How gender affects the newsworthiness of sports news on German TV: An application of the news-factors approach to understanding gender-biased sports news presentation

Abstract

Gender inequalities in sports media are well-documented. This study focuses on sports news composition and how gender influences the prominence of sports news stories. The news-factors approach offers a causal explanation for the lower prominence (i.e. newsworthiness) of women's sports in TV sports reporting. Following this theory's perspective, athletes’ gender is supposed to work as a moderating variable on news values of news factors in sports reports. The content analysis of seven German sports news programs reveals whether the same news factors are treated unequally with regard to women's and men's sports in TV news coverage. The results show that women's sports are presented as less newsworthy than men's sports, although news factors do not differ significantly by gender. However, the moderation effect of gender does not cause lower newsworthiness. That means, e.g., sports women's successes are equally emphasized as the success of male athletes in sports news on TV, and gender does not lower the credits female athlete's success receive in any given news stories. Instead, the results suggest that gender works as a news factor of its own, reducing not the news value of certain news factors but the overall newsworthiness of women's sports in TV coverage. Thus, the results demonstrate that gender inequality in sports media does not necessarily come from journalists perceiving female athletes’ performance as inferior but from presenting women's sports less often and in a far less prominent way than men's sports.

Keywords

gender media sports journalism news factors content analysis

Introduction

Gender inequalities in sports media are a well-documented fact. A myriad of research shows how sports media have marginalized female athletes and women's sports events in past decades (Cooky et al., 2015) through today (Cooky et al., 2021) and in media markets all over the world (Horky and Nieland, 2013). Sports reporting seems to follow and reproduce masculine hegemony (Cooky et al., 2013; Kaiser, 2018; Kian et al., 2008; Vincent, 2004). Meân (2017) lists plenty of research showing how sports media coverage focusing “on sports that re/produce established gender differences as sex-typed (e.g. football for men, gymnastics for women) or gender-neutral sports such as tennis and track" (Meân, 2017: 100) has further established gender differences. In light of these findings, the current patterns of male-dominated sports coverage have their roots not so much in the media but rather in society itself. Bernstein and Kian, 2013: 320 point out that many studies of gender imbalance in sports media are based on the underlying assumption of cultivation theory (Gerbner et al., 2002) that "how members of society see themselves, and how they are viewed and treated by others, is determined to a great extent by their media representations.” That may be even more severe when the lack of media interest "leads to the fact that sponsors likewise take no interest in the sport, which in turn affects the game and development, thus contributing generally to its marginalization" (Pfister, 2010: 241).

On the other hand, the sports journalism profession is dominated by men (Lapchick et al., 2020), making it a masculine and macho-oriented workplace (Bernstein and Kian, 2013: 324). The interwoven interdependencies of sports, media, sponsorship, and public perception of women's sports thus contribute to a culture making it hard for female athletes to succeed in their respective sports and sports as a profession. Moreover, with the large audience of sports in the media, the patterns presented in media sports do likely impact other areas of society. Thus, by examining gender structures in sports coverage, one can also more generally study the emergence and impact of gender inequalities at a somewhat broader societal level.

However, Wenner (2015) suggests that in order “to get at better understandings about how inequities and social problems may be facilitated by mediated sports content, we need to look beyond content to why, how, and under what conditions that content is created […]” (Wenner, 2015: 252). Therefore, this study focuses on the composition of sports news shows revealing how the gender of athletes biases decisions on prioritizing sports news stories. The news factors approach conceptualizes news selection in a way that provides a causal explanation for the difference in importance of women's sports compared to men's sports in TV sports coverage. Furthermore, the two-component theory of news selection as presented by Kepplinger and Ehmig (2006) allows for deconstructing published news stories into the components of news factors and news values of these factors, where the latter are a result of journalists' perception of the reported news. Therefore, this study utilizes the news factors approach to give an additional view on how underlying editorial and journalistic decisions may lead to further imbalances in the representation of women in sports news coverage.

Literature review and theoretical framework

Empirical evidence of gender inequalities in sports media

Regarding sports telecasts, women's sports receive less on-air clock time than male athletes’ competitions, although the gap has narrowed over the last decades when it comes to the Olympics (Billings et al., 2010; Billings and Angelini, 2007; Eastman and Billings, 2000; Tuggle et al., 2002). No more than 10% of coverage in the sports press worldwide is dedicated to women's sports (Beck, 2013; Fredericks and Keim, 2013; Fry and Goh-Leong, 2013; Godoy-Pressland, 2014; Jakubowska, 2013; Le Clair, 2013; Petrica and Ilincescu, 2013). Not even the increasing spread of digital media channels has yet resulted in a significant shift of these persistent patterns of male-dominated sports news (Burroughs et al., 2021; Coche, 2021; Ramon and Rojas-Torrijos, 2021). As for women's sports on TV, longitudinal studies of sports reporting (i.e. on television news and highlight shows) show that coverage proportions have not changed significantly in the last decades (Cooky et al., 2013, 2021; Duncan and Messner, 2005). In the U.S. between 1989 and 2014, reports on women's sports never made up more than 10% of sports coverage (Cooky et al., 2015; Messner, 2002). On German national TV news shows and German regional sports television programs, no more than 10% of sports coverage is devoted to women's sports (Ihle, 2018a, 2018b). The differences between men's and women's sports coverage are not limited to the actual amount of coverage but the sports presentation: Greer et al., 2009 point out that the coverage of male athletes offers more visual variety than coverage of women's sports. Therefore, men's sport is generally presented in more exciting ways than women's sport (Cooky et al., 2015). In summary, with the low amount of coverage of women's sports as well as their lack of prominence in the media compared to men's sports, sports reporting contributes to framing women's sports as being less important than men's sports (Cooky et al., 2013; Godoy-Pressland, 2014; Kaiser, 2018).

The role of news selection for imbalances in sports reporting

News on any media agenda results from selection processes based on prioritization by news agencies, editors, and journalists. However, news selection does not end with choosing some events for publication while rejecting others. News selection also applies to the prioritization of some news stories over others. All stories will not be covered to an equal extent or equally prominent. Journalists and editors decide which events to report on in-depth and which will only be mentioned in the margins. The theory of news factors conceptualizes these underlying processes. As Kepplinger and Ehmig, 2006 point out, newsworthiness is a measure of how prominently news stories are presented compared to other published news. Therefore, the lower newsworthiness of women's sports coverage results from journalists’ and editors’ decisions about which sports news they decide to emphasize more than others.

As conceptualized by Östgaard (1965) and Galtung and Ruge (1965), the theory of newsworthiness suggests that the extent to which certain news factors are present in a news story explains its selection for publication. Thus, it should be possible to find sports media presenting women's sports as equally prominent as men's sports. That is because the underlying structures of sports do not differ radically between women and men – a victory is a victory, and a defeat is a defeat. The importance of an Olympic medal for a domestic athlete in any given sport is the same for a female athlete as for a male athlete. However, according to Schulz (1976), newsworthiness represents how journalists perceive different features of news stories to be more or less relevant. Thus, the newsworthiness of news stories can be measured by observing the length and placement of published news stories. As Kepplinger (1998) suggests, news stories with equal news factors still may be presented differently, according to different importance (i.e. news value) journalists assign to these given news factors. Since sports do differ in their features by disciplines and not so much by gender of athletes, news factors should only differ in between different sports but not in between women's and men's sports. Therefore the empirical evidence of the lower amount and prominence of women in sports media suggests that journalists treat the same sports features differently for women's sports.

Kepplinger and Ehmig (2006) point out that several news stories could feature the same news factors but can still be treated very differently in terms of their newsworthiness in different newspapers or magazines. They argue that different news factors can contribute to the newsworthiness of a news story differently depending on editorial departments’ policies. Thus, measuring the occurrence of news factors, a regression analysis can reveal the actual impact of these factors on journalists’ publication decisions. This impact will describe the news value of the news factors. This approach's actual merit lies in the fact that it uncovers why news with the same news factors may be considered to be of different importance by different editorial departments and journalists. While some journalists tend to see surprising or sensational stories as more relevant for their readers, other journalists may consider stories that affect a wide range of people as more important. This approach has been proven to explain the different news agendas of the tabloid press, national quality press, and regional newspapers (Kepplinger and Ehmig, 2006). The same logic can be applied to assess differences in journalists’ evaluations of different topics that share certain features (news factors). The research interest then relates to different evaluations of similar topics rather than to differences between evaluations by different journalists or newsrooms.

Still, the lower coverage of women's sports is only one possible indicator of journalists and editors treating women's sports as less important than men's sports. Additionally, women's sports news could also be presented as less newsworthy. This would result from news stories being relatively short, being shown not very variegated, and less prominently, which would – in the case of TV sports news mean – being unlikely to be presented as a lead story. E.g., the news of the national men's soccer team coming second in the World Cup is likely to be presented as a lead story followed by an interview and additional reports on fans’ reactions. Thus, that story would be of high newsworthiness. In contrast, the news of the female national hockey team becoming World Champion is more likely to be presented as a short news clip towards the end of a sports news show. Therefore, that story would be of much lower newsworthiness. These two examples illustrate that newsworthiness is not a natural feature of any event. Instead, it results from how journalists edit and present any news story.

This study aims to show if women's sports are presented as less newsworthy and how the news value of several news factors contributes to this outcome. In this regard, the results of this study will shed additional light on the differences in journalists’ publication decisions regarding women's and men's sports. To do that, this study applies the two-component model of news selection as presented by Kepplinger; Kepplinger and Bastian; Kepplinger and Ehmig (1998; 2000; 2006) to TV sports reports. While Kepplinger; Kepplinger and Bastian; Kepplinger and Ehmig (1998; 2000; 2006) ask for differences in news selection of specific media types for the same news content, this study asks for discrepancies between different news content (women's sports vs. men's sports) within the same media types. The empirical approach of assessing the newsworthiness of news stories and news values of news factors presented in this study has been applied to political TV news by Maier (2003b) and Maier et al. (2006).

Application of news factors theory on the field of sports communication

Despite originating in political communication, sports communication scholars have also adopted the news factors approach. Loosen (1998); Loosen (2001); Loosen and Ravenstein (2000) established a set of sport-specific news factors, such as success and failure, irregularities, (short) duration, and the uncertainty of the outcome, to add to already-known news factors, like simplicity, surprise, geographical proximity, and personalization. Rehbach et al. (2016) showed that these news factors have been very consistent over several decades of sports reporting in German newspapers.

However, news factors research, especially in sports communication, has emphasized the need to point out differences between several media outlets rather than focusing on differences between types of sports or between women's and men's sports. Contributing to why different types of sports gain different amounts of media attention, Horky (2009) suggests distinguishing sports in terms of their inherent media suitability. Combining media logic and differences between sports, Horky (2009) points out that not all sports are equally suited for media coverage. Horky (2009) argues that a sport's suitability for media coverage relies on sport-inherent structures, such as movement forms, the degree of organization, and possibilities of identifying with the protagonists. This approach alone cannot answer why there is a lower number of news stories on women's sports than on men's sports. It could only be valid if the sport-inherent structures differed between women's and men's sports. However, the athlete who crosses the finish line first will always be the winner of a race, regardless of whether it is a competition between men or women.

Nevertheless, differences in the media suitability of sports must be considered when explaining the lower newsworthiness of women's sports. Since the actual characteristics of sports usually do not differ between women's and men's sports, it can be assumed that there is little or no difference in news factors in sports news stories when they are split according to gender. Nevertheless, this may only be true for sports with the same sport-inherent structures.

As Kepplinger and Bastian (2000) and Kepplinger and Ehmig (2006) point out, the actual newsworthiness of a news story results from the combined influences of news factors and the news values of these news factors. Figure 1 illustrates this conception of the approach. Therefore, research must distinguish between event characteristics and their journalistic evaluation. The characteristics are both the actual sport and the news factors present in the news story. On the other hand, the journalistic evaluation is the relevance attribution journalists give a specific news story. The approach was initially designed to evaluate the differences in news selection and presentation between different journalists or newsrooms. The idea behind the approach is that the same event is evaluated differently by several journalists, newsrooms, or editors. Thus the very same news story of, e.g., a marriage of two Hollywood actors, will appear as a lead story in one news show, and another news show will only be mentioned as a side note. The reason is that the same news factor (human interest) is considered to have more news value for the journalists of the one news show but less important by the journalists responsible for the latter. In this study, this reasoning is applied to different evaluations of news factors in sports news on women's sports and men's sports. Simply put, the idea here is that a news story on a world championship in any sport will make a headline if the champion is a male athlete but will be less likely to make a headline if the world champion is a woman. In this case, the news factor success is treated differently in news stories on women's and men's sports.

Following Horky (2009), there should be no difference in the media suitability of the same sport with regard to whether it is a women's or men's sporting event. The same should apply to news factors. A sporting event can be either characterized by the fact that a favored team wins or loses or based on any unexpected outcomes (news factors success and failure and surprise). The same is true for local athletes or international stars from abroad (news factors, geographical proximity, and personalization).

Since literature suggests that women's sports news is presented as less important than men's sports, sports news probably features two main characteristics. First, women's sports news is likely to be of lower news value than men's sports. Since the same sports do not differ in media suitability by the gender of the competing athletes, the second assumption is that the occurrence of event characteristics (news factors) actually does not differ significantly between women's and men's sports news.

(H1) The newsworthiness of news stories about women's sports is lower than the newsworthiness of news stories about men's sports.

(H2) The news factors in sports news do not differ significantly between women's and men's sports.

Assuming the above hypotheses are correct, this will leave the open question of why stories’ newsworthiness on women's sports is lower than that of men's sports. As Kepplinger and Ehmig (2006) suggest, the same news factors can be of different news value. Regarding gender differences in sports reporting, the same news factors could be perceived as less relevant when it comes to publication decisions regarding women's sports news. For example, while journalists may decide to make a lead story from a man's world championship, they may decide otherwise for a female world champion. In this case, the gender of the competing athletes could work as a moderating variable that influences the newsworthiness of the present news factor (success) and, consequently, the actual newsworthiness of a news story. So if the same event characteristics (news factors) do not differ in their occurrence but do differ in the impact they have on the newsworthiness of women's and men's sports, gender would work as a moderator on newsworthiness. This moderator would show that the news value of news factors is lowered if sports news covers women's sports.

(H3) The gender of competing athletes works as a moderating variable that influences news factors’ news value.

Method

Content analysis

These hypotheses were tested by conducting a content analysis on a two-year sample of recordings of TV sports news shows of three national and four regional German public-service television networks that were initially broadcasted in 2014 and 2015. In Germany, there are national and regional public service broadcasting networks. On the national level, the two flagship national television channels (Das Erste and ZDF) have been included in this study. On the regional level, a sample of three network stations (MDR, NDR, WDR) was chosen for this study. In 2014, these three stations reached the most viewers nationwide and within their respective broadcast areas, compared to all regional public service broadcasters. Overall, the two main public service programs have a market share of 24 percent of TV viewers, while the regional programs each average about 2 percent of viewers (Zubayr and Gerhard, 2016). While there are commercial TV stations on a national and local level in Germany, there is no competition between commercial and public service stations for sports news shows. That is because German public broadcasters have exclusive broadcasting rights for a large variety of sports in Germany. These rights are based on an agreement with 34 German sports associations representing a majority of German sports. Thus, commercial TV stations broadcast some sporting events for which they hold airing rights but do not feature sports news shows that cover a variety of sports. Therefore, only public broadcasters were included in this study, as sports news programs are not broadcast on commercial television due to the lack of broadcasting rights.

Because the amount of women's sports coverage is much smaller than that of men's sports, a rather sizeable two-year sample was drawn to ensure there were enough reports on comparable sports for both men's and women's sports. So, the sampling units of this content analysis were the individual editions of every program, and the reports within these editions were the coding units.

News selection always reflects on the given events of a day or a week, and therefore news selection decisions are a consequence of choice between all potential stories. The presentation of these stories then reflects on the attributed newsworthiness of these events. The sample includes all units of the population of the selected broadcasts in the period under review. Thus, the seasonally varying news situation is reflected in the sample precisely as it was in the analyzed period. At the same time the rather large sample should diminish influences of specific events which may per se have high newsworthiness, such as world championships or Olympic Games.

The level of journalistic attention (i.e. newsworthiness) was assessed by ranking the reports according to their position in each program's edition. Thus, the first report in a show is the lead story and therefore is ranked first, with the second story ranking second place, and so forth. The coders had to decide whether each report's topic was included in further reports and whether the topic was announced within the broadcast before the actual report. The duration of each report was measured. Each coding unit was assigned to a category from a list of sports and one of four categories based on the gender of the athletes in the competition in question (female, male, mixed, or the report covers several competitions).

News factors were measured using 14 variables based on previous research (Loosen, 1998; Loosen and Ravenstein, 2000; Rehbach et al., 2016): the actuality of the event, geographical proximity, the violation of rules, success, controversy, harm, personal influence, human interest, the level of the elite sportsperson, the unambiguity of the report, surprise, uncertainty, factuality, and the duration of the reported events (Table 1). Each news factor's actual intensity in the report was coded on a four-point scale. The more a report emphasized a given news factor or the more the reported event matched with a particular news factor, the higher the rating of that factor. However, the violation of rules, surprise, and uncertainty factors were coded on a two-point scale (yes/no). The decision to code these news factors as dichotomous variables was based on a pre-test which showed that coding did not reach an acceptable level of reliability if coders had to decide in which intensity these news factors were emphasized in a given news story. But the reliability was on a very good level if they only had to decide if the news factor was present at all. Especially the news factor violation of rules troubled coders because it was hard to distinguish between the intensity of the rule violation and the intensity the violation was emphasized in a report. For example, mentioning a red card in a report on a soccer match would not necessarily increase the intensity of the news factor violation of rules more than telling about a yellow card. That would only be the case if the red card were significantly more prominent in the report.

Table 1.

Definitions of news factors used in this study.

News factor	Definition according to the codebook
Actuality of the event	The actuality of the event refers to the time of the event in relation to the broadcast of the analyzed program.
Geographical proximity¹	Geographical proximity refers to the venue's geographical distance from Germany.
Success²	Success includes not only victories and good rankings, but also positive consequences of sporting activity. Vice versa the lack of a certain result can be described as a failure. Is a result achieved presented as a success or a failure?
Controversy	A controversy is present if the story contains the explicit presentation of differences of opinion. The differences of opinion can be expressed verbally, in writing or even physically.
Harm	Harm is present when the negative consequences of an event are explicitly presented in the news story. The harm can relate to material, non-material and existential circumstances.
Personal influence	Personal influence refers to the importance given to individuals in an event. The more prominent a person is presented relative to institutions or groups, the more personalized the news story.
Human interest	Human interest refers to the level of private information about the athletes as opposed to information about the actual sporting event.
Level of elite sportspersons³	If athletes are referred to as successful athletes, the elites level determines if athletes are successful on international, national or regional level.
Unambiguity of the report	The more a news story focuses on one single event the more unambiguous it is. The more aspects are covered the more ambiguous a news story is.
Factuality	Factuality refers to the presentation of the reported event as an event. Is the event presented as an action/incident/occurrence or is it or classified by interpretations, commenting statements and the like?
Duration of reported events	Does the sporting event last one day or does the report refer to a somewhat long-term process?
Violation of rules is reported	Rule violations are violations of fixed norms and regulations of competition of the sport
Surprise is expressed	An event is surprising if it cannot be anticipated or if it defies expectations.
Uncertainty is expressed	Uncertainty refers to the extent to which an event is reported as not decided for good.

Geographical proximity was measured using the variables event taking place in Germany and regional proximity was derived from mentioned cities or counties. The proximity variable defines whether an event actually took place in the broadcast area of the given network (highest ranking on four point scale) or is linked to the area by a competing team or athlete. If there is no actual link to the region the lowest rank was coded.

The news factor success was originally coded in two variables (success and defeat) but recoded to one four-point scaled variable. It was assumed that it is easier for coders to decide whether success and/or failure occur independently in a report than to decide how the two relate to each other. The recoded variable has four points where 4 means success without any attribution of failure 3 means success and failure are leveled out, 2 means neither success or defeat are attributed and 1 means defeat or failure without any attribution of success.

It should be stressed that only information given within the coded reports was considered for coding. For example, if the world champion of a certain discipline sport participated in a competition, but it was not stated that the athlete was the best in the world, then the level of an elite athlete would not be classified high because the elite level was not mentioned.

Loosen and Ravenstein (2000) pointed out that several news factors could be further analyzed in terms of their interdependencies by principal component analysis (PCA), which could reduce a rather large set of news factors. This study aims to show if and how gender affects the news value of certain established and easily comprehensible news factors. For example, it would be much harder to see how the news value of the news factor success was affected by gender if success was consumed by an abstract PCA computed factor. Therefore a PCA was omitted in favor of more effortless comprehensibility of the results.¹ Instead, the news value of the empirically measured news factors is preferred as the basis for further analysis.

Assessment of the level of newsworthiness

The value of a news report can be measured based on various formal indicators, such as the duration, positioning, and appearance of a report. Based on such variables, an index can be assessed for the news value of the news item, and thus a distinction can be made between less noteworthy items and leading news items (Maier, 2003b: 91). As a measure of each report's actual newsworthiness, an indicator variable was computed. This procedure is similar to the method Ruhrmann et al. (2003) and Maier et al. (2006) established to assess the newsworthiness of reports on television news shows (Maier, 2003b: 91–92; Maier et al., 2006: 58–59). Each report was ranked by the quartile of its position in the single edition of the program (1 point if the report was in the first quartile of the show, 2 if it was in the second, and so on). Similarly, each report was ranked based on its duration. The ranking was based not on the single editions but on every program's dataset as a whole². Reports were given one to four points based on the quartile of each program's report duration. One additional point was given to the coding unit if the report's topic was referenced in at least one teaser within the program. The same logic applied if there was a reference to the topic being covered in another program. Suppose the topic spread across multiple reports within the program; an additional point was scored for each extra report, up to a maximum of three points. These points were summarized for each report and divided by the number of ranking variables (five). To eliminate possible influences of different conventions of program arrangement, the resulting index was then normalized to a scale ranging from 1 (lowest attention level) to 100 (highest attention level). Thus, the normalized index is comparable across the different sports programs. Thus a report from any sampled sports news show would have a value of 100 if it had the highest possible newsworthiness and a value of 1 if it had the lowest newsworthiness. This normalized index is easier to understand than the variables it consists and is at the same time a better approximation to newsworthiness than just the duration of a report.

Reliability

Twenty-one students coded the material. Intercoder reliability was computed using the Lotus coefficient and the SPSS macro developed by Benjamin Fretwurst (Fretwurst, 2015a, 2015b). Although the Lotus coefficient is not as widely adopted as other broader known reliability coefficients, it has the distinct advantage of not being based on pairwise comparisons of coders. Instead, the Lotus coefficient refers to the most common coded value and thus the value implied by the codebook. Therefore, Lotus will avoid that most wrongly coded values contribute positively to the overall reliability. (Fretwurst, 2015a). Above that, the Lotus coefficient and the SPSS macro developed by Benjamin Fretwurst can be used without any restructuring of the coded data as it is often the case with other reliability coefficients. Thus, it was preferred above other established reliability coefficients for this study. The overall intercoder reliability exceeded 90% (Table 2). The reliability of the variables used to measure and compute newsworthiness (duration, teaser, topic, reference) shows overall reliability of 87%. The comparatively low reliability value of the coding of the duration of the coding units (Lotus = .67) is somewhat mitigated in the data analysis because the measured duration of the report does not directly impact the news value. Instead, the reports were ranked by their duration within the respective program, and that ranking was used to compute the newsworthiness (see the previous paragraph). The coding of the news factors reaches an overall reliability score of 91%.

Table 2.

Reliability of coding measured by Lotus coefficient.

Variable	Lotus	Standardized Lotus
Date	.99	.99
Broadcaster	.99	.99
Program (Show)	.99	.99
Duration of the report (in seconds)	.67	.67
Teaser within in the program	.95	.90
Topic spreads over several reports within the program	.91	.89
A reference to that topic is also covered in another program	.95	.91
Competition	.93	.85
Gender of competing athletes^*	.88	.84
Sport	.93	.93
Actuality of the event	.90	.94
Event taking place in Germany	.95	.92
Regional proximity	.82	.84
Violation of rules	.94	.89
Success	.78	.57
Defeat	.78	.74
Controversy^**	.98	.97
Harm^**	.98	.98
Personal influence^**	.94	.92
Human interest^**	.97	.96
Level of an elite sportsperson^**	.87	.82
Unambiguity of the report^**	.95	.94
Surprise	.96	.93
Uncertainty	.94	.88
Factuality^**	.96	.94
Duration of reported events	.89	.85
Overall	.92	.89

Gender of competing athletes in reports was coded in 4 possible categories (female, male, mixed competition, the report covers several competitions).

News factor coded on a 4 points scale. Lotus reliability is calculated with a tolerance of 1 point.

***

Categorical variable of mentioned cities or counties.

Data collection and data preparation

The data collected for this study consist of 7103 reports on sports competitions, including events with male and female athletes, mixed-gender competitions, and mixed coverage of men and women. Of these reports, 504 (7.1%) focus on women's sports.

This study shows how the same sports are treated differently in sports news shows based on whether women or men are competing. In this scope, it is necessary to control the influence of the different distribution of sports and gender over coverage. With soccer being the undisputed number one sport in Germany, it is of little surprise that, by far, the largest share of sports reporting is on (men's) soccer: 5047 (66%) of the coded reports focus on this sport. Therefore, any observations of differences between men's sports and women's sports would be primarily differences between soccer on the one hand and women's sports on the other. To reduce any bias of the overproportioned amount of soccer coverage, a random sample of the reports on men's soccer, consisting of the same number of coding units as those found on women's soccer, replaced the actual men's soccer dataset. Thus, it is made sure women's sport is not primarily compared to only men's soccer (or, in other words, to prevent the data analysis from being soccer-biased).

Another possible cause of bias is that women's sports coverage is not as widely spread across different sports as men's sports coverage. As noted above, Horky (2009) points out that not all sports are equally likely to receive media coverage. Therefore, it may be that the difference in journalistic attention paid to different women's and men's sports is due not only to gender but also to the sports themselves. To control for the impact the media suitability of some sports may have on sports news content, the dataset was further reduced to reports on sports for which coverage of both women's and men's competitions was available. This reduction makes sure that only those sports that are equally likely to be covered on TV coverage are assessed. Thus, this study's results are based on a sample consisting of 1759 sports reports about 64 different disciplines, with 467 (26.5%) of these reports on women's sports (Table 3).

Table 3.

The seven sports news shows and their total number of reports on sports competitions and percentage of reports on women's sports.

	Total no. of reports in 2014 and 2015 and % of reports on women's sports	Sample of reports on the sports with both reports on women and men
ARD-Sportschau (Das Erste)	771 (10.2%)	306 (24.5%)
Das aktuelle Sportstudio (ZDF)	1562 (4.7%)	250 (28.0%)
Die ZDF Sportreportage (ZDF)	1198 (10.5%)	454 (25.8%)
Sport im Osten (MDR)	2244 (7.4%)	480 (31.7%)
Sportclub (NDR)	913 (4.8%)	188 (20.7%)
Sport im Westen (WDR)	350 (4.6%)	70 (18.6%)
Sport inside (WDR)	65 (1.5%)	11 (9.1%)
Total	7103 (7.1%)	1759 (26.5%)

Source: Data analysis of material recorded from TV broadcasts in 2014 and 2015 on German television using onlinetvrecorder.com.

Note: Indications in parentheses specify the broadcasting TV station. "ARD Sportschau" material contains only the Sunday broadcasts, as the Saturday edition of "Sportschau" is devoted solely to the Bundesliga and therefore presents reports on men's soccer only.

Results

Differences in the newsworthiness of stories on women's and men's sports

Hypothesis 1 suggests that the newsworthiness of reports on women's sports is significantly lower than that of news stories about men's sports. The difference in the mean newsworthiness is analyzed using a bootstrapped t-test The standard t-Test as a parametric test assumes normal sampling distribution. Since tests for normality of a distribution may be affected by very large sample sizes, especially if there are outliers, the bootstrapped t-test was preferred over a standard t-test In addition, bootstrapping has proven robust against violations of assumptions of the distributions for standard t-tests (Field, 2018).

The hypothesis is supported by the results of the content analysis conducted. Figure 2 displays the differences in the newsworthiness of reports on female and male athletes. On average, coding units on women's sports receive lower newsworthiness (M = 27.56, SE = 17.57) than those on men's sports (M = 31.30, SE = 18.02). This difference, −3.75, BCa 95% CI [−5.558, −1.849], is significant t(1757) = −3.87, p = .000111; representing a small effect, d = 0.21. The newsworthiness of news stories about women's sports is lower than the newsworthiness of news stories about men's sports.

Differing news factors in sports reporting on women and men

To test Hypothesis 2 it is necessary to understand if the occurrence and intensity of news factors differ between reports on women's and men's sports. Since most news factors were coded on an ordinal scale, the difference cannot be obtained by comparing means but by comparing the observed ranks of the news factors' intensities. Therefore a Mann-Whitney U test was conducted because it is one of the most common tests for the scenario at hand. With ordinal scaled intensities of news factors and two groups to compare (reports on women's sports vs. reports on men's sports), the data meet the assumptions of the Mann-Whitney U test However, there are three news factors that have not been coded on an ordinal scale but as dichotomous variables (news factors violation of rules is reported, surprise is expressed, uncertainty is expressed). For these three news factors, a χ² test was conducted to assess if the occurrence of these news factors differs between reports on women's and men's sports.

News factors are content-related features of news stories. They relate to the actual events but can be emphasized by journalists and editors to a greater or lesser degree. Since sporting competitions’ features do not differ by the gender of the competing athletes, it was assumed that the news factors do not differ either (Hypothesis 2). For example, whether an event occurs nearby or far away does not depend on gender. Therefore, the geographic proximity news factor should not systematically differ between women's and men's events within the same sports. Moreover, this is also applicable to the other news factors.

However, this content analysis's data does not entirely meet this assumption. The chi-square test for the three dichotomous news factors (irregularities, surprise, uncertainty), as well as the Mann-Whitney U test for the other news factors, shows mixed results when it comes to differences between women's and men's sports’ coding units (Tables 4 and 5).

Table 4.

Crosstabulation of the gender of competing athletes and occurrence of news factors violation of rules, surprise, and uncertainty.

		Gender of competing athletes in report
News factor		Female	Male	χ² (df)	p	Φ	Odds ratio
Violation of rules is reported	Yes	461 (.3)	1248 (−.2)	5.59 (1)	.018	.06	2.71
Violation of rules is reported	No	6 (−2.0)	44 (1.2)
Surprise is expressed	Yes	443 (.5)	1185 (−.3)	4.92 (1)	.027	.05	1.67
Surprise is expressed	No	24 (−1.8)	107 (1.1)
Uncertainty is expressed	Yes	452 (.2)	1234 (−.1)	1.41 (1)	.236	.03	1.42
Uncertainty is expressed	No	15 (−1.0)	58 (.6)

Note: Adjusted standardized residuals appear in parentheses below group frequencies. Highlighted rows indicate variables (news factors) that do not differ significantly (p > .05) and therefore meet the assumption of hypothesis 2.

Table 5.

Summary of differences between reports on women's sports and men's sports on Mann-Whitney U test.

	Gender of competing athletes in report				U	z	p	Effect size (r)
	Female (n = 467)		Male (n = 1292)
News factor	Mean rank	Mdn	Mean rank	Mdn
Actuality of the event	872,76	4,00	879,21	4,00	302312,500	0479	0632	0,01
Geographical proximity	823,76	4,00	900,33	4,00	327944,000	3128	0002	0,07
Success	867,15	4,00	884,64	4,00	307683,000	0720	0472	0,02
Controversy	867,26	1,00	224,61	1,00	307633,000	1965	0049	0,05
Harm	856,25	1,00	888,59	1,00	312775,500	2460	0015	0,06
Personal influence	8711	3,00	883,22	3,00	305836,000	0464	0643	0,01
Human interest	889,51	1,00	876,56	1,00	297241,000	−0749	0454	−0,02
Level of elite sportspersons	904,11	1,00	871,29	1,00	290424,000	1328	0184	0,03
Unambiguity of the report	934,04	3,00	860,47	3,00	27644,000	−3080	0002	−0,07
Factuality	909,32	3,00	8694	3,00	28798,500	−1932	0053	−0,05
Duration of reported events	901,36	4,00	872,28	4,00	291706,000	−1541	0123	−0,04

Note: Highlighted rows indicate variables (news factors) that do not differ significantly (p > .05) and therefore meet the assumption of hypothesis 2.

Eight of the fourteen news factors do not differ significantly between reports on women's and men's sports, supporting the assumption that Hypothesis 2 is valid for most news factors. However, six news factors show a statistically significant difference between women's and men's sports. These news factors are violation of rules is reported, surprise is expressed, geographical proximity, controversy, harm, unambiguity of the report. This would mean it is slightly more likely that a report on men's sports mentions rules violations than in reports on women's sports. The same is true for expressing a surprising outcome of a sporting event. The differences between women's and men's sports regarding the news factor geographical proximity say that it is more likely that a report on men's sports is on an event within Germany. That may reflect a broader coverage of local and national men's sports events. The differences in the news factors controversy and harm can be seen similar to the violation of rules news factor, meaning it is more likely that reports on men's sports are slightly more likely to cover controversies of higher intensity than it is the case in women's sports. If a report reflects on the harm done to any athlete (injuries or the like), it is more likely that the results of an injury covered in the report are more severe in reports on men's sports than if the report was on women's sports.

On the other hand, reports on women's sports tend to be more straightforward, as the news factor unambiguity suggests. This could reflect that men's sports are covered more deeply, resulting in reports handling more aspects of one topic resulting in a higher number of more complex reports than on women's sports. Anyway, these statistical findings do not show a systematic difference between women's and men's sports news. Looking at the news factors that differ significantly between female and male athletes’ news coverage, the differences in their mean values are relatively small. None of the news factors differs in their median values. Φ-values smaller than .06 and effect sizes r not exceeding .07 indicate that these differences in the news factors’ means are negligible.

Thus, Hypothesis 2 holds for the majority of the news factors. Therefore, and given the only minimal differences and effect sizes for the remaining significantly differing news factors, hypothesis 2 is supported by the data.

The news value of news factors and gender as a moderator

The gender of the athletes in the reports was assumed to be a moderating variable that explains the existing differences in the overall newsworthiness of reports on women's and men's sports on TV sports news shows. According to the two-component model of news selection, a moderator should influence the news value of several news factors and, therefore, affect a report's overall newsworthiness (Hypothesis 3).

Empirically, news factors’ news values cannot be observed directly (Kepplinger and Ehmig, 2006: 29; Kepplinger and Bastian, 2000: 465–467). This is because news factors are characteristics of news stories, and newsworthiness can only be assessed through the prominence of published news stories. Nevertheless, the actual news value of the news factors featured in news stories can be determined by statistical regression as, Kepplinger and Bastian; Kepplinger and Ehmig (2000; 2006) point out. The same logic of measuring the news values of news factors by statistical regression was successfully applied by Maier (2003a). By further applying gender as a moderating variable in such a regression analysis, it is possible to assess if the gender of competing athletes influences newsworthiness in the way hypothesis 3 suggests. Table 6 displays the regression results of several news factors as predictors of the newsworthiness of the coded reports (see also Footnote 1). The B values of that regression can be interpreted as indicators of the news factors’ news values (Kepplinger and Bastian, 2000: 465).

Table 6.

Regression model of grand mean centered news factors and gender interaction terms as predictors of newsworthiness (level of journalistic attention), with 95% bias-corrected and accelerated confidence intervals (CI). Confidence intervals and standard errors based on 1000 bootstrap samples.

Step 1	B	CI	SE B	β	p
Constant	29.67	28.863,30.483	.41	.00	.000
Actuality of the event	−.05	−1.591,1.489	.79	.00	.948
Geographical proximity	2.70	2.07,3.324	.32	.19	.000
Violation of rules is reported^*	4.94	0.51,9.365	2.26	.05	.029
Success	−.14	−0.866,0.588	.37	−.01	.707
Controversy	−.02	−2.444,2.405	1.24	.00	.988
Harm	−.07	−1.849,1.714	.91	.00	.941
Personal influence	.87	−0.069,1.805	.48	.05	.069
Human interest	3.41	1.744,5.068	.85	.10	.000
Level of elite sportspersons	1.44	0.831,2.054	.31	.11	.000
Unambiguity of the report	−3.80	−5.303,−2.294	.77	−.14	.000
Surprise is expressed^*	2.81	−0.14,5.766	1.51	.04	.062
Uncertainty expressed^*	6.62	2.616,10.63	2.04	.07	.001
Factuality	−4.90	−6.753,−3.043	.95	−.15	.000
Duration of reported events	−2.10	−3.197,−0.993	.56	−.09	.000
Step 2
Constant	29.70	28.885,30.518	.42	.00	.000
Actuality of the event	−.82	−4.03,2.385	1.64	−.03	.615
Geographical proximity	3.02	1.846,4.188	.60	.21	.000
Violation of rules is reported^*	10.07	−3.034,23.176	6.68	.10	.132
Success	.01	−1.347,1.375	.69	.00	.984
Controversy	1.19	−5.949,8.331	3.64	.02	.743
Harm	1.33	−3.079,5.734	2.25	.03	.555
Personal influence	−.25	−2.167,1.673	.98	−.01	.801
Human interest	6.62	3.511,9.733	1.59	.19	.000
Level of elite sportspersons	1.10	−0.107,2.314	.62	.09	.074
Unambiguity of the report	−2.39	−5.283,0.502	1.48	−.09	.105
Surprise is expressed^*	.55	−6.033,7.123	3.35	.01	.871
Uncertainty expressed^*	1.17	−7.293,9.633	4.32	.01	.786
Factuality	−6.30	−10.074,−2.523	1.93	−.19	.001
Duration of reported events	−2.54	−4.903,−0.182	1.20	−.11	.035
Actuality of the event × Gender	1.07	−2.592,4.727	1.87	.03	.567
Geographical proximity × Gender	−.49	−1.884,0.898	.71	−.03	.487
Violation of rules is reported × Gender	−5.69	−19.583,8.198	7.08	−.05	.422
Success × Gender	−.20	−1.811,1.416	.82	−.01	.810
Controversy × Gender	−.94	−8.534,6.66	3.87	−.02	.809
Harm × Gender	−1.60	−6.424,3.216	2.46	−.04	.514
Personal influence × Gender	1.45	−0.752,3.655	1.12	.07	.197
Human interest × Gender	−4.52	−8.204,−0.832	1.88	−.11	.016
Level of elite sportspersons × Gender	.47	−0.933,1.877	.72	.03	.510
Unambiguity of the report × Gender	−1.82	−5.212,1.574	1.73	−.06	.293
Surprise is expressed × Gender	2.74	−4.53,10.008	3.71	.04	.460
Uncertainty expressed × Gender	6.88	−2.657,16.422	4.86	.07	.157
Factuality × Gender	1.79	−2.548,6.136	2.21	.05	.418
Duration of reported events × Gender	.60	−2.079,3.27	1.36	.02	.662

Note: R² = .201 for Step 1; ΔR² = .007 for Step 2 (p = .337). Highlighted cells indicate p < .05.

Dummy coded (0/1) variables that have not been mean-centered.

Ten of the fourteen news factors contribute significantly to explaining the variances in newsworthiness levels. The overall R² of the regression model for the news factors alone (Step 1) is equal to .201. The uncertainty, violation of rules, factuality, and unambiguity news factors affect newsworthiness levels the most Thus, reports on sporting events with some degree of uncertainty regarding the outcome or their impact on future events will be presented more prominently than reports where everything is predictable or already decided. The same is true for events where athletes break the rules.

On the other hand, very factual reports tend to have low newsworthiness. This is plausible because highly factual reports are mostly given as news in brief and are shorter in duration, which results in lower prominence. The same is true for reports with a higher degree of unambiguity because the longer a report on a program is, the more aspects and, therefore, unambiguity will be added.

Concerning the identified news values of the different news factors, a question that remains to be answered is whether the news values of the news factors are affected by the gender of the athletes in the coded reports. Therefore, gender was added to the regression as an interaction term for all the news factors. For most news factors, the interaction term's effect reverses the direction of their effects on the overall newsworthiness (Step 2, Table 6). While, for example, the geographical proximity of an event bolsters the report's newsworthiness, the effect is reversed by the interaction term with gender. This means that in the case of reports on male athletes, reports on events nearby are more likely to be reported prominently, but the same is not true for reports on female athletes. This was expected based on hypothesis 3.

On the other hand, only one interaction term (human interest × gender) is statistically significant. Human interest stories are more likely to be reported prominently when the report is on male athletes. However, overall, the model with the interaction term in it does not significantly add to R² (p = .337), and ΔR² is also very small. Thus, Hypothesis 3 is not supported by the data of this study. Although the newsworthiness of reports on women's sports is lower than that of men's sports, this does not result from a moderation effect of gender on news values of news factors.

Discussion

The findings presented show that the overall newsworthiness of the coverage of women's sports on German television sports programs is lower than that of men's sports. In this regard, the investigated programs are no exception to previously studied sports communication content, and the findings are consistent with previous studies on the inequality in the coverage of women's and men's sports.

The present study attempted to find an additional explanation for the cause of this inequality by applying the news factor theory to the topic of sports coverage. Drawing from the two-component theory of news selection, as presented by Kepplinger (1998), Kepplinger and Bastian (2000), and Kepplinger and Ehmig (2006), it was suggested that the occurrence of news factors is roughly the same between women's and men's sports coverage, with gender working as a moderating variable affecting the news values of these news factors. These assumptions do not hold based on the results of this study. Although it can be said that the absolute expression of the news factors does not differ between women's and men's sports coverage, a moderating influence of gender is not observed. Therefore, including gender as a moderating variable in the equation does not significantly add to the regression of newsworthiness, causing hypothesis 3 to be rejected.

It has been shown that the overall newsworthiness of women's sports coverage is significantly lower than that of men's sports (see above, Figure 2). At the same time, this lower newsworthiness has been shown to not result from a moderation effect of gender, one of the news factors of sports news (see above, Table 6). This leads to the assumption that the impact of gender on the newsworthiness of sports reports’ news coverage is not an indirect effect but a direct one. This additional hypothesis can be examined by adding gender as another predictor in the regression (Table 7, Step 2) rather than as an interaction term. Unlike the previous models, this approach leads to a significant increase in R², while at the same time, the influence of gender itself accounts for a significant impact within this regression model. It is noteworthy that the impact of gender on the newsworthiness of sports coverage is higher than that of seven other news factors when comparing B values. In very simplified terms, on a 100-point scale of newsworthiness, a story ranks 2.33 points higher if the story is about men's sports than if it is about women's sports.

Table 7.

Regression model of news factors and gender as predictors of newsworthiness (level of journalistic attention), with 95% bias-corrected and accelerated confidence intervals (CI). Confidence intervals and standard errors based on 1000 bootstrap samples.

	B	CI	SE B	β	p
Step 1
Constant	44.91	35.031,54.785	5.04	.00	.000
Actuality of the event	.36	−1.192,1.911	.79	.01	.650
Geographical proximity	2.69	2.057,3.315	.32	.19	.000
Violation of rules is reported	6.24	1.789,10.69	2.27	.06	.006
Success	−.05	−0.772,0.682	.37	.00	.903
Controversy	.60	−1.837,3.038	1.24	.01	.629
Harm	.01	−1.775,1.789	.91	.00	.994
Personal influence	.93	−0.003,1.862	.48	.05	.051
Human interest	3.25	1.589,4.901	.85	.09	.000
Level of elite sportspersons	1.35	0.739,1.96	.31	.10	.000
unambiguity of the report	−3.93	−5.43,−2.432	.76	−.14	.000
Surprise is expressed	2.78	−0.172,5.729	1.50	.04	.065
Uncertainty expressed	7.76	3.912,11.6	1.96	.09	.000
Factuality	−5.26	−7.143,−3.367	.96	−.15	.000
Duration of reported events	−1.69	−2.783,−0.589	.56	−.07	.003
Step 2
Constant	43.23	33.287,53.163	5.07	.00	.000
Actuality of the event	.33	−1.224,1.873	.79	.01	.681
Geographical proximity	2.62	1.995,3.253	.32	.18	.000
Violation of rules is reported	5.92	1.471,10.369	2.27	.06	.009
Success	−.09	−0.818,0.635	.37	−.01	.806
Controversy	.52	−1.92,2.949	1.24	.01	.679
Harm	−.10	−1.883,1.678	.91	.00	.910
Personal influence	.90	−0.029,1.833	.48	.05	.058
Human interest	3.35	1.699,5.01	.84	.09	.000
Level of elite sportspersons	1.38	0.771,1.991	.31	.11	.000
unambiguity of the report	−3.85	−5.343,−2.348	.76	−.14	.000
Surprise is expressed	2.59	−0.364,5.534	1.50	.04	.086
Uncertainty expressed	7.71	3.874,11.549	1.96	.09	.000
Factuality	−5.25	−7.139,−3.37	.96	−.15	.000
Duration of reported events	−1.61	−2.706,−0.513	.56	−.07	.004
Gender	2.33	0.605,4.054	.88	.06	.008

Note: R² = .209 for Step 1; ΔR² = .003 for Step 2 (p = .008). Highlighted cells indicate p < .05.

So, it seems as if athletes’ gender does not work as a moderating variable but, in fact, has a direct effect on newsworthiness. This may suggest that when it comes to sports coverage, within a news factors theory framework, gender should be considered a news factor of its own.

From the news value theory perspective, the findings suggest that journalists perceive women's sports as less newsworthy than men's sports, and therefore they are presented less prominently. There is not much difference in what aspects are emphasized in sports coverage between men's and women's sports within the same sports. For example, it makes no difference whether a report is about a female world champion or a male world champion of any sport. The editorial perception of the elite level of being a world champion is not affected by the gender of the athlete. That is why there is neither a difference in the level of elite sportsperson's news factor nor in the interaction term of this news factor and gender. Thus, one might initially assume that sports journalism – like sports themselves – does not establish a hierarchy between genders. A female world champion is on the same level as a male world champion. Both are the world's best However, there is a difference in the observed newsworthiness of women's and men's sports. Summing up, sports journalists do not assess several news factors as being less important with regard to women's sports. Therefore, such differing assessments of specific content attributes between stories on women's and men's sports are not the main cause of the existing inequality in newsworthiness. Instead, in general, women's sports are treated as less newsworthy across different publication decisions, no matter which news factors contribute to the individual news stories. The results of this study clearly show that it is not particular aspects of sports and performance that are presented as less relevant and thereby causing lower newsworthiness of women's sport. Instead, it is sports being played by women that causes the lower newsworthiness in the presentation of sports in TV sports news. In other words, the situation for women's sports in TV sports news coverage is even worse than presumed. For example, if athletes’ successes were treated differently based on gender (moderating effect of gender on a news factor), sports organizations and athletes would at least be left with the option of presenting their sports in a way that would meet more of the news factors that were not treated differently. Nevertheless, as the results of this study show, being female is the only fundamental factor influencing women's sports prominence in sports news.

Conclusion

The results only apply to the TV programs studied. It cannot be entirely ruled out that different results would be observed in other media markets. This is especially true for countries and sports with particularly successful female athletes with a high profile. The possible influence of a particular athlete's prominence should be stressed, especially since, according to Horky (2009), some personal factors contribute to a sport's appeal to the media.

The theoretical framework of news factors extrapolates journalists’ perceptions and editorial decisions from observing and evaluating journalistic output (i.e. news stories and how they are presented). However, this approach cannot distinguish whether the news's presentation results from personal preferences or editorial constraints or is mainly affected by how journalists perceive the audience's favors. Therefore, future research should emphasize actual journalists’ preferences from their individual choices and their decisions regarding sports news. In addition, qualitative approaches such as focus group interviews could provide deeper insight into how editorial decisions on sports news are influenced by personal fondness or professional and male-dominated socialization. The same field of interest could also be accessed experimentally if, for example, the question would be if journalists individually rank several women's sports events as more or equally important than men's sports and how this affects their professional decisions? These questions could be investigated with an online experiment in which the respondents had to rank sports news according to perceived importance. According to the findings of this study, such an experiment could show if women's sports would rank lower than men's sports in journalists’ perception even if any other aspect like sports, level of competition, or geographical proximity were the same. Another aspect could be how journalists perceive their audiences’ preferences and how these affect their decisions regarding sports news compilation. Gaining a deeper understanding of these interrelations between perceptions of the audience could require more qualitative research approaches.

While sports communication academics seem to agree that the discrimination against women's sports exists in sports journalism, little is known about how this discrimination arises within editorial processes. This present study sheds some additional light on that issue. The results clearly show the systematic disadvantage of women's sports in terms of receiving coverage that is equally as prominent as that of their male counterparts. Moreover, this disadvantage is deeply embedded in sports journalism structures as gender directly affects newsworthiness. Future research should further investigate the actual preferences of both journalists and sports audiences.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

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

ORCID iD

Holger Ihle

Notes

References

Aiken

West

(1993) Multiple Regression: Testing and Interpreting Interactions. Newbury Park, Calif. u.a.: Sage.

Beck

(2013) Switzerland. In: Horky

Nieland

J-U

(eds) International Sports Press Survey 2011. Norderstedt: BoD Books on Demand, pp. 81–97.

Bernstein

Kian

(2013) Gender and sexualities in sport Media. In: Pedersen

(ed) Routledge Handbook of Sport Communication. Abingdon, Oxon: Routledge, pp.319–327.

Billings

Angelini

(2007) Packaging the games for viewer consumption: Gender, ethnicity, and nationality in NBC’s coverage of the 2004 summer Olympics. Communication Quarterly 55(1): 95–111.

Billings

Angelini

Duke

(2010) Gendered profiles of Olympic history: Sportscaster dialogue in the 2008 Beijing Olympics. Journal of Broadcasting & Electronic Media 54(1): 9–23.

Burroughs

Malik

Johnson

, et al. (2021) Content analysis of biological sex-specific media coverage of sport: The case of National Collegiate Athletic Association athletic department home webpages. International Review for the Sociology of Sport 4(1): 101269022110296.

Coche

(2021) A new era? How the European ESPN covered the 2019 Women’s World Cup online. International Review for the Sociology of Sport 8(2): 101269022199224.

Cooky

Council

Mears

, et al. (2021) One and done: The long eclipse of women’s televised sports, 1989–2019. Communication & Sport 9(3): 347–371.

Cooky

Messner

Hextrum

(2013) Women play sport, but not on TV. Communication & Sport 1(3): 203–230.

10.

Cooky

Messner

Musto

(2015) “It’s Dude Time!”: A quarter century of excluding women’s sports in televised news and highlight shows. Communication & Sport 3(3): 261–287.

11.

Duncan

Messner

(2005) Gender in Televised Sports: News and Highlights Shows, 1989-2004. Los Angeles: Amateur Athletic Foundation of Los Angeles.

12.

Eastman

Billings

(2000) Sportscasting and sports reporting. Journal of Sport and Social Issues 24(2): 192–213.

13.

Field

(2018) Discovering Statistics Using IBM SPSS Statistics. Los Angeles, London, New Delhi, Singapore, Washington DC, Melbourne: Sage.

14.

Fredericks

Keim

(2013) South Africa. In: Horky

Nieland

J-U

(eds) International Sports Press Survey 2011. Norderstedt: BoD - Books on Demand, pp. 187–205.

15.

Fretwurst

(2015a) Lotus manual: Reliability and accuracy with SPSS. Version 1.0 2015. Available at: https://www.iakom.ch/Lotus/LotusManualEng.pdf (accessed 6 January 2021).

16.

Fretwurst

(2015b) Reliabilität und Validität von Inhaltsanalysen: Mit Erläuterungen zur Berechnung des Reliabilitätskoeffizienten ‘Lotus’ mit SPSS. [Reliability and validity of content analyses: With explanations for the calculation of the reliability coefficient 'Lotus’ with SPSS]. In: Sommer

Matthes

Wettstein

Wirth

(eds) Qualitätskriterien in der Inhaltsanalyse: [Methods and Research Logic in Communication Science. Quality Criteria in Content Analysis]. Köln: Herbert von Halem, pp.176–203.

17.

Fry

Goh-Leong

(2013) Singapore. In: Horky

Nieland

J-U

(eds) International Sports Press Survey 2011. Norderstedt: BoD - Books on Demand, pp. 220–225.

18.

Galtung

Ruge

(1965) The structure of foreign news. Journal of Peace Research 2(1): 64–90.

19.

Gerbner

Gross

Morgan

, et al. (2002) Growing up with television: cultivation processes. In: Bryant J and Zillmann D (eds) Media effects: Advances in theory and research. Mahwah, NJ: Erlbaum, pp. 43–67.

20.

Godoy-Pressland

(2014) ‘Nothing to report’: A semi-longitudinal investigation of the print media coverage of sportswomen in British Sunday newspapers. Media, Culture & Society 36(5): 595–609.

21.

Greer

Hardin

Homan

(2009) “Naturally” less exciting? Visual production of men’s and women’s track and field coverage during the 2004 Olympics. Journal of Broadcasting & Electronic Media 53(2): 173–189.

22.

Hair

Anderson

Tatham

, et al. (1995) Multivariate Data Analysis with Readings: With Readings. Englewood Cliffs, N.J.: Prentice Hall.

23.

Horky

(2009) Was macht den Sport zum Mediensport? Ein Modell zur Definition und Analyse von Mediensportarten. [What turns sports into media sports? A model to define and analyze types of media sports]. Sportwissenschaft 39(4): 298–308.

24.

Horky

Nieland

J-U

(2013) Comparing sports reporting from around the world - numbers and facts on sports in daily newspapers. In: Horky

Nieland

J-U

(eds) International Sports Press Survey 2011. Norderstedt: BoD - Books on Demand, pp. 22–41.

25.

Ihle

(2018a) Im Abseits oder aus der Tiefe des Raumes - Wie vielfältig berichtet das Fernsehen über regionales Sportgeschehen? Eine Analyse der Sport-Magazinsendungen in MDR, NDR und WDR [The diversity of sports journalism on regional television in Germany. A content analysis of German television sports newsmagazines of three publicservice broadcasting television channels]. Publizistik 63(1): 97–123.

26.

Ihle

(2018b) Sport in den Fernsehnachrichten: Ergebnisse einer Inhaltsanalyse der Sportberichterstattung in den deutschen Hauptnachrichtensendungen. [Sports in television news shows: Results of a content analysis of sports coverage in German prime-time news programs]. Media Perspektiven 10: 486–495.

27.

Jakubowska

(2013) Poland. In: Horky

Nieland

J-U

(eds) International Sports Press Survey 2011. Norderstedt: BoD - Books on Demand, pp. 98–112.

28.

Kaiser

(2018) Women’s and men’s prominence in sports coverage and changes in large-, medium-, and small-city newspapers, pre- and post-Title IX. Communication & Sport 6(6): 762–787.

29.

Kepplinger

(1998) Der Nachrichtenwert der Nachrichtenfaktoren. [The news value of the news factors]. In: Holtz-Bacha

Scherer

Waldmann

(eds) Wie die Medien die Welt Erschaffen und wie die Menschen Darin Leben: Für Winfried Schulz. Opladen: [How the media Create the World and how People Live in it]. Westdeutscher Verlag, pp.19–38.

30.

Kepplinger

Bastian

(2000) Der prognostische Gehalt der Nachrichtenwert-Theorie. [The prognostic substance of the news values theory]. Publizistik 45(4): 462–475.

31.

Kepplinger

Ehmig

(2006) Predicting news decisions. An empirical test of the two-component theory of news selection. Communications 31(1): 25–43.

32.

Kian

Vincent

Mondello

(2008) Masculine hegemonic hoops: An analysis of media coverage of march madness. Sociology of Sport Journal 25(2): 223–242.

33.

Lapchick

Bowman

Ewing

, et al. (2020) The 2021 Sports Media Racial and Gender Report Card: Associated Press Sports Editors (APSE). Available at: https://www.tidesport.org/_files/ugd/138a69_e1e67c118b784f4caba00a4536699300.pdf.

34.

Le Clair

(2013) Canada. In: Horky

Nieland

J-U

(eds) International Sports Press Survey 2011. Norderstedt: BoD - Books on Demand, pp. 163–177.

35.

Loosen

(1998) Die Medienrealität des Sports: Evaluation und Analyse der Printberichterstattung. [The media Reality of Sports: Evaluation and Analysis of Coverage in Print media]. Wiesbaden: Deutscher Universitäts-Verlag.

36.

Loosen

(2001) „Das wird alles von den Medien hochsterilisiert“: Themenkarrieren und Konjunkturkurven der Sportberichterstattung. [Topic careers and cyclical curves in sports reporting.]. In: Roters

Klingler

Gerhards

(eds) Sport und Sportrezeption. Baden-Baden: Nomos, pp.133–147.

37.

Loosen

Ravenstein

(2000) Sportberichterstattung im Fokus der Nachrichtenwertforschung: Fußball-WM '98 und Sport in der Printmedienberichterstattung. [Sports coverage in the focus of news value research: World cup '98 and sports in print Media coverage]. In: Schorr

(ed) Publikums- und Wirkungsforschung: Ein Reader. Wiesbaden: Westdeutscher Verlag, pp.191–204.

38.

Maier

(2003a) Analysen deutscher Fernsehnachrichten 1992-2001. [Analyses of German television news 1992-2001]. In: Ruhrmann

Woelke

Maier

Diehlmann

(eds) Der Wert von Nachrichten im Deutschen Fernsehen: Ein Modell zur Validierung von Nachrichtenfaktoren. [The Value of News on German Television. A Model for the Validation of News Factors]. Opladen: Leske+Budrich, pp. 61–98.

39.

Maier

(2003b) Operationalisierung der Inhaltsanalyse. [Operationalization of the content analysis]. In: Ruhrmann

Woelke

Maier

Diehlmann

(eds) Der Wert von Nachrichten im deutschen Fernsehen: Ein Modell zur Validierung von Nachrichtenfaktoren. [The Value of News on German Television. A Model for the Validation of News Factors]. Opladen: Leske+Budrich, pp.51–98.

40.

Maier

Ruhrmann

Klietsch

(2006) Der Wert von Nachrichten im deutschen Fernsehen: Ergebnisse einer Inhaltsanalyse 1992 - 2004. [The Value of News on German Television: Results of a Content Analysis 1992–2004]. Düsseldorf: Landesanstalt für Medien Nordrhein-Westfalen (LfM).

41.

Meân

(2017) Sport as gender/feminist studies. In: Billings

(ed) Defining Sport Communication. New York, NY: Routledge, pp.92–106.

42.

Messner

(2002) Taking the Field: Women, men, and Sports. Minneapolis: University of Minnesota Press.

43.

Östgaard

(1965) Factors influencing the flow of news. Journal of Peace Research 2(1): 39–63.

44.

Petrica

Ilincescu

(2013) Romania. In: Horky

Nieland

J-U

(eds) International Sports Press Survey 2011. Norderstedt: BoD - Books on Demand, pp. 129–147.

45.

Pfister

(2010) Women in sport – gender relations and future perspectives. Sport in Society 13(2): 234–248.

46.

Ramon

Rojas-Torrijos

(2021) Public service media, sports and cultural citizenship in the age of social media: An analysis of BBC sport agenda diversity on Twitter. International Review for the Sociology of Sport 39(7): 101269022110439.

47.

Rehbach

Ihle

Nieland

J-U

(2016) Olympische Prinzipien und gesellschaftliche Werte in der Printberichterstattung: Ausgewählte Befunde einer Pilotstudie. [Olympic principles and social values in print reporting: Selected findings from a pilot study]. In: Schürmann

Mittag

Stibbe

Nieland

J-U

Haut

(eds) Bewegungskulturen im Wandel: Der Sport der Medialen Moderne - Gesellschaftstheoretische Verortungen. [Changing Cultures of Movement. Sports in “Mediale Moderne” - Socio-Theoretical Positionings]. Bielefeld: Transcript, pp.225–249.

48.

Ruhrmann

Woelke

Maier

, et al.2003) Der Wert von Nachrichten im deutschen Fernsehen: Ein Modell zur Validierung von Nachrichtenfaktoren. [The Value of News on German Television. A Model for the Validation of News Factors]. Opladen: Leske + Budrich.

49.

Schulz

(1976) Die Konstruktion von Realität in den Nachrichtenmedien: Analyse der aktuellen Berichterstattung. [The Construction of Reality in the News media. Analysis of Current Coverage]. Alber: Freiburg im Breisgau.

50.

Tuggle

Huffman

Rosengard

(2002) A descriptive analysis of NBC’s coverage of the 2000 summer olympics. Mass Communication and Society 5(3): 361–375.

51.

Vincent

(2004) Game, sex, and match: The construction of gender in British newspaper coverage of the 2000 Wimbledon championships. Sociology of Sport Journal 21(4): 435–456.

52.

Wenner

(2015) Communication and sport, where art thou? Epistemological reflections on the moment and field(s) of play. Communication & Sport 3(3): 247–260.

53.

Zubayr

Gerhard

(2016) Tendenzen im Zuschauerverhalten: Fernsehgewohnheiten und Fernsehreichweiten im Jahr 2015. [Trends in viewer behavior: Viewing habits and TV ratings in 2015]. Media Perspektiven 3: 142–155.