Substantive Attributes and Sub-attributes

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

44

and Sever (1999), this study uses a quantitative content analysis, from the perspective of second-level agenda-setting theory, to analyze the relationship of the rank of substantive attributes and their sub-attributes in the corporate news release and the printed news coverage in Taiwan.

5.1 Substantive Attributes and Sub-attributes

The first two research questions addressed in this study is in order to discuss the relationship of substantive attributes and sub-attributes between the news release and the news coverage. What is the relationship between substantive sub-attributes in corporate news release and in news coverage? What is the relationship between substantive attributes in corporate news release and in news coverage? Through the rank of attributes and

sub-attributes determined by the frequency in the agenda of company reputation and comparing the news release with the news coverage, results indicate that the rank of

sub-attributes in the news release is positively correlated with the news coverage though the sequence of rank and places are different. However, the rank of attributes in the news release is not correlated with the news coverage.

In the news release, the company emphasizes on those sub-attributes about the intrinsic products and service, and company images and appeals; however, the news coverage in the newspaper focuses on company's stance, product quality, investment, profits, and future opportunities, mostly related to the financial performance. And the same situation is reflected

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

45

on the results of attributes, which the news coverage put more focus on the financial performance. Meanwhile, substantive attribute and sub-attribute related to the financial performance are relatively neglected in the news release, for example, the attribute "financial performance" ranks fourth out of total six places, and its sub-attributes, "Company has record of being profitable" and "Company is low risk investment" ranks thirteenth and seventeenth out of total twenty places. And this may be explained that journalists set the media agenda of HTC reputation mainly in the aspect of financial performance, such as high and low risk of investment, or records of being profitable and unprofitable, which is not the key messages company tends to deliver to the public primarily. Instead, the priority the company tends to deliver is the product and brand appeals intrinsically. Above all, with different focuses, this fact may also explain that journalists in Taiwan may still prefer to use the information they gathered independently while selecting news and to use only parts of news releases (Pincus, Rimmer, Rayfield, and Cropp, 1993; Curtin, 1999).

Nevertheless, regarding to the rank of last five sub-attributes and last two attributes in both the news release and news coverage, there are still similarities and consistency. For both news items, attributes about workplace and environment and social responsibility, and also their sub-attributes, such as fair or unfair rewards and supporting good causes or not, are few.

And this could be reflected the fact that not only does the company not focus on delivering these two kinds of messages to the public, but also do journalists has less desire and are not

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

46

autonomic to access to these subjects or gather related news. In other words, they may simply completely use the news release as the news coverage.

5.2 Affective Tones and Agenda Valence

The last research question addressed in the study is that what the relationship is between the agenda valence in corporate news release and in news coverage. The content analysis shows the results that positive and neutral tones of attributes in the news release are positively correlated with ones in the news coverage. Determined by the affective tones of twenty sub-attributes, results of agenda valence indicate that most of news releases were positive and only a few are neutral, and there was no negative ones. However, for the news coverage, though most of news articles are still positive, there are also 50 negative news articles about HTC reputation, and more neutral ones than ones in the news release. And this could be also explained that journalists may prefer to gather the information in the news selection process from various news sources, including positive, negative, and neutral sources, instead of completely accept the news release only.

Furthermore, although there is no negative news release in the sample, there are still negative news coverage in the newspapers. The sources of those 50 negative news articles in the sample can be divided into three categories: banks/financial consultancy companies (the most, e.g. Golden Sachs, CLSA Asia-Pacific Markets, J.P. Morgan, etc.); consumers/opinion

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

47

leaders (second most, e.g. politician, celebrity, etc.); HTC itself (e.g. employees, press conferences, the conference call, etc.).

5.3 Limitations

As is the case of most research, there are limitations. First, this study focuses on the agenda of company reputation. Based on Reputation Quotient, the content analysis could further analyze the data and find results. However, while coding, there were some sample articles irrelevant to the company reputation and the coders simply coded them as missing variables. Moreover, this study included all sort of news article related to HTC reputation, and there may be repeated or highly similar news articles in the news coverage sample.

Second, some statistics used in this study are very simple. By using descriptive analysis, such as frequencies and percentage, results could only explain simple data descriptions and part of the relationship, but could not explain and clarify the complete relationship. And also the number of values while analyzing the relationship of substantive attributes is a concern which may influence the result. Lastly, the affective tone is only used to judge the valence of the agenda. Instead, this study does not clarify the causality between the affective tones of attributes and the agenda valence clearly.