Comparing the Five Measures of Media Reputation Attributes in Local and National Newspapers

Abstract

Researchers have developed multiple measures to assess media reputation as a driving force of corporate reputation. This study compared five measures that have been used to examine the seven attributes of media reputation. These five measures of media reputation have different fundamental assumptions such as linearity, inclusion of neutral tone items, and negative items. A content analysis of 2,817 news articles regarding nine big food corporations revealed significant differences in the media reputation attribute scores between elite national newspapers and local newspapers using the five measures. The findings indicate that these measures cannot substitute for one another, and using elite national newspapers versus local newspapers would lead to significantly different assessments of media reputation.

Keywords

measurement media reputation corporate reputation attribute content analysis

Food plays a crucial role in our lives and affects our well-being. People are concerned about food safety and seek relevant information to know more about food quality and safety. There are two major ways through which people get knowledge about food: direct experience such as tasting and seeing, and indirect experience from others’ experience via interpersonal communication or mass media (Emler, 1990; Grauman & Kruse, 1990). Government agencies such as the United States Department of Agriculture (USDA) frequently recall food with quality/safety problems. Many of these recalls became high profile issues in news media. Research showed that news reports on food safety issues have significant impacts on consumers’ perceptions and choices of food products (Frewer et al., 1997; Jonge et al., 2010).

The outbreak of COVID-19 across the world caused infections among employees of food companies, which resulted in the closure of many food processing plants. For example, after 180 employees were tested positive for COVID-19 in April 2020, Tyson Foods, which is the largest processor of chicken, beef, and pork in the United States, closed several meat and cheese processing plants across the country. As usual, this food safety issue quickly became a high-profile event in news media. During this pandemic period, people have heightened concerns about the COVID-19 exposure risk related to meat processing. A recent study conducted by International Food Information Council Foundation showed that 24% of respondents ranked the COVID-19 related food safety issue as their number-one concern (Food Insight, 2020).

The food safety crises have tremendous impacts on food companies and industry. The average direct cost of a food recall is over $10 millions (Food Safety Magazine, 2012). The indirect cost in terms of the reputation damage would be much larger. A survey conducted by U.S. Grocery Supplier found that 65% respondents said they will not buy any spinach and peanut butter after knowing these products were recalled (Keith, 2019). Therefore, food safety is the biggest threat to the profitability of food industry (Food Safety Magazine, 2012).

As food recalls and other food safety issues have become a news staple, news media serve as the public’s major information source about food safety. The agenda-setting theory argues that news media play a crucial role in people’s perceptions of public issues in general, and of food safety issues in particular. Even in normal time, news media are also the major information sources for people to know about food and food companies.

Literature showed that news media are one of the driving forces of corporate reputation (e.g., Fombrun & Shanley, 1990; Kiousis et al., 2007; Meijer & Kleinnijenhuis, 2006). To further explore the influence of news media on corporate reputation, Deephouse (2000) proposed a new concept—media reputation—and defined it as the “the overall evaluation of a corporation presented in the media” (p. 1097). Consequently, scholars in both communication and management disciplines have developed multiple measures of media reputation, and used these different measures to investigate the relationship between media reputation and corporate reputation (e.g., Kiousis et al., 2007; Meijer, 2004; Meijer & Kleinnijenhuis 2006; Zhang, 2016, 2018a), as well as the relationship between media reputation and corporate financial performance (e.g., Deephouse, 2000; Kiousis et al., 2007; Pollock & Rindova, 2003). These researchers applied the agenda-setting theory as the theoretical foundation in such investigations, especially in the investigation of the relationship between media reputation and corporate reputation (e.g., Kiousis et al., 2007). However, unlike the agenda-setting studies in the political news setting, the results on the relationship between media reputation and corporate reputation were inconsistent and contradict one another. For example, Fombrun and Shanley (1990) did not find significant correlation between media visibility and corporate reputation at the overall level, whereas Wartick (1992) found the positive and significant correlation between media visibility and corporate reputation in the groups of companies with “good” and “average” reputation levels. The inconsistent findings were also shown in the recent multiple cross-country empirical studies compiled by Carroll (2013).

Zhang (2016) compared different measures of media reputation and found that they might be the reason for the inconsistent findings. That empirical study only focused on the various measures of a company’s overall media reputation, not the media reputation on the specific attributes, such as products and services, leadership, and financial performance. The second level agenda-setting theory (McCombs et al., 1998) argues that the salience of attributes in media coverage and the tones of media coverage influence people’s perceptions of these attributes. This theory lays the theoretical ground for exploring the relationship between media reputation and corporate reputation at the attribute level. Nevertheless, the various measures of media reputation and inconsistent findings in the literature obstructed the application of agenda-setting theory in the business news setting. When researchers attempt to assess the media reputation of a company with a variety of measurement options, they would ask whether these different measures generate the same results or significantly different ones. In practice, when a crisis happens, for example, a food recall is announced by USDA, the food company wants to assess the loss of its reputation in the news media during this crisis period, how to assess its media reputation at the attribute level using all these existing measures? Hence, it is necessary to investigate whether different measures of media reputation for the attributes of a company are significantly different from each other.

Application of the Second-Level Agenda-Setting Theory and Priming Theory in Studying Business News

The second-level agenda setting effect focuses on the influence of the attributes and tones of the news media coverage on audiences’ perceptions. This effect has two dimensions: the cognitive (substantive) dimension regarding the attribute characteristics that describe the object, and the affective (evaluative) dimension regarding the tone (e.g., positive, negative, neutral) of these characteristics (McCombs & Reynolds, 2009). Put simply, the second-level agenda-setting effect hypothesis argues that by emphasizing different attributes of the objects, and by expressing the tones toward these objects as well as their attributes, media can influence not only the public’s perceptions of the salience of these attributes, but also their emotional perceptions and evaluations of these objects and their attributes (McCombs, 1997; McCombs & Ghanem, 2001).

The second level agenda-setting works as a theoretical foundation to explore the relationship between media coverage and corporate reputation. Ponzi et al. (2011) defined corporate reputation as the assessment of people’s various emotional and cognitive perceptions of a corporation. Fombrun and Shanley (1990) argued media coverage is one of the driving forces of corporate reputation. Past research on the relationship between media coverage and corporate reputation frequently focused only on crisis situations and negative framing (Fan et al., 2013; Mason, 2019; Xu & Wu, 2020; Yeo et al., 2020), but overlooked the positive coverage or other aspects of routine news coverage of a company. In the measurement of corporate reputation, researchers employed frameworks consisting of multiple attributes. For example, Reputation Institute’s (RI) RepTrak^® Pulse consists of seven attributes of corporate reputation: (1) products and services, (2) leadership, (3) financial performance, (4) innovation, (5) citizenship, (6) governance, and (7) workplace. The ratings of these attributes depend on people’s perceptions of them. According to the second level agenda-setting theory, the cognitive (substantive) dimension of the second level agenda-setting posits that the salience of the reports regarding the attributes of a corporation influences the salience of these attributes in people’s perceptions. The affective (evaluative) dimension of the second level agenda-setting posits that the tones of the reports regarding the attributes of a corporation influence people’s emotional perceptions of these attributes. If these theoretical statements are true, it is important for company managers to know how news media report the attributes of their companies as well as the tones of these attributes. The measurement of the media coverage of the attributes and their tones will provide a valuable guideline for managers to formulate strategies of media relations by identifying which specific aspects of the companies they need to improve in news coverage.

The second level agenda-setting effect is a common theoretical framework to examine the influences of attributes of political figures in the news media on the public’s perceptions of these attributes, and the influences of the media tonalities on the public’s emotional feelings toward these politicians (e.g., Becker & McCombs, 1978; Kim & McCombs, 2007; McCombs et al., 1998). In recent years, several scholars have applied this theoretical framework to investigate the relationship between the attributes of corporations in the news media and corporate reputation in the corresponding attributes. For example, Kiousis et al. (2007) applied the framework of The Harris Poll Reputation Quotient (RQ) to categorize business news into six attribute categories: vision & leadership, social responsibility, emotional appeal, products and services, workplace environment, and financial performance, and examined the affective (evaluative) dimension of second-level agenda-setting effect on these six attributes. However, unlike the studies of the second level agenda-setting effect of political news, these studies did not yield robust and consistent results.

There is another theory—priming theory—in the media effects literature. This theory argues that media content affects the standards people use to make evaluations or judgments (Iyengar & Kinder, 1987). A few studies used this theory to explore the recency effect of business news on corporate reputation (e.g., Wartick, 1992). There has been an attempt to combine this theory and the second level agenda-setting theory. Sheafer (2007) proposed a new concept—affective priming—to describe the process in which people attach affective attributes to an object, and argued this is an affective-evaluative component in the priming process. However, this attempt has been only limited to the political news field. Although priming theory and agenda-setting theory have different perspectives, the empirical research of both theories requires the evaluations of media content. Researchers need to know what measures and which news samples they should use, and whether different measures generate consistent results to be used reliably.

Zhang (2016) investigated the predictive power of overall media reputation for overall corporate reputation using regression estimation. His study found large gaps of R² values among different measures ranging from 0.01 to 0.23 in local newspaper and that the predictive power of the measure on local newspapers was higher than that of the same measure on elite newspapers. Six out of seven indexes have higher R² values when the local newspapers were assessed than when the elite newspapers were assessed. These findings suggested various measures of media reputation and news sources used in the measurement caused the inconsistent findings between media reputation and corporate reputation. However, that study only assessed the media reputation at the overall level. To advance the application of the agenda-setting theory as well as priming theory in the business news setting, it is necessary to investigate whether various measures at the attribute level are also significantly different from each other using different news sources.

Measures of Media Reputation

Scholars from either the management discipline or the communication discipline have proposed various measures of media reputation. Only five media reputation measures that can be used to assess the media coverage at the attribute level (see Table 1).¹ These five measures use different formulas, which can be broadly classified into two groups based on linearity assumptions: three are linear function measures and two are non-linear function measures. Linear function measures include Fombrun-Shanley index, Wry index, and Kiousis index. Non-linear function measures include Deephouse media endorsement index and Deephouse media favorability index.

Table 1.

Comparison of Existing Measurements of Media Reputation.

Index name	Studies used the index	Formula	Whether neutral coding units are included in the denominator
(1) Kiousis index	Kiousis et al. (2007)	$f$	No
(2) Fombrun- Shanley index	Fombrun and Shanley (1990)	$f + n$	Yes. The proportion of neutral coding units is included.
(3) Wry index	Wry et al. (2006); Vogler at al. (2016)	$f - u$	Not directly
(4) Deephouse media endorsement index	Deephouse (1996); Bansal and Clelland (2004)	$y = {\begin{cases} \frac{f^{2} - f u}{r^{2}}, i f f > u \\ 0 i f f = u \\ \frac{f u - u^{2}}{r^{2}}, i f u > f \end{cases}$	No
(5) Deephouse media favorability index	Deephouse (2000), Pollock and Rindova (2003); Tong (2013); Agnihotri (2014); Ouyang et al. (2017)	$y = {\begin{cases} \frac{f^{2} - f u}{r^{2}}, i f f > u \\ 0 i f f = u \\ \frac{f u - u^{2}}{r^{2}}, i f u > f \end{cases}$	Yes. The number of neutral coding units is included in r.

Note. f = proportion of positive items, n = proportion of neutral items, u = proportion of negative items, r = number of relevant units. All these five measures fall into the scale range (−100, 100).

Fombrun and Shanley (1990) created the Fombrun-Shanley index to examine the correlation between media favorability and corporate reputation. But they did not explain why the linear function was used and why only positive and neutral coding units were included in the function. Wry et al. (2006) created the Wry index to investigate the media reputations within and among strategic groups which are sets of industry competitors that have similar characteristics. There was no explanation for the linear function as well as the components of media reputation in this study. Vogler et al. (2016) also used the Wry index to examine the influence of stakeholder groups on media reputation during the crisis periods. Kiousis et al. (2007) used the Kiousis index to estimate the influence of media coverage on public opinion of corporations and their financial performance. However, that study did not use the term of media reputation or media favorability, but instead used the positive tone of media coverage as the independent variable. The percentage of positive tone is also one of the measures of media reputation.

The non-linear measures are based on the Janis-Fadner imbalance coefficient function. Janis and Fadner (1943) created this coefficient to measure wartime propaganda. There is a slight difference between the two non-linear function measures. Although they use the same formula, the denominators of the formulas are different. The denominator of Deephouse media endorsement index only includes positive units and negative units.² Deephouse (1996) used this index to measure the public endorsement without explaining why Janis-Fadner imbalance coefficient was used and why only positive and negative units were included in the calculation. Bansal and Clelland (2004) also used the Deephouse media endorsement index to examine the correlation between unsystematic risks and corporate environmental legitimacy in the context of natural environment.³

The denominator of Deephouse media endorsement index includes positive units, negative units, and neutral units. Deephouse (2000) created this index to measure media reputation and examine the correlation between media reputation and corporate financial performance. This study did not explain why this index differs from the Deephouse media endorsement index.⁴ Other researchers have used the Deephouse media endorsement index to examine the relationship between media reputation and a firm’s stock market performance. For example, Tong (2013) used it to examined the impacts of media reputation on IPO share price; Agnihotri (2014) used it to assess the correlation between media reputation and a firm’s market evaluation; And OuYang et al. (2017) used it to test the correlation between media reputation and post-crisis stock return.

The differences in the formulas of these measures and the lack of discussion on the rationales behind these differences make it necessary to do the empirical analysis to compare these measures. When researchers attempt to measure media coverage of companies and have multiple choices of measures, it is essential to know whether they produce the same results or significantly different results, and which measure generates higher or lower score in certain attributes than others. The answers to these questions would further our understanding on the measurement of media reputation at the attribute level.

Another important issue is to determine what kinds of news samples should be used. Scholars have used different news samples in measuring media reputation. For example, Kiousis et al. (2007) selected both The New York Times and The Wall Street Journal. Lee and Carroll (2011) used two national U.S. newspapers (The New York Times and USA Today) and seven major metro newspapers (The Los Angeles Times, The Washington Post, The Chicago Sun-Times, The Atlanta Journal Constitution, The Seattle Times, The Boston Globe, and The Houston Chronicle). These different news samples may yield different results in the measurement of media reputation. The New York Times and The Wall Street Journal are national elite newspapers which have been demonstrated to have the agenda-setting power on other news media (Zhang, 2018b; Gilbert et al., 1980; Golan, 2006; Reese and Danielian, 1989). Therefore, there was an argument that they should be used to measure media reputation (e.g, Kiousis et al., 2007). On the other hand, since Pew Research Center (2012) showed that local newspapers are the major information sources for most Americans. Therefore, there is a rationale that local newspapers should be used as the news sample to measure media reputation. As there are different approaches to sample newspapers, we need to know how different newspaper samples affect the measurement of media reputation.

Based on the literature review, we proposed the following research questions. The first research question focuses on the differences between the measures when the elite newspapers are used as the news sample. The second research question focuses on the differences between the measures when the local newspapers are used as the news sample. The third research question compares the impact of newspaper choice on the measurement of media reputation.

RQ1: Will the existing measures of media reputation on company attributes be significantly different from each other when only elite newspapers are used as the news sample?

RQ2: Will the existing measures of media reputation on company attributes be significantly different from each other when only local newspapers are used as the news sample?

RQ3: Will the same measure of media reputation on company attributes generate significantly different results when elite and local newspapers are separately used as the news samples?

Method

Corporation Sample Selection

The comparison of media reputations of companies in a single industry can be on the same footing as news media have different interests toward different industries. We choose the food industry as the sample industry since people are concerned about food information which may relate to their health. The outbreak of COVID-19 pandemic makes people have more concerns with COVID-19 related food safety issues (Food Insight, 2020). Bonaiuto et al. (2012) found that the food company’s reputation had impacts on consumer’s food choice. Sheiham et al. (1987) argued that people’s evaluations of foods are based not only on their direct experience but also on their indirect experience including media communication. Furthermore, the information asymmetry problem is serious in food markets, and the public has a strong need for the food information to address this problem (Antle, 1999).

Reputation Institute (RI) consistently tracks the reputation of companies through extensive national surveys with large samples. Based on these surveys, RI developed the RepTrak^® Pulse index to measure the emotional perceptions of the public about companies. The authors were granted access to the attribute measures of the RepTrak^® Pulse index. We chose the food corporations on the ranking list of RI’s “Reputations of the 150 Largest Public U.S. Companies” as the sample of corporations. There were only nine food corporations in the sample, including: General Mills, Kraft Foods, Kellogg, Sara Lee, H. J. Heinz, Dean Foods, ConAgra Foods, Tyson Foods, and Archer Daniels Midland. Nine companies consist a small sample size. To increase the number of observations in the data anlaysis, we used the panel data (also known as longitudinal or cross-sectional time-series data) of nine corporations within a 5-year period (n = 45).

News Sample Selection

We included both elite and local newspapers in the news population. Since The New York Times has been a benchmark for media agenda (Gans, 1979), and has significant influence on inter-media agenda-setting process (Vliegenthart and Walgrave, 2008), we chose it as the representative of elite newspapers. We also chose The Wall Street Journal as another representative of elite newspapers because it has significant influence on business news (Kiousis et al., 2007). We included local metro newspapers in the news population since the Pew’s national survey showed that local metro newspapers are the major information sources for most Americans (Pew Research Center, 2012).

News articles published in the time period from 2007 to 2011 constituted the sampling frame of the news population. Since The Wall Street Journal news data cannot be obtained from Lexis-Nexis, we used Factiva to obtain news articles in these two elite newspapers. However, because Lexis-Nexis contains the data of more than 2,500 newspapers (LexisNexis Academic, 2021), we used it as the source for the local news population.

Lexis-Nexis and Factiva both have a search tool that helps locate the central theme of the new articles: search the keywords in the headline and lead paragraph. We used this search tool to search the names of corporations. However, the two databases have different search tools to screen the research results for specific purposes. For the Lexis-Nexis, we used the “ticker,” which is the code used to uniquely identify a publicly traded company on a stock market, to identify a corporation’s presence in the article. We combined this search tool and the keywords (corporations’ names) tool to search news articles about the corporations. Factiva does not have the “ticker” filtering tool. Instead, it has the industry filter (food + beverages/tobacco) which was used as the alternative tool to screen news articles. We combined this filter and the keywords (corporations’ names) tool to search news articles about the corporations. In the second stage of searching, we used the brand names belonging to these corporations as the keywords for searching the news articles. We also combined the ticker, the industry filter, and the keywords filter to search news articles about the brands.

We downloaded all the news articles obtained by these search methods. Then, we screened irrelevant articles from the sample. Articles which fall into the following categories were excluded from the final news sample: story not refer to the corporation/brand, corrections/clarification, obituary, news forecast with the headlines “Looking Ahead” and “Coming Up,” and duplicate articles. The screened unduplicated news articles constituted the news sampling frame.

Then, we used an online sample size calculator which is a free program developed by Creative Research Systems (Sample Size Calculator, 2021) to calculate the sample size of news articles for each company. We calculated the sample size using the following method: for the smaller population, a larger subset of population will be included in the sample size; for the larger population, a smaller subset of population will be included. Finally, we obtained a total of 2,817 articles in the news sample.

Content Analysis

We developed the codebook that guided the coding process. We conducted a pilot coding to train coders and build consensus at the earlier stage, and created and revised the coding sheet at this stage. In the next stage, the first author and another coder worked together to code 300 news articles. The coders read the first five paragraphs and used the information in these paragraphs to answer the coding questions. When the coders felt the first five paragraphs did not provide sufficient information to code, they read a few more paragraphs. We revised the codebook multiple times during the training process so that the coders’ coding approaches were calibrated. This ensured the coders coded the content in the same way. In the next coding stage, we randomly selected 300 news articles, which constituted 10.2% of the articles in the news sample, and coded them independently without any discussion or collaboration. Moreover, we applied the blind coding technique which requires that another coder does not know the purpose of the research (Neuendorf, 2002).

We used the framework of Reputation Institute’s (RI) RepTrak^® Pulse as the frame reference to categorize the seven attributes of corporate news. During the coding process, the coders identified one or more of the seven attributes in the news articles. We used the brief definitions of these seven attributes provided by Fombrun (2006): (1) Products or services: the perceptions of the quality, value, and reliability of the company’s products and services; (2) Leadership: how much the company demonstrates a clear vision and strong leadership; (3) Financial performance: the perceptions of the company’s profitability, prospects, and risk; (4) Innovation: how the firm makes or sells innovative products or innovates in the way it does business; (5) Citizenship: the perceptions of the company as a good citizen in its dealings with communities, employees, and the environment; (6) Governance: whether a firm behaves ethically and is open and transparent in its business dealings; (7) Workplace: the perceptions of how well the company is managed, how it is to work for, and the quality of its employees.

Once one attribute was identified, the coders coded for its tone that could be one of the four tonalities: positive, negative, neutral, or mixed. According to Lee and Carroll (2010), positive news refers to the content which indicates the company has positive emotional appeal, is an object of admiration and respect, or particularly trustworthy. Negative news refers to the content indicating the company generates negative emotional appeal, or is portrayed as unworthy of admiration, respect, or trust. Neutral news refers to the content that does not have any positive and negative connotations. Mixed news refers to the content that relates to the company that has roughly equivalent positive and negative connotations in the story. The coders applied these definitions of media tonalities in the coding process.

Inter-coder reliability was tested for the coding work. The minimum acceptable level of the inter-coder reliability coefficients was selected as 0.80 as suggested by Lombard et al. (2002). Following the previous studies (Kiousis et al., 2007; Ragas, 2010), we examined two inter-coder reliability indexes: Holsti’s (1969) formula and Scott’s Pi. The coding and inter-coder reliability test processes were repeated until the values of both indexes reached the level of 0.80 and higher for all variables in the coding sheet. The inter-coder reliability test coefficients ranged from 0.80 to 0.98. In the final stage, the first author coded the rest of the news articles in the news sample due to the high cost of human coding and limited budget of this project (Melloni et al., 2017).

Results

Table 2 reported the number and percentage of articles that contained the seven attributes in the two elite newspapers and local newspapers. These numbers and percentages showed that some attributes, such as products and services and financial performance, were reported more frequently than other attributes, such as citizenship, governance, and workplace. Another pattern was that even for the same attribute, its frequency of coverage and proportion in the news articles varied across time both in the elite newspapers and local newspapers. The means and standard deviations of the five media reputation measures of seven attributes were reported in Table 3. The scores of all these five measures ranged from 0 to 100. It showed Fombrun-Shanley index media reputation ratings, which did not count negative items, were higher than all other linear and non-linear measures in all attributes except for innovation attribute. For the non-linear measures, Deephouse media endorsement index media reputation ratings were higher than those of Deephouse media favorability index.

Table 2.

Number of Articles of Seven Attributes in the Elite and Local Newspapers.

		The New York Times					The Wall Street Journal					Local newspapers
		2007	2008	2009	2010	2011	2007	2008	2009	2010	2011	2007	2008	2009	2010	2011
P&S	Number	25	10	28	30	31	48	30	37	38	33	174	102	251	208	187
P&S	%	51.02	29.41	48.28	52.63	54.39	53.33	53.57	40.22	40.43	50.77	56.13	37.23	41.35	41.11	39.96
LEAD	Number	9	0	5	10	15	31	11	14	17	22	75	34	103	80	98
LEAD	%	18.37	0.00	8.62	17.54	26.32	34.44	19.64	15.22	18.09	33.85	24.19	12.41	16.97	15.81	20.94
PERF	Number	23	18	36	29	29	47	28	63	57	38	136	107	207	182	166
PERF	%	46.94	52.94	62.07	50.88	50.88	52.22	50.00	68.48	60.64	58.46	43.87	39.05	34.10	35.97	35.47
INNO	Number	3	4	3	7	3	6	1	1	4	2	23	46	38	35	23
INNO	%	6.12	11.76	5.17	12.28	5.26	6.67	1.79	1.09	4.26	3.08	7.42	16.79	6.26	6.92	4.91
CITI	Number	5	0	4	3	2	3	6	5	10	5	57	26	111	88	67
CITI	%	10.20	0.00	6.90	5.26	3.51	3.33	10.71	5.43	10.64	7.69	18.39	9.49	18.29	17.39	14.32
GOV	Number	5	6	5	6	4	4	8	3	11	6	37	19	88	62	38
GOV	%	10.20	17.65	8.62	10.53	7.02	4.44	14.29	3.26	11.70	9.23	11.94	3.93	14.50	12.25	8.12
WORK	Number	1	2	2	2	0	3	0	3	3	2	18	19	38	42	25
WORK	%	2.04	5.88	3.45	3.51	0.00	3.33	0.00	3.26	3.19	3.08	5.81	6.93	6.26	8.30	5.34

Note. The percentage (%) number in the table was calculated by the number of articles containing a attribute divided by the total number of articles in NYT, WSJ and Local newspapers in each year. P&S represents the attribute of products and services, LEAD represents the attribute of leadership, PERF represents the attribute of financial performance, INNO represents the attribute of innovation, CITI represents the attribute of citizenship, GOV represents the attribute of governance, and WORK represents the attribute of workplace.

Table 3.

Descriptives of Five Measures of Media Reputation of Seven Attributes.

	Measures	P&S	Lead	PERF	INNO	CITI	GOV	Work
Elite newspapers	Kiousis index	4.27 (7.88)	58.27 (38.25)	31.49 (27.90)	48.65 (50.67)	51.81 (47.06)	11.96 (29.07)	7.21 (23.75)
	Fombrun- Shanley index	95.35 (11.03)	77.29 (37.84)	49.90 (29.86)	48.65 (50.67)	56.31 (46.66)	17.59 (32.54)	14.41 (34.73)
	Wry index	−.38 (12.99)	55.05 (39.51)	12.23 (37.34)	48.65 (50.67)	45.72 (53.49)	−11.57 (52.26)	−4.50 (40.94)
	Deephouse media endorsement index	7.05 (50.66)	67.90 (44.43)	14.06 (59.04)	48.65 (50.67)	47.56 (54.82)	−13.89 (63.93)	−5.41 (40.45)
	Deephouse media favorability index	.02 (3.37)	46.43 (41.18)	8.40 (26.75)	48.65 (50.67)	43.97 (53.65)	−10.26 (50.46)	−5.11 (40.53)
Local newspapers	Kiousis index	12.03 (10.13)	76.89 (21.34)	43.97 (21.65)	84.76 (32.70)	70.38 (30.44)	21.84 (24.24)	32.59 (34.33)
	Fombrun- Shanley index	89.33 (13.54)	95.33 (8.91)	67.63 (21.13)	86.04 (32.21)	72.26 (30.17)	40.47 (33.70)	46.15 (39.80)
	Wry index	3.62 (16.14)	74.13 (22.42)	24.08 (35.17)	83.86 (33.66)	54.21 (47.30)	−12.34 (46.65)	8.84 (55.12)
	Deephouse media endorsement index	22.46 (56.40)	89.72 (23.21)	34.50 (46.07)	87.08 (33.28)	55.37 (47.57)	−11.24 (65.10)	15.19 (69.52)
	Deephouse media favorability index	.43 (5.68)	61.44 (28.45)	16.87 (21.39)	81.65 (36.30)	48.41 (46.04)	−11.65 (38.80)	4.53 (45.28)

Note. The numbers in the table are the mean of media reputation ratings. The numbers in the parentheses are the standard deviations. Number of observations = 37. P&S represents the attribute of products and services, LEAD represents the attribute of leadership, PERF represents the attribute of financial performance, INNO represents the attribute of innovation, CITI represents the attribute of citizenship, GOV represents the attribute of governance, and WORK represents the attribute of workplace.

RQ1 asked whether the existing measures of media reputation generate ratings that are significantly different from each other when only the elite national newspapers are used as the news sample. We performed one-way ANOVA analysis for the elite newspapers and local newspapers separately. The results for the elite newspapers showed that the five measures were significantly different from each other (F = 18.6, sig. <.001). RQ2 asked whether the existing measures of media reputation generate ratings that are significantly different from each other when only the local metro newspapers are used as the news sample. The results for the local newspapers showed that the five measures are also significantly different from each other (F = 31.6, sig. <.001). RQ3 asked whether the same measures of media reputation generate reputation ratings that are significantly different from each other when the elite and local newspapers are separately used as the news samples. The results of two-way ANOVA showed that the same index generates significantly different ratings when the two different news samples are used separately (F = 73.7, sig.<.001).

To further pinpoint exactly how these five measures are different from each other, we performed the Bonferroni post hoc multiple comparison procedures. The results of multiple comparisons of five measures were reported in Table 4. The results showed that Fombrun-Shanley index was significantly different from all the other four measures. Kiousis index was significantly different from Deephouse media favorability index, Fombrun-Shanley index, and Wry index. Deephouse media favorability index was significantly different from Deephouse media endorsement index, Fombrun-Shanley index, and Kiousis index. Deephouse media endorsement index was significantly different from Deephouse media favorability index and Fombrun-Shanley index. Wry index was significantly different from Kiousis index and Fombrun-Shanley index.

Table 4.

Post Hoc Bonferroni Multiple Comparisons in Attribute Media Reputation Score in Combined Elite and Local Newspaper Sample.

(I) Index	(J) Index	Mean difference (I–J)	Sig.	95%CI
(I) Index	(J) Index	Mean difference (I–J)	Sig.	Lower bound	Upper bound
DEEF	DEEP	−8.78*	0.03	−17.18	−0.38
	FOMB	−37.14*	0.00	−45.54	−28.74
	KIOU	−15.75*	0.00	−24.15	−7.35
	WRY	−3.26	1.00	−11.66	5.15
DEEP	DEEF	8.78*	0.03	0.38	17.18
	FOMB	−28.36*	0.00	−36.76	−19.96
	KIOU	−6.97	0.20	−15.37	1.43
	WRY	5.52	0.65	−2.88	13.92
FOMB	DEEF	37.14*	0.00	28.74	45.54
	DEEP	28.36*	0.00	19.96	36.76
	KIOU	21.39*	0.00	12.99	29.79
	WRY	33.88*	0.00	25.48	42.28
KIOU	DEEF	15.75*	0.00	7.35	24.15
	DEEP	6.97	0.20	−1.43	15.37
	FOMB	−21.39*	0.00	−29.79	−12.99
	WRY	12.49*	0.00	4.09	20.89
WRY	DEEF	3.26	1.00	−5.15	11.66
	DEEP	−5.52	0.65	−13.92	2.88
	FOMB	−33.88*	0.00	−42.28	−25.48
	KIOU	−12.49*	0.00	−20.89	−4.09

DEEP denotes Deephouse media endorsement index. DEEF denotes Deephouse media favorability index. FOMB denotes Fombrun-Shanley index. WRY denotes Wry index. KIOU denotes Kiousis index.

The mean difference is significant at the 0.05 level.

Discussion

Previous studies used different measures of media reputation in the study of the impacts of media coverage on corporate reputation and financial performance. Zhang (2016) compared them at the overall level and found the ignificant differences among the measures and between local and national elite newspapers. This study compared five different measures at the attribute level based on the affective dimension of the second-level agenda-setting theory, and found the significant differences at the attribute level as well.

In particular, the examinations of both elite newspapers and local newspapers found that the five measures of media reputation were significantly different from each other. This finding suggested that significantly different results would be generated when different measures are used to measure the attributes in the media coverage of companies in either elite newspapers, local newspapers or both. This has the methodological implication for future research. Researchers need to be noted that when they choose one of these measures to conduct research, their findings would be influenced by their choice of the measure of media reputation.

For example, when researchers choose the Fombrun-Shanley index, which excludes negative items and includes neutral items, they should be aware that their findings would be significantly higher than those that used the other four measures. When they choose the Kiousis index, which includes only positive items, their findings would be significantly lower than other indexes, especially the Deephouse media favorability index and the Fombrun-Shanley index. When they choose the Wry index, which includes negative items, their findings would be significantly lower than those that use the Kiousis index or the Fombrun-Shanley index. When researchers choose Deephouse media favorability index which includes negative items and neutral items and counts relevant units only, their findings would be significantly lower than those that use Deephouse media endorsement index, Fombrun-Shanley index, or Kiousis index. When they choose Deephouse media endorsement index, their findings would be significantly higher than those that use Deephouse media favorability index, but lower than those that use Fombrun-Shanley.

Another important finding is that elite newspapers and local newspapers generated different results even when the same measure was used. This finding also has the methodological implication for future research. Many previous studies argued that The New York Times has inter-media agenda-setting capability, thus, it is reasonable to use it as the news source for assessing media reputation. This argument is appealing since the coding work for a few national elite newspapers will be much less than that for many local newspapers. However, the finding of this study showed that when these national elite newspapers are used as the news sample to measure media reputation, they generate significantly different results from those when the local newspapers are used. In other words, the elite newspapers could not be used as the proxy for the local newspapers to measure media reputation.

The corporate reputation ratings, such as The Harris Poll Reputation Quotient (RQ) and The RepTrak^® Pulse index, were generated by the national survey of general public. Pew Research Center (2012) showed that most Americans use local newspapers as their major news source rather than the national elite newspapers. Therefore, local newspapers should be a better news sample to examine the second-level agenda-setting effect for business news than elite newspapers. In addition, because of the geographical proximity issue, some companies may be covered more by local newspapers if their headquarters are in the local newspaper markets. The dataset of this study indeed showed this pattern.

Media reputation is one of the driving forces of corporate reputation (Fombrun & Shanley, 1990), and a strategic resource for a firm (Deephouse, 2000). There are multiple studies that suggested media reputation has significant influence on corporate financial performance and stock prices (e.g., Agnihotri, 2014; OuYang et al., 2017; Tong, 2013). Moreover, media reputation has also been used as the measure of public endorsement and corporate legitimacy, indicating the wide application of this concept in the business world. The role of media reputation in the future research of business communications should be more prominent as the business and communication activities are further integrated. As there are multiple measures of media reputation in the literature, researchers need to know whether these measures generate the same assessments or significantly different ones.

Zhang (2016) argued that the difference in media reputation measurement is one of the reasons for the inconsistent findings in the relationship between media coverage and corporate reputation in the previous studies. He found the measures of media reputation created significantly different results at the overall level. This study found this is also the case at the attribute level. Both studies found that the news sample also affects the media reputation measurement. Hence when different measures are used either at the overall or attribute level, the research findings would be also very different. These findings have important implications for the application of the second level agenda-setting theory and priming theory in the business communication field. The application of a theory in a certain field cannot advance if there are always inconsistent results caused by different measures. In other words, the validity of a theory cannot be confirmed if some results verify it while others do not.

There are two common causes for inconsistent results: (1) the theory is not valid in this certain field; (2) there are some issues/problems in the operationalization of the key concepts in the theory. This study found that the issues in the operationalization of media reputation, are the cause of the inconsistent results in the previous studies. Therefore, we suggested the need to develop a consensus in the measurement of media reputation for future research. If a consensus on the measurement is formed, the validity of the second level agenda-setting theory and priming theory explaining how business news coverage affects the public’s perception of a company can be further examined. In this sense, this study is an indispensable step in advancing the validation of the media effects theories in the business communication setting.

The advancement of the application of the second level agenda-setting theory and priming theory in the business communication setting has practical implications for brand managers and marketing communication professionals. As news media influence people’s perceptions of companies and associated brands, public relations or corporate communication practitioners should consistently monitor the media coverage of their companies and brands. Quantifying the media coverage would help them find the patterns and trends of their companies’ media reputations. However, they confront the same question as the academic researchers do: whether these measures generate the essentially same results or significantly different ones? Based on what we found in this study, practitioners will know how different measures of media reputation would generate significantly different assessments of media reputation at the attribute level, and different news sources (local vs. national) would also render significantly different assessments. Almost all local newspapers resulted in higher reputation scores for the seven attributes except for product and services in the Forbrum-Stanley index. Therefore, companies need to use multiple measures and local news media to assess the media reputations of their companies rather than one single measure or just national news media without solid rationale. Of course, these significantly different assessments would render confusions in interpreting the implications of the results for the business communication strategies, which is very detrimental to the application and validity of media reputation in practice. Therefore, developing a consensus in the measurement of media reputation is needed in the profession as well.

As the first study comparing the existing measures of media reputation at attribute level, this study has several limitations. The first limitation is the generalizability of the findings across different industries as these findings were from the data of the companies in the food manufacturing industry. Although this study is especially relevant to the current COVID-19 pandemic when people have more concerns with food safety, caution should be taken when generalizing the conclusions to other industries, as Schudson (1984) suggested media messages may have different levels of influence on people’s attitudes toward different products/services. The small number of sample firms also limited the application of multivariate analysis.

Another limitation of the study is that we did not survey consumers’ responses to the media coverage of the attributes as did in other agenda-setting studies. We suggested future researchers to test the perceptions of media reputation attributes and corporate reputation attributes to determine which measurement of media reputation best predicts the consumers’ perceptions of corporate reputation.

The data used in this study also have limitations. We obtained the news data from two large news databases: Lexis-Nexis and Factiva. Although they are very large databases, there will still be omission of some newspapers/articles in them (Weaver and Bimber, 2008). Thus far, some relevant articles may not be obtained through the search tools provided by these datasets.

Finally, this study only examined whether different measures generate different ratings in newspapers only. With the increasing importance of social media in business communication, it is necessary to measure companies’ social media reputations. Several very recent studies demonstrated that social media have significant impacts on organizational reputation (e.g., Xu & Wu, 2020; Yeo et al., 2020). Future research on media reputation should apply the machine learning approach to process the large quantity of social media content regarding different reputation attributes. Researchers should also compare the five different measures or create a new measure that best captures the attribute ratings. Assessing the relative importance of social media coverage and traditional media coverage in forming media reputation is another important future research topic.

In closing, media reputation has become a significant topic in business communication. This is the first study that compared multiple measures of media reputation at the attribute level. The findings showed the five existing measures of media reputation are significantly different from each other, and news sample choice also affects the results. The findings of this study call for the careful selection of measures and the understanding of how different measures can tremendously affect the media reputation ratings.

Footnotes

Authors’ Note

This manuscript is original and is not under consideration or published elsewhere.

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

Xiaoqun Zhang

Notes

Author Biographies

Xiaoqun Zhang is Associate Professor in the Department of Media Arts at University of North Texas.

Louisa Ha is Professor in the School of Media and Communication at Bowling Green State University, USA. She is the Editor-in-Chief of Journalism and Mass Communication Quarterly.

References

Agnihotri

(2014). Mass-media-based corporate reputation and firms’ market valuation: Evidence from emerging markets. Corporate Reputation Review, 17(3), 206–218.

Antle

(1999). The new economics of agriculture, American Journal of Agricultural Economics, 81, 993–1010.

Bansal

Clelland

(2004). Talking trash: Legitimacy, impression management, and unsystematic risk in the context of the natural environment. Academy of Management Journal, 47(1), 93–103.

Becker

L. B.

McCombs

M. E.

(1978). The role of the press in determining voter reactions to presidential primaries. Human Communication Research, 4(4), 301–307.

Bonaiuto

Caddeo

Carrus

Dominicis

S. D.

Maroni

Bonnes

(2012). Food reputation impacts on consumer’s food choice. Corporate Communications: An International Journal, 17(4), 462–482.

Carroll

C. E.

(2013). The handbook of communication and corporate reputation. John Wiley & Sons, Inc.

Deephouse

D. L.

(1996). Does isomorphism legitimate? Academy of Management Journal, 39(4), 1024–1039.

Deephouse

D. L.

(2000). Media reputation as a strategic resource: An integration of mass communication and resource-based theories. Journal of Management, 26(6), 1091–1112.

Emler

(1990). A social psychology of reputation. European Review of Social Psychology, 1(1), 171–193.

10.

Fan

Geddes

Flory

(2013). The Toyota recall crisis: Media impact on Toyota’s corporate brand reputation. Corporate Reputation Review, 16(2), 99–117.

11.

Fombrun

C. J.

(2006). The Reptrak system. Presented at the 10th Anniversary Conference on Reputation, Image, Identity and Competitiveness. New York, NY, May 25-28.

12.

Fombrun

C. J.

Shanley

(1990). What’s in a name? Reputation building and corporate strategy. Academy of Management Journal, 33, 233–258.

13.

Food Insight. (2020, June 26). COVID-19 and food safety concerns: Results from the 2020 food and health survey. https://foodinsight.org/covid-19-and-food-safety-concerns-results-from-the-2020-food-and-health-survey/

14.

Food Safety Magazine (2012). Recall: The food industry’s biggest threat to profitability. https://www.foodsafetymagazine.com/signature-series/recall-the-food-industrys-biggest-threat-to-profitability/

15.

Frewer

L. J.

Howard

Hedderley

Shepherd

(1997). The elaboration likelihood model and communication about food risks. Risk Analysis, 17(6), 759–770.

16.

Gans

H. J.

(1979). Deciding what’s news: A study of CBS evening news, NBC nightly news Newsweek and Time. Pantheon.

17.

Gilbert

Eyal

McCombs

M. E.

Nicholas

(1980), The state of the union address and press agenda, Journalism Quarterly, 57(4), 584–88.

18.

Golan

(2006). Inter-media agenda setting and global news coverage: Assessing the influence of The New York Times on three television news evening programs. Journalism Studies, 7(2), 323–33.

19.

Grauman

Kruse

(1990). The environment: Social construction and psychological problems. In Himmelveit

H. T.

Garkell

(Eds.), Societal psychology (pp. 212–229). Sage.

20.

Holsti

(1969). Content analysis for the social sciences and humanities. Addison-Wesley Publishing Company.

21.

Iyengar

Kinder

D. R.

(1987). News that matters. University of Chicago Press.

22.

Janis

I. L

Fadner

(1943). The coefficient of imbalance. Psychometrica, 8, 105–119.

23.

Jonge

J. D.

Trijip

H. V.

Renes

R. J.

Frewer

L. J.

(2010). Consumer confidence in the safety of food and newspaper coverage of food safety issues: A longitudinal perspective. Risk Analysis, 30(1), 125–142.

24.

Jowell

Witherspoon

Brook

(Eds) (1987) British social attitudes: The 1987 report. Cower Pub Co.

25.

Keith

(2019, 18 February). Reputation management tips for food companies. https://blog.flexxray.com/blog/reputation-management-tips-for-food-companies

26.

Kim

McCombs

M. E.

(2007). News story descriptions and the public’s opinions of political candidates. Journalism & Mass Communication Quarterly, 84(2), 299–314.

27.

Kiousis

Popescu

Mitrook

(2007). Understanding influence on corporate reputation: An examination of public relations efforts, media coverage, public opinion, and financial performance from an agenda-building and agenda-setting perspective. Journal of Public Relations Research, 19(2), 147–165.

28.

Lee

S. Y.

Carroll

C. E.

(2010). The emergence, variation and evolution of CSR on the media and public agenda. Journal of Business Ethics, 104(1), 115–131.

29.

LexisNexis Academic. (2021). https://www.lexisnexis.com/pdf/LNA-Brochure-053013.pdf

30.

Lombard

Snyder-Duch

Bracken

C. C.

(2002). Content analysis in mass communication: Assessment and reporting of intercoder reliability. Human Communication Research, 28, 587–604.

31.

Mason

(2019). Media frames and crisis events: Understanding the impact on corporate reputations, responsibility attributions, and negative affect. International Journal of Business Communication, 56(3), 414–431.

32.

McCombs

M. E.

(1997). Building consensus: The news media’s agenda-setting roles. Political Communication, 14(4), 433–443.

33.

McCombs

M. E.

Ghanem

(2001). The convergence of agenda setting and framing. In Reese

Gandy

Grant

(Eds.), Framing public life (pp. 67–81). Erlbaum.

34.

McCombs

M. E.

Llamas

J. P.

Lopez-Excobar

Rey

(1998). Candidate’s images in Spanish elections: Second-level agenda-setting effects. Journalism & Mass Communication Quarterly, 74(4), 703–717.

35.

McCombs

M. E.

Reynolds

(2009). How the news shapes our civic agenda. In Bryant

Oliver

M. B.

(eds) Media effects: Advances in theory and research. (pp. 1–16) Lawrence Erlbaum Associates, Inc.

36.

Meijer

(2004). Does success breed success? Effects of news and advertising on corporate reputation. Aksant Academic Publisher.

37.

Meijer

Kleinnijenhuis

(2006). News and corporate reputation: Empirical findings from the Netherlands. Public Relations Review, 32, 341–348

38.

Melloni

Gaglio

Perogo

(2017). Saying more with less? Disclosure conciseness, completeness, and balance in integrated reports. Journal of Accounting and Public Policy, 36(3), 220–238.

39.

Neuendorf

K. M.

(2002). The content analysis guidebook. Sage Publications.

40.

OuYang

Wei

Liu

(2017) Information asymmetry and investor reaction to corporate crisis: Media reputation as a stock market signal. Journal of Media Economics, 30(2), 82–95.

41.

Pew Research Center. (2012, 13 June). 72% of Americans follow local news closely. https://www.pewresearch.org/internet/2012/04/12/72-of-americans-follow-local-news-closely/.

42.

Pollock

T. G.

Rindova

V. P.

(2003). Media legitimation effects in the market for initial public offerings, Academy of Management Journal, 46(5), 631–642.

43.

Ponzi

L. J.

Fombrun

C. J.

Gardberg

N. A.

(2011). RepTrak ™ pulse: Conceptualizing and validating a short-form measure of corporate reputation. Corporate Reputation Review, 14(1), 15–35.

44.

Ragas

M. R.

(2010). Agenda-building and agenda-setting in corporate proxy contests: Exploring influence among public relations efforts, financial media coverage and investor opinion. [Unpublished dissertation], University of Florida.

45.

Reese

S. D.

Danielian

L. H.

(1989). Intermedia influence and drug issue: Converging on cocaine. In Shoemaker

P. J.

(Ed.), Communication campaigns about drugs: Government, media, and the public. (pp. 29–45). Erlbaum.

46.

Sample Size Calculator. (2021). https://www.surveysystem.com/sscalc.htm

47.

Schudson

(1984). Advertising, the uneasy persuasion: Its dubious impact on American society. Basic Books, Inc., Publishers.

48.

Sheafer

(2007). How to evaluate it: The role of story-evaluative tone in agenda setting and priming. Journal of Communication, 57(1), 21–39.

49.

Sheiham

Marmot

Rawson

Ruck

(1987). Food values: Health and diet. In Jowell

Witherspoon

Brook

(Eds), British Social Attitudes: The 1987 Report. Cower Pub Co.

50.

Tong

(2013). Media reputation in initial public offerings: A study of financial news coverage in Hong Kong. Public Relations Review, 39(5), 470–483.

51.

Vliegenthart

Walgrave

(2008). The contingency of intermedia agenda setting. A longitudinal study in Belgium. Journalism and Mass Communication Quarterly, 85(4), 860–877.

52.

Vogler

Schranz

Eisenegger

(2016). Stakeholder group influence on media reputation in crisis periods. Corporate Communications: An International Journal, 21(3), 322–332.

53.

Wartick

S. L.

(1992). The relationship between intense media exposure and change in corporate reputation. Business & Society, 31, 33–42.

54.

Weaver

D. A.

Bimber

(2008). Finding news stories: A comparison of searches using LexisNexis and Google News. Journalism and Mass Communication Quarterly, 85(3): 517–532

55.

Wry

Deephouse

McNamara

(2006). Substantive and evaluative media reputations among and within cognitive strategic groups, Corporate Reputation Review, 9(4), 225–242.

56.

(2020). Countering reactance in crisis communication: Incorporating positive emotions via social media. International Journal of Business Communication, 57(3), 352–369.

57.

Yeo

S. L.

Pang

Cheong

Yeo

J. Q.

(2020). Emotions in social media: An analysis of Tweet responses to MH370 search suspension announcement. International Journal of Business Communication, 57(2), 194–211.

58.

Zhang

(2016). Measuring media reputation: A test of the construct validity and predictive power of seven measures. Journalism and Mass Communication Quarterly, 94(3), 884–905.

59.

Zhang

(2018a). Estimating the weights of media tonalities in the measurement of media cov erage of corporations. Communication Research, 45(7), 987–1011.

60.

Zhang

(2018b). Intermedia agenda-setting effects in corporate news: Examining the influence of The New York Times and The Wall Street Journal on local newspapers. Journal of Applied Journalism and Media Studies, 7(2), 245–263.