Chicago Tribune and the Washington Post, found that all three newspapers had increased their emphasis on science news from 1966 to 1990. The absolute number of science sections rose from 18 in the period 1966–1970 to 56 in 1986–1990 and the percentage of science articles from 0.42% to 2.04% (p. 57). A later study of all science-related content in the New York Times’ science section from 1980 to 2000, performed by Clark and Illman (2006), observed a continuation of this upward trend. They reported that the amount of editorial science content in Science Times had more than tripled in the period, with the largest increase occurring from 1995 to 2000 (Clark and Illman, 2006: 504).

Sources and triggers

Traditionally, science journalists as well as general journalists tend to favour national events as triggers for their news stories. Bauer et al. (1995), for example, found that more than two-thirds of the British news devoted to science in the period from 1946 to 1986 was based on national events. As science journalists increasingly adapt web-based tools for getting and sharing information about events in science all over the globe, the national emphasis in science news stories may tend to be replaced by a more international outlook. From around 2000, using the web to search for material and get in touch with sources became standard practice for science journalists across the globe (Trench, 2007).

Many longitudinal studies of science news have looked specifically at newspaper articles that are triggered by the completion of scientific research printed in journals. Studies have found that journal articles account for about half of the science news found in the United States and European newspapers (Clark and Illman, 2006; De Semir et al., 1998; European Commission, 2007; Suleski and Ibaraki, 2010). However, Weitkamp (2003) in an earlier study from 2000 and 2001, when online tools were not predominant, found only 15% of science news in the United Kingdom to be triggered by a journal article. Recently, Kiernan (2014) in a content analysis of the New York Times’ science section from 1998 to 2012 found an increase in the use of journals as well as in the number of different journals cited, attributing the observations to the increased online availability of journals.

Elmer et al. (2008: 882) made a useful distinction between scientific triggers, that is, events within the world of research, such as publication of papers, congresses and announcements from scientific institutions; non-scientific triggers, such as epidemics and political decisions; and combined scientific and non-scientific triggers, such as rocket launchings or policy-making related to scientific issues. They found that about half of the news articles in dedicated science sections were triggered by scientific events, whereas scientific triggers only accounted for 12%–15% of the science news in other sections of the newspapers (Elmer et al., 2008: 883).

Science news in Denmark

In 2003, the Danish Government introduced a new Act on Universities, establishing public communication of science and technology and knowledge transfer as a third mission of the university (Nielsen, 2005). Kjærgaard (2008), looking at news about nanotechnology published in Danish newspapers between 1996 and 2006, suggests that the framing of science news is closely linked to the national agenda provided by government policy-making and research initiatives. Nanotechnology, mainly due to its novelty as a research field and the hype attached to it, admittedly is a special case, but Kjærgaard’s (2008) conclusion about the embeddedness of science news in regional and national news contexts appears to be applicable across countries (for other studies of nanotechnology in the news, see Anderson et al., 2005; Kulve, 2006; Stephens, 2005).

The science media ecosystem in Denmark increasingly uses experts with academic credentials as sources of information and commentary (Albæk et al., 2003). Albæk et al. (2003) analysed 1286 articles sampled in 1 month in 1961, 1971, 1981, 1991 and 2001, respectively, to find that, in absolute and relative terms, the number of articles mentioning researchers has grown. In particular, social scientists have come to play a dominant role. As in other countries, knowledge and information seem to flow regularly and smoothly from Danish researchers and research institutions to the media (Albæk, 2011; Peters et al., 2008).

As mentioned, Lund et al. (2009) and Lund (2013) mapped the flow of news in the general media ecosystem in Denmark. Their methodology was used as the foundation for this study, and so, their results provide an important context for our findings. All news items appearing in Danish news media in a November news week in 1999, 2008 and 2011, respectively, were collected and coded. Surprisingly, given that the number of news-producing journalists has remained steady over the years, the total amount of news increased from just below 32,000 articles in 1999 to just above 75,000 in 2008 (Lund et al., 2009: 9). One of the reasons why Danish journalists have become so much more productive in terms of the number of articles they put out is that the amount of non-original material, defined as news stories originating directly from another news media or news agency, also increased. The tendency to publish ‘readymade news’ was most discernible online. A comparable figure resulted from Van der Wurff and Lauf’s (2005) comparative study of newspapers in 16 countries, where it was found that 70% of the content in online news articles stemmed from news agencies.

Drawing on the above literature, we focus on certain characteristics of science news in Danish newspapers in 1999 and 2012, such as the total amount of science news in print and online, types of trigger events for science news, references to scientific journals and sources, length of articles, amount of ‘readymade news’ and more. The year 1999 is identical to the first year in which the study of general news-flow in Denmark was carried out, and 2012 was the most recent year available at the time of sampling. Importantly, the two years 13 years apart provide insights into the situation before and after the emergence of online newspapers in Denmark. Comparing our results from the 1999 and 2012 samples, respectively, we hope to be able to discern traces of recent changes to the small science media ecosystem in Denmark.

3. Sampling and coding method

This study deals with nine national newspapers in Denmark, all of which can be accessed and searched on the leading Danish provider of media intelligence, InfoMedia: Jyllands-Posten (online: jyllands-posten.dk), Politiken (online: politiken.dk), Berlingske (online: b.dk), Information (online: information.dk), Kristeligt Dagblad (online: kristeligt-dagblad.dk), BT (online: bt.dk), Ekstrabladet (online: eb.dk), MetroXpress (online: mx.dk) and Weekendavisen (online: weekendavisen.dk). All but Weekendavisen (literally The Weekend Newspaper) are published on a daily basis. BT and Ekstrabladet are the Danish tabloids, while MetroXpress (established in 2001) is a free newspaper. The remaining six constitutes the quality press in Denmark. We also included the dedicated, online science news service Videnskab.dk. Jyllands-Posten was the first news organization in Denmark to publish an online edition in 1994; the remaining newspapers had an online presence by the turn of the millennium, except for MetroXpress (online edition opened in 2007).

In 1999, there was just one science section in all of the nine newspapers, published on a weekly basis in Berlingske. In 2012, the number was near five: two sections in Weekendavisen, two weekly ones in Jyllands-Posten and Politiken dedicating a couple of pages in its Sunday edition to science.

Using InfoMedia, which does not include online articles from 1999, we sampled science articles from the printed editions of the newspapers in the period 15–21 November 1999 and all science articles, printed as well as online ones, in the corresponding period 12–18 November 2012. We recognize that using seven consecutive days instead of a constructed week reduces the generalizability of our results (Krippendorff, 2013). However, using consecutive days has a number of advantages for this study. First, it allows us to track the developments of stories over multiple days and locate sources and triggers not present in one isolated article. Also, we can account for the number of duplets in the form of identical articles published online and in print by the same media. Using identical weeks to the Lund et al. (2009) study also allows for direct comparison. Krippendorff (2013) does not explicitly warn against all use of consecutive weeks, only cautions not to base too general claims on this method. Riffe et al. (1993) cites numerous reliable studies using consecutive weeks.

We define science news as editorial content focusing on scientific knowledge, findings, methods, processes, opinions, events or institutions within all scientific fields. Our broad approach to what constitutes science news is similar to other studies such as Bauer et al. (2006), Bauer et al. (1995) and Elmer et al. (2008). We operationalized our science news definition using the following search terms to extract articles from Infomedia: researcher, science, journal, research, knowledge, survey, study, expert, professor and university (in Danish: forsker, videnskab, tidsskrift, forskning, viden, undersøgelse, studie, ekspert, professor, universitet – note that the Danish term for science, ‘videnskab’, as the German ‘Wissenschaft’, includes all faculties). In 1999, Infomedia did not include articles from Kristeligt Dagblad, which had to be searched using microfilm. In addition to the automated search, all news headlines from all selected newspapers were screened manually on lists generated by Infomedia to ensure that all science news indeed had been collected. We then went through all articles, removing those few ones that did not fit our definition of science news. Pragmatically, if neither the headline, nor the subheading, nor the introduction made reference to scientific aspects, the article was discarded. A total of 693 articles (44 duplets, 15% of online science news was identical to print editions) were selected and coded.

Each article was coded manually using 10 categorical variables in addition to basic information about the name and type of the newspaper (quality press or tabloid) and the article’s page number, section, length, publication date and platform (print or online; see Table 1).

Table 1.

Coding frame.

Variable	Categories	Clarifying definitions
Academic field	Natural sciences, social sciences, humanities, health, engineering, science in general	Natural sciences: Physics, chemistry, astronomy, biology and so on
		Social sciences: Political science, economy, law, sociology and so on
		Humanities: History, communication, culture, linguistics, archaeology, religion and so on
		Health: Medicine, psychology, sport, gastronomy and so on
		Engineering: Inventions, technological and so on
		Science in general: Research policy, research ethics and so on
		Multiple fields: Applied if a primary academic field is not identifiable
Triggers	Non-scientific, scientific, non-scientific and scientific	Scientific: Journal article, scientific conference, opinion expressed by scientist(s) and so on
		Non-scientific: Natural events, policy-making, public events and so on
		Non-scientific and scientific: Rocket launch, research policy initiatives and so on
Geographical origin of trigger	Danish, Foreign, Danish and foreign	Danish: Research from Danish institutions, events in the Danish science community and so on
		Foreign: Research from foreign institutions, international events and so on
		Danish and foreign: International research with Danish contribution or international events with a specific Danish news angle
Type of scientific trigger	Journal article, new research without journal reference, conference, report or release, other	Journal article: Peer review articles published in journals
Not applied if trigger is non-scientific		New research without journal reference: New research without explicit journal reference
		Conference: Scientific conferences
		Report or release: Publication or announcement of non-peer reviewed surveys and articles, PhD theses, book release, videos and so on
Number of interview subjects		Subjects named and quoted in the article that have a scientific affiliation
Nationality of interview subjects	Same as ‘geographical origin of trigger’
Media source	Original, Danish media, foreign media, news agency, other	Original: No media source identified, indicating that this may be an original news story not previously published by other news media
		Danish/foreign media: The story is based on an existing news item found in Danish or foreign media

The code ‘academic field’ characterizes the academic field most prominent in the article. To code academic fields, we classify disciplines according to main faculties most commonly used in Denmark: natural science, health science, social science and the humanities. The code ‘trigger’ captures the type of events that demonstrably led to the publication of the article. If the article makes explicit reference to scientific events, such as the publication of new research, it is coded as ‘scientific trigger’. In this case, the supplementary code ‘scientific trigger’ is used to retrieve more detailed information. The code ‘geographical origin of the trigger’ is used to recover information about whether the article originates from Danish or international events, or both. The number of named and quoted interview subjects is coded, as well as their nationality. The code ‘media sources’ captures stories that have been produced by other media or news agencies, but not press releases produced by universities or scientific journals.

One author (Gunver Lystbæk Vestergård) examined each of the articles, while another (Kristian H. Nielsen) co-coded 10%, the amount required for testing inter-coder reliability (Krippendorff, 2013). A calculation of Krippendorff’s alpha coefficient resulted in inter-coder reliability values between .81 and .93, which is generally acceptable.

4. Results

Since the consecutive days sampling method used in this study has a potential bias due to in-between week differences, we searched our material for specific events that may have resulted in unusually high or low science coverage in our two November weeks. One such event, the so-called Dandy case was found in the 1999 sample. However, this case appears to have affected only our ‘academic field’ variable. When we eliminate articles about this event, the values of all other variables do not change significantly (Tables 2 and 3).

Table 2.

Results: 1999 and 2012.^a

	1999		2012		Print 2012		Web 2012
Number of news articles	5697		11,437		5303		6076
Number of science news articles^b	211	4%	482	4%	184	3%	298	4%
Average word count per article	549		505		557		472
Academic field
Natural sciences	36	17%	69	14%	22	12%	47	16%
Social sciences	37	18%	129	27%	55	30%	74	25%
Humanities	30	14%	102	21%	43	23%	59	20%
Health	56	27%	159	33%	53	29%	106	36%
Engineering	10	5%	13	3%	4	2%	9	3%
Science in general	41	19%	6	1%	4	2%	2	1%
Multiple fields	1	0%	4	1%	3	2%	1	0%
Total	211	100%	482	100%	184	100%	298	100%
Triggers
Scientific	125	60%	246	51%	91	49%	155	52%
Non-scientific	75	36%	229	48%	89	48%	140	47%
Non-scientific and scientific	9	4%	7	1%	4	2%	3	1%
Total^c	209	100%	482	100%	184	100%	298	100%
Geographical origin
Danish	148	72%	313	65%	117	64%	196	66%
Foreign	48	23%	151	31%	57	31%	94	32%
Danish and foreign	10	5%	18	4%	10	5%	8	3%
Total^c	206	100%	482	100%	184	100%	298	100%
Scientific trigger
Journal	8	6%	96	39%	35	38%	61	39%
Unpublished new research	25	20%	67	27%	24	26%	43	28%
Conference	4	3%	2	1%	1	1%	1	1%
Other scientific report or release	25	20%	50	20%	17	19%	33	21%
Other	63	50%	31	13%	14	15%	17	11%
Total	125	100%	246	100%	91	100%	155	100%
Interview subjects
Number of subjects	165		491		182		309
Number of subjects per article	0,8		1,0		1,0		1,0
No subjects	83	41%	119	25%	52	28%	65	22%
One subject	82	41%	253	52%	88	48%	165	55%
Two subjects	28	14%	97	20%	39	21%	60	20%
More subjects	8	4%	13	3%	5	3%	8	3%
Total^c	201	100%	482	100%	184	100%	298	100%
Nationality of subjects
Danish	140	85%	405	82%	153	83%	254	82%
Foreign	24	15%	86	18%	31	17%	55	18%
Total^c	164	100%	491	100%	184	100%	309	100%
Media origin
Own research	164	81%	279	58%	139	76%	139	47%
Other media	13	6%	113	23%	27	15%	87	29%
News agency	26	13%	84	17%	15	8%	69	23%
Other	0	0%	6	1%	3	2%	3	1%
Total^c	203	100%	482	100%	184	100%	298	100%

Chi-square tests were all significant at p < .05 or lower.

Relative frequency based on total number of news items. Confidence interval (CI) (alpha = .05): 1999 = 3.0%–4.4%, 2012 = 2.8%–4.4%. Videnskab.dk was excluded from the calculation of relative frequency for Web 2012 as they only publish online science news and therefore risk skewing the result.

Some articles from 1999 could not be coded in all categories, resulting in sums lower than total number of articles.

Table 3.

High-quality papers versus tabloids.

	HQ 1999		Tabloid 1999		HQ print 2012		Tabloid print 2012		Web HQ 2012		Web tabloid 2012
Number of science news articles	199		12		145		22		127		72
Average word count per article	548		560		608		267		484		408
Academic field
Natural sciences	32	16%	4	33%	20	14%	1	5%	20	16%	4	6%
Social sciences	36	18%	1	8%	47	32%	2	9%	42	33%	18	25%
Humanities	30	15%	0	0%	37	26%	4	18%	33	26%	9	13%
Health	52	26%	4	33%	33	23%	12	55%	29	23%	39	54%
Engineering	10	5%	0	0%	3	2%	1	5%	2	2%	2	3%
Science in general	39	20%	2	17%	4	3%	0	0%	1	1%	0	0%
Multiple fields			1	8%	1	1%	2	9%	0	0%	0	0%
Total	199	100%	12	100%	145	100%	22	100%	127	100%	72	100%
Triggers
Scientific	116	58%	9	75%	69	47%	12	55%	54	43%	36	50%
Non-scientific	72	36%	3	25%	72	50%	10	45%	72	57%	35	49%
Non-scientific and scientific	9	5%	0	0%	4	3%			1	1%	1	1%
Other	2	1%
Total	199	100%	12	100%	145	100%	22	100%	127	100%	72	100%
Geographical origin
Danish	140	72%	8	67%	94	65%	11	50%	90	71%	42	58%
Foreign	44	23%	4	33%	46	32%	7	32%	32	25%	30	42%
Danish and foreign	10	5%			5	3%	4	18%	5	4%	0	0%
Total	194	100%	12	100%	145	100%	22	100%	127	100%	72	100%
Scientific trigger
Journal	8	7%	0	0%	27	40%	1	25%	18	33%	16	44%
Unpublished new research	20	18%	5	56%	20	29%	3	25%	18	33%	11	31%
Conference	4	3%	0	0%	1	1%	0	0%	1	2%	0	0%
Other scientific report or release	23	20%	2	22%	10	15%	3	25%	12	22%	7	19%
Other	59	52%	2	22%	10	15%	1	25%	5	9%	2	6%
Total	114	100%	9	100%	68	100%	12	100%	54	43%	36	100%
Interview subjects
Number of subjects	158		7		151		16		153		67
Number of subjects per article	0,8		0,6		1		0,7		1		1
No subjects	75	40%	8	67%	40	28%	8	36%	14	11%	18	25%
One subject	80	42%	2	17%	65	45%	12	55%	77	61%	43	60%
Two subjects	27	14%	1	8%	35	24%	2	9%	32	25%	9	13%
More subjects	7	4%	1	8%	5	3%	0	0%	4	3%	2	3%
Total	189	100%	12	100%	145	100%	22	100%	127	100%	72	100%
Nationality of subjects
Danish	133	84%	7	100%	126	83%	13	81%	129	84%	54	81%
Foreign	25	16%			25	17%	3	19%	24	16%	13	19%
Total	158	100%	7	100%	151	100%	16	100%	153	100%	67	100%
Media origin
Own research	154	81%	10	83%	118	81%	11	50%	62	49%	21	29%
Other media	11	6%	2	17%	20	14%	6	27%	26	20%	33	46%
News agency	26	14%		0	5	3%	4	18%	39	31%	17	24%
Other	0	0			2	1%	1	5%	0	0%	1	1%
Total	191	100%	12	100%	145	100%	22	100%	127	100%	72	100%

HQ: high qulaity.

Amount of science news

We find that the absolute number of articles more than doubled from 211 in 1999 to 482 in 2012. In relative terms, excluding Videnskab.dk that publishes only science news, the amount of science news relative to the total news production remained steady at 3.7%, or rather 3.0%–4.4% in 1999 and 2.8%–4.4% in 2012 (alpha = .05). We found that in 2012 more articles were published online than in print (298 online, 184 in print), which means that the absolute rise in the amount of science news can be solely accredited to the emergence of online newspapers. In particular, by 2012, the tabloids were publishing more than three times as many science stories online as in their printed editions.

With respect to the academic fields covered, we observe small shifts from 1999 to 2012. The social sciences, the humanities and the health sciences received more coverage in 2012 (the tabloids in particular favoured health news) compared to 1999, whereas there was a decline in the relative frequency of articles covering the engineering and natural sciences. The fact that the category of science in general decreased dramatically from 1999 to 2012, we put down to the Dandy case mentioned above.

Non-scientific versus scientific triggers

The number of science news articles triggered by events outside of science increased in relative importance from 36% to 48%, while scientific triggers decreased from 60% to 51%. Combined non-scientific and scientific triggers remained low at 4% and 1%, respectively.

Use of scientific journals

We find that scientific journal articles are more often used as triggers for science news in 2012 compared to 1999. In 1999, just 6% of the science news articles included references to scientific journals; in 2012, the corresponding figure was 39% (of all articles triggered by a scientific event. If applying the variable to all science news, regardless of trigger event, an increase in journal articles as trigger is still visible, from 4% in 1999 to 20% in 2012). We specifically note that, in 1999, only scientific research published in a few, well-known, high-prestige journals, namely, Science, Nature and JAMA, would trigger science news in Danish newspapers. In 2012, articles from a wide variety of journals from Toxicology through Sexual Medicine to Nature Materials, Paediatrics and PLOS ONE were used to trigger science news stories. We sampled and coded just 12 science news articles in total in the two Danish tabloids in 1999, none of which made explicit reference to scientific journals. In 2012, the tabloids would mention scientific articles as triggers in 25% (print) and 44% (online) of their science stories triggered by a scientific event.

Counterweighing the increasing use of scientific journals was the decrease in other scientific triggers from 50% in 1999 to 13% in 2012. The value ‘Other’ covers feature stories, descriptions of new lab initiatives, research projects, controversies and other coverage of science in the making.

National versus foreign triggers

In 1999, 72% of all science news articles were triggered by a Danish event; in 2012, the figure was 65%. The corresponding figures for foreign trigger events were 23% and 31%, respectively. More than 80% of all interview subjects with a background in research were affiliated with Danish research institutions. This holds for our 1999 and 2012 samples. Thus, somewhat contrary to our expectations, science news in Danish newspapers remains closely connected to the national research system and national agendas. However, when we looked only at news articles specifically triggered by scientific events, we did see a growth in the use of international science events as triggers: The amount of articles triggered by national science events was 70% in 1999 and 54% in 2012, while the same numbers for articles triggered by international science events were 22% and 41%, respectively.

Demographic of science news

The average length of science news articles was 549 words in 1999 and 505 in 2012 (about 8% in difference). The difference mainly is due to the emergence of online articles, which tend to be shorter. The average length of online articles in 2012 was 472 words, while printed articles in 2012 averaged 557 words. A science news article in a quality press newspaper (print edition) had 608 words in 2012, but just 548 words in 1999.

Along similar lines, we find that the number of interview subjects is about the same in 1999 as in 2012. This goes for printed as well as online articles. The number of articles using no interview subjects with a scientific affiliation was 41% in 1999 and 25% in 2012, whereas the same numbers for articles with one interview subject were 59% and 75%.

Amount of original material

In agreement with Lund et al. (2009), we find that there are fewer original science news articles in 1999 compared to 2012. In 1999, original science news articles accounted for 81% of all articles, and in 2012 the same figure was down to 58%. About 53% of online science news articles in 2012 were found to be based on previously published news stories from other media sources. Especially online tabloid news was found to only contain 29% original material. We find another tendency to be prevalent in quality press: More than 80% of all science news articles printed in high-quality newspapers in 1999 and 2012 were based on original material. Of all the non-original science news articles in 2012, 23% could be attributed to other media and 17% to news agencies. Foreign media seemed to have grown in importance as sources of inspiration for Danish journalists: In 2012, 8% of all science news articles cited foreign media, up from just 2% in 1999. Among the most quoted international media, we find the Norwegian science portal Forskning.no, the online media Livescience and the online edition of British media such as The Guardian, The Daily Mail, The Telegraph, and BBC. Videnskab.dk is credited as media source in 9% of all 2012 science news articles, which makes it the most-used source of information and a good example of the potential impact of a Government-sponsored, online outlet for science news stories.

5. Discussion

Longitudinal studies of science news in other European countries, that is, Great Britain, Bulgaria, Italy and Germany, have found an increase in the amount of science news in the later decades of the twentieth century and into the 2000s. These studies looked only at printed science news articles, and none are as recent as this study. We could not confirm the global trend of increased science news publication in the Danish news media. Relative to the entire news production in Denmark, the percentage of science news was the same in our two sample periods, 15–21 November 1999 and 12–18 November 2012, namely, about 4%. However, our conclusion corresponds to the finding by Groves et al. (2015) that the amount of science news has only increased in absolute terms but not in relative terms. Incorporating Bauer et al.’s (2006) conclusion that the historical development of the intensity of science coverage follows a waved cubical curve, we speculate that the amount of science in mass media have remained relatively stable over decades independent of cultural factors.

The growing importance of online news media for science news in Denmark means that science stories are becoming available to new audiences, in particular those who prefer tabloid newspapers to quality ones. Our results show that the two Danish tabloids, Ekstrabladet and BT, featured very few science stories in their printed 1999 editions. In the 2012 sample, the number of science stories in the two tabloids was considerably higher, and their online editions carried by far most of them. Unfortunately, tabloids rarely feature in studies of science news. If the increased presence of science stories in tabloid media, due to the rise in online news environments, is more than a Danish phenomenon – and we would expect so – tabloids certainly merit more attention from science communication scholars.

Like Elmer et al. (2008), we note that science news quite often is triggered by non-scientific events. However, the most noticeable difference in triggers for science news, we find, is the more pronounced tendency to use scientific publications as triggers for science news in 2012 compared to 1999. We speculate that this has to do with the increased availability of scientific journals and papers online and also, more importantly, with the emergence of online news services, such as EurekAlert and AlphaGalileo, dispatching press releases providing journalists with easy access to information about new scientific publications.

We also suggest the availability of online material influenced the observed differences between the amount of original versus non-original science stories featured in 1999 and 2012, respectively. The number of original science news was higher in 1999 than in 2012, and in 2012 online newspapers carried the least amount of original material. This could be due to the availability of readymade science news online, which is easy to cite and use. In the 2012 sample, original in-depth articles appear more frequently in the print editions compared to online. We conclude that newspapers have begun to differentiate between their print and online science news services with the printed editions featuring most of the original (and longer) science news stories, a finding which is comparable to Lund’s (2013) results. This differentiation deserves more attention in future studies.

It has been argued that due to the emergence of web-based information and communication technologies, the science media ecosystem has become intrinsically global. On the other hand, news systems traditionally are primarily national in their outlook. Our 1999 and 2012 results substantiate these two somewhat conflicting trends. First, the geographic origin of all trigger events, the origin of scientific triggers and the use of media sources in 2012 are more international compared to 1999, indicating that journalists more often have to negotiate the traditional news value of cultural proximity up against online availability of scientific material from international sources. However, the nationality of interviewed sources was predominantly Danish in both 1999 and 2012. One interpretation is that the media emphasize a Danish ‘spin’ on science news through the use of local sources – regardless of the geographical origin of the story.

Although the generalizability of the comparative results from 1999 and 2012 is low, we do suggest our study points towards certain trends within the science news culture. In 1999, science news was predominantly found in elite newspapers and based on national scientific events. In 2012, science news was broadly available online and in tabloids, triggered by scientific as well as non-scientific, often political, events and more frequently based on international sources. Therefore, as exemplified by the breaking of the discovery of the Higgs particle, we suggest the variety of audiences exposed to science news has increased with consequences for the public understanding of science that remain understudied. These empirical findings correspond with the idea of an ongoing medialization of science stating that scientific debates in the media have become more pluralized, diversified and egalitarian (Schäfer, 2009).

Footnotes

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author biographies

Gunver Lystbæk Vestergård is an industrial PhD student at the Experimentarium Science Centre and Centre for Science Studies, Aarhus University, Denmark. She is doing quantitative and qualitative studies of science news in Denmark and Great Britain.

Kristian H. Nielsen is an Associate Professor at the Centre for Science Studies, Aarhus University, Denmark. His research interests include history of science, science in society and science communication.

References

Albæk

(2011) The interaction between experts and journalists in news journalism. Journalism 12: 335–348.

Albæk

Christiansen

Togeby

(2003) Experts in the mass media: Researchers as sources in Danish daily newspapers, 1961–2001. Journalism and Mass Communication Quarterly 80: 937–948.

Anderson

Allan

Petersen

Wilkinson

(2005) The framing of nanotechnologies in the British newspaper press. Science Communication 27: 200–220.

Bauer

Petkova

Boyadjieva

Gornev

(2006) Long-term trends in the public representation of science across the ‘Iron Curtain’: 1946–1995. Social Studies of Science 36: 99–131.

Bauer

Ragnarsdottir

Rudolfsdottir

Durant

(1995) Science and Technology in the British Press, 1946–1990. London: Science Museum.

Boczkowski

(2005) Digitizing the News: Innovation in Online Newspapers. Cambridge, MA: MIT Press.

Bucchi

Mazzolini

(2003) Big science, little news: Science coverage in the Italian daily press, 1946–1997. Public Understanding of Science 12: 7–24.

Clark

Illman

(2006) A longitudinal study of the New York Times science times section. Science Communication 27: 496–513.

De Domenico

Lima

Mougel

Musolesi

(2013) The anatomy of a scientific rumor. Scientific Reports 3: 2980.

10.

De Semir

Ribas

Revuelta

(1998) Press releases of science journal articles and subsequent newspaper stories on the same topic. JAMA: Journal of the American Medical Association 280: 294–295.

11.

Elmer

Badenschier

Wormer

(2008) Science for everybody? How the coverage of research issues in German newspapers has increased dramatically. Journalism & Mass Communication Quarterly 85: 878–893.

12.

European Commission (2007) European Research in the Media: What Do Media Professionals Think? Brussels: European Commission. Available at: http://ec.europa.eu/research/conferences/2007/bcn2007/journalists_en.pdf

13.

Fahy

Nisbet

(2011) The science journalist online: Shifting roles and emerging practices. Journalism 12: 778–793.

14.

Gerhards

Schäfer

(2010) Is the internet a better public sphere? Comparing old and new media in the USA and Germany. New Media & Society 12: 143–160.

15.

Granado

(2011) Slaves to journals, serfs to the web: The use of the internet in newsgathering among European science journalists. Journalism 12: 794–813.

16.

Groves

Figuerola

Quintanilla

MÁ

(2015) Ten years of science news: A longitudinal analysis of scientific culture in the Spanish digital press. Public Understanding of Science. Available at: http://pus.sagepub.com/content/early/2015/04/02/0963662515576864.abstract

17.

Habel

Liddon

Stryker

(2009) The HPV vaccine: A content analysis of online news stories. Journal of Women’s Health 18: 401–407.

18.

Hyde

(2006) News coverage of genetic cloning when science journalism becomes future-oriented speculation. Journal of Communication Inquiry 30: 229–250.

19.

Kahle

(2012) The Higgs Boson: Treated like a rock star. CERN Bulletin, Issue 30–31. Available at: http://cds.cern.ch/journal/CERNBulletin/2012/31/News%20Articles/1462248

20.

Kiernan

(2014) Source diversity among journals cited in Science Times. Public Understanding of Science. Available at: http://pus.sagepub.com/content/early/2014/07/24/0963662514542908.abstract

21.

Kjærgaard

(2008) Making a small country count: Nanotechnology in Danish newspapers from 1996 to 2006. Public Understanding of Science 19(1): 80–97.

22.

Krieghbaum

(1941) American newspaper reporting of science news. Kansas State College Bulletin XXV(5): 1–73.

23.

Krippendorff

(2013) Content Analysis: An Introduction to Its Methodology, 3rd edn. London: SAGE.

24.

Kulve

(2006) Evolving repertoires: Nanotechnology in Daily newspapers in the Netherlands. Science as Culture 15(4): 367–382.

25.

Logan

Zengjun

Wilson

(2000) Science and medical coverage in the Los Angeles Times and the Washington post–A six-year perspective. Science Communication 22(1): 5–26.

26.

Lund

(2013) Mangfoldighed i dansk dagspresse [Diversity in the Danish daily press]. Copenhagen: Møller forlag.

27.

Lund

Willig

Blach-Ørsten

(2009) Hvor kommer nyhederne fra? Den journalistiske fødekæde i Danmark før og nu [Where does the news come from? The journalistic food chain in Denmark then and now]. Aarhus: Ajour.

28.

Nielsen

(2005) Between understanding and appreciation: Current science communication in Denmark. Journal of Science Communication 4(4). Available at: http://jcom.sissa.it/archive/04/04/A040402/

29.

Pellechia

(1997) Trends in science coverage: A content analysis of three US newspapers. Public Understanding of Science 6(1): 49–68.

30.

Peters

Brossard

De Cheveigné

Dunwoody

Kallfass

Miller

. (2008) Science-media interface: It’s time to reconsider. Science Communication 30: 266–276.

31.

Riffe

Aust

Lacy

(1993) The effectiveness of random, consecutive day and constructed week sampling in newspaper content analysis. Journalism & Mass Communication Quarterly 70(1): 133–139.

32.

Schäfer

(2009) From public understanding to public engagement: An empirical assessment of changes in science coverage. Science Communication 30(4): 475–505.

33.

Stephens

(2005) News narratives about nano S&T in major U.S. and Non-U.S. newspapers. Science Communication 27(2): 175–199.

34.

Suleski

Ibaraki

(2010) Scientists are talking, but mostly to each other: A quantitative analysis of research represented in mass media. Public Understanding of Science 19(1): 115–125.

35.

Trench

(2007) How the internet changed science journalism. In: Bauer

Bucchi

(eds) Journalism, Science and Society: Science Communication between News and Public Relations. New York, NY; London: Routledge, pp. 133–142.

36.

Van der Wurff

Lauf

(eds) (2005) Print and Online Newspapers in Europe: A Comparative Analysis in 16 Countries. Amsterdam: Het Spinhuis.

37.

Weitkamp

(2003) British newspapers privilege health and medicine topics over other science news. Public Relations Review 29(3): 321–333.

38.

Williams

Clifford

(2009) Mapping the Field: Specialist Science News Journalism in the UK National Media. Cardiff: Cardiff University School of Journalism, Media and Cultural Studies. Available at: http://www.cardiff.ac.uk/jomec/resources/Mapping_Science_Journalism_Final_Report_2003-11-09.pdf

39.

Wilson

Bonevski

Jones

Henry

(2009) Media reporting of health interventions: Signs of improvement, but major problems persist. PLoS ONE 4: e4831.