. aura”. . , £3 0’ '1- add”. pm. .7 SHn0\fi08 LIRPA RV Michigan. state University This is to certify that the dissertation entitled U.S. PRESS CDVERAGE OF THE ISRAELI-PALESTINIAN CONFLICT: AN INTEGRATIVE THEDRETICAL NDDEL OF INFLUENCEOFNEWSTREAB/IENTOFNEWSSOUICES presented by Kuang-Kuo Chang has been accepted towards fulfillment of the requirements for Ph .D . degree in Mass Media ;) _£¢ 2-» Major prof r Date 4-11-03 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 4 v— - A’- . 4-— ‘ v _ v v __. '- —v‘iw 4-_7~ PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. I DATE DUE DATE DUE DATE DUE ”MAY ' ' 1 4 2006 6/01 cJCIRC/DateDuep65-p. 15 U.S. PRESS COVERAGE OF THE ISRAELI-PALESTINIAN CONFLICT: AN INTEGRATIVE THEORETICAL MODEL OF INFLUENCE OF NEWS TREATMENT OF NEWS SOURCES By Kuang-Kuo Chang A DISSERTATION Submitted to Michigan State University In partial fulfilhnent of the requirements for the degree of DOCTOR OF PHILOSOPHY School of Journalism 2003 ABSTRACT U.S. PRESS COVERAGE OF THE ISRAELI-PALESTINIAN CONFLICT: AN INTEGRATIVE THEORETICAL MODEL OF INFLUENCE OF NEWS TREATMENT OF NEWS SOURCES By Kuang-Kuo Chang Highly involved in the prolonged Israeli-Palestinian conflict includes four major U.S. newspapers through their uses of news sources who might have framed this issue. This dissertation endeavors to scrutinize the adoption and resulting impacts of news sources with an integrative theoretical path model that employs concepts of hierarchy of influences, the international commmlication world system and framing analysis. Findings from data analyses showed that Israeli officials were treated more favorably than their counterpart Palestinian officials as a result of greater press access in Israel and a friendlier foreign policy toward that country. However, more favorable news treatment of Israeli officials doesn’t lead to selection of Palestinian terrorism or peace process as the dominant news frame, while resulting in thematic nature of news content. Neither does better treatment of Israeli officials assign, directly or indirectly, causal responsibility to Palestine. That Israeli casualties are capable of directly assigning causal responsibility to Palestine provides an additional prospective for future framing research. Therefore, theoretically, (more favorable news treatment of Israeli official) news sources have only marginal bridging roles between contextual variables of international communication (press access and foreign policy) and framing research (nature of news content) as proposed in the framework and path model. © Copyright by Kuang-Kuo Chang 2003 For the peace and health of my family iv ACKNOWLEDGEMENTS Since the moment I received the admission e-mail from Professors Stephen Lacy and Charles Salmon that welcomed me to the “Michigan State Family,” I have been part of this proud and sweet group during my four-year journey. I really believe that I am destined, after a two-year detour at Oregon, to graduate a tough Spartan. Go Green! Go White! As always. These four years, particularly when approaching the end of my studies, would’t have been easy without the great caring and encouragement from a number of people. During the past two years, I experienced the ups and downs of my life as a doctoral student and as a human being, mostly as a result of my frustrating job search. Now, with the title of PhD. being attached to my last name, I feel exuberant and inspired because a bright future is ahead of me. This new career wouldn’t have been realized without tremendous support from my dissertation committee members and several other good fiiends. Reading through a chunk of material is not a fim task for anyone, especially if he has to do it more than twice and provide insightful comments and suggestions for revision. I wholeheartedly appreciate my adviser and chairperson, Professor Lacy, for his invaluable efforts and advice on making this dissertation a better work and making me a better scholar. Dr. Lacy is simply the role model for me and perhaps for a lot of other people. Appreciations also go to Professors Salmon and Fred Fico and Assistant Professor Dan Shaver for providing useful suggestions for improving this dissertation. Professors Salmon and Fico also play an important role in nurturing me to be a better researcher and helping me land a job at Nanyang Technological University. Particular thanks are indebted to Eric Freedman, who took pains to proofread this manuscript, and to Geri and Flora, who helped me with intercoder reliability testing. Thanks also go to Nancy Ashley for helping me resolve those unexpected problems, so that I could concentrate on completing and defending this dissertation. My gratitude also is owed to many good friends, such as Wayne Wanta, Tien-Tsung Lee, Professor James Detjen, Hugh Martin and Arvind Diddi, for providing their spiritual support and friendship throughout these years. Indebtednesses particularly are owed to my aging parents, who’ve allowed me to study in the United States for more than a decade without me worrying too much. It’s payback time for me, not only to take good care of them but also to make them proud of their son. Appreciation and love also go to my newlywed wife, who quit her job and came to this icy-cold place to take care of me. We will soon begin our new life journey together. Many thanks also are owed to my younger brother and sister-in-law for offering me a very comfortable home to live and work while in Taiwan and for taking good care of our parents while I have been away. Finally, I am especially indebted to my adviser, Dr. Lacy. If not for his caring, understanding, encouraging, patience, support, tolerance and advice, this dissertation may not have been completed by now. Had more self-disciplined, I could have become a whole lot better scholar than sooner. I’m glad that I have learned the tools and ways of thinking fi'om him to improve myself as a scholar and as an individual. I owe you so much, and thank you very much, Sir. vi TABLE OF CONTENT List of Tables ................................................................................ xi List of Figures ............................................................................... xiii Chapter 1. Introduction ............................................................................ 1 Purpose of the Study .............................................................. 1 Historical Review of the Conflict .............................................. 3 2. Theoretical Framework ............................................................... 8 3. Literature Reviews .................................................................... 16 World News Coverage ........................................................... 16 Foreign Policy ............................................................. 16 New Sources (officials) .................................................. 17 Organizational ............................................................. 19 Event-Oriented (media routines) ....................................... 21 Context-Oriented ......................................................... 23 Framing Analysis ................................................................. 26 Foreign Policy ............................................................. 27 New Sources (officials) .................................................. 29 News Sources (non-officials) ........................................... 3O Organizational ............................................................. 3 1 Media Routines ........................................................... 32 Foreign Policy on Israel, Palestine and the Crisis ............................ 34 Press Coverage of Israel and/or Palestine or Arab ........................... 35 4. Hypotheses .................................... 39 5. Methods ................................................................................. 44 Research Designs ................................................................. 44 Conceptual & Operational Definition of Variables .................. 47 vii 6. Results Data Analysis ..................................................................... Independent Variables ............................................. Newspaper ...................................................... News Provider ................................................. Resource Allocation .......................................... Foreign Policy ................................................. Press Access ................................................... Percentage of Population of the Ethnic Groups ........... Level of Casualty ............................................. Type of News Source .......................................... Dependent Variables .............................................. News lead ...................................................... Frequency of Full Quote ..................................... Placement ...................................................... Number of Words in the First Full Quote .................. Frequency of News Sources in the Full Story ............. Construct of News Treatment of Type of New Sources ... News Treatment ............................................... News Frames ................................................... Episodic vs. Thematic ........................................ Responsibility Assignment .................................. Control Variables .................................................. Statistical Analysis ............................................................... Assumptions of Multiple Regression Analysis ....................... Preliminary Analysis of Data ........................................... Statistical Procedures .................................................... OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Hypotheses Testing ............................................................... The Path Model ................................................................... viii 7. Discussion oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo Press Image in the Media War .................................................. Theoretical Framework Construction .......................................... International Communication .......................................... Uses of News Sources as Dependent Variable ................ News sources .................................................. Press Access ................................................... Foreign Policy ................................................. Casualty ......................................................... News Provider ................................................. Resource Allocation .......................................... Ethnic Population ............................................. Summary ....................................................... Framing Analysis ......................................................... Palestinian Terrorism as Dependent Variable .................. Uses of News Sources ........................................ News Provider ................................................. Nature of News Content as Dependent Variable .............. Uses of News Sources ........................................ News Provider ................................................. Responsibility Assignments as Dependent Variable ......... Uses of News Sources ........................................ Nature of News Content ...................................... Casualty ......................................................... Summary ........................................................................... 8. Conclusions Conclusions ....................................................................... Hierarchy of Influence Hypotheses .................................... International Communication World System ......................... Framing Analysis ......................................................... ix 102 Limitations and Suggestions .................................................... 1 10 Data set .................................................................... l 10 Variable Operations ...................................................... 111 Appendices A. Tables and Figures ................................................................... 114 B Summary of Hypotheses ............................................................ 151 C. Coding Protocol ....................................................................... 152 D. Coding Instructions .................................................................. 156 E. Intercoder Reliability Test Outcomes ............................................. 162 References .................................................................................... l 66 Table H PWSQMPPJN i—‘ I—fi H O m o 11c. 11d. lle. 12a. 12b. 12c. 12d. 12e. 13a. 13b. LIST OF TABLES Frequency of Stories in Different Datelines in the Newspapers ................. Frequency of Stories Prepared by News Provider in the Newspapers .......... Mean Score for Each Type of News Source in News Lead ...................... Placement of First Full Quote for Each Type of News Source .................. Mean Score of Number of Word in the First Full Quote ......................... Mean Score of Number of Full Quote for Each Type of News Source ........ Mean Score of Frequency for Each Type of News Source ....................... Indexed Score for Each Type of News Source .................................... Mean Score of Irnmediacy of Each of News Theme .............................. Mean Score of Causal and Treatment Responsibility Assignment .............. . Multiple Regression Analysis of Israeli Officials News Treatment Overall 11b. Multiple Regression Analysis of Israeli Officials News Treatment in The New York Times ........................................................................ Multiple Regression Analysis of Israeli Officials News Treatment in Washington Post ....................................................................... Multiple Regression Analysis of Israeli Officials News Treatment in The Los Angeles Times ..................................................................... Multiple Regression Analysis of Israeli Officials News Treatment in Houston Chronicle ..................................................................... Multiple Regression Analysis of Peace Process Overall ......................... Multiple Regression Analysis of Peace Process in The New York Times Multiple Regression Analysis of Peace Process in Washington Post ........... Multiple Regression Analysis of Peace Process in The Los Angeles Times Multiple Regression Analysis of Peace Process in Houston Chronicle ........ Multiple Regression Analysis of Palestinian Terrorism Overall ................. Multiple Regression Analysis of Palestinian Terrorism in The News York Times .................................................................................... xi 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 13c. 13d. 13c. 14a. 14b. 14c. 14d. l4e. 15a. 15b. 15c. 15d. 15c. 16. Multiple Regression Analysis of Palestinian Terrorism in Washington Post .. Multiple Regression Analysis of Palestinian Terrorism in The Los Angeles Times .................................................................................... Multiple Regression Analysis of Palestinian Terrorism in Houston Chronicle Logistic Regression Analysis of Nature of News Content Overall ............. Logistic Regression Analysis of Nature of News Content in The News York Times .................................................................................... Logistic Regression Analysis of Nature of News Content in Washington Post ....................................................................................... Logistic Regression Analysis of Nature of News Content in The Los Angeles Times .......................................................................... Logistic Regression Analysis of Nature of the News Content in Houston Chronicle ................................................................................ Multiple Regression Analysis of Causal Responsibility Assignment Overall . Multiple Regression Analysis of Causal Responsibility Assignment in The New York Times ....................................................................... Multiple Regression Analysis of Causal Responsibility Assignment in Washington Post ....................................................................... Multiple Regression Analysis of Causal Responsibility Assignment in The Los Angeles Times .................................................................... Multiple Regression Analysis of Causal Responsibility Assignment in Houston Chronicle ..................................................................... Summary of Hypotheses Testing .................................................... xii 136 137 138 139 140 141 142 143 144 145 146 147 148 149 LIST OF FIGURE Figure 1. Path model of the theoretical framework for overall newspapers xiii CHAPTER I INTRODUCTION Purpose of the Study It was written in 1963 that though small in geographic area, Israel is critical in world political arenas and therefore in American politics (Safi'an, 1963). Over the past nearly 40 years, the roles of Israel and Palestine in international and American political milieu clearly have escalated to an unprecedented level of importance, under the constant spotlight of world politics and news media After decades of military battles that resulted in great loss of human life, Israel and Palestine in recent years began an effort to reconcile their long-lasting conflict by trying to recognize each other, with the United States as an active intermediary. Nonetheless, during the negotiation process whenever a slim hope of peace was imminent, severe military confrontations occurred and casualties ensued. More recently, hundreds of citizens from both sides died in a series of skirmishes after pledges of peace had been made. A reporter of The Los Angeles Times wrote: “If a picture is worth a thousand words, a thousand pictures just might win the war... Every photograph—and indeed, every word—is a weapon in the media war” (Oct. 16, 2000). Directly involved in this media war are the U.S. media that relay, create or frame the press images of both Israelis and Palestinians battling for international public opinion in support of their causes and pointing the blames to the other side. Currently, with more Jews residing in the United States than elsewhere, including Israel (2001), the media war between Israel and Palestine for more favorable press treatment and images in the U.S. press is even more intense as coverage of Israel will likely affect how Americans and world society perceive Israel. Similarly, with a growing number of Arabic immigrants in the United States, depictions of Palestine and the Muslim world are increasingly likely to influence American politics. One particularly significant issue is how The New York Times, a flagship domestic and international newspaper, selects news frames about the two political entities and their prolonged conflicts. The aim of this research is to content analyze how Israel and Palestine are framed in four mainstream U.S. newspapers, partly through their uses of news sources from the two sides in the all-out war for favorable press images. Specifically, this project will content analyze news articles in four mainstream U.S. newspapers—The New York Times, the Washington Post, The Los Angeles Times, and Houston Chronicle—from Jan. 21, 2000, to Jan. 21, 2002. This study applies the flaming concept within a theoretical framework based on the hierarchy of influences model to scrutinize news coverage of the two political entities in the mainstream U.S. newspapers. F urthennore, it allows us to expand the framing concept, which developed from more domestic-oriented issues, to international issues. More importantly, this research would provide a strong case study that adopts some elements of the hierarchy of influences model in framing analysis of international news issues. Review of the Current Conflict As the current research limits itself to a two-year time frame, it seems more appropriate to review the conflict from a more recent time line. For a more comprehensive historical review, please refer to the following studies (Arielli, 1975; Bickerton & Klausner, 2002; Bleaney, 1994; Boulding, 1975; Cohn-Sherbok & El-Alami, 2001; Eban, 1975; Frisch, 1998; Hart, 1987; Kapitan, 1997; Lapidoth, 1975; Lukacs, 1991; Peretz, 1996; Quandt, 2001; Quester, 1975; Shamir, 1975) that examine the issue from a variety of perspectives. Founded in 1964 by Yasser Arafat, the Palestinian Liberation Organization gained popularity after the 1967 Six-Day War during which the first military agreement between the United States and Israel was reached. The objective of the PLO and its military unit, Fatah, was to diminish the Jewish state by way of military force; Fatah carried out its first military attack on Israel in 1965 (Peretz, 1996). The PLO proclaimed in its charter that liberation of Palestine is “a national duty.” The charter stated, “The Palestine Arab people... through the armed Palestine resolution, rejects any solution that would substitute for the complete liberation of Palestine” (cited by Peretz, 1996), and thus, the establishment of the Israeli state was deemed void (Peretz, 1996). By 1968, the PLO had turned itself into the representative of Palestine and Palestinians, particularly after Yasser Arafat was elected the leader of the political entity (Hart, 1987; Peretz, 1996); it was recognized by the United Nations in 1974 (Lukacs, 1991; Peretz, 1996). While the PLO has gained international recognition over the years, Israel and the United States still refused to recognize it. They label it a terrorist organization as a result of its attacks on Israeli civilians (Bickerton & Klausner, 2002; Quandt, 2001 ). The Reagan administration was thought to have privately allowed the Israeli invasion of Lebanon in 1982 led by Ariel Sharon, hoping that the dysfunction of the PLO would diminish Arab militants’ antagonism to a peace agreement (Bickerton & Klausner, 2002; Peretz, 1996; Quandt, 2001). This U.S. wish to establish “Peace for Galilee” appeared to come to an end when the government witnessed Israel’s formidable stand during the Lebanon War (Peretz, 1996). Many believe that sovereignty for Palestine underlay the core of the current issue, and it accordingly prioritized the peace negotiation in 1982 during the Reagan administration, which sought a middle ground by refusing a proposal for an independent state of Palestine and permanent Israeli control of the occupied territories (Frisch, 1998; Peretz, 1996; Quandt, 2001; Segal, 1997). Both Israel and the PLO rejected the proposal as well. Displeased with lack of support from its Arab cohorts during the Lebanon War, the Palestinians experienced mounting frustration and fear that the increased amounts of land garnered by Israel in Palestine would put them in a more disadvantageous situation. They felt more concerned particularly after the Likud party controlled the Israeli government, which publicly endorsed greater Jewish settlement and denounced the notion of land for peace (i.e., McKenna, 1997; Peretz, 1996). Over the years, youthful Palestinians grew impatient and hostile toward Israeli controls and, from time to time, challenged its authority (Peretz, 1996). In late 1987, violence between Jewish settlers and the Palestinian refugees surged and the first Intifada, or uprising, was spoiling. “By December 1987, tension within the territories was palpable. Only a spark was needed to set off an explosion” (Peretz, p. 89, 1996). The first Intifada erupted on Dec. 8, 1987, when four Palestinians laborers returning from Israel after work were slain and seven others seriously injured by the Israeli army (Lukacs, 1991 ). Tremendous protests held during the funeral exploded into uncontrollable violence (Bickerton & Klausner, 2002; F risch, 1998; Peretz, 1996). Unlike the previous unrests, the Intifada was politically motivated. To avoid inflicting civilian casualties, the Israeli government conducted more moderate measures to handle the unexpected uprising (Bickerton & Klausner, 2002; F risch, 1998; Peretz, 1996). But after those measures failed and more severe tactics were used, the resistance turned more organized and effective (Peretz, 1996). Underground militant groups such as Hamas, which called for the demise of Israel and resisted any negotiation, were in control of the uprising (F risch, 1998). The Intifada, which reunited different factions of the region, foregoing internal disagreements seen in past struggles with Israel, was used as a strategy to form the state of Palestine (F rish, 1998). The Intifada seemed also to bring the Palestinians to realize no one, not even the PLO, could be trusted to resolve the sovereignty issue and their self-regarded natural right of return to the place (Peretz, 1996) The first Intifada that lasted for six years ended in late 1993 when the Oslo peace negotiation between Israel and Palestine was reached (Bickerton & Klausner, 2002; Cohn-Sherbok & El-Alami, 2001; Frisch, 1998; Lukacs, 1991; Peretz, 1996; Quandt, 2001). The historic settlement that won the Nobel Peace Prize for Perez and Arafat set a 1999 dateline for an entire peace agreement. The two sides in the Oslo declaration acknowledged: “It is time to put an end to decades of confrontation and conflict, recognize their mutual legitimate and political rights, strive to live in peaceful coexistence and mutual dignity and security and achieve a just, lasting and comprehensive peace settlement and historic reconciliation through the agreed political process” (Israel-PLO Declaration of Principles, 1993). The declaration that proposed an independent state of Palestine did not stop clashes between the two sides. Fights and skirmishes continued throughout the years, and peace negotiations were on and off. Arafat and Palestinians’ ultimate goal of establishing an independent state was put off again and again after threatening to declare it unilaterally and drawing fire from both the Israeli and the U.S. governments (Peretz, 1996; Quandt, 2001). As recently as late 2000 at Camp David, the peace negotiation convened by President Clinton failed after a deal about the sovereignty of Jerusalem, an independent state of Palestine and a right of return for Palestinian refugees fell through between the two camps. In the negotiation, the Arab nations demanded a complete Israeli withdrawal from occupied territories, compensation for displaced Palestinian refirgees and above all, the establishment of a PLO-govemed Palestinian state (i.e., Quantd, 2001). Unyielding to pressure at the negotiation table, Arafat was hailed as a hero when he returned to the West Bank. On the contrary, Prime Minister Barak was criticized as a traitor for willing to give up more land to Palestinians. Courting his hawkish peers, Sharon, the mastermind of the Lebanon invasion, visited the holy site of Temple Mount, despite hostile warnings from the Palestinian militants. Immediately after his tour, the second Intifada broke out on Sept. 28, 2000, in which at least six people were killed and more than 200 hurt. The two political entities blamed each other for the eruption of the on-going uprising, which has caused thousands of deaths, most of them Palestinians, over 18 months. While Palestinians carried out suicide bombings and other terrorist attacks on Israeli civilians, the Israeli government targeted execution killings of identified terrorists. Again, while military truces and peace negotiations, with international interrnediacy, have been on and off, the battles are still on. CHAPTER II THEORETICAL FRAMEWORK The hierarchy of influences model (Shoemaker & Reese, 1996) provides a strong basis for constructing the theoretical framework of the current research. Integrated within this broader framework is the world system model of international communication (Chang, 1998) and framing analysis (Entrnan, 1993; Iyengar, 1991, 1993). Three levels of influence of the hierarchy model adopted are extra-media, organization and media routines. Each higher level is a constraint for the lower levels. Chang’s international communication world system model reflects Shoemaker and Reese’s hierarchy framework to a great degree. The world system concept was initiated by Wallerstein (1974, 1979), who categorized world society into three hierarchies of nations, from the most to the least influential: core, semi-periphery and periphery nations. Chang’s first filter of his integrative world communication model—the world power of the (reported) nation—parallels the extra-media sources in terms of political and economic clout of the participant in the environment. The second filter, where the conceptual dimension of international communication is categorized into either event- or context-oriented, resembles the media routines of the hierarchy model in that media routines address to a great degree news natures and criteria. The interaction of the second filter and the third filter—the nature of attribute of either internal attributes or international interaction—also mirrors that between organization and media routines of the hierarchy model, which leads journalists to cover domestic news and world news in a similar way. Entrnan’s notion of news frames and Iyengar’s idea of nature of news frames (episodic vs. thematic) and ensuing responsibility assignments also relate to the theoretical framework being proposed as it involves news sources. This integrative theoretical framework provides a basis for the current research. Located at the second highest level of the hierarchy are extra-media influences. While the extra-media level includes several factors, it focuses on the sources of the information, or news sources, that affect news content. Shoemaker and Reese (1996) define news sources as “external suppliers of raw material” by ways of interviews, speeches, hearings or corporate reports; sources have tremendous impact on news content because journalists often count on them for information. News sources also often are news framers who offer or define themes or symbols that help journalists to make sense the environment (Friedland & Zhong, 1996; Gitlin, 1980; Pan & Kosicki, 1993; Tuchman, 1978). Uses of news sources in general and official sources in particular have been widely examined as a dependent variable as a fimction of a number of factors (see the diagram of the integrative theoretical framework in Figure l, p. 150). Chang’s (1998) international interaction dimension in his world system model suggests foreign policy or relations can be a force in influencing news coverage and ensuing content in that foreign policy can decide on the range and types of uses of news sources. Research on international relations (i.e., Berry, 1990; Cohen, 1963) suggests U.S. foreign policy has an impact on the uses of official government news sources. Framing analysis has also reported similar effect in foreign policy. Although elite theory argues that the elite dominates as the primary source because of strong power and influence (i.e., Reese, Grant, & Danielian, 1994), the organizational approach, such as uncertainty theory, suggests sources are sought out because of organizational constraints. Reese, Grant and Danielian (1994) used network analysis to show that the handful of political news sources, most of them either current or former government officials who have appeared on the major network news or related talk shows, is a closely knitted political elite network. Like the six-degrees of separation of Kevin Bacon, the network of the news sources can be similarly traced from the fact that the same familiar prominent faces kept showing up on “CBS News,” “Nightline,” “MacNeil/Lehrer” or “This Week with David Brinkley.” By contrast, individuals, when used as sources, were often mistreated due to their lack of influence (Walters & Horrrig, 1993; Whitney et al., 1985). In a similar vein, in terms of influence and power, the core or elite nations in the world system are more likely to be covered in the news media of non-core countries than vice versa. These core nations, with great clout in world politics and economies, are the prime world market players and newsmakers and are much more likely to be covered even without going through the remaining filtering process (Chang, 1998). However, for non-core nations, such processes are a must that can help boost them to become world news, particularly in the U.S. news media. The influence of official news sources is an important factor in sourcing patterns, but organizational goals and media routines can also exercise constraints on its process and outcome (see Figure 1). Theoretically, organizational aims are fulfilled through the pursuit of high efficiency that helps organizations trim operational costs and optimize profits. In the news business, the beat system (Fishman, 1980) allows a journalist to seek out news in a more effective manner. Gandy (1982) and Dimmick (1974) suggested news production efficiency is usually acquired by resorting to suitable news sources— 10 government officials or non-govemmental people, depending on the circumstances—for appropriate information. Scholars (Cho & Lacy, 2000; Dimmick, 1974; F ico, 1984; Candy, 1982; Johnson, 1997; Lacy, Chang, & Lau, 1989) examining organizational influences have studied several key variables and found strong influences on news content: circulation size, use of wire services and resource allocations for news story. Lacy and associates (1989) and Johnson (1997) reported a positive correlation between circulation size and quantity of international news; newspapers per se as news organizations then can lead to different outcomes in news coverage that reflects the influence of the size of their market. Lacy and associates (1989) didn’t find any relationship between the proportion of foreign-bom Americans and foreign news coverage, but such a news selection process can be related to the percentage of different American ethnic groups as news demanders, which is a better measure of the market. The theoretical framework suggests this relation. Lacy and colleagues (1989) noted that the number of foreign staff on assignment could affect how news is covered; the nature and content of news are partly caused by choices of news sources. Past research also reported fi'equent uses of wire service would raise the dependency on governmental official sources because of greater time constraints facing wire service journalists (Fico, 1984). More importantly, in fulfilling organizational goals, resource allocations and uses of wire service and official news sources through routine channels have also been built into media routines (Shoemaker & Reese, 1996), partly because of the beat system (Fishman, 1980). Shoemaker and Reese defined media routines as “those patterned, routinized, repeated practices and forms that media workers use to do their jobs” (p. 105). 11 The routine channels include official proceedings, press releases, press conferences and non-spontaneous events (Sigal, 1973). Time and budget constrain the access of journalists, particularly wire service writers, to news sources. Sigal (1973) noted that media routines forced domestic journalists and foreign correspondents to rely heavily on U.S. and foreign government official sources, respectively. Moreover, it is logical that wire service reporters facing a greater time constraint will seek out and rely more heavily on official news sources (Fico, 1984), usually more accessible and able to provide information considered more authoritative. However, a greater number of reporters covering any story can probably improve their uses of official along with unofficial news sources. Hence, the theoretical framework suggests that the sources of news (staff vs. wires) and the number of journalists reporting on a story can also affect the decision on uses of official news sources. Press access is another key factor leading to reliance on official news sources. Theories on international news determinants suggest press access is a strong predictor of international news selection for time-stringent television reporting (Chang, et al., 1987; Chang & Lee, 2001). These theories then imply that wire services sharing a similar level of time constraint with television or news media highly reliant on wire services will be significantly influenced by insufficient accessibility of the press to information. Limited or lack of access to information and sources will reduce the scope of data and news sources available to journalists. As reporters struggle with the lowered accessibility and options of sources, they can become highly dependent on only a handful of news sources, particularly government officials (i.e., Kim, 2000). Thus, this theoretical framework suggests press access can have a direct impact on official news source uses. 12 Similar interactive mechanisms of organizational goals and media routines can be as strong and are also reflected in Chang’s (1998) model. Two approaches explore how non-core nations can become international news (Chang, 1998): a context-oriented perspective interacting with international interactions, or an event-oriented perspective interacting with news criteria. The context-oriented approach (Ahem, 1984; Chang, et al., 1987) examines the interactive relationship of the foreign nation to occurring events with such contextual elements as economic relations, political affiliation, geographic proximity and press access. The international interaction thus can be considered an integrative force of the context-oriented factors and, more importantly, is more similar to the foreign relation or policy factor. Sourcing pattern studies suggest foreign policy- driven news tends to rely on governmental officials in foreign lands as major sources (i.e., Chang, 1999b; Sigal, 1973). Also, it is reasonable that greater international interaction of a foreign country with the United States can indicate and lead to its improved newsworthiness in U.S. news media, whose foreign correspondents in residence can become more familiar with its political milieu and reliant on its political personnel. Accordingly, the theoretical framework suggests U.S. foreign policy would affect the decision on uses of official news sources. The event-oriented approach (Ahem, 1984; Chang, et al., 1987) suggests that regardless of those external contextual factors, some internal characteristics inherent in global events, such as the negative nature of the events, can determine whether such events get covered. It then suggests world news coverage will follow conventional news judgments, and journalists will seek out news sources accordingly. Prominence, a widely accepted news criterion as part of the media routine, is conceptually referred to as the 13 magnitude of impact of a news story. At point is the degree and prominence of casualty in any event; human losses are more prominent than property losses (Shoemaker & Reese, 1996). International communication research has also suggested that level of casualty is a very important factor in determining the selection of international news (Belle, 2000; Chang & Lee, 2001; Chang & Lee, 1992). An important event-oriented criterion and variable, casualty can influence decisions on source selections. Also at issue are the ensuing casualties and therefore causal and treatment responsibility for the delay of peace process of the long-lasting and sometimes intensified conflict between Israel and Palestine. Thus, this theoretical framework suggests that casualties as a media routine variable will have a role in official news sources uses, particularly when dealing with conflicts in foreign countries. It also indicates that casualty can lead to the assignments of causal and treatment responsibilities. Use of official news sources can also be treated as an independent variable (see Figure 1, p. 150). Framing research suggests one of the major tasks of the analysis is to identify the sources of news frames. (Friedland & Zhong, 1996; Pan & Kosicki, 1996). Framing research indicates that official news sources are able to set the types and nature of news flames in a way that constructs social reality (Tuchman, 1978) and assigns responsibilities (Iyengar, 1991, 1993; Iyenagar & Kinder, 1987) to either individuals or society as a whole. The theoretical framework thus posits these relationships between uses of official news sources and types of news frames (see Figure 1). Moreover, the organizational factors in the influence hierarchy model can directly affect the nature of news frames in part because wire services are likely to report conflicts in their international news coverage (Cho & Lacy, 2000; Hart, 1961; Lacy, et al., 1989). Episodic 14 frames place the focus on the outstanding dramatic elements of concrete incidents. Accordingly, this framework postulates the path relations between uses of official news sources and the organizational variable and nature of news frames (see Figure l, p. 163). On the other hand, the thematic frame underscores the background information of news and perhaps the newsmakers. Organizational factors in the framework result in that newspaper reporters, who are less constrained by time, are likely to provide more thematic frames because they are able to look for more information and present more thorough coverage that gives context. Hence, organizational factors lead wire service reporters to provide more episodic oriented stories than do their newspaper counterparts. The relations are outlined in the framework (see Figure 1). Uses of official news sources are the central element of the theoretical framework that connects the hierarchy of influence model, international communication world system model and framing analysis. A number of factors affect the choices of official news sources through which news fi'ames are determined. This framework suggests foreign policy (international interactions), press access (extra-media) and uses of wire services, number of reporters on assignment (organizational variables) and casualty (news criteria of media routine) can be strong predictors of how official news sources are used and treated. Likewise, these factors and official news sources can play important roles in deciding the nature of news frames, which makes responsibility assignments in the news reports. The theoretical framework is presented in a diagram to show the possible path of influence on official news sources and news frames. 15 CHAPTER III LITERATURE REVIEWS The present research involves the perspectives discussed in the building of the broader theoretical framework that integrates elements of the influence hierarchy model, international communication world system model and framing research. In this chapter, related literature on international affairs coverage and framing analysis is reviewed. World News Coverage International affairs coverage is one of the most researched areas in mass communication. Much research has been devoted to examine how the U.S. news media cover world events, particularly after the New International Information Order, or NIIO, was brought out by the Third World nations and led to heated debates in the early 19703. In the span of some 30 years, numerous conceptual terms or theoretical works have surfaced, from a number of different perspectives, to try to explain the nature of world news coverage. Foreign Policy. Chang’s (1998) international interactions factor of his world system model explains the role of foreign policy in world news coverage. International interactions, by definition, include economic and foreign relations and policy. Foreign policy has been documented to be influential in determining how a nation is covered. Many international communication researchers (i.e., Chang, 1988, 1993; Chang, 1999a, 1999b; Cohen, 1963; Kim, 2000; Kriesberg, 1946; Lee & Yang, 1995; 16 Mowlana, 1984; Rachlin, 1988; Srebemy—Mohammadi, 1990; Wang & Lowry, 1999; Welch, 1971) have shown the U.S. media were often covering and interpreting world affairs in accordance with U.S. foreign policy or simply used it as the benchmark. However, much of the research failed to provide an adequate operational definition of the concept of foreign policy. Most of the studies were descriptive and qualitative. More recently, a “rallying around the flag” (McCartney, 1994) phenomenon was documented (Gutierrez-Villalobos, Hertog, & Rush, 1994; Livington & Eachus, 1995; Zaller & Chiu, 1996). This hypothesis suggests that press support of the govemment’s foreign policy becomes greater and more prominent during international crises or national emergencies. Lee and Yang (1995), for instance, compared the Associated Press’s and Japan’s Kyodo New Agency’s reporting about the Tiananmen democratic movement and concluded that both wire services were consistent with their country’s foreign policy. While the student movement was described by the Associated Press reports as a battle for democracy and human right against a communism regime, Japan’s economic interests prevented the agency from sounding adverse to the Chinese government. Similar outcomes were disclosed in Kim’s (2000) comparisons of mainstream U.S. media coverage of the Tiananmen movement and South Korea’s Kwangu democratic movement in 1979. The New York Times and the Washington Post were found to report the two similar events consistent with U.S. foreign policy of two administrations. While the Bush government supported the pro-democratic movement in China, the Carter regime condemned the violence in South Korea. News Sources (officials). Having not been the main focus of international communication research, the role of news sources had not been vigorously examined or 17 reported in these types of studies until more recently. For instance, Kim (2000) unveiled the varied uses of news sources as a result of contrasting foreign policy exercised by the first Bush and Carter administrations. While the U.S. media used student demonstrators as their prime sources in covering the Tienanmen protest, they relied mostly on Korean government officials for information in covering the democratic movement there. Kim (2000) explained it was probably due to the fact that the world media corps converged on Beijing as a result of Gorbachev’s scheduled China visit and witnessed the demonstration first hand. On the other hand, the smaller scale of democratic protest in Korea was not presented with similar luck in timing, and more important, news media’s access to protesters was tightly controlled by Korea’s military government, whose officials became predominant sources. But, it is likely the differences resulted from organizational factors other than government controls. Without focusing on international communications, Sigal (1973) in his scrutiny of reporters-officials relationship argued that foreign correspondents of The New York Times and the Washington Post, to a greater degree than their Washington bureau counterparts, rely even more heavily on government officials in the reported country as their news sources, mostly through routine channels. He reported that foreign government official sources accounted for more than 60 percent of the total sources in foreign news coverage while U.S. official sources accounted for 23 percent. Berry (1990) observed similar patterns of uses of foreign governmental official sources in The New York Times’ coverage of U.S. foreign policies. Three important groups of variables more or less fit under the umbrella of Chang’s world system international communication model. They are organizational 18 variables, event-oriented (media routines) variables and context-oriented variables. The idea of an organizational approach can be traced back as early as to Hart’s (1961) and Ostgaard’s (1965) work. The notion of context-oriented and event-oriented variables originated with Ahem (1984). Organizational. Organizational variables can have strong constraints and effects on the production of news content. The influences apply not only to selections of domestic issues but certainly to choices of international news. Despite their great influence, little research has adopted this approach to examine the nature of world news transmissions. Ostgaard (1965) is probably the first author to suggest examination of the impact of ownership on world news coverage. He argued that internal or external information controls by way of media organizational goals, self-censorship or governmental regulations could effectively undermine message flow across borders. For instance, the easy and frequent uses of student protestors in Tienanmen (Kim, 2000) were the results not only of the government’s failure to control the news flow but also of the immediate access foreign correspondents had to the protesting students. The timely presence of these world journalists also made the coverage very efficient in terms of cost cutting. The organizational variables were first termed by Lacy, Chang and Lau (1989). In principle, the organizational approach indicates that factors internal to news institutions would influence the nature of media coverage accessible to the newsroom and the choices of news for publication from the available reporting. In a national study of 114 American newspapers, they concluded that circulation size and level of reliance on wire services were the two strongest predictors of the percentage of space assigned to world news. Not 19 only was a positive correlation found between the two independent variables and the dependent variable, but also the two explanatory variables accounted for more than 25 percent of the variance. They also found dependence on wire service was positively associated with the amount of conflict or disaster news; it explained 15 percent of the variance in selection of such news. Their findings confirmed evidence reported by Hart in 1961. Hart argued that the vast majority of world news provided by wire services was about conflict, crime and disaster and that the Associated Press and United Press International accounted for nearly two-thirds of international news in the newspapers examined. Comparable results were disclosed in Cho & Lacy’s (2000) national studies of Japanese local press coverage of world affairs. They nevertheless found a negative correlation between the number of wire services and number of conflict and disaster news. They offered two possible reasons for the difference: the allocation of newsroom resources and the existence of only two news services in Japan. Editors either assigned less resource to a second agency or they selected a story with fewer conflict or disaster elements from available stories provided by the two agencies, which further limits the purview of choice. The impact of circulation size on the amount of international news reporting was also supported in Johnson’s (1997) content analysis study of 515 articles from 34 U.S. newspapers’ coverage of Mexico. Circulation size was not only the second most important predictor but also the predominant factor in deciding the length of the stories. Similar results were, however, not reported in Cho and Lacy’s (2000) study. It is possibly due to the nature of the sample. While using a nationwide sample, only local newspapers were analyzed, and as a result the small variance in circulation size among those local 20 newspapers failed to generate its predicted effect as occurred in the study of Lacy and colleagues (1989) and Johnson (1997), which included both national and local newspapers. Cho and Lacy reported a mean circulation of some 215,000 for those local newspapers; all four Japanese national newspapers, which were not included for analyses, have circulation over 1 million and rank among the highest worldwide. E vent-Oriented (media routines). Ostgaard (1965) and Rosengren (1974) both pledged to use extra-media materials to better gauge the impact of news determinants on international news reporting. Since then, a variety of strong efforts have been provided. Notable was the work by Chang, Shoemaker and Brendliger (1987). They further theorized the determinants of world news coverage by examining patterns in mainstream U.S. newspapers and on ABC, CBS and NBC. Integrating previous theoretical notions and adopting extra-media information, Chang and associates conceptualized seven independent variables as either context-oriented or event-oriented factors to examine if certain events were covered or not in those American news outlets. The event-oriented approach (Ahem, 1984; Chang, et al., 1987) suggests that regardless of some external contextual factors, certain internal characteristics inherent in global events, such as the negative nature of events, can determine whether such events get covered. They found event-oriented factors, such as normal deviance of issues, relevance to the United States, and potential for social change, to be strong determinants, while contextual variables were mostly insignificant. Chang and colleagues’ influential study, however, suffered some methodological drawbacks. Their uses of dichotomous variables for some independent variables were not statistically ideal, and the outcomes might have been different if they had employed all continuous independent variables. 21 control variable, was found the greatest predictor variable. Belle, however, (mis)treated it as a context-oriented variable, instead of an event-oriented variable used in most studies. Context-Oriented. The context—oriented approach (Ahem, 1984; Chang, et al., 1987) examines the interactive relationship of the foreign nation to occurring events with such contextual elements as economic relations, political affiliation, geographic proximity and press access. This approach has created varying results, mostly because of different operational definitions of the concepts and methods. The key contributors of this orientation include Galtung and Ruge (1965), Rosengren (1974), Hester (1971, 1973) and Wu (2000). Galtung and Ruge investigated 12 factors of international issues reporting. They concluded that events were more likely to receive coverage if they were frequent, innovative, relevant, consonant, scarce, or referred to elite nations. Regardless of their pioneer contribution, Galtung and Ruge were criticized for having difficult to measure factors because most of them are psychological variables (Hur, 1984). Hester (1971, 1973), who adopted an international relations approach, defined the hierarchy of the nation as trade volume, economic association as economic development, and cultural association as shared language. Hester’s national hierarchy notion defined in terms of trade volume was nonetheless not equivalent to Chang’s (1998) world system concept expressed in term of three echelons of world power—core, semi-periphery and periphery. The operational definitions of trade volume has been used mostly in the sense of a reported nation’s total amount of trade exchange with all trade partners (Chang et al., 1987; Hester 1971, 1973; Wu, 2000), instead of only with the United States, and levels of trade relations, instead of real dollar term, were used. As a result, the likely predictive power was not present. 23 Chang and Lee (2001) adjusted this problem by using the trade relation with the United States but still could not generate any predictive power from it. Two explanations are likely. The first again lies in the conventional use of a trade relation scale, instead of the real dollar number of the trade volume. The smaller variance may have mitigated the potential influence. Second, the extra-media documents—Keesing '5 Contemporary Archives: Record of World Events—used by Chang and Lee (2001) and their forerunners’ (Chang, et al., 1987) may not have provided the most ideal reference basis for comparative analysis. While comprehensive, this record book is unlikely to garner all possible information in an exhaustive manner because the constraints faced by American media are also exerted on foreign newspapers’ data collection process in their homeland. The selected events in this document were based on those reported in each nation, and not all events in the nation become news. For instance, negatively oriented media routines may have prevented some major “positive” stories from being reported by local press and thus to the world press, and accordingly the great amount of negative stories may have crowded out trade volume from becoming a strong indicator. In a very large-scale study, Wu (2000) categorized and examined nine systemic determinants of world news coverage from 38 countries. He found the United States was the country dominating the world media spotlight; it was the most covered nation in 23 of the 37 (excluding itself). Also, based on the patterns revealed in the 10 most covered nations, the economic and political influences were found to be the two primary factors in getting covered in another nation’s news media. More important, overall, trade volume and presence of world news agencies, newspapers and wire services alike, were the two greatest predictors. The role of trade relations thus seems clear. It appeared to be a strong 24 determinant when only context-oriented variables were examined, but when event- oriented variables were controlled for, its significance disappeared. While Wu’s study confirmed previous research that showed U.S. predominance in world news outflow, he failed to provide a genuine systemic approach in analyzing international news flow. He suffered from an incomplete analysis that probably had led to incorrect conclusions because the nine systemic determinants he examined were indeed those dubbed context-oriented variables, and none of the so-called event-oriented variables were included. In fact, the event-oriented factors, which were not tested, were only qualitatively explained as confounding variables. Such important events as NATO’s involvement in the Bosnia crisis and the UN. Women’s Conference in Beijing undoubtedly could have affected or changed the quantity or nature of world news coverage during the period studied. Wu’s limited study demonstrated the complexity of research on international news coverage. In summary, the nature of world affairs (event- or context-oriented), press access (extra-media), news media’s organizational factors (circulation size and dependence on wire services) and foreign policy (international interactions) play important roles in deciding whether matters become news in a newspaper. Also, it is likely that government officials of the reported countries are the dominant news sources because of the constraints of organizational goals and media routines on the journalists. Accordingly, these foreign government officials fi'equently become the news framers in world news coverage. 25 Framing Analysis Before starting to review the literature on framing analysis, it seems appropriate to discuss briefly its conceptual development because it relates to the theoretical and methodological approaches to be adOpted in the current research. Many flaming researchers (Entman, 1991; Gamson, 1992; Gitlin, 1980; Gofflnan, 1974; Huang, 1996; Neuman et al., 1992; Scheufele, 1999) have distinguished two types of news flames— media flames and individual flames (also often referred to as audience frames). While media flame has been defined as a central idea adopted by journalists or the news organizations to present, organize and interpret information for a news story (i.e., Entrnan, 1993; Gamson, 1989; Gitlin, 1980; Pan & Kosicki, 1993), audience flame is considered “conceptual tools” that audiences “rely on to convey, interpret and evaluate information” (Neuman et al., 1992, p. 60). This distinction makes the theory more complete in examining the formation process of media content and public opinions. While media flames are investigated by content analysis, examination of individual flames requires polling public opinion to examine and measure the potential flaming effects. That is, the media flames are treated here as a dependent variable, and the elements (or sources of influence) of the proposed theoretical flarnework are the independent variables. In a more thorough flaming process and analysis, the media frames will become an independent variable to examine its impact on the individual flames as the dependent variable. However, due to budget limitations, the current research examines only the content-based media flames, without polling the public to examine the audience flames. 26 Another noteworthy theoretical underpinning is the broad definition of news flames. Over the years, scholars have used and defined the term in a loose manner reaching no universally accepted definition. News frame has been defined from as restricted as a central idea (i.e., Gamson, 1989) to as broad as episodic or thematic flame (Iyengar, 1991, 1993), and flom as vague as one that can be conveyed through metaphors or catchphrase (Gamson, 1989) to as clear-cut as one that can be identified through simple types or categories of news flames or themes (Entman, 1993; Sylvester, et al., 1999). For the possible task of reliability testing and future replications, this research adopts the definition by Entman (1993, p, 52), who treated news flames as "some aspects of a perceived reality [italic added] " made more salient in “a communicating text... to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation.” The definition not only indicates that a news story can convey and present more than one news frame or theme but also is capable of identifying and assigning the causal and treatment responsibility as suggested by Iyengar. Several studies have examined the role of news sources in the flaming process, but they tended to adopt an elite perspective. Fewer researchers have investigated the roles of organizational factors and media routines in the production of media content. This literature section also reviews the studies in a hierarchical way in terms of the importance of influences as was in the international issues coverage. Foreign Policy. Norris ( 1995) suggested foreign policy can be operationally defined in terms of fiiends and enemies, and such evidence has been found in several investigations (i.e., Entman; 1991; Peh & Melkote, 1991). Entman (1991) and Peh and Melkote (1991) found an impact of foreign policy on media coverage of two comparable 27 events—the downing of a Korean Airlines passenger jet by the Soviets in 1983 and of an Iran Air passenger jet by the U.S. military in 1988. Both studies disclosed that the former Soviet Union, an enemy, was depicted as evil and malicious, while the American action was not. Similar influence also was found in Wall’s (1997a, 1997b) examinations and comparisons of three major U.S. news magazines coverage of the parallel crises in Bosnia and Rwanda. The coverage of Bosnia (Wall, 1997b) may have supported Norris’s (1995) argument that news flames can shift because of the change in world politics and ensuing foreign policy. The perceived threat to U.S. society has drifted from communism to international terrorism, and it is highly likely that the news flames in the international news in U.S. newspapers will be dominated in the future by international terrorism. Foreign policy also seemed to influence the media adoption of nature of news flames. While the crisis in Bosnia was depicted with a thematic flame that offered contextual backgrounds of the issue, Rwanda’s was painted episodically and focused primarily on the details of daily events with little background information about the crisis (W all, 1997a, 1997b). The tribal violence in this Afiican nation was then described as irrational and incomprehensible, but conflicts in Bosnia were considered understandable and justifiable although the ensuing violence and casualty could be greater. Accordingly, while Rwanda was responsible for its own misfortune, the mishap in Bosnia was attributed to world society. The difference, Wall indicated, was due mostly to foreign relations, political affiliation as well as physical proximity of Bosnia to the Western Europe and by extension the United States. Her argument seemed backed up by strong evidence. First, while news magazines cited more Western officials in the Bosnia crisis, 28 they used Rwandan officials more flequently in that conflict. Moreover, while Bosnian coverage had a larger number of stories, Rwandan coverage had a greater number of words per story, which should have enabled the journalist to provide more contextual information to the crisis. But the fact that it was not the case seemed to indicate that foreign policy was at least an important factor in deciding the news frames and content. The responsibility-assignment impact of the episodic and thematic flames was, however, not shown in Semetko and Valkenburg’s (2000) study of news coverage of the EU. meeting in Amsterdam. Iyengar (1991) argued that episodic flames broadly used by television newscast will lead the audience to assign responsibility to individuals, instead of to the government, but Semetko and Valkenburg’s results suggested that television news can be episodic and flame the government, not individuals, as responsible. They indicated the variance could have arisen flom cultural or political differences. However, it is also plausible that uses of news sources, which were not examined, could have been another significant determinant. News Sources (officials). Internationally, based on Wall’s (1997a, 1997b) findings, it seemed foreign policy and organizational factors could influence the selection of official news sources. The closer foreign relations and physical distance of Bosnia to Western Europe led American journalists to rely more heavily on Western governmental officials than on Serbian officials as primary sources. By contrast, lack of such advantages plus lack of press access forced reporters to use Rwandan officials as their major sources, at nearly double the flequency of Western officials. Rwanda is listed as one of the countries with the tightest government control in the world (Sussman & Guida, 2000). 29 Domestically, Entman and Rojecki (1993) reported that the framing judgments adopted by journalists were greatly influenced by elites, who were capable of mitigating political or social pressures flom nuclear freeze movement supporters flamed as irrational or extreme. Examining the framing process of the movement, Entman and associate concluded, “The media in general belittled the public and its involvement, whereas critiques of elite opinion were rare, muted, and inconsistent ...” (p. 157). All news media, as predicted by Entman and Rojecki, used government officials or the social elite of the anti-fleeze movement considerably more often than movement organizers or supporters. Borrowing Chan and Lee’s protest paradigm (1984), McLeod and Hertog (1998) and McLeod and Detenber (1999) suggested strong source dependences in general and dependency on official sources in particular in news coverage. Similarly, Eilders and Luter (2000) re-confirmed Bennett’s (1990) indexing thesis that news media tend to rely on government sources. Huang (1996) noted that the flame or attribute salience is an outcome of the crowding-out effect (Hilgartner & Bosk, 1988) of news flames. The likely interlocking relationship between political news sources, news organizations and the journalist was described by Woo (1996), who pointed out a cozy tie between journalists and political sources. Some prominent political journalists enter politics after service in news organizations. A similar phenomenon is also seen in the United States where a handful of reputable journalists turn themselves into press secretaries after successful journalistic careers. News Sources (non-officials). Liebler and Bendix (1996) revealed comparable effects in their examination of networks’ reporting about forestry clear-cutting issues. They found timber industry supporters cited more frequently than clear-cutting 30 Opponents, and as a result, public opinions were more in favor of the pro-cut standpoint. Parallel outcomes and sourcing patterns were reported in Chang’s (1999a) study of news coverage of Oregon’s clear-cutting measure and voting. Chang found that the news flame used mostly in the newspapers was “It’s too extreme,” a political advertising campaign by the forest industry. This finding confirmed the claim by Neuman and colleagues (1992) that news flames established by news sources are “conceptual tools which media and individuals rely on to convey, interpret and evaluate information” (p. 60) or issues. Both Liebler and Bendix’s and Chang’s studies combined content analysis and public opinion poll data. Organizational. Studies of influences of organizational variables on news content have examined circulation size, uses of wire services, reliance on wire services (Lacy, Chang, & Lau, 1989) and political affiliation of the news organization (Gans, 1973). Content analyzing how 29 newspapers nationwide covered and flamed risks facing whites and blacks, Gandy and associates (1997) reported that circulation size played a role in selecting certain news flames. Similar findings were unveiled in an earlier study by Goshore and Gandy (1995). Gandy and colleagues (1997) also found that market size of the newspapers was positively correlated with the frequency of negative news flames used, but market pluralism in terms of the percentage of black population in the market helped reduced the tendency. Although Gandy and associates treated the percentage of ethnic population as part of market structure, the concept of market pluralism is an extra-media variable in Shoemaker and Reese’s model, where the audiences as news sources have certain influences on news content. 31 Candy and associates (1997) uncovered a significant role of the proportion of black population in using negative flames, but an early study (Goshore & Candy, 1995) found an opposite result: the proportion of black population increased employment of negative news flames. This sharp distinction clearly lay in methodological differences. The 1995 study adopted news stories about one event on one day flom 49 newspapers, but the 1997 study looked at news coverage of a wide variety of issues published in a time span of four years between 1989 and 1993. Moreover, the 1995 study failed to define the key concepts while the 1997 study offered detailed information on them. The implication of the finding for the present research is that the percentage of population of Jewish Americans and Arab Americans could influence how news flames are used within the market of periphery of the four newspapers by means of choosing news sources. To gain and retain the readership of any large ethnic group, the newspaper need to provide relevant information its readers need and want to read. If these readers are dissatisfied with the reports, they can discontinue their subscription to the newspapers (Lacy, Chang, & Lau, 1989). When these readers are used as news sources, they can then affect the way news flames will be employed. Media Routines. Media routines also are influential in deciding news content. Lawrence (2000) suggested that the nature of the issue could result in selecting different news frames. Gandy and colleagues (1997) agreed when they concluded that news subjects played a significant role in using the discrimination flame to depict the different plight between blacks and whites. Subjects or types of news usually reflect the nature of the news, such as prominence and conflict, and as discrimination connotes conflict, it can result in particular news frames. This reasoning then supports the observation that 32 strategic flames (Cappella & Jamieson, 1997), which are interchangeable with game flames (Lawrence, 2000), were more likely to be used during political campaigns and election and political conflicts because of the dramatic effects of the news. The concept of strategic flames was initially derived flom Cappella and Jamison (1997). The authors claimed that conventional news criteria make journalists pursue and present news stories flom an event-oriented and episodic approach focusing on the conflict or competition nature of the news. The strategy-based coverage contains a number of dimensions that include: (1) winning and losing as the central concern; (2) the language of wars, games and competitions; and (3) heavy weighing of polls and the candidates’ standing in them. Since the appearance of this concept, it has been exclusively applied to news analysis of political campaigns or election because of its characteristics and assumptions. As a result, the strategic flame notion is more limited than Iyengar’s episodic frame conception in that the latter can be applied more broadly to areas other than political campaigns or elections. The impact of news criteria also was found in Ashley and Olson’s (1998) investigation of The New York Times, Time and Newsweek coverage of the women’s movement between 1966 and 1986. They argued that media usage of undercounting, among other factors, in newswriting led to the effect of de-legitimizing feminism advocates while promoting anti-feminism groups. One noteworthy thing is that the authors purposefirlly rejected using a random sample because they claimed that would not be representative of all news media on this issue. Instead, they garnered their data by identifying stories flom The New York Times Index and The Reader’s Guide to Periodical Literature. However, they didn’t clarify whether the 499 articles were exhaustive. Also, 33 while Ashley and Olson (1998) did examine the potential impact of the gender of reporters on coverage, they ignored the potential influence of sources of the news and who wrote the reports (staff versus wire service reporters) in the Times. Neither did they investigate the likely difference between the newspaper and the magazines and between the two magazines. As a result, the findings could have varied had these potential factors been included for analysis. In summary, while news sources are capable of flaming news, organizational factors (circulation size, market pluralism and proportion of ethnic population as audience) and media routines can also affect choices of news flames. Moreover, it is reasonable, although untested, that uses or reliance on wire services can also influence the nature of news flames. Foreign Policy on Israel, Palestine and the Crisis The significance of the Israeli and Palestine conflict in particular and Middle East issues in general to U.S. foreign policy was best reflected by Colin Powell’s choice of his first oversea travel to the region as U.S. Secretary of State. Over the years, public opinion polls have persistently shown that the majority of the public has considered this issue as either vital or very important to U.S. foreign policy and national interests and has said finding a solution to the crisis should be a major priority (Gallup, 2001). The Chicago Council on Foreign Relations survey (1998) also reported international terrorism was considered the No. 1 threat to U.S. vital interests; combating international terrorism ranks third as an important foreign policy goal; and that respondents believe the U.S. 34 government should deploy troops if Arab forces invade Israel, a scenario whose importance ranked behind to the scenario where Iraq invades Iran. While the public thinks this issue is crucial to the United States, it holds varying attitudes and opinions, in terms of favorableness and sympathy, toward the two political entities. According to Gallup surveys (2001), the American citizenry has consistently favored Israel (63 percent) more than Palestine (22 percent). While the former is treated as a fliend, the latter is considered more of an enemy. The 2001 terrorist attacks in the United States also worsened the already unfavorable U.S. opinion faced by Palestinians. Moreover, the sympathies of the U.S. public with Israelis jumped flom 41 percent to 55 percent after the Sept. 11 attacks; by contrast, the sympathies of the American people toward Palestinians dropped flom 13 percent to only 7 percent (Gallup, 2001). Moreover, the percentage of U.S. residents who think the government should side with Israel in the crisis rose flom 16 percent to 27 percent after the attacks, while the percentage suggesting the government side with Palestine remains marginally at 1 percent (Gallup, 2001). More importantly, more people (flom 52 percent to 63 percent) thought the U.S. should either increase or keep the same amount of economic or military aid to Israel after the attacks (Gallup, 2001). Judging flom these poll results, the U.S. government and public support Israel more than Palestine. As a result, the U.S. government holds a more amicable foreign policy on Israel than its counterpart. Press Coverage of Israel and/or Palestine or Arab This section reviews studies that examined news coverage of Israel and/or Palestine or Arab nations. Several studies examined the depiction of Arabs in the U.S. 35 press (i.e., Belkaoui, 1979; Daughterty & Warden, 1979; Mousa, 1987; Suleiman, 1974; Terry & Mendanhall, 1974; Trice, 1979). In general, Arabs had been depicted negatively in the American press since the establishment of Israel as an independent country in 1948, and there existed a consistent anti-Arab and pro-Israeli bias. Mousa (1987) content analyzed the Arabs’ image in The New York Times between 1916 and 1948, a time flame the author considered the cornerstone in the development of the Arabs’ images. Mousa examined the sourcing and attribution patterns in the newspaper and found that pre-Israel Jewish sources, along with France and England, were most often cited. These groups of sources were also most likely to provide unfavorable attributions to the negative depiction of Arabs. She noted that it was the period when the pre-Israel Zionists were embarking on establishing a country in Palestine and when the Arabs were fighting the French who was occupying their territories. Mousa reported that the coverage tended to be conflict- and event-oriented, or episodically flamed. The episodic flame might have led Americans and world society to perceive the Arabs, who were considered responsible for their own fate, in an unfavorable fashion. Mousa also suggested that unbalanced coverage by The New York Times probably contributed to misconceptions about the Arabs before 1948. On the other hand, the more favorable description of Jews in the American press was somehow reflected in public opinions about American Jews. This group was considered by white people to be more socially fliendly; they are more willing to have social contact with Jews than with all other minority groups, including blacks and Asian Americans (Wilson, 1996) 36 Directly related to the present research was Sylvester, Wu and Hamilton’s (1999) (media) framing analysis of the Philadelphia Inquirer’s coverage of Israeli/Palestinian conflicts. They content analyzed 423 regular news stories, news briefs, photos, cartoons and editorials, flom January through October 1998. Sylvester and associates identified five flames for their analyses: Palestinian terrorist/terrorism, the Palestinian independence movement, Israel’s security/safety, Israeli (fundamentalist) religious/r1 ght wing tenacity, and peace process/negotiations. Based on their selection of flame categories, they seemed to have biases similar to those found in the U.S. press. Their failing to list accused Israeli terrorism actions as a category implied they thought that only Palestine would conduct terrorism. This research corrects this potential problem by listing it as a flame category, along with several new categories to make the analysis more complete and accurate. Sylvester and colleagues (1999) found that the Inquirer’s news reports basically provided a balanced coverage and depiction of both sides. The peace process was found to be the predominant flame in the news, accounting for 50 percent of the total number of flames. However, Israeli sources were used more flequently than were Palestinian sources in the news stories. It is possibly because more journalists reside in Israel, which is safer and have more press access. Moreover, editorials and opinion columns offered considerable insight and background information about the conflict between the two sides, but whether similar efforts were given in the regular news stories was left unanswered. That is to say, the researchers did not specifically examine the uses of episodic vs. thematic flames and the assignment of responsibility in the news articles. The current research makes an adjustment to this drawback as well. Sylvester and 37 associates (1999), however, reported that Associated Press articles, which accounted for 36 percent of the regular stories, were more negative about and critical of Israel and its actions than staff-written stories. They nevertheless didn’t explicate why, although the group considered the negative reportage by wire services had supplemented the more positive coverage of Israel by Inquirer staff and had made the news coverage as a whole more balanced. One possible explanation may lie mostly in time constraints. While staff writers of the Inquirer were less restricted by time pressure, Associated Press reporters were consistently pressed by stringent and constant deadlines. To provide constantly updated reports to its subscribers, wire services reporters exercise conventional news judgment criteria to the greatest degree, and accordingly, these reports tend to be more negative than positive. 38 CHAPTER IV HYPOTHESES According to the framework and literature, several factors can influence the selection of news sources as a dependent variable. In general, government officials at home or abroad usually predominate as news sources who can flame the issues. First, the foreign policy of the United States plays an important role in press coverage of world issues or nations. Academicstudies (i.e., Daughterty & Warden, 1979; Mousa, 1987; Suleiman, 1974; Terry & Mendanhall, 1974) and opinion polls indicate that the United States and its citizenry are more favorable toward Israel than Palestine. Press access and organizational factors such as use of wire services and resource allocations are influential in determining the choices of official news sources and news flames. Also, news organizations per se can have an impact because of varying circulation size and readership. Casualty as media routine also is a plausible determinant. More important, uses of official news sources can become an independent variable and affect the decision on the types and natures of news flames and assign responsibilities directly. Stories written by wire services and newspaper writers can vary in terms of nature of news flames. Accordingly, a series of hypotheses (see Appendix B) are posed based on the proposed integrative theoretical flamework. H1: The diflerence between favorable treatment of Israeli and Palestinian official news sources was greater during the period when U.S. foreign policy was more favorable toward Israel. 39 Rationale (H1): Past research showed that foreign policy could considerably influence how the U.S. news media covered international news and treated the reported nations. This policy toward a nation then can affect the magnitude of international interactions and interdependence of this country with its counterparts. Foreign policy and relations suggests the level of importance or relevance of the nation to the American politics. The closer and greater diplomatic interactions of the United States with a nation, the more likely the U.S. press will pay attention to that nation and would probably have reporters in residence to cover news there. And as journalists become more familiar with the political environment and personnel in that country, they would be more likely to use those news sources more flequently and efficiently. Over the decades, the U.S. government and its foreign policy have been more supportive of Israel than of Palestine. The level of governmental support of Israel varied flom administration to administration, and it is likely that the level of favorableness of news treatment of Israel would also vary flom administration to administration. H2: As press access in Israel is greater than that in Palestine, Israeli ofi‘icial news sources would receive more favorable news treatment than Palestinian oflicial sources. Rationale (H2): Press access has been a good determinant of world news coverage (Chang, et al., 1987; Chang & Lee, 2001). The greater press access a country offers, the greater the access of news sources foreign correspondents are likely to have in coverage (Kim, 2000) and therefore they are likely to be used more flequently and prominently. H3: Wire service reporters would be less likely than newspaper stafl writers to give favorable news treatment to Israeli oflicial sources than Palestinian official sources. 40 Rationale (H3): Wire service reporters facing more stringent and constant deadline pressure will rely more heavily on Palestinian official news sources as predominant sources than their newspaper counterparts, and as that reliance becomes more routine, wire service reporters would become more familiar with Palestinian official sources and give them more favorable news treatment. H4: If the proportion of American Jews in the city population is greater than that of American Arabs in the newspaper ’s readership radius, then Israeli official news sources would receive more favorable news treatment than Palestinian oflicial news sources. Rationale (H4): Past research showed that media or message pluralism, in terms of percentage of the ethnic group, could affect the choice of sources (Gandy et al., 1997). H5: The more journalists covering a story, the more likely Israeli oflicial sources would receive more favorable news treatment than Palestinian oflicial sources. Rationale (H5): More reporters covering the story can help lessen the time constraint because these reporters can cooperate in a way to seek out more diverse news sources for more comprehensive information. Reliance on Palestinian official news sources will be lessened, and as a result Israeli official news sources can be used more often and given more favorable news treatment. H6: The greater the level of casualties among Israelis, the more likely Israeli oflicial sources would receive more favorable news treatment than Palestinian official sources, and vice versa. Rationale (H6): When casualties occur flom a battle or conflontation, reporters will likely seek necessary information flom more directly related authorities about the 41 occurrence, causes, aftermath and potential impact of the violence. These related authorities accordingly are likely to be used more flequently and prominently. H7 to H13 examine and test the dependent variables of the types and natures of media flames, as well as the assignments of responsibility suggested in the news articles. H7: The more favorable news treatment Israeli oflicial sources has received, the more likely the peace process would be the dominant news frame. H8: Wire service reporters would be less likely than newspaper staff writers to use the peace process as the dominant news frame. Rationale (H7 & 8): While Palestinians are sometimes considered as terrorists, Israelis are thought to have promoted peace with its counterparts and the Arab world as a whole (Gallup, 2001). Theories have suggested uses of wire services have led to more conflict coverage of international affairs (Cho & Lacy, 2000; Hart, 1961; Lacy, et al., 1989). While the peace-process news flame can also be conflict and negative (i.e. delay of peace talks), violence or terrorism and the ensuing loss of life could be more flequently used as the dominant news flame in wire service stories. H9: The more favorable news treatment Israeli oflicial sources has received, the more likely the story would be more thematic than episodic. H10: News stories written by wire service reporters would be less thematic than episodic compared to stories written by newspaper stafl writers. Rationale (H9 & 10): Being treated more favorably as news sources, Israeli official sources could be given more opportunities to discuss the issue and therefore could probably provide more background information, which could be adopted and reported by American journalists. International communication theories suggested wire 42 services-prepared stories tended to be event-oriented (Cho & Lacy, 2000; Hart, 1961; Lacy, et al., 1989) and therefore episodically flamed. H11: The more favorable news treatment Israeli oflicial sources have received, the more likely both causal and treatment responsibilities would be assigned to Palestine. H12: The more thematic the news frames are, the more likely both causal and treatment responsibility would be assigned to Palestine. H13: The greater the level of Israeli casualties, the more likely both causal and treatment responsibilities would be assigned to Palestine. Rationale (H11, 12 & 13): As Palestine is criticized by Israeli official sources and therefore assigned as the culprit of the conflict, it would also be more likely to be considered responsible for its own wrongdoing and action and to make necessary changes in the way it handles the crisis. Iyengar (1991, 1993) suggested different natures of news flames would lead to varying responsibility assignments. Moreover, when Israeli casualties are higher than its counterparts’, news coverage would be more likely to focus on the violent actions and ensuing mishaps caused by Palestinians, who are also considered responsible for averting conflict, partly by controlling terrorist acts by Islamic fundamentalist groups. 43 CHAPTER V METHODS Research Design The present research content analyzes The New York Times, the Washington Post, The Los Angeles Times and the Houston Chronicle during the time flame of Jan. 21 , 2000, through Jan. 21, 2002. This time flame covers two administrations (Clinton and the second Bush) and is evenly divided into two one-year sub-periods. Thus, this design allows a comparison of foreign policy between two administrations. The rationales of using the four newspapers encompass: (1) All four newspapers have a high circulation (Merrill, 1968; Stempel, 1961, 1965; Weber, 1990); (2) All but the Chronicle put a strong emphasis on international news coverage; (3) All of the cities where the newspapers are headquartered are major metropolitans where both Jewish Americans and Arab Americans reside. For instance, New York City, headquarters of The New York Times, has the largest city of Jewish Americans and the second largest of Arab Americans (2000); Houston, housing the Houston Chronicle, is one of the few cities with a relatively greater ratio of Arab Americans than other major U.S. cities. This would allow an examination of the potential impact of an ethnic population on news coverage of the issue. The population of the analysis includes all regular news stories appearing on the flont pages, international pages and national pages during the period. A total of 549 stories were found and used, with 228 stories in the first period (the Clinton administration), and 321 in the second period (the Bush administration). In breakdown, 175 (83 in the first period, 92 in the second period) stories were from The New York Times; 124 (57, 67) from the Washington Post; 182 (69, 113) from The Los Angeles Times; and 68 (19, 49) flom the Chronicle. Stories flom all four newspapers were selected based on the Nexis electronic database, using the key words of Israel AND Palestine contained in the articles. That indicates that only news items with both Israel AND Palestine in the headline or context would be selected; only those relevant to the conflict to be the focus of the present study, which analyzes how the two sides were covered and reported together. This research command and strategy also enables the researcher to limit the study to a workable extent for content analysis under the constraints of time and budget. A coding protocol (see Appendix C) and corresponding coding instruction (see Appendix D) were created for analyzing the articles, which were the unit of analysis. To obtain a 95 percent level of probability and 90 percent of agreement, 85 units were randomly selected for the inter-coders reliability tests (Riffe, Lacy, and Fico, 1998). Two additional coders were used in two phases to perform the task. Scott’s Pi was used for categorical variables, Spearrnan’s R for ordinal variable(s) and the Pearson correlations for interval or ratio variables. The three coders were the author and a graduate of the same doctoral program and a close associate of the author, who possesses proficient command of the English language. The two extra coders were trained through a series of stages suggested by Riffe, Lacy and Fico (1998). First, the author farniliarized the other coders and increased their comfort level with content of interest by providing 20 news articles not published in 45 any of the four newspapers analyzed. At this stage, no coding effort was made; instead, the coders were trained to expect what the news story and content analysis would be like. Then the author provided a copy of the coding protocol and instructions to them and explained to and discussed with them the definition of each variable and each item within a variable. Considerable time was spent to allow them to become familiar with the coding protocol and instructions before starting to practice some coding. With this complicated coding scheme, during the manipulation check period, the number of news stories coded was limited to a maximum of six within a session of an hour. During this practice stage, news articles flom not-to-be—analyzed newspapers were used again. More than two weeks were given for practice on materials flom other newspapers, and during the week, discussions on the coding protocol and instruction were held for possible further revisions to obtain acceptable inter-coder reliability level. To ensure the coders in two different phases were on the same page in their understanding of the coding protocol and instruction, another set of coding practice took place, using materials not being coded before. After the outcomes flom the practices achieved high agreement in coding, real materials flom the four newspapers were coded for inter-coder reliability testing. Three weeks were allowed for the real testing, partly due to the busy schedule of two other coders; the materials for reliability testing were randomly selected. The reliabilities range flom 85.7 to 100 percent (for the reliability for each tested variable, see Appendix E). 46 Conceptual & Operational Definition of Variables About 45 variables including content-unrelated ones were used as either independent or dependent variables for the content analysis. Independent Variables Newspaper: Each of the four newspapers analyzed is given an ID number. News Provider: The “news provider,” a categorical variable, is used as an organizational variable. Past research shows that wire service stories can differ flom those prepared by newspapers’ staff writers largely because of different degree of deadline pressure (i.e., F ico, 1984). The differences lie mostly in the uses of news sources and flames of news story. The news provider for each item will be identified through the byline given in the story and recorded accordingly. Eleven categories, including the “non- available,” of the variable are listed. They are: (1) news staff/correspondent, (2) news staff/wire service, (3) Associated Press, (4) other U.S. news wire service, (5) foreign news wire service, (6) combined wire services, (7) other U.S. newspaper(s), (8) Israeli non-wire news media, (9) Palestinian non-wire news media, (10) others, and (99) N/A. Wire news service is defined here as a news organization, U.S. or foreign, that provides and sells news stories to newspapers, magazines, radio, television or others media subscribing to its service. By this definition, news organizations like Associated Press, Reuters, UPI, AF P and Tass are considered as wire services. Similarly, The New York Times news service, the LA. T imes- Washington Post news service and Cox news service are treated as wire news services when such bylines start the story and will be coded as “4= other U.S. news wire service” when used individually. More important, to be clear, if The News York Times News Service, not a staff writer’s name, is listed as the 47 byline of the story in the Times, the news provider will still be coded as “4,” not “l= news staff/correspondent.” The same rule is applied to the LA. Times, Washington Post and Houston Chronicle. News provider will be used in the analysis as a dummy variable, with wire service= 0, and staff: 1, and all others= missing. Resource Allocation: Resource allocation is operationally defined in terms of the number of journalists—newspaper staff or wire service reporters—writing a story. The number of journalists writing the story is calculated by the number of reporters appearing in the byline as well as used at the end of the story as a contributor. It is arbitrary here that if no name appears in the byline, as is likely in some stories provided by a wire service, then the number of journalists will be coded as “2” because such cases usually indicate a collaborate efforts of two or more reporters. Also arbitrarily, “2” will be assigned to the story that uses “combined wire services” in that there is no way to know exactly how many journalists flom those wires were writing about the event. Moreover, when a story is a combined effort flom staff writer(s) and wire service reporter(s), the number of journalists is counted by adding the numbers for staff and wire service writer. If no countable number is given as in “by staff and wire service,” then “3” will be arbitrarily coded. Resource allocation is an interval variable. Foreign Policy: Foreign policy is essentially a reflection of a president’s decision on his goals sought flom a foreign country at the world political arena. While foreign policy may vary among different administrations, it seems quite consistent between Clinton’s and Bush’s policy toward Israel, although the Clinton administration appeared to be less hostile to Arafat and Palestine. For instance, weekly presidential documents showed that Clinton has invited Arafat to the United States for peace negotiation 48 meetings or personal visit since 1993, the first year of his two-term presidency, but Bush refused to invite the Palestinian leader during his first year as president. Moreover, Clinton has been considered and hailed as the U.S. president being most friendly to the PLO and Palestinians (Weekly Compilation of Presidential Documents, 2000). Accordingly, foreign policy is operationally defined here as the policy operated by the two administrations. More specifically, the concept is operationalized by recoding the year of the “date” variable into a dichotomous variable, where Clinton= O= year 2000, and Bush= 1= 2001. Press Access: Press access has long been considered an important factor in a country’s control of domestic as well as international information flow. This concept is operationally defined by the assessment of governmental control of internal and external information flow, a yearly report conducted by Sussman and Guida. Based on their annual reports, Israel is more press accessible than Palestine. The press access variable is represented by and recoded flom the dateline variable, with Israel= l and Palestine= O. This variable is used as a nominal or dichotomous variable. Percentage of P0pulation of the Ethnic Groups: The percentage of the population of Jewish Americans and Arab Americans in the newspaper’s readership radius are obtained flom the 1990 census data and other available resources, such as American Jews Yearbook 2001. As this variable is not directly linked to news content per se, it is not directly coded; it is examined and counted by the researcher himself. This is an ordinal- level variable. This variable will be recoded into a ratio variable (number of American J ews/number of American Arabs) for each of the newspapers analyzed. Then, the magnitude of the ratios for the four cities where the newspapers headquarter were 49 compared to the order of news treatment of Israeli officials in the newspapers for testing corresponding hypothesis. Level of Israeli Casualty: Casualty is operationally defined as including Israelis or Palestinians killed or wounded. To measure this concept more precisely, casualty is differentiated between wounded and killed, which are assigned a different scale. Killed is assigned a score of 2; wounded is assigned a score of 1. That is to say, if there are four Israelis killed and five Israelis wounded, then a total score of 13 will be coded for Israeli casualties. Similarly, if two Palestinians are killed and seven are injured, then a total score of 11 will be entered for Palestinian casualties. The level of Israeli casualty is operationally defined as Israeli casualty divided by Israeli casualty plus Palestinian casualty, (Israeli casualty/Israeli casualty + Palestinian casualty). Type of News Source: Shoemaker and Reese (1996, p. 127) defined new sources as “external suppliers of raw material, whether speeches, interviews, corporate reports or government hearings.” Type of new sources is thus defined here as the type of people, agents, organizations or governments quoted or cited in the news stories by ways of speeches, interviews, reports, press releases or hearings. The type of news sources is measured by its flequency of being quoted or cited. In each news story, the identity of any source is given, even when an unnamed source is used with such a phrase as “a government official said under the condition of anonymity,” so we are able to decide what, if not who, the source is. In most cases, the identity of a source is given by the title that precedes or follows the name of the source. This design is different flom the LSU project and is more appropriate. In that project, the team of scholars counted any source only once even if it appeared more than once. This research corrects this problem. 50 The selected ten types of news sources are: (l) Israeli prime minister (Barak or Sharon); (2) other Israeli authority: (3) Israeli non-authority; (4) Yasser Arafat; (5) other Palestinian authority; (6) Palestinian non-authority; (7) American authority; (8) American non-authority; (9) non-Palestinian Arab nations; and (10) others. The 10 types of news sources were chosen to fit the need for investigating sourcing patterns, for one, in these newspapers’ coverage of the conflict. The first six individuals or groups are likely to be used more flequently because of their direct involvement in the issue; American officials have also been found to be important news sources in foreign affairs, due mostly to their high involvement. Similarly, non-American officials, such as U.S. citizens traveling in the regions, non-Palestinian Arab officials and others including Britain were also selected for their connection to the region and conflict. The operational definition for each of the 10 types of news sources is provided in Appendix B. The types of news source, primarily Israeli and Palestinian official news sources, provide a basis for measuring and testing the news treatment of (the percentage of) Israeli official news sources as dependent and independent variables. Dependent Variables News Lead: The operational definition of news lead is adopted as given in the news articles; it is the first paragraph. The news lead, conventionally, is the most essential part of a story in that it leads the news story by providing the underlying information and construct of the story. Frequency of Full Quotes: is a ratio-level variable. It is used for measure because it shows how important the opinions of news sources are; the more a news source is deemed significant to the story, the more flequently the source would be quoted. A full 51 quote is defined as in the form of a complete sentence within the quotation marks or more (like a whole paragraph). For instance, the sentence—“We did not expect the withdrawal to be completely without dangers or difficult moments,” Mr. Barak said—is a full quote. Any quoting patterns other than one with a complete sentence will not be considered as a full quote. For instance, sentences like——Both sides “should immediately implement an unconditional cessation of violence,” the committee said; or Palestinian security officials should ensure that Palestinian workers with permits to work in Israel are “flee of connections to organizations and individuals engaged in terrorism” are not considered a full quote, even though the first example misses only the subject to be a complete sentence. Note that interrupted quotes are counted as many. If three full quotes are used for Israeli official news sources in a story, then “3” will be coded and entered accordingly. Also, if Palestinian official source is fully quoted only once in a story, then “1” will be coded. Placement: When a source is considered more important, he or she could also be quoted more prorrrinently, in a higher place in the story, so that his or her vieWpoints can be expressed more quickly. Placement is operationally defined in terms of the percentile of the first full quote and is represented by the simple formula, l-[(Xth-l)/N], where Xth is where the first full quote appears and N is the total paragraphs of a story. Using (Xth-I) allows the first full quote appearing in the news lead to always have a score of l or 100 percent. Moreover, the same quote—“We did not expect the withdrawal to be completely without dangers or difficult moments,” Mr. Barak said—was placed at the 9th paragraph of the story composed of 28 paragraphs and is coded first as “9.” Then, the placement of this quote in terms of the percentile will turn out to be about 71.43 percent {= 1-[(9- 52 1)/28)]}. In case no full quote from, say, a Palestinian official news source, is used, the percentile of the full quote will be arbitrarily decided as “0.” Number of Words in the First Full Quote Given to the Type of News Source: It is also logic that when news sources and their arguments are deemed more significant, they would be allowed to address their thoughts in more detail than vice versa. Number of words is a ratio-level variable and is by definition decided by the words count of the full quote that excludes descriptions used around the quotation mark and punctuation marks. For instance, the full quote—“As the word peace echoes in our hearts, how can we not think of the tensions and conflicts which have long be troubled the region of the Middle East,” the pope said at the airport reception. The quote—“So often hopes for peace have been raised, only to be dashed by new waves of violence.”—is counted as having 46 words. Frequency of News Sources Used in the Full Story: When news sources are considered more essential to the story, they are more likely to be cited or paraphrased more flequently. This is a ratio-level variable. The uses of news sources include sources being paraphrased, partially quoted or fully quoted throughout the story. The flequency is derived flom name count based on the categories of news sources operationally defined in the coding protocol; it is counted and coded into a certain type of source each time the name or identity of the source is mentioned. Construct of News Treatment of Type of News Sources: This construct is revised based on the definitions created by Stempel and Culbertson (1984). They distinguished between prominence and dominance. The former was defined as the frequency of mention of news sources, and the latter referred to the likelihood of being quoted instead 53 of merely being reported about. However, it is clear that an interaction effect can exist between the two operational concepts. News sources can be not only quoted but also quoted more frequently in a more prominent place, such as a quote placed in the news lead. Thus, the present research integrates and measures the uses of quotes in several ways. News Treatment is an important concept that cannot be sufficiently measured with the individual sub-concepts defined earlier. It is an integrative notion of prominence and dominance and is equal to the total uses of news sources in five individual measurements. Accordingly, news treatment is operationally defined by an index that is constructed with the previously defined five sub-concepts of uses of types of news sources in the news lead; placement of the first full quote of type of news source; number of word in the first full quote given to the type of news source; flequency of full quotes of the type of news source; and frequency of the type of news source used in the story. To ensure an index was creatable for each type of news sources, the following steps were completed or assured: (1) As the scale of measurement of the variables are not identical, each was converted into a Z-score; (2) the internal reliability among the variables was found moderately positive, with the Cronbach’s alpha greater than .70 but less than .90; (3) the variables were coded into the same direction as the sign of the alpha was positive; (4) none of the variables was redundant as the alpha is not greater than .90. No redundant measure was detected for any of the ten types of news sources. The Cronbach’s alpha for each of the ten types of news sources are as follows: Israeli prime minister (alpha =. 8278), other Israeli authority (.765 7), Israeli non-authority (.8301 ), Arafat (.8616), other Palestinian authority (.8616), Palestinian non-authority (.7025), 54 American authority (.8765), American non-authority (.8462), Non-Palestinian Arabic nations (.8078), and others (.8313). The index was then created by summing up the means of five variables associated with each type of the news source. That is, for instance, Israeli authority index: mean (Israeli authority in news lead + flequency of full quote of Israeli authority + placement of the first full quote + number of word of the first full quote + flequency of citation in the full story). News Frames: Entman (1993, p. 52) considered news flames as “some aspects of a perceived reality” made more salient in a news story. Here news flames are operationally defined as more clear-cut types of news flames or themes, as opposed to those hidden metaphors, catchphrases or symbolic devices suggested by Gamson (1989). A total of 11 types of news flames are created in a way to correct the possible pitfalls found in the LSU project. The LSU project, however, provided a basis for generating the 11 types of news flames. These 11 flames for the current research are: (l) Israeli sovereignty; (2) Israeli religious tenacity; (3) Israeli terrorism; (4) Palestinian terrorism; (5) Muslim religious tenacity; (6) Palestinian independent movement; (7) peace process; (8) world terrorism; (9) overall world peace; (10) American interests; and (11) others (list). Here the news themes were measured in the same way placement of news sources was conducted, with the formula of 1-[(Xth-l)/N]. That is, the percentile of each of the 11 news flames was used in a way to examine the importance of each flame through its immediacy of appearance in the news article. For instance, if a news flame of the peace process appeared the first time in the news lead of a story, then 1 was coded accordingly. Likewise, if a news theme, say world peace, was not used at all in the story, then it was arbitrarily coded as “0.” This was used as a ratio-level variable. 55 Episodic vs. Thematic: The notion and definition of episodic and thematic frame are borrowed from Iyengar’s (1991, 1993), who dubbed the two terms. He states: “The episodic news flame takes the form of a case study or event-oriented report and depicts public issues in terms of concrete instances. The thematic flame, by contrast, places public issues in some more general or abstract context and takes the form of a ‘takeout,’ or ‘backgrounder,’ report directed at general outcomes or condition.” Iyengar also suggested that a mixed flame is not uncommon in news coverage, but in most cases, there exists a predominant flame that can be distinguished. Accordingly, an episodic flame is more event-oriented, descriptive and focuses on the dramatization of the issue development, while thematic flame is more issue-oriented, analytical and pays more attention to the political, cultural, historical or religious background and cause of the news issue. This is a categorical variable. Responsibility Assignment: Borrowed flom psychology, the causal responsibility concept focuses on the origin of a problem. In contrast, treatment responsibility pays attention to who or what has the power to alleviate (or forestall alleviation of) a problem. Responsibility assignments are decided on the overall reading of the story. It is likely that news sources flom both sides blame each other for delaying peace talks or for any violent action, but the decision can be made through reading the headline, the quotes and comments used. The decision on the assignment of treatment responsibility can be similarly made. Both causal and treatment responsibility assignment variables are created with a scale for five dichotomous groups of political entities: (1) Israel (assigning responsibility to Palestine) or Palestine (assigning responsibility to Israel); (2) United States; (3) the 56 Arab world; (4) the Western society; and (5) the United Nations. If none of the five groups assigned causal or treatment responsibility to either Israel or Palestine, then “none” will be checked. Israel or Palestine could be assigned with causal and/or treatment responsibility by any of the five political entities, or by none of the five entities in any given story. For instance, if in a story Palestine was assigned with causal responsibility by Israel and the United States, then it was assigned a causal-responsibility score of 2. On the other hand, if Israel was assigned with the responsibility only by Palestine, then a score of l is recorded. If neither party was assigned with the responsibility, the column “none” is checked. Then, a ratio variable of causal or treatment responsibility assigned to Israel, represented by the (total causal or treatment responsibility assigned to Israel)/(total causal or treatment responsibility assigned to Palestine), was generated. In bivariate analysis, the test value of this ratio variable is chosen at 1. Put another way, a causal-responsibility- assignment score of within three standard deviation of 1 indicates both sides were equally blamed (or not blamed) for the increasing of violence or delay of peace negotiation. A score greater than (positive) three standard deviation of 1 suggests causal responsibility was assigned more to Israel than to Palestine, and a score less than (negative) three standard deviations of 1 means the causal responsibility was assigned more to Palestine than to Israel. The rules are likewise applied to examining treatment- responsibility- assignment between Israel and Palestine. Control Variables When evaluating any association between independent and dependent variables, it is essential to decide whether the relation is genuine or spurious by controlling for 57 potential confounding variables and removing their impacts (Tabachnick & F idell, 2001). A spurious relationship occurs when initial influences of the independent variables disappear after confounding variables are controlled. To examine the true effect, if any, of the independent variable, spurious relationship should be removed as much as feasible by controlling for as many confounding variables as possible. In path analysis, it is necessary that all other independent variables not needed for examining the dependent variable for each phase be used as control variables. That is to say, for instance, in testing against type of news theme as a dependent variable, foreign policy, press access, resource allocation and casualty will be controlled for. Also, foreign policy, press access, news provider, resource allocation and percentage of American Jews will be controlled to test against causal and treatment responsibility assignments as dependent variables. Statistical Analysis Assumptions of Multiple Regression Analysis Because this research investigates a population, inferential statistics is not required. Standard multiple regression analysis and logistic regression analysis, along with a variety of bivariate statistics, were used to offer a more comprehensive analysis of the study. While multiple regression analysis techniques allow researchers to examine situations under which a dependent variable is influenced simultaneously by more than one independent variable, bivariate statistics is a method that can show only binary associations between one independent variable and one dependent variable (Tabachnick & Fidell, 2001). 58 For multiple regression analysis to be more valid, six assumptions need to be examined against the data prior to testing the hypotheses because any violation of the assumptions can distort statistical outcomes. The six assumptions are: randomization of sampling, normality of data distribution, linearity, homoscedasticity of relationships among variables, absence of multicollinearity, and an acceptable ratio of cases to independent variables (Tabachnick & F idell, 2001). Randomization of sampling is an essential element in any inferential statistical methods because the value of the inferences relies greatly on how satisfactory the sample represents the population. It is usually achieved by using a random sample. The assumption of normality can be examined by looking at the standardized z- scores to detect potential outliers. It is suggested (Tabachnick & F idell, 2001) that standardized score greater than 3.29 (p< .001 for two-tailed test) are likely outliers, and as the size of the population data is quite large, with more than 500 cases, the researcher decided to adopt a more stringent benchmark that has any cases with a standardized score larger than 3 as outliners. If outliners exist, they have to be transformed; otherwise this smaller but more influential group of cases could distort the outcome of the analysis (Tabachnick & Fidell, 2001). The assumptions of linearity and homoscedasticity of relationship among variables can be diagnosed by examining the bivariate scatterplots of the residuals against the dependent variable and independent variables. The linearity assumption suggests a straight-line relation between two variables (Tabachnick & Fidell, 2001). If nonlinearity is detected and is curved, transformation of data by using a logarithm should be considered. The homoscedasticity assumption means that a roughly constant standard 59 deviation exists in all continuous variables (Tabachnick & F idell, 2001). If this assumption related to the normality assumption failed, the residuals of the variables will deviate away flom zero, lacking a consistent variation between the examined variables. The fifth assumption, multicollinearity, which happens when variables are highly correlated and can inflate standard errors of the estimated regression coefficients. Multicollinearity exists when bivariate correlations between independent variables reach .90 or greater. The sixth assumption of case ratio requires an acceptable ratio between number of cases and of independent variables; for statistical regression, a cases-to- independent variable ratio of 40 to l is usually recommended because such regression analysis can create an outcome that fails to generalize if the sample is too small (Tabachnick & F idell, 2001). The present research used unstandardized regression coefficients, standardized regression coefficients and squared part correlations to test the hypotheses using multiple regression analysis, and used unstandardized regression coefficients, wald test and odd ratios for hypotheses using logistic regression analysis of dichotomous dependent variables. Squared part r (flom the semi-partial correlations) provides the percentage of variance in the dependent variable explained by each independent variable controlling for all other independent variables (Tabachnick & Fidell, 2001). As logistic regression analysis does not provide semi-partial for evaluating individual predictor’s strength on dependent variable, odd ratios are used; the greater the ratio (flom 1), the more important the independent variable to the dependent variable (Tabachnick & F idell, 2001). Wald test or ratio is used to evaluate the contribution of an explanatory variable to a model, and as the wald test tends to be more conservative and could lead to a greater chance for Type 60 11 error, rejecting alternative hypothesis when it is true (Tabachnick & F idell, 2001), the wald test is used in a more liberal manner, with this population data set. Preliminary Analysis of Data Prior to testing the hypotheses, preliminary analysis of the data is needed to decide how satisfactory the data fit the assumptions of the multiple regression analysis methods. This section reports the outcomes of the preliminary analysis. The first assumption of randomization of sample was met by using a population, and the remaining four were analyzed as follows: First, the normality of data distribution was examined with flequency splits for dichotomous variables or standardized z-scores for continuous variables (Tabachnick & F idell, 2001) to detect possible outliners for each of 11 variables used in the analysis. None of the dichotomous variables—foreign policy, press access, news provider, the nature of news theme—revealed a 90-10 split between categories, and therefore no outliner exists. As for continuous variables, casualty, percentage of American Jews, percentage of news treatment of Israeli officials, and the peace process (or Palestinian terrorism) have no outliners. Resource allocation in terms of number of reporting journalists was found to have four cases of outliners; causal responsibility assignment has 17; and treatment responsibility assignment has two. Each outliner was reassigned the value of three standard deviation flom the mean. For resource allocation (ntunber of reporter), the mean was 1.2514 and the standard deviation was .6121; each of the four outliners was then reassigned to 3 flom 8, 5 or 43. For causal responsibility assignment, the mean was .9945 and the standard deviation was .1416; each of the 17 outliners was corrected to 1.41 flom 61 2, or .57 from .50. For treatment responsibility assignment, the mean was .9982 and the standard deviation was .030; both outliners were altered to .90 flom .50. Second, scatterplots of residuals were checked to examine whether the assumptions of linearity of relationship among the variables and homoscedasticity, were violated. No apparent patterns to the residuals were found and therefore neither non- linearity nor heterosedasticity occur. In terms of multicollinearity, bivariate correlations among independent variables were checked and no correlation greater than .90 was found. The greatest bivariate correlation was .393 between news provider (as a dummy) and percentage of American Jews. Thus, multicollinearity was not a concern. As for the case ratio, because the examination of the proposed theoretical model is based upon the total number of cases for all four newspapers, the data satisfy the 40-to-1 case ratio recommended by T abachnick and F idell (2001), even though the ratios for individual newspapers don’t satisfy the requirement. In general, the data gathered for the research met all six assumptions of the linear multiple regression model, and the preliminary examination suggests the model fit the data quite well. Statistical Procedures Hypothesis 1 measures news treatments of Israeli and Palestinian official news sources in the news articles. The statistical procedure contains several stages: (1) for the independent variable, foreign policy is recoded flom the year of the “date” variable into Bush= 1= year 2001, and Clinton= 0= year 2000; (2) creating a percentage variable of news treatment of Israeli official sources—news treatment of Israeli official 62 sources/(news treatment of Israeli official sources + news treatment of Palestinian official sources); (3) testing it in a standard multiple regression analysis against the recoded foreign policy as independent variable. Moreover, in bivariate analysis, the breakeven point of this percentage variable is chosen at the three standard errors of the mean, which is .5730, with its standard error of about .0154. That is, .5268 is selected as the benchmark. This selection seems practically and statistically appropriate in that it allows some leeway and avoid being entrapped into the black-or-white argument were .5 score used. Therefore, a score equal to .5268 indicates no particular preference exists for either Israeli officials or Palestinian officials. A score great than .5268 suggests Israeli officials received more preferable news treatment than Palestinian officials; the greater the score (greater than .5268), the more favorable the news treatment the Israeli officials received. On the other hand, a score less than .5268 means Palestinian officials received more preferable treatment than their counterparts, and the smaller the score (less than .5268), the more preferable the news treatment Palestinian officials received. Variables combined for the news-treatment index for Israeli or Palestinian officials as news sources as a dependent variable and independent variable include: -- Official sources in the news lead: Variable 11, 12, 14 & 15; -- Placement of the first quote of news sources: Variables 21, 22, 24 & 25; -- Number of words in the first full quote: Variables 31, 32, 34 & 35; -- Frequency of full quote flom sources in the story: Variables 41, 42, 44, & 45; -- Frequency of news sources used in the whole story: Variables 51, 52, 54 & 55. 63 * Hypothesis 2 fUrther examines the potential impact of press access on news treatments of Israeli official news sources and Palestinian official news sources. The statistical procedures are identical to those in H2, by replacing the independent variable with press access, which is recoded from “dateline” into Israel= 1 and Palestine= 0. * Hypothesis 3 examines the potential impact of news provider (staff vs. wire service) on the news treatments of Israeli official news sources and Palestinian official news sources. The procedures similar to those used in H2 and H3 includes: (1) using the recoded dichotomous variable of news provider; (2) using the percentage variable of news treatments of the Israeli official sources—news treatments of Israeli official sources/(news treatment of Israeli official sources + news treatment of Palestinian official sources); (3) testing it against the dichotomous independent variable. * Hypothesis 4 examines the possible influence of the percentage of the two ethnic groups on news treatments of Israeli and Palestinian official sources. The statistical procedures include (1) creating a ratio variable of American Jews for each newspaper, the number of American Jews divided by that of American Arabs for each corresponding city under examination; (2) using the percentage variable of news treatments of Israeli official source, news treatments of Israeli official sources/(news treatments of Israeli official sources + news treatments of Palestinian official sources) to test the independent variable. * Hypothesis 5 examines the likely influence of the number of reporters writing for a story on the different news treatments between Israeli official sources and Palestinian official sources. The percentage variable of news treatment of Israeli official sources is used to test against the interval independent variable. 64 * Hypothesis 6 examines the potential impact of the magnitude of casualty on news treatment of Israeli and Palestinian official news sources. The statistical procedures include: (1) recoding the level of casualty of Israelis and of Palestinians into a percentage variable of Israeli casualty, casualty of Israelis/(casualty of Israelis + casualty of Palestinians), and using it as the independent variable; (2) using the percentage variable of news treatment to test the independent variable. * Hypothesis 7 examines the influence of (percentage variable of) news treatment of Israeli official sources on the uses of type of news flame. The statistical procedures include: (1) using the percentage variable of news treatment of Israeli official sources as the independent variable; (2) running multiple regression analysis against the peace- process news flame as a ratio-level variable. * Hypothesis 8 examines the potential impact of the dichotomous independent variable of news provider (staff vs. wire) on uses of peace process news flames by running a multiple regression analysis. * Hypothesis 9 examines the potential influence of (percentage variable of) new treatment of Israeli official sources on the nature of news flame as dichotomous dependent variable, with thematic= 1 and episodic= 0. The statistical procedures include: (1) using the percentage variable of news treatment as the independent variable; (2) running a logistic regression against the dichotomous dependent variable. * Hypothesis 10 examines the potential impact of the dichotomous independent variable of news provider (staff vs. wire service) on nature of news flames in a multiple regression analysis. Also, cross-tabs are used to examine the binary relationship between the two variables. 65 * Hypothesis 11 examines the potential influence of (percentage variable of) news treatment of Israeli official sources on the assignment of responsibility (causal and treatment) as two separate dependent variables. The statistical procedure is using the percentage variable of news treatment of Israeli official sources as the independent variable to test the dependent variable by running multiple regression analysis. * Hypothesis 12 examines the potential effect of nature of news flames on assignments of causal and treatment responsibility with a multiple regression analysis. * Hypothesis 13 examines the potential impact of the level of Israeli casualty on assignments of causal and treatment responsibility. The statistical procedures include: (1) re-coding the level of casualty of Israelis and of Palestinians into a percentage variable of Israeli casualty, casualty of Israelis/(casualty of Israelis + casualty of Palestinians), and using it as independent variable; (2) running multiple regression analysis against the dependent variables. 66 CHAPTER VI RESULTS Data Analysis Before getting into the results of the 13 hypotheses, it is both beneficial and necessary to present the basic information (see Tables 1 to 10) about the variables used and tested in the hypotheses. Consistently, more stories were written during the Bush government than during the Clinton administration. One striking fact is that the number of stories reported in the Chronicle during the Bush regime is considerably greater than during the Clinton administration, more than twice as many. The outcome may lie in the fact that Bush has roots in Houston and was governor of Texas before winning the presidency and, therefore, the issue and Bush’s accompanying political actions and policy are given considerably more attention. Table 1 (p. 114) shows that all four newspapers filed more of their news stories flom Israeli or Israeli-occupied territories than flom its Palestinian counterpart. Overall, 252 of 549, or 45.9 percent, news articles had a dateline in Israel while 166, or 30.27 percent, were datelined Palestine. Individually, The New York Times (81 vs. 43) has nearly twice as many articles datelined flom Israel than flom Palestine, and the Chronicle (34 vs. 16) has more than double as many flom the former than flom the latter. Dateline- wise, Israel has only a slight advantage in the Washington Post (46 vs. 42). And The New York Times has the greatest news reports, in terms of both number and percentage (28 or 67 16 percent), from the United States, where peace negotiations were held during the final days of the Clinton administration. The Houston Chronicle has the fewest number of stories but second largest in percentage in this category. The combined number for territories or political entities, such as the United Nations, European countries and other world locations, is very marginal. The Washington Post has 17 un-datelined stories, accounting for nearly 14 percent of total articles. Table 2 (p. 115) reveals the flequency of news articles prepared by different types of news providers. All together, staff or foreign correspondents accounted for 475, or 86.5 percent, of articles; Associated Press provided 45, or 8.20 percent, of them, with 16, or 2.9 percent, provided by other U.S. news wire services. As shown, all but the Chronicle (17 or 25 percent of the total 68) prominently ran articles written by their own staff writers. Each of the three national newspapers used at least 93 percent stories written by their own staff. On the other hand, the Houston newspaper adopted stories dominantly provided by Associated Press (37 or 54.4 percent) and other U.S. (8) or foreign (2) wire services. Altogether, the Chronicle used 48 or 70.6 percent of news stories written by wire services; only 17 stories were written by its own staff, mostly about the peace talks in the country. The Times, the Post and the LA. Times had marginal numbers of A.P.-prepared stories: one, three and four, respectively. The Houston daily also adopted three articles prepared by other U.S. newspapers, while the other three national newspapers have none. No story by Israeli or Palestinian non-wire news media or others was picked up by any of the four newspapers. Table 3 (p. 116) discloses a pattern in newspapers’ selection of news sources in the leads. Other Israeli officials, American officials and other Palestinian officials are the 68 three types of news sources most likely to be used in the leads overall and in all but the Houston newspaper. More importantly, overall, other Israeli authorities (with a mean of .1111) are the top choice, followed by American authorities (.0874) and other Palestinian authorities (.0783). A mean score of .1 l 11 indicates that in more than 11 of 100 times, other Israeli authority would appear in the lead of a story. In the same vein, nearly nine of 100 times, American authorities were used in the lead, and other Palestinian authorities were adopted in the lead nearly eight of 100 times. Between the uses of the Israeli premier and Yasser Arafat in leads, the former (.0310) had an overall edge over the latter (.0219). Also, Israeli officials in total (.1421) were more flequently employed than Palestinian officials in total (.1002), so were total Israelis (.1467) than total Palestinians (.1257). However, the advantages were not presented in the Chronicle, which used Arafat (.0294) twice as more flequently than Barak and Sharon combined (.0147), total Palestinian officials (.1176) over total Israeli officials (.0882), and total Palestinians (.1617) over total Israelis (.0882). Also noteworthy are that both The Los Angeles Times and the Chronicle used Palestinian non-authority in a high flequency (.0385 and .0441), which made up somehow the difference for the uses of official sources between Israel and Palestine, and that the Post used “others (.0726)” sources considerably more often. American non-authority and non-Palestinian Arab authority were rarely used overall and by any of the four newspapers. In the uses and placement of the first full quote of types of news sources, Table 4 (p. 117) reveals a similar pattern where other Israeli officials, Palestinian non-officials and other Palestinian officials are among the top three in individual newspapers and overall; American authorities followed immediately. In three out of four papers (except 69 for the Post), other Israeli officials are the group that gets the first full quote of the stories most flequently. Instead of having Israeli officials (26.82) as their top choice, the Post went to Palestinian non-authorities (30.18) most flequently. A mean score of 26.82 suggests the average first quote of other Israeli authority was ranked at that percentile in a story, and a score of 30.18 indicates the average first quote of other Palestinian authority was placed as high. It then shows that other Palestinian officials, on average, were given more favorable placement than their Israeli counterparts. Similar to the choice between Arafat and Israeli prime minister, the Houston Chronicle is the only newspaper that used the Palestinian leader (12.25) more often than his Israeli counterparts (7.21), though overall, the Israeli premier has the advantage over Arafat. Moreover, the Chronicle was the sole paper that used other Palestinian officials more prominently than Palestinian non- officials. The selection of non-official news sources between Israel and Palestine apparently makes up the difference in the official news sources between the two groups. While total Israeli officials were used more prominently than total Palestinian officials overall and in all newspapers but the Chronicle, total Palestinian sources were used more prominently than total Israeli sources in all and particularly in the Chronicle. The New York Times has the smallest disparity between total Palestinian and Israeli sources. As in the use of leads, American non-authority and non-Palestinian Arab authority were not quoted in a prominent position of the story. Similarly, Table 5 (p. 118) shows that other Israeli officials (12.581), other Palestinian officials (10.00) and Palestinian non-officials (8.4291) were the three types of news sources used most generously (in number of words) in their first full quotes in the stories. A mean score of 12.581 simply means that on average as many words were used 70 in the first full quote of other Israeli officials in a story; 10 words were used in the first quotes of other Palestinian officials. American authorities (5.7832) ranked fourth again. And again, the Chronicle, in contrast to the other three dailies, gave Arafat (5.1324) nearly twice as much space to Israeli prime ministers (2.6029). Moreover, the Houston daily also cited American authorities (6.0588) more generously than Palestinian non- authorities (4.7647), as contrasted with the other three newspapers. Furthermore, the Chronicle (1.5294) is the newspaper that cited Israeli non-officials least flequently compared to its counterparts (5.76, 5.0968, 5.6099, respectively for The New York Times, the Post and The Los Angeles Times). The employment of non-official news sources also made up the difference in the official news sources between Israel and Palestine. After non-Palestinian and non-Israeli officials were taken into account, the Times and the LA. Times in particular and the four newspapers in general still used Israeli sources as a whole more dominantly than Palestinian sources, but the uses of total Palestinians have outweighed those of total Israelis in the Post and the Chronicle. The gap in the uses between total Israeli (21.9675) and total Palestinians (17.52) in The New York Time remains quite large. Again, overall, American non-authority (1.9071) and non-Palestinian Arab authority (1.3625) were given the least space, in number of words, in their first full quotes. Table 6 (p. 119), which reports the mean scores of the number of full quotes used for each type of news sources, also mirrors the pattern revealed in Table 4 & 5. Overall, other Israeli authorities (1.0091), other Palestinian authorities (.7905) and Palestinian non-authorities (.7905) dominated the news sources in the flequency of being fully quoted. A mean score of 1.0091 suggests on average other Israeli authority was fully 71 quoted about once in a story while other Palestinian authority was fully quoted less than once in a story. The nearly identical similarities also are reflected in the Houston Chronicle’s preferred uses of Arafat to Israeli prime ministers and its selections of Israeli non-authorities. That is to say, in the Chronicle, Arafat (.2941) was quoted nearly twice as flequently as Israeli premiers (.1471), and Israeli non-authorities were selected considerably less often than the other three newspapers. Also, American authorities ranked immediately behind the other three types of news sources. Noteworthy is that The New York Times, shown also in Table 4, was the one that quoted other Israeli officials (1.16) and total Israeli officials (1.4229) most generously, while the Chronicle the least (.5882 and .7353). But the LA. Times (.4781) had a greater difference in uses between total Israeli officials and total Palestinian officials than did the Times (.3886). In the makeup effect of quoting non-official news sources, the Post presented a large discrepancy (.8790), the greatest among the four newspapers, in the uses of total Palestinians and total Israelis as sources, compared to .6167 for the Chronicle, .4339 for the LA. Times, and .1714 for the Times. Table 7 (p. 120) shows that pattern again where other Israeli officials (4.6066), Palestinian non-officials (3.1093) and other Palestinian officials (3.0182) topped the list in terms of the flequency of citations or paraphrases given to each type of news sources, and similarly American authorities (2.0091) ranked immediately behind. A mean score of 4.6066 indicates that, on average, other Israeli authority was cited or paraphrased as many times in a story. Not shown in the previous four tables, other Israeli officials were the unanimous first choice across the four newspapers in getting cited or paraphrased although the Houston Chronicle used it the least generously. The Chronicle has the 72 identical sourcing pattern between Arafat and Israeli prime ministers. Not only was Arafat selected more frequently within the newspaper but across the four newspapers. More specifically, within the paper, Arafat (1.0441) was cited more than twice as often as the Israeli premier (.4853), and he was cited or paraphrased more flequently than in the Times (.6229), the Post (.4677) and the LA. Times (.6647). Expectedly, the Chronicle also had the largest difference in the uses between total Palestinians and total Israelis, even though the Houston daily selected Palestinian non-officials less flequently than the other three newspapers. Noteworthy here are that the uses of total Israeli sources (7.9286) remained greater than those of total Palestinian sources (7.4725) in the LA. Times and that the Times employed total Palestinian sources (6.5886) only slightly more flequently than total Israeli sources (6.4572). Again, American non-authority (.3206 overall) and non-Palestinian Arab authority (.2847) were least cited in any story for all newspapers. Table 8 (p. 121) shows the degree of news treatments for each type of news sources in terms of indexed mean score. Through the outcomes of the previous five tables (3 through 7) regarding different ways of uses of types of news sources, it is not difficult to reckon or expect that other Israeli officials (18.571), other Palestine officials (14.114), and Palestinian non-officials (13.512), along with American officials (8.5274), are the four types of news sources that would receive the most preferable news treatment in news coverage, and that Israeli officials would obtain more favorable treatment than Palestinian officials. The greater the indexed score, the more favorable news treatment the corresponding news source has received. As the same logic goes, while Israeli premiers (5.3973) receive an overall more favorable news treatment than Arafat (3.8962), they (3.3221) lose the preference to Arafat (6.6225) in the Houston Chronicle while 73 retaining it in the other three newspapers. The Houston daily also gives more favorable treatment to Arafat than do the other three. The Chronicle (12.7007) not only ranks last among the four newspapers in news treatment of other Israeli officials, but within the newspaper, it gives only slightly more preference to other Palestinian officials (12.2297). American non-authority and non-Palestinian Arabic authority were treated least favorably overall and by each individual newspaper. While other Israeli officials also receive the greatest preference within individual newspaper and across, it is The Los Angeles Times (20.6505), not The New York Times (18.4228), ranked behind even the Post (18.9456), that offers the most favorable news treatment to them. The same assessment holds when the added score (23.654) for news treatment of Israeli officials that include both the Israeli premier (5.2307) and other Israeli officials (18.4228). Nonetheless, when it comes to the total percentage of news treatment of Israeli officials (= total Israeli officials/total Israeli officials + total Palestinian officials), The New York Times (.6003) emerges as the newspaper that gives the most favorable news treatment to Israeli officials; LA. Times ranks second (.5958), the Post third (.5480). The reason is that the Times (16.108) gives the least favorable news treatment to Palestinian officials (= Arafat + other Palestinian officials). Houston Chronicle (.4947< .5268) not only ranks the last but also gives more favorable news treatment to Palestinian officials than to Israeli officials. However, overall, the newspapers (.5730) were more favorable toward Israeli officials. Moreover, when non- officials were included, only The New York Times still provided clear-cut better news treatment to Israeli sources as a whole, and the LA. Times gave only very marginal 74 preference to Palestinian news sources (34.948 vs. 34.861 for Israeli sources). The Post treated Palestinians (36.115) more favorably than Israelis (30.894). The data analyses up to this point clearly reveal several major characteristics: First, Israeli officials tended to be used or treated more favorably than their Palestinian counterparts in all newspapers but the Chronicle. Second, the uses of Palestinian non- officials, alongside their Israeli non-officials counterparts, have considerable makeup impacts on how total Israeli and total Palestinian news sources were used or treated. Third, the Houston Chronicle is unique flom the other three newspapers. The striking dissimilarity probably lies in the fact that most of its news stories were written by wire- service journalists, in contrast to the other newspapers whose articles were prepared mostly by staff. Fourth, American non-authority and non-Palestinian Arab authority were the two types of news source used least flequently or prominently. While the Houston Chronicle differentiates itself flom the other three newspapers, it is consistent with them in the selection of news themes or flames. Table 9 (p. 122) discloses the mean scores of the immediacy of the 11 news themes mentioned in terms of the order in the stories in the publications. A score of 100 suggests a particular news theme was used in the lead; the greater the percentage score, the earlier the theme was mentioned and the more significant the news theme was in the story. Overall and individually, peace process or negotiation (with an overall mean of 64.28) is the leading news theme selected by the papers. And interestingly, the Chronicle (69.30) is indeed the one that gave the theme the greatest attention, more than that of the Times (67.49), the Post (61.60) and the LA. Times (61.14). Other themes given high attention include Palestinian terrorist/terrorism (with an overall mean of 37.24), others (34.75), and 75 Palestinian security or independent movement (25.51), Israeli sovereignty or security (13.21), Israeli terrorists/terrorism (10.87) and world terrorism (10.31). Several noteworthy things deserve extra attention and explanations: First, the news theme of Israeli terrorism, excluded in the LSU project, not only surfaces in the current research but also receives considerable attention. Intriguing but not surprising, this theme is used least prominently by the Times (5.06) and most prominently by the Chronicle (19.30). One reason could be the relative use of A.P.- and other wire service- written news stories. While the Times used Israeli terrorism least significantly, it adopted the Israeli-security theme in the highest immediacy. Moreover, the great attention paid to “others” theme is more of an unexpected finding. Of the aspects coded as “others” are different impacts of the conflict, such as its great influence on Palestinian economy, civil life and cuniculums and its plea for religious support flom the Pope. For instance, the papers tum their focus onto how the Palestinian Liberation Operations (P.L.O.) and Palestinian Authority preach hatred toward Israel in their revised textbooks and train their youth in military skills in summer camps. Also noticeable are the examination and sensational presentation of the impact of press images by the LA. Times. Furthermore, the appearance of the world peace theme can be expected as the study covers the time flame during which the “9-1 1” terrorists attack occurred. It is reasonable to see that the Washington Post (14.90) and The New York Times (10.82) are the two paying the highest attention because Washington, DC, and New York City were the locations of the attacks. Another interesting thing is that the correlation between the themes of world terrorism and world peace in the four newspapers is almost reversed. The Post never mentions world peace in its coverage. 76 Table 10 (p. 123) reports the mean scores of the scale of causal and treatment responsibility assignments to Israel or Palestine in the newspapers. Overall, bivariate statistics results suggest that causal responsibility (mean: .9943) is assigned to Palestine; the p-value for one-tailed, one-sample t-test is less than .05, with 1 as the test value and benchmark. The treatment responsibility (mean= .9996), however, is not assigned to either party, with a p-value great than .05. Individually, only the Times assigned causal responsibility to Palestine; p-value is less than .05. The other three newspapers have no difference in either causal or treatment responsibility assignments to either political entity. Hypotheses Testing This section presents the outcomes of the 13 hypotheses testing. In addition to the findings flom multiple regression analysis, results flom binary statistics are also reported to offer a more thorough analysis. Table 11a (p. 124) report the findings flom multiple regression analysis for hypothesis 1 through 6. The six independent variables accounted for only 8.9 percent of the variance of the dependent variable, with a constant B- coefficient of .292, and a standard error of .082, and degree of fleedom of 400. Hypothesis 1 states that the difference between favorable treatment of Israeli and Palestinian official news sources was greater during the period when U.S. foreign policy was more favorable toward Israel. Outcomes flom multiple regression analysis (see Table 11a) suggest that the hypothesis is accepted. The unstandardized regression coefficient is positive, at .106, and the beta equals to .152. The fact when a story was written—during the Bush or Clinton administration—explained 2.3 percent of the total variance for the 77 news treatment of Israeli officials. It suggests, although weakly, that when a story was reported during the Bush government, Israeli officials would receive 10.6 percent more favorable news treatment. Furthermore, binary statistics findings reveal that the Bush administration has a mean score of .6095, compared to Clinton’s .5225. The scores indicate that the newspapers in general offer more favorable treatment of Israeli officials than their Palestinian counterparts, particularly in the Bush government, because the mean is greater than .5268, the arbitrary breakeven point. Hypothesis 2 states that because press access in Israel is greater than that in Palestine, Israeli official news sources would receive more favorable news treatment than Palestinian official news sources. This hypothesis also is supported (see Table 11a). Press access appears to be the strongest predictor of the dependent variable, with a beta weight of .244; it also accounts for 5.9 percent of the total 8.9 percent of the variance. The B- coefficient indicates that when a news article was datelined in Israel, Israeli officials would receive 17.2 percent more preferable news treatment. Outcomes flom binary statistics reveal similar outcomes as in the multiple regression analysis. When articles are datelined Israel, Israel officials (.6524) receive considerably more preferable news treatment. And while Palestinian officials (.4948 <. 5268) obtain slightly better treatment when stories are datelined Palestine, it is Israeli officials who get overall more favorable news treatment. Hypothesis 3 states that staff writers would be more likely than wire-service reporters to give favorable news treatment to Israeli official sources than Palestinian official sources. This hypothesis is unsupported (see Table 11a). While the direction of 78 the relationship is positive as predicted, its strength of association is too weak for the hypothesis to be accepted. News provider as an independent variable accounts only for .5 percent of the variance, with a low beta weight of .081 and unstandardized coefficient of .083. Also, results flom binary statistics show no significant difference in the news treatment of Israeli officials between staff and wire-service writers. The mean score for staff writers is .5841, compared to .5131 for wire service writers. Hypothesis 4 postulates a positive association between the ratios of American Jewish population to American Arab population in the newspaper readership radius and the ratios of preference of news treatment of Israeli official sources in each individual newspaper. This hypothesis also is not supported (see Table 11a, p. 124). The beta weight for ratio of American Jews is even smaller than that of news provider in hypothesis 3, even though the positive relationship is found as predicted. It has an unstandardized regression coefficient of .003, with a beta weight of .071; it explains only .4 percent of the variance of the dependent variable. Hypothesis 5 posits a positive relationship between the number of reporters writing the stories and the favorableness of news treatment of Israeli officials. The hypothesis is not supported, either (see Table 11a). The strength of association between resource allocation in terms of number of reporters on assignment and the level of news treatment of Israeli officials are very trivial. Similarly, hypothesis 6, which states that the greater the level of casualties occurring to Israelis, the more likely Israeli official sources would receive more favorable news treatment than Palestinian official sources, is not supported (see Table 11a). The 79 outcomes report not only a negative but also an invisible association, and therefore, the independent variable has no impact on the dependent variable. Table 123 (p. 129) report outcomes flom multiple regression analysis for hypotheses 7 & 8. Controlling for five variables, the two independent variables— percentage of Israeli officials news treatment and news provider—explained less than 1.6 percent of the dependent variable, uses of peace process as dominant news flame. The r- squared is .075, contributed by a total of seven independent or controlling variables, with a constant B-coefficient of .564, a standard error of .095 and a degree of fleedom of 400. However, press access seems to be an important variable; it explains 3.6 of the total 7.5 percent of variance, with an unstandardized coefficient of .158 and a beta weight of .199. Hypothesis 7 assumes a positive association between the percentage of news treatment of Israeli officials (IV) and uses of peace process/negotiation as the dominant news theme (DV). The hypothesis (see Table 12a) is not supported. This independent variable has an unpredicted negative association (B-coefficient= -.146) and a standard error of .057, although it accounts for 1.5 percent of the variance. While the peace process, as reported in Table 10, is the dominant news theme in all four newspapers, it may be used by sources other than Israeli officials. In reality, Israeli officials, particularly Ariel Sharon, seem to cite terrorism or accuse Palestinian leaders or militants of terrorists more flequently when terrorist attacks, such as suicide bombings executed by Palestinian militants, hit the country. Moreover, in responding to Israeli accusations, Arafat and other Palestinian officials tend to refer to peace negotiation to avert criticism flom international society. The phenomenon is more evident since the still ongoing second intifada, or uprising, beginning on Sept. 28, 2000. 80 Additionally, treatment of Israeli officials was positively associated with the news themes of Palestinian terrorism and Israeli security. Hypothesis 8 posits staff writers would be more likely than wire-service writers to use the peace process as the dominant news theme. The hypothesis is unsupported (see Table 12a). While the direction of association is as predicted to be positive, the strength of association is, however, too weak, with a beta weight of .019 and a squared part r of less than .001. Binary statistics results also reveal similar insignificance of the influence of news provider on selecting peace process as the dominant news flame. A mean score of 62.91 for the news flame for wire service writers was recorded compared to 64.41 for staff writers. With explanations concerning flequent uses of Palestinian terrorism, it seems suitable to replace the dominant flame of peace process with Palestinian terrorism and examine the feasible impact of the two predictors on uses of Palestinian terrorism flame. Results (see Table 13a, p. 134) flom multiple regression analysis suggest that Palestinian terrorism could be a slightly more suitable dependent variable in a sense. The overall r-squared improves flom .075 to .194. Nonetheless, hypotheses 7 and 8 are not supported with the replacement dominant flame as dependent variable. While the direction of association for both independent variables is positive as predicted, their strengths are not strong enough for their corresponding hypothesis to be supported, using the same criteria in evaluating the peace process. However, Foreign policy as a control variable apparently provided the decisive factor in the choice of this particular news flame; it accounted for 15.3 percent of the variance, with a B-coefficient of .321 and a beta weight of .40. 81 Hypothesis 9 states that the more favorable news treatment Israeli official sources have received, the more likely the story would be more thematic than episodic. The hypothesis (see Table 14a, p. 139) is accepted. The unstandardized regression coefficient for the independent variable of news treatment of Israeli officials is equal to .623, with a constant of —2.419; this independent variable has a wald or t-ratio of 3.332 and an odd ratio of 1.865. The greater the odd ratio, the stronger the predictability of an individual explanatory variable is. Binary statistics obtains similar outcomes. Hypothesis 10 postulates that news stories written by wire service writers would be less thematic than episodic compared to stories written by newspaper staff writers. The hypothesis (see Table 14a) is supported. The independent variable has a positive B- coefficient and very sizeable wald and odd ratios. And indeed, it has the greatest odd ratio compared to the other independent variable and all of the controlling variables used for the logistic regression analysis. Binary statistics also reports similar outcomes. Stories written by wire reporters are considerably more episodic (with a cross-tabs association of 85 percent) than thematic (15 percent). In contrast, stories prepared by staff are slightly less episodic (49) than thematic (51). Hypothesis 11 posits positive correlations between the level of favorable news treatment of Israeli officials and ratios of both causal and treatment responsibility assignment to Israel. The hypothesis (see Table 15a, p. 144) is not supported. While a positive association does exist in causal responsibility assignment, the beta weight and squared part r is very small. Note that the responsibility-assignment ratio (= responsibility assigned to Israel/responsibility to Palestine) is in an opposite direction to the percentage . variable of news treatment of Israeli officials. Thus, for the hypothesis to be supported, 82 the coefficients have to be negative. Binary statistics reveals similar outcomes in terms of direction of association between the variables. Hypothesis 12 assumes that thematic-oriented articles are more likely than episodic-oriented articles to have both causal and treatment responsibility assigned to Palestine. In causal responsibility assignment, the hypothesis (see Table 15a) is unsupported. The independent variable not only has a small beta weight and accounts insignificantly for the contribution of the variance, it has an opposite relationship of association; instead of a positive correlation, a negative association is found. Hypothesis 13 states that the greater the level of the Israeli casualties, the more likely both causal and treatment responsibility would be assigned to Palestine. In causal responsibility assignment, the hypothesis (see Table 15a, p. 144) is supported. The unstandardized regression coefficient is negative, indicating the independent variable is positively associated with the dependent variable because the responsibility assignment ratio is in an opposite direction to the percentage variable of news treatment of Israeli officials. The beta weight and squared part r are very small. This explanatory variable explained 1 percent of total 1.7 percent of the variance of the dependent variable. As treatment responsibility assignment is presented in only two of the four newspapers, multiple regression analysis was not conducted to examine the potential impact of these three independent variables on treatment responsibility assignment. 83 The Path Model Figure l (p. 150) shows the path model of the theoretical framework for the four newspapers as a whole. Only foreign policy and press access, as hypothesized, have a positive impact on giving more preferable news treatment of Israeli officials. Not as predicted, neither news treatment of Israeli officials nor news providers has any influence on selection of Palestinian terrorism as the replacement dominant flame. But both predictors do have an effect on selecting nature of news content. In terms of causal responsibility assignment, only Israeli casualty has an influence on causal responsibility assignment to Palestine, while the other two explanatory variables do not possess similar predictability. In summary, treatment of news sources connects only three variables between international communication and flaming research. That is, press access and foreign policy as contextual variables examined in international communication research could have indirect influences on choice of nature of news content in flaming studies. The more crucial role of news sources, direct or indirect, in assigning responsibility in flaming analysis was lacking. This theoretical path model is thus limited because of the constrained bridging role of (Israeli official) news sources that tie variables between international communication and flaming research. Summary Outcomes flom multiple regression analysis, along with bivariate statistics, suggest that press access (with a beta weight of .244) and foreign policy (.152) are the only two factors that can determine news treatment of Israeli officials. The constant has an unstandardized coefficient of .292. A summary of the hypotheses testing is provided in 84 Table 16 (p. 149). Of the six independent variables, the top two—press access and foreign policy— have their corresponding hypothesis accepted. The results suggest that news treatment of Israeli officials could be more favorable than that of Palestinian officials when news stories were prepared in Israel and when reported during the Bush administration. In the regression analysis of the peace process as dependent variable, hypotheses on the potential impact of percentage of Israeli officials news treatment and news provider were not supported. For news treatment of Israeli officials, the association was in the opposite direction; instead of a positive relationship, a negative one was detected. In regards to the potential effect of news provider, the independent variable has a very small beta weight and squared part r, though with a positive correlations. Moreover, the two hypotheses are not supported, either, with the Palestinian terrorism as replacement dominant news flame and dependent variable. In testing hypotheses 9 & 10 concerning uses of nature of news theme, logistic regression analysis outcomes suggest both the percentage of Israeli official news treatment and news provider have some influence with sizeable wald ratio and odd ratio. As for the causal responsibility assignment as dependent variable, only hypothesis 13 associated with the effect of Israeli casualty was supported. 85 CHAPTER VII DISCUSSION Press Image in the Media War “Who” and “what” in press coverage are very important because these two elements can bring about the press image of concern to those news sources and organizations they represent in the coverage. As pinpointed by Israeli government spokesman Nachman Shai, “The media war is not only the coverage of real war, maybe the real war is the media war” (quoted by Marjorie Miller of The Los Angeles Times, Oct. 16, 2000). As a result, Miller observed: “Both sides want to mobilize their compatriots, to rally national support for their leaders and to sway international public their way. Neither wants to be blamed for the violence, and each wants to make the point that it is the victim of the other’s wrath... That is why the Israel Defense Forces are arming their soldiers with video cameras as well as guns...” Israeli officials, at least during these two years, seem to have an edge over Palestinian officials in the media war battling for more favorable press images. Descriptive statistical results (see Tables 3 through 8) show, in general, Israeli officials received more favorable news treatment—in the “who” component—than their Palestinian counterparts in every way of news sources use. They were used both more prominently and dominantly, getting higher and more quotes and citations with more space. Results concerning press image and sources usage are consistent with those of previous studies (i.e., Daughterty & Warden, 1979; Mousa, 1987; Sylvester, Wu, & 86 Hamilton, 1999; Suleiman, 1974; Trice, 1979). Both reported either a pro-Israeli sources (Mousa, 1987) or Israeli official sources (Sylvester, et al., 1999). But what they said about each other can be more meaningfirl than who said it. In this regard, Israeli officials also seem to have the edge. Most notable is how the press flames the image of the leaders of the two parties. For instance, as data showed in Table 9, the news flame of Palestinian terrorism (with a score of 30.72) by way of suicide bombings was cited more prominently than the news flame of Israeli terrorism (5.06) in the form of target killing in the U.S. news media. In news reports, for instance, Sharon has referred to Arafat as “the snake’s head,” “an arch-terrorist” (The Los Angeles Times, June 28, 2001), “a murderer of women and children” (The LA. Times, Nov. 2, 2000), “the Bin Laden of Palestine” (The Times, Sept. 14, 2001) and “the greatest obstacle to peace and security in the Middle East” (Washington Post, Dec. 20, 2001), and accused . him of being the leader of a “gang of thugs” (The LA. Times, Oct. 12, 2001), whose “fingerprints are all over this [bombing] dripping in blood” (The LA. Times, June 2, 2001). As a result, his credibility is not worth as much as “a Swiss clock” (The Times, Sept, 1, 2001). So too has the Palestinian Authority been criticized for advocating and sponsoring terror (The New York Times, Oct. 18, 2001; The LA. Times, Dec. 7, 2001) and as untrustworthy (The LA. Times, June 28, 2001). While the U.S. press published such harsh criticisms of Arafat, similar verbal attacks on Sharon were not printed. Moreover, data and news reports also reveal that Arafat and the Palestinian Authority were accused of utilizing terrorist attacks as leverage to delay the peace negotiations and to try to steer both domestic and international public opinions to their favor. “He has stopped the security cooperation with Israel, which has increased his 87 popularity, and is perceived to be allowing this intifada to continue... Arafat has deliberately used the violence as a tool to manipulate world opinion and to intemationalize the conflict, seeking to break away from what he saw as the ineffectual structure of the American-brokered peace negotiation” (The Times, Dec. 7, 2000). Consequently, the Israeli government argued that Arafat is not carrying out negotiation in good faith (The Post, July 21, 2000) partly because he has failed to or has been unwilling to tackle Palestinian militant terrorists. One of the most terrifying suicide-bombing attacks occurred on a busy day in Jerusalem, Aug. 9, 2001. The Times described its seriousness: “The timing and location of the bombing was significant: lunch hour at the intersection of J affa and King George Streets. This was equivalent to a midday attack in the heart of Times Square” (Aug. 10, 2001). The press not only specified how prominent terrorists identified by Israeli government were treated while in jail, but went so far as to cite President Bush’s concern that “Palestinian jails. .. are still built with bars in flont with revolving doors at the back” (The LA. Times, Dec. 7, 2001). Obviously, these criticisms of Arafat and the Palestinian Authority would not create preferred press images of them. By acquiring more favorable news treatment, Israeli officials seem to garner more positive press images. With a brief descriptive discussion of the uses of news sources (“who”) and resulting press images (“what”), it is time to turn to the theoretical perspectives of these findings (“why”). 88 Theoretical Framework Construction As the data analysis yielded unpredicted results that fail to support most of the hypotheses and thus the proposed theoretical model, due to much variance across the four newspapers in the tested variables, the researcher decided to conduct additional data analyses for each newspaper. However, statistical outcomes for individual newspapers do not see much improvement except for the Washington Post in building the theoretical model. In this section, the researcher discusses and explains the implication of the model in past research with an effort to integrating outcomes for individual newspapers. International Communication Uses of News Sources as Dependent Variable Hypotheses 1 through 6 have the percentage of news treatment of Israeli officials as the dependent variable to test against six independent variables: foreign policy, press access, news provider, percentage of Israeli casualty, ratio of Israeli population and newsroom resources in term of number of reporters on the assignment. Data support only two hypotheses associated with potential impact of foreign policy and press access. News Sources. The findings in terms of usage of news sources are compatible with those of past studies. First, data show (see Table 3 through 8) that all four newspapers tend to use foreign officials instead of U.S. officials as the major news sources (Bennett, 1990; Berry, 1990; Eilders & Luter, 2000; McLeod & Hertog, 1998; McLeod & Detenber, 1999; Sigal, 1973) even though this dissertation employed a different indicator, a news treatment index other than simple flequency. Across the four newspapers, Israeli officials (Israeli primer and other Israeli officials) and Palestinian 89 officials (Arafat and other Palestinian officials) were used most frequently than any other types of news source. Together, Israeli and Palestinian officials were used more often than all other sources combined, in all sorts of use except for use in the total number of full quote. Moreover, in the breakdown, Israeli officials were used more frequently than Palestinian officials; general Israeli officials were used more often than Israeli prime ministers, as were general Palestinian officials other than Yasser Arafat, and Israeli premiers were used more prominently than Arafat, except in the Chronicle. This is probably due to the impact of accessibility to the news sources. In addition, Palestinian civilians or non-officials were also adopted often, probably because of accessibility as well. Moreover, the wide uses of Palestinian non-officials, alongside Israeli non-officials, have a strong makeup effect that helps offset the greater use of Israeli officials over their Palestinian counterparts. U.S. officials were the groups used the fourth most flequently; it is likely due to its involvement, particularly during the Clinton government. Press Access. As reported in the results chapter, press access and foreign policy are the only two influential factors in deciding on the choices of news sources. International communication research (Chang, Shoemaker, & Brendliger, 1987; Chang & Lee, 1992; Chang & Lee, 2001) has shown that press access has gained its importance in predicting who would get covered. Surveys by Sussman and Guida show Israeli press enjoys greater access than its Palestinian counterpart. One particular occasion could perhaps shed some light on the point: When two Israeli soldiers were lynched by a West Bank mob, Palestinian officials endeavored, but failed, to confiscate all journalists’ fihn recordings of the incident, and as a result, this brutal image invited considerable criticism flom international society. By contrast, when a Palestinian boy was shot to death, the 90 Israeli military did not attempt to cover up the incident by seizing the films (Miller, LA. Times, Oct. 16, 2000). Press access, as shown in Table lla, predicts that either Israeli officials or Palestinian officials would receive better news treatment. The effect of press access appears not only in the newspapers as a whole but also in each individual newspaper (see Table 1 lb through 1 1e, pp. 125-128). Of the four newspapers, the LA. Times seems to be influenced most by this independent variable, which explains 11.2 percent of its variance. Foreign Policy. Furthermore, it is highly documented, while mostly descriptive, that the U.S. press essentially adopts U.S. foreign policy in covering world affairs (i.e., Chang, 1988, 1993; Cohen, 1963; Kim, 2000; Lee & Yang, 1995; McCartney, 1994; Mowlana, 1984; Wall, 1997a, 1997b; Wang & Lowry, 1991). The outcomes here also are in agreement with past research. The impact of foreign policy exists not only in the four newspapers as a whole but also in all (see Table 11b through 1 1e) but the Post, which has a very small B-coefficient, beta weight and squared part r. As the Bush government has been friendlier toward Israel than the Clinton government had been, the U.S. press has offered corresponding favorable news treatment of Israeli authorities. Most evident is that while Clinton invited Arafat to the United States soon after he assumed leadership, Bush has still been reluctant to meet him personally and still brands him a terrorist. The striking impact of foreign policy in the Bush administration is more evident in the case of Houston Chronicle. This daily shifted its attitude toward and treatment of Israeli officials significantly after Bush was sworn into the office. Before, the Chronicle was indeed more favorable toward Palestinian officials. 91 Casualty. Literature in the past tended to find that event-oriented variables have greater impacts than context-oriented, such as press access and foreign policy, on choice of foreign affairs and news sources. While a number of studies (i.e., Belle, 2000; Chang & Lee, 1992; Chang & Lee, 2001) reveal that level of casualty can influence such decisions, results flom multiple regression analyses for this research do not support the hypothesis and those in the past. After controlling for other factors, this event-oriented variable, in terms of percentage of Israeli casualty, has an overall negative, not a positive, relationship with the level of news treatment of Israeli officials. The direction of association was found various across the four newspapers (see Table 11b through lle): A positive relationship was found in The New York Times and The Los Angeles Times, while a negative association was shown in the Post and the Chronicle. This inconsistency may be due to four factors or their combined effects. First, the operational definition of the concept may not be appropriate and has led to measure errors. Second, each newspaper may have reported different attacks or violence in this prolonged conflict. Third, the counting of casualty for both sides could be erroneous by accident. Fourth, the notion of undercounting addressed by Ashley and Olson (1998) could also have been employed. They argue that the press can use the technique to downplay one side while promoting the other. It is shown that when newspapers (the LA. Times and the Times) reported Israeli casualties greater than the average, they were more likely to treat them more favorably and vice versa. Although the empirical results seem to reveal a bias toward Israel, they cannot suggest the undercounting tactic was adopted. News Provider. While past research also shows that organizational variables, notably uses of wire services, could be influential, none of the organizational factors was 92 found important in the present research. First, the results on the effect of news provider— between staff writers and wire services—are incongruent to those in the past (i.e., Fico, 1984). The independent variable is found unimportant across the four newspapers (see Table 11b through lle, pp. 125-128) and can explain no more than .05 percent of the variance for any of the four newspapers. Two explanations are feasible. First, scarce uses of wire-service reporters and stories, due most likely to not employing a more representative sample as done by Lacy and associates (1989), probably may have led to a small variance in the measurement and thus the unimportance of the factor. Also, the distribution of news reports by staff and wire service is skewed greatly toward staff for the three elite newspapers, while it is skewed toward wire service for the Chronicle. Second, the accessibility factor in terms of press access could have crowded out the potential impact of news providers, staff and wire service reporters alike, when facing stringent time constraints. However, noteworthy is that, in the additional bivariate statistics outcomes of individual newspapers, the Chronicle did not echo Sylvester and associates’ (1999) claims that Associated Press articles were more negative toward Israeli officials. Instead, AP-written stories treated Israeli officials more favorably, and it was articles prepared by non-AP wire services that treated Palestinian officials more favorably. Resource Allocation. The potential influence of resource allocation, in terms of number of reporters on assignment, is invisible, and its corresponding hypothesis was not supported. Moreover, extra analyses of individual newspapers (see Table 11b through 1 1e) all of which have the hypothesis rejected show that the LA. Times and the Post that use more reporters have a positive relationship, while the other two employing fewer staff 93 have a negative correlation. Two reasons are plausible for the rejection of the hypothesis: First, it may lie in the smaller variance in the Times and the Chronicle, both of which have either one or two reporters on assignment. It could imply that when only two reporters were working on a same story, the additional one could probably be using non- Israeli official sources more flequently. In the cases of the Post and the LA. Times, both of which occasionally had more than two reporters on assignment, the correlation is either positive or nearly irrelevant. This outcome can be interpreted in two ways: First, with more reporters on assignment, different news sources, including Israeli officials, can obtain extra attention, or second, these extra reporters may simply pay their attention to Israeli officials. Second, using the number of journalists on assignment as an indicator of resource allocation may not be appropriate for the measure because for individual reporters, allocating limited time to different news assignments to meet their daily workload is probably a better indicator for the variable. Compounding this uncertainty is the control for the ratio of Israeli population in the multiple regression analysis. The negative association in the correlations became positive, indicating additional reporter(s) would lead to even more favorable news treatment of Israeli officials if the newspaper ignores the significance of readership of American Jews and American Arabs in the city. Underlying the confusion could also be the potential individual journalist’s characteristics and orientation toward the issue and their personal ties with the news sources, which however is not in the domain of the dissertation. 94 Ethnic Population. As analyzed, the ratio of Jewish population does not have an influence in deciding uses of Israeli official news sources, even though the association is positive as predicted. This finding is compatible with those in the past. For instance, Lacy and associate (1989) didn’t find associations between the proportion of foreign-bom Americans and foreign news coverage. The non-existence of the predictability of the independent variable might imply that while the economic factor is important to the growth of the newspaper, other more crucial political considerations may just be weightier in their decision on selection of news and news sources in a way to fulfill their other goals. No additional analysis is conducted for individual newspapers because of the constant value of the independent variable for each newspaper. Summary. In summary, Israeli officials overall would be treated more favorably with weak effects of context-oriented variables, namely press access and foreign policy that explained very small variance of the dependent variable. That is to say, when stories were datelined in Israel or published during the Bush regime, Israeli officials would receive more preferable news treatment. On the other hand, the event-oriented variable (Israeli casualties) and the organizational variables (news providers, resource allocation and ethnic population) have no impact on choosing and treating news sources of Israeli officials or Palestinian officials. And to reiterate, The New York Times is most favorable toward Israeli officials, followed by The Los Angels Times, the Washington Post; the Houston Chronicle is slightly more favorable to Palestinian officials but changes its policy when Bush becomes president. 95 Framing Analysis Palestinian Terrorism as Dependent Variable Entman (1993, p. 52) defined news flame as “some aspects of a perceived reality” made more salient in “a communication text... to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation.” Thus, there may exist one particular news flame dominating the story. The dominant flame in question is the peace process or negotiation, which was used most prominently in terms of immediacy of the theme appearing the first time in the story. This section discusses the outcomes regarding peace process as the dominant news flame as a result of possible uses of news sources and news providers as the two independent variables. Uses of News Sources. Literature on flaming research suggests that news sources are capable of setting news flames to promote their aims. However, hypothesis 7 regarding the potential impact of Israeli officials on employing the peace process as the dominant flame is not supported. Instead of a positive correlation, a negative association was discovered in the four newspapers as a whole and in each of the four newspapers (see Table 12b through 12e, pp. 130-133). The rejection of the hypothesis probably lies in the failure of not coding ‘tho said what” as one unit; it was dealt as two separate elements— who (the news sources) and what (the news flames). As binary Pearson correlations shows, uses of Israeli officials were positively correlated with the news themes of Palestinian terrorism/terrorist, the hypothesis testing thus was twisted by replacing Palestinian terrorism as the dependent variable. Using Palestinian terrorism in reality seems more appropriate in that it is coded unidirectional; that is, only Israelis or non-Palestinians would use this term to refer to Palestinians, who 96 will not regard and dub themselves as terrorists. By contrast, the peace process is coded multidirectional. First, the peace-process news flame contains both positive and negative meanings and perspectives, that is, promotion or delay of the process. Second, both Israeli and Palestinian officials, as well as non-officials, could have mentioned this term in the news and therefore could confuse the data analysis. Results flom multiple regression analysis also fail to support the hypothesis, for which the Palestinian terrorism flame was dependent variable. Analysis of individual newspapers (see Table 13b through l3e, pp. 135-138), however, shows that the hypothesis is accepted for the Post, with a very weak strength of association between the news treatment of Israeli officials and news flame of Palestinian terrorism. In a way, this finding that arguably slightly supports Tuchman’s claim (1978) that news sources are capable of setting up news flames tends to reveal the existence of large variance among the four newspapers. News Provider. It is well documented that wire service writers tend to cover conflict or disasters in world affairs (Cho & Lacy, 2000; Hart, 1961; Lacy, et al., 1989) because of such organizational constraints as time and accessibility. However, findings do not support hypothesis 8, which is accepted for the Times in analyses of individual newspapers (see Table 12b through 12e). The identical effect of news provider on selecting Palestinian terrorism (see Table 13b through l3e) as the dominant flame also showed only in the Times. The rejection of the hypothesis is most likely due to the measurement error of the dependent variable. Either the peace process or the Palestinian terrorism flame represents only part of news content that can be either conflict or negative. 97 In summary, uses of preferred news sources (Israeli officials) and the type of news provider (staff writers or wire service writers) have no role in determining on Palestinian terrorism as the replaced dominant news flames. However, the former has a vague impact in the Post, and the latter in the Times. Nature of News Content as Dependent Variable According to Iyengar (1991, 1993), an episodic flame is more event-oriented, concrete and descriptive, whereas a thematic flame is more issue-oriented, general and analytical. Literature on uses of news sources and news provider of organizational variables indicates both factors have a role in choosing the nature—episodic or thematic—of news flames. This section discusses the implication of the findings about the two variables in existing theory. Uses of News Source. Framing analysis points out that news sources not only could determine the choice of news flames but also affect the selection of nature of news content. Findings here suggest more favorable news treatment of Israeli officials could have led to significantly more flequent uses of thematic news content in the newspapers as a whole. Similar outcomes flom additional analysis of individual newspapers (see Table 14b through 14c, pp. 140-143) are reported in all newspapers but the Post. The two Times have sizeable wald ratios and odd ratios; the Post has a negative association, with small ratios. The results are in concord with Wall’s (1997a, 1997b) studies that reveal different employment of news sources led to contrast nature of news flame in news coverage of Bosnia and Rwanda. The reason underlying both cases is that when certain news sources, Israeli officials in question, receive better news treatment, they obtain 98 greater opportunities to offer different perspective of the issue, which lead to a more contextual, analytical and thematic news flame. On the other hand, as Palestinian officials are not treated as favorably, they have limited chances to present views other than the most urgent and concrete fact, such as the number of casualty inflicted upon Palestinian civilians flom the battles, and accordingly to a more episodic news flame. In a way, Israeli officials crowd out their Palestinian counterparts in setting the tone and nature of the news. News Provider. Organizational concept suggests that wire service reporters facing considerable deadline pressure, particularly in war zones, would be more constrained by conventional media routines, that is, get the basic information quick and depth later. Adopting concrete and descriptive information as the death toll and how the killing occurs, stories by wire service journalists would naturally become more event-driven and episodic. However, statistically, the impact of who wrote the story—the news provider factor—exists only in the LA. Times, and it is relatively stronger than news treatment of Israeli officials for the newspaper in terms of predictability; news provider has a greater wald ratio (8.039) and odd ratio (24.45). The findings might have suggested that the impact of the independent variable in the LA. Times is so strong that the overall impact is present though with invisible influence in the other three newspapers (see Table 14b through 14e). Therefore, even though the overall outcomes support the hypothesis, the true omnipotent effect of news provider does not exist. One likely source for this confounding result is the control of foreign policy, which seems to be a significant confounding variable in the analysis. It is suggested conservatively that wire service 99 reporters are more likely than staff writers to use episodic frames more frequently than thematic flames. In summary, uses of news sources and of the type of news providers could affect the nature of news frame, either thematic or episodic. And in the current study, more favorable treatment of Israeli officials likely receiving more chances to offer additional contextual information led to more thematically oriented news coverage of the issue. In a more conservative manner, the greater time constraints might force wire service reporters to select news elements that are more specific and lead to more episodically driven news reporting. Responsibility Assignments as Dependent Variable Past research also suggests that uses of news sources, casualty of media routines and nature of news flames can influence how responsibilities are assigned. While the original notion adopted an approach that has responsibility assigned to either individuals or overall society, this dissertation examines flom a different angle and focuses on the assignments between rivals—Israel or Palestine—ignoring the more general system, society or the whole world. Uses of News Sources. Framing scholars, notably Iyengar (1991, 1993) and Entman (1993), argue that news sources can help assign causal and treatment responsibility to certain parties. Results flom multiple regression analysis do not suggest such an impact on causal responsibility assignment. Extra analysis of individual newspaper (see Table 15b through 15e, pp. 145-148) indicates such an effect in the Washington Post, though very weak and vague, that more favorable news treatment of Israeli officials would lead to greater causal treatment responsibility assignment to their 100 Palestinian rivals for the delay of peace negotiation or increased violence in the conflict. The Times and LA. Times have similarly small beta weight and squared part r; the former has a negative association. While the Post has a beta weight of (negative) .100 and a squared part r of .009, the Chronicle has a beta weight of (negative) .109 and a squared part r of .008. The outcomes based on those for Washington Post are only weakly congruent to those in the past studies. Nature of News Content. Iyengar (1991, 1993), Wall (1997a, 1997b) and Mousa (1987) proclaim that episodic flame would help assign at least causal responsibility to individual and thematic flame to society or the world as a whole. Statistical findings do not support this argument. However, additional investigations of individual newspapers (see Table 15b through 15e) find such an influence in The New York Times, which has a visible beta weight and a squared part r of .014 (or 1.4 percent of the variance). It is then arguably that thematically flamed news articles in the Times are more likely than episodically flamed reports to assign causal and treatment responsibilities to Palestine, as suggested by Iyengar, Wall and Mousa. Casualty. While casualty was found here not important in deciding selections of news sources as a result of several factors, it does have an impact on causal responsibility assignments. The results here disclose that Israeli casualty is by and large associated positively with the tendency of assigning causal responsibility to Palestine. In general, the greater Israeli casualty is reported in the news, the more likely Palestinians are assigned causal responsibility, with the Times as a unique case (see Table 15b through lSe). The New York Times not only has a less noticeable beta weight and squared part r, it has a negative association between the independent and dependent variables. One possible 101 explanation is offered here for the presence of a negative sign in the Times: The foreign correspondent at the newspaper could suggest or convey ideas that blame the Israeli government for not efficiently protecting its civilians. And indeed, such orientations do appear in several articles where lawmakers as well as settlers blame Sharon government’s failure to safeguard the life and property of their citizens and let the conflict drag on. Summary From the data analyses and the path models, it is clear that some variations and similarities exist among the four newspapers. First, except for the Washington Post, all other newspapers have both foreign policy and press access as contextual variables for which the corresponding hypothesis was supported; the Post has only press access accepted as hypothesized. The reason foreign policy was not an important predictor was likely that press access (with a squared part r of .091 and a beta weight of .320) explains the bulk of the variance of the dependent variable of news treatment of Israeli officials and crowded out the impact of foreign policy (squared part r= .003 and beta= .054), alongside other predictive factors. Second, the relative strengths between the two important factors are not the same across the four newspapers. While press access is relatively much stronger for the Post and the LA. Times, foreign policy is weightier for the Times and particularly the Chronicle. It thus implies that while the Post is affected mostly by accessibility constraints, the Chronicle is greatly influenced by the U.S. foreign policy steered by the incumbent U.S. president. The greater pressure flom press access felt by a newspaper may lead it to use and assign more reporters to cover the story; this relation is revealed in the Post, 102 which once assigned eight staff writers to cover a single story. Moreover, it is reasonable to suspect foreign policy would have the greatest impact on the relatively smaller newspaper, considering that the circulation size is positively associated with the amount of international news (Lacy et al., 1989; Johnson, 1997). Furthermore, when used as a controlling variable, foreign policy indeed has the strongest impact across all four newspapers, in a positive way, on selecting Palestinian terrorism as the replacement dominant news flame. In this regard, foreign policy seems to have crowded out the potential impact for news treatment of Israeli official news sources for the other three newspapers, for which the difference in the beta weight between the former and the latter is fairly large. In contrast, the disparity is much smaller for the Post, which has the hypothesis accepted. Moreover, the influence of foreign policy also appears in the causal responsibility assignment in the Times and the Chronicle. The major similarity in the separate path models for individual newspapers was the ineffectiveness of news treatment of Israeli officials on causal responsibility assignment, which seems explained by the influence of Israeli casualty for all but the Times. The Washington Post is the only newspaper for which a weak influence of news treatment of Israeli official is present. This responsibility assignment in The New York Times, instead, was decided by the more thematic nature of the content in the newspaper. Of the four publications, The Los Angeles Times model is identical to the overall model where more favorable news treatment of Israeli officials connects three variables between international communication and flaming research. However, the Washington Post could be considered more complete because (more favorable treatment of) Israeli officials can directly assign causal responsibility to Palestine. 103 CHAPTER VIII CONCLUSIONS Conclusions Research on sourcing patterns has persistently revealed that government officials are most likely to be used flequently than other types of news sources (i.e., Sigal, 1973). This observation applies equally well to world news coverage that foreign government officials are not only adopted more flequently than non-officials news sources but also more flequently than U.S. officials. The significance of news sources is detailed in Shoemaker and Reese’s (1996) hierarchy of influence hypothesis, declaring that news sources are extremely influential in news content because journalists depend highly on them for needed information. How news sources are selected is one of the main subjects in communication research in general and international communication research in particular. Domestically, such concepts as media routines (Fishman, 1980; Tuchman, 1978), and uncertainty theory (Dimmick, 1974) and information subsidy (Gandy, 1982) flom organizational perspectives, as well as elite theory (Reese & Danielian, 1994), were proposed or applied to examine its impact. Internationally, theories of international communication world system (Chang, 1998) and contextual vs. event-driven notions (Chang, et al., 1987; Chang & Lee, 2001; Wu, 2000) also were presented. Moreover, the capability of news sources to flame the story is highly documented in flaming research (i.e., Entman, 1993; Gamson, 1989; Gitlin, 1980; Iyengar, 1991, 1993; Pan & Kosicki, 1993; Tuchman, 1978). Tuchman claims that news sources are able 104 to construct social reality through presentation of his highlighted viewpoints. Entman (1993) and Iyengar (1991, 1993), among others, argue that news sources can flame press images by underscoring certain perspectives of the issue while downplaying others and accordingly assign responsibility to a certain group, individual or society as a whole. While these two large blocs of research are related to uses and influences of news sources, they are hardly integrated into a more thorough theoretical flamework to scrutinize the roles of news sources, which was examined as one of the main purposes of this dissertation. Findings flom data analyses seem able to provide some, though limited, insights to existing research on the two areas and its integrity as a whole. Data flom descriptive analyses are in support of outcomes regarding the impact of contextual variables in the past studies of international communication. Not only do foreign government officials predominate over U.S. government officials as news sources, foreign non-official news sources——Palestinian non-authorities in question—— could also be used more prominently than U.S. authorities. Furthermore, Palestinian non- official sources also are highly used to the extent they make up the disparity in the uses of Palestinian official news sources. Past research that examines factors for selections of news and news sources focuses more on cross-nations or cross-issues data. Across nations or issues, varying degrees of effects of contextual variables such as foreign policy or press access, event- driven variables such as casualty or abnormality, and organizational variables such as uses of wire services and ethnic population have been found. However, the crowding-out effects in terms of sourcing also function within issues. 105 The present research discloses that Israeli officials were treated more favorably overall than their Palestinian counterparts in every possible way examined in the news coverage of the conflict. They were quoted and cited more prominently and dominantly with greater immediacy and space. Nonetheless, only foreign policy and press access of contextual variables in international communication research have a predicted positive effect on adoption and treatment of news sources. These outcomes ostensibly echo the arguments and theories in the past. Greater access would allow greater convenience and efficiency for the journalist to collect necessary data. Consequently, concepts of information subsidy (Gandy, 1982) and uncertainty theory (Dimmick, 1974) that are highly illustrated in domestic news reporting can be equally applicable to news coverage of foreign affairs. Press access as a control variable also is effective in choosing the peace process as the dominant news flame. Moreover, reporters in a foreign nation with cozier ties to the United States would become more familiar with local politics and politicians and therefore garner their information more easily. In addition, foreign policy as a controlling variable has impacts on newspapers’ choice of type of news flame of Palestinian terrorism, a more appropriate dependent variable than the peace process, and on causal responsibility assignments. Thus, the impact of foreign policy on the press disclosed in the past studies (i.e., Cohen, 1963; Sigal, 1973) is also revealed in the current research. The small number of newspaper used and the use of unique newspapers may have led to the disappearance of influence in the oft-found significant organizational factors in deciding uses and treatment of Israeli official sources in the study. Other feasible explanations for the failure to find their predictability include inappropriate use of 106 operational definition of some concepts. For instance, the deficient effect in resource allocation is probably due to the inappropriate measure of the concept in terms of number of journalists. That could also imply that the information subsidy or uncertainty theory is at work because the journalist, usually only one, is trying to collect as many documents as possible flom whoever the sources most available to him or her, and that could be one reason that Palestinian non-officials are treated quite nicely. The theoretical importance of news sources in flaming research is quite weak and vague flom the outcomes of the data analyses. In general, neither more favorable treatment of Israeli officials nor uses of staff writer have led to the choice of either peace process or Palestinian terrorism as the dominant news flame. Although additional analysis of individual newspapers does find the impact of better treatment of Israeli officials in the Post, it cannot be argued that such an effect does exist. It is because such separate influence is not only very small but is shown only in one newspaper. Moreover, Israeli officials treated more favorably in the Times are able to indirectly assign causal responsibilities to Palestine, in part by their immediate influences on nature of news content. The evidence flom separate path models, therefore, is conservatively in agreement with the statement of Entman (1993, p. 52): News sources can underscore certain aspects of a perceived reality by making them more prominent in a text to “promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation.” Thus, sources as news flamers could somewhat construct the reality to their particular purposes, as argued by Tuchman (197 8). In terms of the path models based upon the theoretical flamework that integrates concepts of hierarchy of influences, international communication and flaming analysis, 107 (Israeli official) news sources, as measured in the dissertation, play a limited theoretically bridging role between contextual variables in international communication and flaming research. Treatment of news sources is determined by press access and foreign policy, and treatment of news sources weakly influences the choices of dominant news flame, nature of news content or responsibility assignments in different path models. While The Los Angeles Times is identical to the overall model, the Washington Post model seems to be much more useful. Additional Assessment of the Theories Hierarchy of Influence Hypotheses. While Shoemaker and Reese (1996) propose that foreign relations or policy as an ideological variable would have a stronger influence on selection of news sources than does press access as an extra-media variable, data results flom the current research do not support their proposition. Overall and in the LA. Times, press access bears more weights than foreign policy in such uses; foreign policy indeed is unimportant for the Post. And foreign policy appears to be only slightly stronger than press access in the Times; it carries relatively more weights only in the Chronicle. The failure of their hypothesis is possibly due to the fact that their analyses are based dominantly on research on domestic news coverage. As the precarious world society continues to shift, it is difficult to predict precisely the corresponding changes in the relative strengths of predictability of the two factors. Therefore, their proposition remains at best skeptical. The relative strengths in influence of the two variables can possibly be better decided by examining multi-national data, with more refined operational definitions of the two concepts. 108 International Communication World System. Chang’s (1998) world system model seems to be more valid than does Shoemaker and Reese’s, based on the data findings. On one hand, the interaction between foreign policy and press access as two contextual variables does have visible influences on choices of news sources. The two explanatory variables account for the bulk of variance for the four newspapers as a whole and for each individual newspaper. On the other hand, the interaction between event- oriented variables does not yield any impact on the decision of news sources. Uses of news providers and casualty as a media routine variable do not have any impact on such choices. Therefore, based on the findings flom the present research, it appears that the international interaction dimension—foreign policy and press access—can better predict decision on news sources than does the event-oriented dimension with interaction between such variables. However, due to the very small number of newspaper and limited operational definitions of news provider and casualty, it seems dangerous to conclude that the event-oriented dimension is not effective in its predictability. Framing Analysis. The flaming effects of news sources seem inconclusive, and the flaming analysis needs more research upon. The data results do not quite support the argument that sources can highlight a certain news flame to their favor and assign corresponding responsibilities. The reasons likely lie in the difficulty of coding and pairing “who says what” together. This barrier could probably explain why no efforts are attempted to tie type of news flame and responsibility assignments. It seems that flaming researchers tend to confine themselves to investigating more logic relations between news sources uses and elements in flaming analysis, while ignoring more latent factors that can play a role in production of news content. The current research does find and add 109 additional elements to this research area. Notable is that both contextual, organizational and media routine variables can directly influence the assignments of responsibility. Foreign policy, in particular, seems to have an important role in such decisions. Accordingly, this dissertation does provide some insights into and more directions for future flaming analysis. Limitations and Suggestions Constraints of time, budget and individual effectiveness and their interaction effect undoubtedly have limited the scale and overall quality of the research. These negative constraints and influences can be found and addressed in two major aspects. Data Set. Adopting a more conventional way, the researcher selected the three most commonly used newspapers—The New York Times, the Washington Post and The Los Angeles T imes—plus the less often studied Houston Chronicle to fit the special needs of the study. While the selection decision is justifiable due to the constraints, it can be improved. For one, the Chicago Tribune could have been included except for its recent exclusion flom the Nexis electronic database. Second, either the Detroit News or the Detroit Free Press could also be used except for its lack of news stories in the database. The Detroit area has the second-largest Arab-American population in the United States, and inclusion of either Detroit newspaper would have offered more insights into the research. Of course, a random sample is the best option as it can truly represent the genuine population of newspapers. Also, in expanding the scope of the study, other content, namely those in the editorial pages, could have been included. Such additions would have provided different 110 perspectives on the issues regarding the newspapers’ position and coverage of the issue. Furthermore, “if a picture is worth a thousand words, a thousand pictures just might win the war.” It is an intriguing statement that deserves scrutiny. In numerous content analysis projects, photos were either included or were the sole focus. It would serve greatly in supplementing visual images to verbal images to similar research, and thus it is suggested to conduct a separate study concentrating on the pictorial images in the U.S. press and potential effects. Even if a thousand pictures cannot win a war, it may have won the international public opinion. Additionally, the time flame of the study can be longitudinally extended. Instead of investigating only two years, this study could have been traced back to, say 1948, the birth of the Israeli state, and examined the issue at critical points across several transitions of governments. That way, the American Jewish population ratio variable for each individual newspaper can be also improved with more longitudinal census data. Rather than being a constant, this variable would have more variance for each daily examined. Together, with a much broader and complete data, it is suggested that probability sample be used and that more functional sampling methods, notably the constructive weeks, be used. In so doing, generalizability and predictability will be considerably enhanced using existing and additional variables. Variables Operations. The constraints were vividly reflected by the fact that the researcher was advised to simplify the coding protocol for the content analysis. The concepts of prominence and dominance of news sources are not easy to measure, and while the index has captured the most essences of the two interactive notions, it is not 111 nearly perfect. For one, the news headline was left off the analysis. Second, the index could have included the placement of and number of words in the second full quote for each type of news sources. More important, although the index did convey the concept of “who say,” it did not encompass the more valuable component of “what.” Therefore, additional efforts need to be taken to examine what has been said by those news sources and in what circumstances. Furthermore, the operation of news sources is conducted in a broader and more general term. Except for “Israeli premier” and “Yasser Arafat,” which were added later, news sources are examined in terms of types, not individually. In so doing, the chance for generating more specific outcomes is foregone even though it would not affect the purpose of the dissertation. The operational definitions of other concepts, particularly those used as independent variables in testing the news treatment ratio as the dependent variable, are not ideal, either. Both the concepts of foreign policy and press access are more complex than the ones defined and adopted in this study. Lack of more and detailed data forced the researcher to employ more simplified operational definitions of the two concepts. Take press access: While reporters and researchers acknowledge the fact that Israeli media enjoy more fleedom and access than Palestinian press concerning domestic issues, the level of accessibility may not be the same for foreign affairs and correspondents. The simplified definition may not have completely captured the true effect of the concept on the uses of news sources. Moreover, as briefly discussed earlier, the population ratio variable operated as a constant for each individual newspaper even though it has some variances for overall analysis. The problem could have been slightly lessened if yearly or quarterly data for the 112 two examined ethnic groups have existed. Unfortunately, not only has it not, precise figures were difficult to obtain. The census data provide only information on “ancestry”; and instead of the category of “Jewish,” “Israel” is adopted. Accordingly, the number of population reported in the census data and in the external publication between the two categories for cities being examined exists an astronomical discrepancy. Furthermore, even though the American Jews Yearbook is published each year, the related data are not updated. Other deficiencies include the argument by the American Arab Institute that American Arabs are undercounted partly because some of their peers either were not interviewed or did not return the census questionnaires. The institute does not issue any publication addressing the population matter in the United States, either. And as suggested earlier, adoption of more longitudinal data could ease the problem by giving more variance for overall analyses and analyses for each individual newspaper. Also, the non-existing effect of resource allocation is probably due to the far-flom perfect measure. For one, this concept is more complex and broader than merely number of journalists on assignment. For that, it should, if possible, measure the journalist’s time spent on covering the story and on interviewing with certain news sources asking specific and detailed question. It can possibly encompass the personal orientation of the journalist toward the issue and his or her personal relationship with the news sources after being on assignment for a long period of time. 113 Appendix A: Tables and figure TAB LE 1 newspapers Dateline Israel or other Israeli occupied Territories Palestine or Palestinian Territories Other Mid-East Political entities United States All Other World Locations N/A Total NY. Times‘ 81 (46.29) 43 (24.57) (5.14) 28 (16) (5.14) (2.86) 175 WP.2 46 (37.10) 42 (33.87) 2 (1.60) 1 1 (8,87) 6 (4.84) 17 (13.71) 124 Note: 1. The New York Times (n= 175) 2. Washington Post (n= 124) 3. Los Angeles Times (n= 182) 4. Houston Chronicle (n=68) Newspaper LA. Times3 91 (50) 65 (35.71) (1.65) 17 (9.34) (3.30) (0) 182 HQ“ 34 (50) 16 (23.53) 3 (4.41) 8 (11.76) 3 (4.41) 4 (5.88) 68 Frequency and percentage (%) of stories in different datelines in the A115 252 (45.90) 166 (30.27) 17 (3.10) 64 (11.66) 24 (4.37) 26 (4.74) 549 5. The New York Times + Washington Post + Los Angles Times + Houston Chronicle 114 TABLE 2 providers in the newspapers News Provider News staff/ Correspondent Staff/wire service Associated Press Combined U.S./ Foreign Wire Services Other U.S. Newspapers (Staff) (Wire services) Total NY. Times 168 (96) (1.14) (.57) (2.28) (0) 168 (2.86) 175 WP. 119 (95.98) 0 (0) 3 (2.42) 2 (1.61) (0) 119 (4.03) 124 115 Newspaper LA. Times 171 (93.96) 1 (.55) 4 (2.20) 6 (3.30) (0) 171 10 (5.49) 182 H.C. 17 (25) 0 (0) 37 (54.41) 11 (16.18) (4.41) 17 48 (70.59) 68 Frequency and percentage (%) of stories prepared by different news All 475 (86.52) 3 (.55) 45 (8.20) 23 (4.19) 3 (.55) 475 67 (12.20) 549 'TAdBLIE3 story News Source Israeli Premier (Barak or Sharon) Other Israeli Authority Israeli Non-authority Yasser Arafat Other Palestinian Authority Palestinian Non-authority American Authority American Non-authority Non-Palestinian Arab authority Others Israeli authority Palestinian authority Total Israelis Total Palestinians NY. Times .0286 .1029 .0057 .0171 .0800 .0114 .0914 .0057 .0057 .0171 .1315 .0971 .1372 .1085 WP. .0242 .0887 .0081 .0161 .0726 .0161 .1290 0.00 .0161 .0726 .1129 .0887 .1210 .1048 Newspaper 116 LA. Times .0440 .1484 .0055 .0275 .0769 .0385 .0440 .0055 .0055 .0110 .1924 .1044 .1979 .1429 H.C. .0147 .0735 0.00 .0294 .0882 .0441 .1176 0.00 0.00 .0147 .0882 .1176 .0882 .1617 Mean score of each of the following news sources used in news lead in the All .0310 .1111 .0055 .0219 .0783 .0255 .0874 .0036 .0073 .0273 .1421 .1002 .1467 .1257 TABLE 4 used in terms of mean percentile of order in the story News Source Israeli Premier Other Israeli Authority Israeli Non-authority Yasser Arafat Other Palestinian Authority Palestinian Non-authority American Authority American Non-authority Non-Palestinian Arab authority Others Israeli authority Palestinian authority Total Israelis Total Palestinians NY. Times 9.92 28.30 8.93 6.14 21.81 23.02 14.50 3.62 2.86 13.10 38.22 27.95 47.15 49.76 WP. 9.90 26.82 10.05 4.23 27.87 30.18 13.98 4.98 3.25 13.42 36.72 32.10 46.77 62.28 Newspaper 117 LA. Times 11.86 26.35 9.21 9.19 21.07 24.96 15.18 3.79 1.70 8.36 38.21 30.26 47.42 55.22 H.C. 7.21 21.54 3.68 12.25 20.03 16.10 14.59 .39 1.32 7.40 28.75 32.28 32.43 52.31 Placement of the first full quote of each of the following news sources All 10.22 26.48 8.63 7.47 22.71 24.42 14.62 3.59 2.37 11.09 36.70 30.18 45.33 54.60 TABLE 5 the following news sources in the story News Source Israeli Premier Other Israeli Authority Israeli Non-authority Yasser Arafat Other Palestinian Authority Palestinian Non-authority American Authority American Non-authority Non-Palestinian Arab authority Others Israeli authority Palestinian authority Total Israelis Total Palestinians NY. Times 4.08 12.1257 5.76 2.3486 8.5143 6.6571 5.4914 1.6914 1.12 4.7314 16.2057 10.8629 21.9657 17.52 Newspaper WP. LA. Times 3.6855 4.9725 13.7177 13.8187 5.0968 5.6099 1.7661 3.5879 13.7823 9.5165 10.7823 9.9541 6.3790 5.5549 2.1371 2.4176 3.00 .6868 6.6774 5.8187 17.4032 18.7912 15.5484 13.1044 22.50 24.4011 26.3307 23.0585 118 H.C. 2.6029 8.3676 1.5294 5.1324 8.2206 4.7647 6.0588 .6765 .8088 3.7794 10.9705 13.3530 12.4999 18.1177 Mean score of the number of words used in the first quote given to each of All 4.1038 12.581 5.0364 2.9727 10.00 8.4291 5.7832 1.9071 1.3625 5.4135 16.685 12.973 21.721 21.402 TABLE 6 Mean score of the number of full quotes of each of the following news sources in the story Newspaper News Source NY. Times WP. LA. Times H.C. All Israeli Premier .2629 .2177 .2912 .1471 .2477 Other Israeli 1.16 1.00 1.0275 .5882 1.0091 Authority Israeli .4286 .5242 .5440 .0882 .4463 Non-authority Yasser Arafat .1886 .0806 .1758 .2941 .1730 Other Palestinian .8457 .9919 .6648 .6176 .7905 Authority Palestinian .9886 1.5484 1.4560 .5294 .7905 Non-authority American .4686 .5403 .51 10 .50 .5027 Authority American .1029 .2016 . 1429 .0441 .0311 Non-authority Non-Palestinian .1543 .0968 .0934 .0294 .1056 Arab authority Others .3600 .4597 .3187 .2059 .3497 Israeli authority 1.4229 1.2177 1.3187 .7353 1.2568 Palestinian authority 1.0343 1.0725 .8406 .9117 .9635 Total Israelis 1.8515 1.7419 1.8627 .8235 1.7031 Total Palestinians 2.0229 2.6209 2.2966 1 .441 1 1.7540 119 TABLE 7 or paraphrased in the story News Source Israeli Premier Other Israeli Authority Israeli Non-authority Yasser Arafat Other Palestinian Authority Palestinian Non-authority American Authority American Non-authority Non-Palestinian Arab authority Others Israeli authority Palestinian authority Total Israelis Total Palestinians NY. Times .7600 4.7543 .9429 .6229 3.2114 2.7543 2.1086 .3543 .4253 1.4286 5.5143 3.8343 6.4572 6.5886 WP. 1.0806 3.8710 .1129 .4677 2.8226 3.1452 2.2339 .4355 .3387 1.6129 4.9516 3 .2903 5.0645 6.4355 120 Newspaper LA. Times 1.1044 5.3956 1.4286 .6648 2.9341 3.8736 1.6209 .2912 .1538 1.4231 6.5000 3.5989 7.9286 7.4725 H.C. .4853 3.4559 .2353 1.0441 3.1029 1.9118 2.3824 .1029 .1765 .9265 3.9412 4.1470 4.1765 6.0588 Mean score of the frequency of each of the following news sources quoted All .9126 4.6066 1.0546 .6539 3.0182 3.1093 2.0091 .3206 .2847 1.4062 5.5192 3.6721 6.5738 6.7814 TABLE 8 News Source Israeli Premier Other Israeli Authority Israeli Non-authority Yasser Arafat Other Palestinian Authority Palestinian Non-authority American Authority American Non-authority Non-Palestinian Arab authority Others Israeli authority Palestinian authority Percentage of Israeli officials Total Israelis Total Palestinians Percentage of Total Israelis NY. Times 5.2307 18.4228 7.2265 3.2385 12.8695 10.6414 8.3010 2.1905 1.7415 6.6681 23.654 16.108 . 6003 30.880 26.750 .5358 Newspaper WP. LA. Times 5.1070 6.5307 18.9456 20.6505 6.8424 7.68 2.3729 4.5479 17.9480 13.40 15.7937 16.9968 9.4420 7.8826 2.8240 2.8951 3.4841 .9565 8.9568 7.6585 24.052 27.181 20.321 17.951 .5480 .5958 30.894 34.861 36.115 34.948 .4610 .4994 121 Indexed score for the following news sources in the story H.C. 3.3221 12.7007 1.8897 6.6225 12.2297 7.4110 9.2048 .8275 1.0279 5.0005 16.023 18.852 .4947 17.913 26.263 .4055 All 5.3973 18.571 6.6291 3.8962 14.114 13.512 8.5274 2.3983 1.7865 7.3068 23.968 18.010 .5730 30.597 31.522 .4926 TABLE 9 terms of percentile of order News Theme Israeli sovereignty/ Security Israeli religious/ Right-wing tenacity Israeli terrorists/ Terrorism Palestinian terrorist/ Terrorism Palestinian (Muslim) Religious/right-wing Tenacity Palestinian security/ Independence Movement Peace process/ Negotiation World terrorism World peace American interest Others NY. Times 16.80 .83 5.06 30.62 1.87 27.22 67.49 10.82 .16 .27 22.48 WP. 14.38 0.00 9.96 35.96 .47 32.20 61.60 14.90 0.00 .60 37.73 Newspaper 122 LA. Times 9.63 1.17 13.92 43.62 6.16 20.59 61.14 7.05 .76 .73 44.33 H.C. 11.41 0.00 19.30 39.52 7.32 22.08 69.30 9.38 .98 .11 35.21 Mean score of the immediacy of news themes mentioned in the story in All 13.21 .65 10.87 37.24 3.65 25.51 64.28 10.31 .42 .48 34.75 TABLE 10 Mean score of the scale" of casual and treatment responsibility assignments to Israel or Palestine in the story Newspaper NY. Times WP. LA. Times H.C. All Responsibility Assignment__ Causal Responsibility .9771 1.0081 1.0275 1.0147 1.0055 Assigned to Israel Causal Responsibility 1.0114 1.0161 1.0220 1.0441 1.02 Assigned to Palestine Treatment Responsibility 1.00 1.00 1 .00 1 .00 1 .00 Assigned to Israel Treatment Responsibility 1.0057 1 .00 1.00 1.0147 1.0036 Assigned to Palestine Ratio of Causal Responsibility .9877 .9964 l .0020 .9871 .9943 Assigned to Israel Ratio of Treatment Responsibility .9994 1 .00 1 .00 .9985 .9996 Assigned to Israel Note 6: Responsibility assigned to either Israel or Palestine by Palestine/Israel, United States, the Arabic world, the Western society, and United Nations, or None/NA. 123 TABLE 1 la Outcomes of multiple regression analysis of percentage of Israeli officials news treatment against the independent variables of press access, foreign policy, news provider, ratios of American Jews to American Arabs population, percentage of Israeli casualty, and number of reporters on assignment, in the four newspapers as a whole Independent Variable Press Access (Israel) Foreign Policy (Bush) News Provider (Staff writers) Ratio of American Jews Percentage of Israeli casualty Number of Writers Constant DF= 400 R2= .089 B- Coefficient .172 .106 .083 .003 -.O44 .007 .292 SE. .034 .034 .055 .002 .069 .033 .082 124 Beta .244 .152 .081 .071 -.031 .011 Squared Part r .059 .023 .005 .004 .001 <.001 P-value <.001 .001 .066 .088 .261 .412 TABLE 11b Outcomes of multiple regression analysis of percentage of Israeli officials news treatment against the independent variables of press access, foreign policy, news provider, ratios of American Jews to American Arabs population, percentage of Israeli casualty, and number of reporters on assignment, in The New York Times Independent Variables Press Access (Israel) Foreign Policy (Bush) News Provider (Staff writers) Percentage of Israeli casualty Number of Writers Constant DF= 118 R2: .038 B- Coefficient .098 .102 .040 .009 -.006 .446 SE. .069 .065 .170 .140 .134 .236 125 Beta .134 .146 .022 .006 -.004 Squared P-value Part r .017 .080 .021 .060 .001 .407 <.001 .474 <.001 .482 TABLE 110 Outcomes of multiple regression analysis of percentage of Israeli officials news treatment against the independent variables of press access, foreign policy, news provider, ratios of American Jews to American Arabs population, percentage of Israeli casualty, and number of reporters on assignment, in the Washington Post Independent Variables Press Access (Israel) Foreign Policy (Bush) News Provider (Staff writers) Percentage of Israeli casualty Number of Writers Constant DF= 85 R2: .097 B- Coefficient .223 .038 .115 -.233 .019 .381 SE. .078 .075 .191 .169 .078 .223 126 Beta .320 .054 .065 -.155 .025 Squared Partr .091 .003 .004 .022 .001 P-value .003 .307 .275 .082 .407 TABLE 11d Outcomes of multiple regression analysis of percentage of Israeli officials news treatment against the independent variables of press access, foreign policy, news provider, ratios of American Jews to American Arabs population, percentage of Israeli casualty, and number of reporters on assignment, in The Los Angeles Times B- S.E. Beta Squared P-value Independent Coefficient Part r Variables Press Access .232 .054 .336 .112 <.001 (Israel) Foreign Policy .108 .056 .154 .023 .028 (Bush) News Provider .081 .112 .057 .003 .236 (Staff writers) Percentage of .083 .106 .061 .004 .220 Israeli casualty Number of .018 .042 .033 .001 .338 Writers Constant .25 7 . 140 DF= 145 R2: .137 127 TABLE 1 le Outcomes of multiple regression analysis of percentage of Israeli officials news treatment against the independent variables of press access, foreign policy, news provider, ratios of American Jews to American Arabs population, percentage of Israeli casualty, and number of reporters on assignment, in the Houston Chronicle B- S.E. Beta Squared P-value Independent Coefficient Part r Variables Press Access .109 .097 .157 .024 .133 (Israel) Foreign Policy .266 .102 .365 .128 .007 (Bush) News Provider .055 .112 .074 .005 .313 (Staff writers) Percentage of -.204 .169 -.l70 .028 .1 16 Israeli casualty Number of -.061 .092 -.097 .008 .257 Writers Constant .3 77 . 190 DF= 48 R2: .187 128 TABLE 12a Outcomes of multiple regression analysis of uses of peace process as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the four newspapers as a whole Independent Variables Percentage of Israeli officials News treatment News Provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Ratio of American Jews Percentage of Israeli casualty Number of Writers Constant DF= 400 R2: .075 B- Coefficient -.l46 .022 .158 -.O97 .0006 .024 .067 .564 SE. .057 .063 .040 .039 .002 .079 .038 .095 Beta -.129 .019 .199 -.122 .013 .015 .088 Squared Part r .015 <.001 .036 .014 <.001 <.OOl .007 P-value .006 .361 <.001 .007 .405 .380 .038 TABLE 12b Outcomes of multiple regression analysis of uses of peace process as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in The New York Times Independent Variables Percentage of Israeli officials News treatment News Provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Percentage of Israeli casualty Number of Writers Constant DF= 118 R2: .067 B- Coefficient -.080 .208 .110 -.O47 .053 .246 .191 SE. .102 .185 .076 .072 .153 .145 .261 130 Beta -.O72 .106 .136 -.O6l .032 .155 Squared Part r .005 .010 .017 .004 .001 .024 P-value .219 .133 .076 .257 .365 .047 TABLE 12c Outcomes of multiple regression analysis of uses of peace process as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the Washington Post B- S.E. Beta Squared P-value Independent Coefficient Part r Variables Percentage of Israeli officials -.258 .127 -.223 .045 .023 News treatment News Provider -.018 .218 -.009 <.001 .467 (Staff writers) Press Access .198 .094 .246 .049 .019 (Israel) Foreign Policy -.171 .084 -.212 .044 .025 (Bush) Percentage of -.008 .194 -.004 <.001 .484 Israeli casualty Number of .027 .089 .031 .001 .383 Writers Constant .736 .258 DF= 85 R2: .137 131 TABLE 12d Outcomes of multiple regression analysis of uses of peace process as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in The Los Angeles Times B- S.E. Beta Independent Coefficient Variables Percentage of Israeli officials -. 193 .100 -.166 News treatment News Provider .092 .133 .056 (Staff writers) Press Access .183 .068 .230 (Israel) Foreign Policy -.118 .067 -.146 (Bush) Percentage of -.023 .126 -.015 Israeli casualty Number of .068 .050 .111 Writers Constant .522 .168 DF= 145 R2: .106 132 Squared Part r .024 .003 .047 .020 .001 .012 P-value .028 .246 .004 .040 .429 .088 TABLE 12e Outcomes of multiple regression analysis of uses of peace process as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the Houston Chronicle B- S.E. Beta Squared P-value Independent Coefficient Part r Variables Percentage of Israeli officials News Treatment News provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Percentage of Israeli casualty Number of Writers Constant DF= 48 R2: .047 .001 .087 .083 -.015 .117 .105 .431 .184 .136 .119 .132 .207 .112 .240 133 .001 .106 .108 -.018 .089 .153 <.001 .009 .011 <.001 .007 .020 .497 .257 .245 .457 .287 .176 TABLE 13a Outcomes of multiple regression analysis of uses of Palestinian terrorism as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the four newspapers as a whole Independent Variables Percentage of Israeli officials News treatment News Provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Ratio of American Jews Percentage of Israeli casualty Number of Writers Constant DF= 400 R2: .194 B- Coefficient .063 .032 -.103 .321 -.001 .115 .070 .061 SE. .054 .059 .038 .037 .002 .074 .036 .090 Beta .055 .027 -.128 .400 -.024 .071 .095 Squared Partr .003 .001 .015 .153 <.001 .005 .009 P-value .123 .293 .004 <.OOl .318 .062 .019 TABLE 13b Outcomes of multiple regression analysis of uses of Palestinian terrorism as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in The New York Times Independent Variables Percentage of Israeli officials News treatment News Provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Percentage of Israeli casualty Number of Writers Constant DF= 118 R2: .339 B- Coefficient -.080 .438 -.270 .313 .204 -.225 .090 SE. .085 .154 .063 .060 .128 .121 .218 135 Beta -.073 .225 -.337 .410 .124 -.143 Squared Partr .005 .048 .107 .161 .015 .030 P-value .196 .003 <.001 <.001 .056 .033 TABLE 13c Outcomes of multiple regression analysis of uses of Palestinian terrorism as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the Washington Post B- Independent Variables Percentage of Israeli officials .1 97 News treatment News Provider .030 (Staff writers) Press Access -.156 (Israel) Foreign Policy .311 (Bush) Percentage of .020 Israeli casualty Number of .098 Writers Constant .012 DF= 85 R2= .248 Coefficient SE. .114 .196 .084 .077 .174 .080 .232 136 Beta .177 .015 -.202 .400 .012 .120 Squared Partr .028 <.001 .033 .156 <.001 .014 P-value .044 .439 .034 <.001 .405 .112 TABLE 13d Outcomes of multiple regression analysis of uses of Palestinian terrorism as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in The Los Angeles Times Independent Variables Percentage of Israeli officials News treatment News Provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Percentage of Israeli casualty Number of Writers Constant DF= 145 R2= .178 B- Coefficient .114 -.046 .012 .320 .072 .104 .032 SE. .097 .130 .066 .065 .123 .049 .163 137 Beta .097 -.027 .015 .389 .045 .167 Squared Part r .008 .001 <.001 .142 .002 .027 P-value .122 .362 .426 <.001 .281 .018 TABLE l3e Outcomes of multiple regression analysis of uses of Palestinian terrorism as the dependent variable against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the Houston Chronicle B- S.E. Beta Squared P-value Independent Coefficient Part r Variables Percentage of Israeli officials -.121 .185 -.097 .008 .260 News Treatment News provider -.1 15 .137 -.124 .013 .203 (Staff writers) Press Access -.049 .120 -.057 .003 .342 (Israel) Foreign Policy .426 .133 .471 .184 .001 (Bush) Percentage of .176 .209 .1 18 .013 .202 Israeli casualty Number of .072 .1 13 -.O93 .007 .263 Writers Constant .248 .242 DF= 48 R2: .238 138 TABLE 14a Outcomes of logistic regression analysis of nature of news as dependent variables against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the four newspapers as a whole Independent Variables Percentage of Israeli officials News treatment News Provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Ratio of American Jews Percentage of Israeli casualty Number of Writers Constant N= 549; DF= 7 R2: .129 B- Coefficient .623 1.932 .078 -.324 -.006 .103 .356 -2.419 SE. .342 .434 .231 .229 .015 .417 .227 .617 139 Wald 3.332 19.79 .115 2.005 .145 .061 2.449 Odd Ratio 1.865 6.902 1.081 .723 .994 1.109 1.427 P-value .034 <.001 .368 .352 .352 .402 .059 TABLE 14b Outcomes of logistic regression analysis of nature of news as dependent variables against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in The New York Times Independent Variables Percentage of Israeli officials News Treatment News provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Percentage of Israeli casualty Number of Writers Constant N=111;DF=8 R2=.108 B- S.E. Coefficient .925 .658 6.96 15.865 .315 .440 .308 .411 -.743 .953 .294 15.929 -8.335 -15.929 140 Wald 1.973 .193 .512 .560 .895 .095 Odd Ratio 2.521 1054 1.371 1.360 .476 1.342 P-value .080 .331 .237 .227 .172 .379 TABLE 14c Outcomes of logistic regression analysis of nature of news as dependent variables against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the Washington Post Independent Variables Percentage of Israeli officials News Treatment News provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Percentage of Israeli casualty Number of Writers Constant N= 79; DF= 8 R2: .097 B- SE. Coefficient -.792 .742 1.113 1.219 .607 .515 -.811 .482 -.336 .982 -.207 .503 -.257 1.415 141 Wald 1.140 .834 1.387 2.834 .117 .169 Odd Ratio .453 3.044 1.835 .444 .715 .813 P-value .143 .181 .120 .046 .366 .341 TABLE 14d Outcomes of logistic regression analysis of nature of news as dependent variables against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in The Los Angeles Times Independent Variables Percentage of Israeli officials News Treatment News provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Percentage of Israeli casualty Number of Writers Constant N= 138; DF= 8 R2: .238 B- Coefficient 1.134 3.197 -.120 -1.302 .627 .104 -2.464 SE. .623 1.127 .432 .480 .741 .323 1.328 142 Wald 3.316 8.039 .077 7.366 .716 .104 Odd Ratio 3.109 24.45 .887 .272 1.872 1.110 P-value .035 .003 .391 .004 .194 .374 TABLE 14e Outcomes of logistic regression analysis of nature of news as dependent variables against percentage of Israeli officials news treatment and news provider as independent variables, controlling for other variables, in the Houston Chronicle Independent Variables Percentage of Israeli officials News Treatment News provider (Staff writers) Press Access (Israel) Foreign Policy (Bush) Percentage of Israeli casualty Number of Writers Constant N= 46; DF= 7 R2= .143 B- Coefficient 2.044 .497 -1.021 .327 1.504 .793 -4.234 S.E. 1.584 1.036 .847 1.265 1.487 .879 2.311 143 Wald 1.665 .230 1.454 .067 1.024 .815 Odd Ratio 7.719 .1643 .360 1.387 4.502 2.210 P-value .099 .316 .114 .398 .116 .184 TABLE 15a Outcomes of multiple regression analysis of causal responsibility assignment to Israel against percentage of Israeli officials news treatment, percentage of Israeli casualty, and nature of news as independent variables, controlling for other variables, in the four newspapers as a whole Independent Variables Percentage of Israeli officials News treatment Percentage of Israeli casualty Nature of News theme Press Access (Israel) Foreign Policy (Bush) News Provider (Staff writers) Ratio of American Jews Number of Writers Constant DF= 391 R2: .017 B- Coefficient -.009 -.031 .0004 -.001 -.009 .003 -.005 .0004 1.011 SE. .012 .016 .008 .008 .008 .013 <.001 .008 .019 144 Beta -.042 -.101 .003 -.007 -.059 .016 -.004 .029 Squared Part r .002 .010 <.001 <.001 .003 <.001 <.001 .001 P-value .217 .025 .478 .446 .131 .394 .473 .286 TABLE 15b Outcomes of multiple regression analysis of causal responsibility assignment to Israel against percentage of Israeli officials news treatment, percentage of Israeli casualty, and nature of news as independent variables, controlling for other variables, in The New York Times Independent Variables Percentage of Israeli officials News treatment Percentage of Israeli casualty Nature of News theme Press Access (Israel) Foreign Policy (Bush) News Provider (Staff writers) Number of Writers Constant DF= 115 R2: .054 B- Coefficient .009 .035 -.018 .18 -.15 -.022 .014 .976 SE. .020 .029 .014 .015 .014 .036 .028 .050 Beta .042 .112 -.120 .120 -.105 -.060 .049 Squared Partr .002 .013 .014 .013 .010 .003 .002 P-value .332 .118 .107 .109 .139 .272 .303 TABLE 15c Outcomes of multiple regression analysis of causal responsibility assignment to Israel against percentage of Israeli officials news treatment, percentage of Israeli casualty, and nature of news as independent variables, controlling for other variables, in the Washington Post Independent Variables Percentage of Israeli officials News treatment Percentage of Israeli casualty Nature of News theme Press Access (Israel) Foreign Policy (Bush) News Provider (Staff writers) Number of Writers Constant DF= 80 R2= .073 B- S.E. Beta Coefficient -.019 .023 -.100 -.O36 .034 -.127 -.012 .016 -.088 .004 .017 .028 .007 .015 .054 .005 .039 .015 -.028 .016 -.205 1.050 .046 146 Squared Partr .009 .014 .007 .001 .003 <.001 .042 P-value .202 .147 .228 .414 .321 .449 .038 TABLE 15d Outcomes of multiple regression analysis of causal responsibility assignment to Israel against percentage of Israeli officials news treatment, percentage of Israeli casualty, and nature of news as independent variables, controlling for other variables, in The Los Angeles Times Independent Variables Percentage of Israeli officials News treatment Percentage of Israeli casualty Nature of News theme Press Access (Israel) Foreign Policy (Bush) News Provider (Staff writers) Number of Writers Constant DF= 145 R2= .055 B- S.E. Beta Coefficient -.015 .019 -.071 -.060 .023 -.216 .007 .013 .050 .003 .012 .025 -.005 .012 -.033 .0005 .025 .002 .002 .009 .016 1.030 .030 147 Squared Part r .004 .046 .002 .001 .001 <00] <00] P-value .216 .006 .291 .391 .352 .493 .425 TABLE 15e Outcomes of multiple regression analysis of causal responsibility assignment to Israel against percentage of Israeli officials news treatment, percentage of Israeli casualty, and nature of news as independent variables, controlling for other variables, in the Houston Chronicle B- S.E. Beta Squared P-value Independent Coefficient Part r Variables Percentage of Israeli officials -.O34 .053 -.109 .008 .264 News treatment Percentage of -.094 .057 -.249 .055 .056 Israeli casualty Nature of .040 .039 .166 .022 .155 News theme Press Access -.032 .033 -.145 .020 .169 (Israel) Foreign Policy -.O44 .037 -.195 .030 .117 (Bush) News Provider -.010 .037 -.043 .002 .394 (Staff writers) Number of .037 .031 .189 .003 .116 Writers Constant 1.033 .065 DF= 46 R2: .199 148 TABLE 16 Summary of the hypotheses testing for each newspaper and for overall Newspaper Hypothesis N. Y. Times W.P. LA. Times H. C. Overall H1 Yes No Yes Yes Yes H2 Yes Yes Yes Yes Yes H3 No No No No No H4 -- -- -- -- N0 H5 No No No No No H6 No No No No No H7a No No No No No H7b No No No No No H8a Yes Yes No No No H8b Yes No No No No H9 Yes No Yes Yes Yes H10 No No Yes No Yes H1 1a No Yes No No No H1 lb No -- -- No -- H12a Yes No No No No H12b Yes -- -- No _- H13a No Yes Yes Yes Yes H13b Yes -- -- No -- 149 AEmtobog adage—«B 255 man: no out? 80.58 3650 segue—am 3 9530M Aosmfionb 1 8988 3050 :25“ «COHGOO maoc MO Ogflz MNO. "Q ”o: MO wfiDEHNOHw W302 N . . sq Po .6 228$ 9 62585 a. 3.0488 x. “58%me tam—Emmcoqgm VVN. ”Q “or: £380 mBoH 505:2. mo ..\o 5:32? 8508M CREE 8235 96 Z scams @8on 32¢ 33$ mason ammouom mama—«Quin: :29»: you #33089... 1360.82: 95 no $62: 55— : «...-w:— 150 Appendix B: Summary of the 13 hypotheses H1: The difference between favorable treatment of Israeli and Palestinian official news sources was greater during the period when U.S. foreign policy was more favorable toward Israel. H2: As press access in Israel is greater than that in Palestine, Israeli official news sources would receive more favorable news treatment than Palestinian official news sources. H3: Wire service reporters would be less likely than newspaper staff writers to give favorable news treatment to Israeli official sources than Palestinian official sources. H4: If the proportion of American Jewish population is greater than that of American Arabs in the newspaper readership radius, then Israeli official news sources would receive more favorable news treatment than Palestinian official news sources. H5: The more journalists covering a story, the more likely Israeli official sources would receive more favorable news treatment than Palestinian official sources. H6: The greater the level of casualties occurring to Israelis, the more likely Israeli official sources would receive more favorable news treatment than Palestinian official SOUICCS. H7: The more favorable news treatment Israeli official sources have received, the more likely the peace process would be the dominant news theme. H8: Wire service reporters would be less likely than newspaper staff writers to use peace process as the dominant news theme. H9: The more favorable news treatment Israeli official sources have received, the more likely the story would be more thematic than episodic. H10: News stories written by wire service reporters would be less thematic than episodic compared to stories written by newspaper staff writers. H11: The more favorable news treatment Israeli official sources have received, the more likely both causal and treatment responsibility would be assigned to Palestine. H12. The more thematic the news themes are, the more likely both causal and treatment responsibility would be assigned to Palestine. H13: The greater the level of the Israeli casualty is, the more likely both causal and treatment responsibility would be assigned to Palestine. 151 Appendix C: Coding protocol of the content analysis 1. Story ID: __ 2. Date: (month/day/year) 3. Newspaper ID: (1) NYT (2) Washington Post (3) LA. Times .(4) Houston Chronicle 4. Total number of paragraphs: 5. Length of the news story: (words) 6. Section: (1) Front page (2) National page (3) International front page (4) International inside page 7. Dateline: (1) Israel or Israeli occupied territory (2) Palestine or Palestinian territories (3) Other Middle-East political entities (i.e., Jordan, Egypt, or Lebanon) (4) United States (5) United Nations (6) European countries (7) Other locations of the world (99) N/A 8. News providers: (1) News staff/correspondent (2) Staff/wire service (3) Associated Press (4) Other U.S. news wire services (5) Foreign news wire services (6) Combined wire services (7) Other U.S. newspapers (8) Israeli non-wire news media (9) Palestinian non-wire news media (10) Others (99) N/A 9. Last name of the reporting journalist(s): 10. Number of journalist(s) writing for the story: ** [New sources are defined as people, agents, organizations or governments quoted or paraphrased in the news stories. A given source cannot be counted only once if that source has been quoted or paraphrased several times] * Frequency of each of the following news sources used in the news lead (V. l 1~20): 11. Israeli Prime Minister: 12. Other Israeli authority: 13. Israeli non-authority: l4. Yasser Arafat: 15. Other Palestinian authority: 16. Palestinian non-authority: 17. American authority: 18. American non-authority: 152 l9. Non-Palestinian Arab nations: 20. Others: * Placement of the first full quote of each of the following news sources (V. 21~30): 21. Israeli Prime Minister: 22. Other Israeli authority: 23. Israeli non-authority: 24. Yasser Arafat: 25. Other Palestinian authority: 26. Palestinian non-authority: 27. American authority: 28: American non-authority: 29. Non-Palestinian Arab nations: 30. Others: * Number of words used in the first full quote given to each of the following news sources (V. 31~40): 31. Israeli Prime Minister: 32. Other Israeli authority: 33. Israeli non-authority: 34. Yasser Arafat: 35. Other Palestinian authority: 36. Palestinian non-authority: 37. American authority: 38. American non-authority: 39. Non-Palestinian Arab nations: 40. Others: * Number of full quotes of each of following news sources (V. 4l~50): 41. Israeli Prime Minister: 42. Other Israeli authority: 43. Israeli non-authority: 44. Yasser Arafat: 45. Other Palestinian authority: 46. Palestinian non-authority: 47. American authority: 48: American non-authority: 49. Non-Palestinian Arab nations: 50. Others: 153 * Frequency of each of the following news sources quoted or paraphrased (V. 51~60): 51. Israeli Prime Minister: 52. Other Israeli authority: 53. Israeli non-authority: 54. Yasser Arafat: 55. Other Palestinian authority: 56. Palestinian non-authority: 57. American authority: 58. American non-authority: 59. Non-Palestinian Arab nations: 60. Others: 61/62. Number of Israeli casualty the day before: (killed)/ (wounded) 63/64. Number of Palestinian casualty the day before: (killed)/ (wounded) * Immediacy of frames mentioned in the story in terms of order of paragraph (V. 65~75): 65. Israeli sovereignty/security/z 66. Israeli religious/right-wing tenacity: 67. Israeli terrorists/terrorism: 68. Palestinian terrorist/terrorism: 69. Palestinian religious/right-wing tenacity: 70. Palestinian independence movement/security: 71. Peace process/negotiation: 72: World terrorism: 73. World peace: 74. American interest: 75. Others: 76. News story is: (1) More episodic (depiction of the event) (2) More thematic (overview of the issue or other similar events) (3) More balanced in episodic and thematic 77. The causal responsibility of the conflict is assigned to Palestine by: (l= yes, 0= no) Israel United States The Arab world The Western society United Nations None/N A llllll 154 78. The causal responsibility of the conflict is assigned to Israel by: (1= yes, 0: no) Palestine United States The Arab world The Westem society United Nations None/NA Hill! 79. The treatment responsibility of the conflict is assigned to Palestine by: (1= yes, 0= no) Israel United States The Arab world The Western society United Nations None/NA Ill!!! 80. The treatment responsibility of the conflict is assigned to Israel by: (1= yes, O= no) Palestine United States The Arab world The Western society United Nations None/NA llllll 155 Appendix D: Coding instructions for the coding protocol of the content analysis VARIABLE: 1. Each news item will be numbered and assigned with a story id. 2. The date of each item will be identified and recorded. 3. Each of the four newspapers analyzed is given an id as in the coding list. 4. The total number of paragraph for each news item will be counted and recorded. 5. The length of each news item is measured in terms of the number of word given in each story. 6. Each story is coded into one of the four types. 7. The dateline given in each news item, if any, will be identified and recorded. * To avoid any potential confusion about coding the dateline in terms of nations or political entities, particularly on “Israel or Israeli occupied territory” and “other Middle- East political entities,” a thorough list of the Israeli, Palestinian and other Arabic nations or territories will be provided to the coders. 8. The news provider for each news item that is decided by the byline given in the story will be identified and recorded accordingly. Wire news service is defined here as a news organization, U.S. or foreign, that provides and sells news stories to news media subscribing to its service. By the definition, news organizations like Associated Press, Retuers, UPI, AFP and Tass are considered as wire services. A complete list of wire services will be provided to the coders. Similarly, The New York Times news service, the LA T imes- Washington Post news services and Cox news service are treated as wire news services when such bylines start the story and will be coded as “4= other U.S. news wire service” when used individually. More important, to be clear, if The News York Times News Service, not a staff writer’s name, is listed as the byline of the story in the Times, the news provider will still be coded as “4,” not “l= news staff/correspondent.” The same rule is applied to the LA. Times, Washington Post and Chicago Tribune. 9. The last name(s) of reporting joumalist(s) is recoded and written down as given, if any. 10. The number of journalists writing for the story is identified by the number of reporters appearing in the byline of the story, as well as at the end of the story as a contributor. It is arbitrary here that if no name appears in the byline likely in some stories provided by a wire service, then the number of journalist will be coded as “2” in that a good proportion of stories were written by more than one writer. Also arbitrary, “2” will be assigned to the story that uses “combined wire services” in that there is no way to 156 know exactly how many journalists from those wires were writing on the event. Moreover, when a story is a combined effort from staff writer(s) and wire service reporter(s), the number of journalists is counted by adding the numbers for staff and wire service writer. If no number is given as in “by staff and wire service,” then “3” will be arbitrarily coded. This variable is used to measure if the number of journalist on assignment would affect the sourcing patterns. ** News sources: Shoemaker and Reese (1996, p. 127) defined new sources as “external suppliers of raw material, whether speeches, interviews, corporate reports, or government hearings.” New sources are defined here as people, agents, organizations or governments quoted or paraphrased in the stories. A given source is counted each time it appears. In each news story, the identity of any source is given, even when an unnamed source is used, such phrase as “a government official said under the condition of anonymity” is used, so we are still able to decide what, if not who, the source is. In most cases, the identity of a source is given by the title that precedes or follows the name of the source. The definition of each type of news source used in the news lead, which is the first paragraph of a story, is defined as follows. The definition is also used for the groups of variables, 21~3O (placement of the first full quote), 3 l~40 (number of words in the first full quote), 4l~50 (number of full quote), 51~60 (frequency of news sources cited or quoted in the whole story), respectively. 11. Israeli Prime Minister: is defined as given. It would be either Ehud Barak or Ariel Sharon holding the title. 12. Other Israeli authority: is defined as Israeli citizens holding elected or appointed position(s) of an kind in Israeli government at the time of news reporting, other than as the prime minister; retired authority from that particular position is still considered an authority if he or she is appointed for a different position. For instance, Israeli Foreign Minister Perez is an Israeli authority. 13. Israeli non-authority: is defined as Israeli citizens holding no elected or appointed governmental position(s) of ay kind in Israel at the time of news reporting. For instance, former Prime Minister Ehud Barak is treated as an Israeli non-authority in that he doesn’t hold any governmental position after stepping down. Similarly, Shlomo Avineri, an Israeli political scientist, is considered an Israeli non-authority. Likewise, Nahum Zarnir, a bus driver, is an Israeli non-authority. 14. Yessar Arafat: is defined as given; he is the leader of the Palestinian Authority. 15. Other Palestinian authority: is defined as Palestinian citizens holding elected or appointed position(s) of any kind in Palestinian Authority or Palestine Liberation Organization at the time of news reporting; retired authority from that particular position 157 is still considered an authority if he or she is appointed for a different position. Members of Fatah, the main faction of the PLO, are also considered as a Palestinian authority. 16. Palestinian non-authority: is defined as Palestinian citizens holding no elected or appointed governmental position(s) of ay kind in Palestinian Authority or Palestine Liberation Organization at the time of news reporting. For instance Palestinian Red Crescent Society and its member are coded as Palestinian non-authority. Likewise, Joseph Thalgieh, father of Johnny Thalgieh killed by Israeli gunman, is considered a Palestinian non-authority. Similarly, the Rev. Theophanis, a Greek Orthodox priest is a Palestinian non-authority. Moreover, members of Popular Front for the Liberation Palestine, Hamas or Islamic Jihad are also Palestinian non-authority. 17. American authority: is defined as American citizens holding elected or appointed position(s) of any kind in the U.S. government at the time of news reporting; retired authority from that particular position is still considered an authority if he or she is appointed for a different position. For instance, U.S. Secretary of State Colin Powell is an American authority. Also, George J. Mitchell, former U.S. Senator, appointed by Clinton to lead a committee, is also considered an American authority. 18. American non-authority: is defined as American citizens holding no governmental position(s) of any kind in the United States at the time of news reporting. Retired government officials holding no other elected or appointed position is considered non- authority. For instance, Dennis Ross, who retired last year after serving for more than a decade as a U.S. mediator in the Middle East and now holding no governmental position, is considered a non-authority, same as Martin Indyk, former ambassador to Israel under President Bill Clinton, if used as a news source today. Likewise, Professor Shibley Telhami of the University of Maryland is a non-authority. l9. Non-Palestinian Arab nation authority: is defined as non-Palestinian Arab person holding governmental position(s) of any kind in their own nation at the time of news reporting. For instance, Syrian president, Bashar aI-Assad, is a non-Palestinian Arab people holding governmental position, so is Egypyian President Hosni Mubarak. 20. Others: is defined as including the rest of possible political entities or personnel not listed in the aforementioned, or other news agencies. It can include such political entities as members of EU, which will be coded as “Others.” Also, when Agence F rance-Presse is used as news source, it is considered and coded as “others.” Or when Pope John Paul II is used, he is also coded as “Others. 21~30. Variables 21 to 30 are used to measure the dependent variable of news placement of the first full quote. A full quote is defined as in the form of a complete sentence or more (like a whole paragraph). For instance, the sentence-—“We did not expect the withdrawal to be completely without dangers or difficult moments,” Mr. Barak said—is a fill] quote. 158 On the other hand, such quoting patterns in the sentences as—Both sides “should immediately implement an unconditional cessation of violence,” the committee said; or Palestinian security officials should ensure that Palestinian workers with permits to work in Israel are “free of connections to organizations and individuals engaged in terrorism” are not considered as a full quote, even though the first case misses only the subject to be a complete sentence. Placement is defined in terms of the percentile of the first full quote and is represented by the simple formula, l-[(Xth-1)/N], where Xth is where the first firll quote appears and N is the total paragraphs of a story. Using (Xth-1) allows the first full quote appearing in the news lead to always have a score of l or 100 percent. Moreover, the same quote— “We did not expect the withdrawal to be completely without dangers or difficult moments,” Mr. Barak said—was placed at the 9th paragraph of the story composed of 28 paragraphs and is coded first as “9.” Then, the placement of this quote in terms of the percentile will turn out to be about 71.43 percent {= l-[(9-l)/28)]}. In the case of no full quote of, say, Palestinian official news source, is used, the percentile of the full quote will be arbitrarily decided as “O.” 31~40. This group of variables are used to measures the dependent variable of number of words in the first full quote given to news sources The number of word by definition is decided by the words count of the full quote that excludes punctuation marks and those descriptions used around the quotation marks. For instance, the news space of the full quote—“As the word peace echoes in our hearts, how can we not think of the tensions and conflicts which have long be troubled the region of the Middle East,” the pope said at the airport reception. “So often hopes for peace have been raised, only to be dashed by new waves of violence.”——is counted as 46 words. 4l~50. Variables 41 to 50 are used to measure the dependent variable of the number of full quote in the whole story. The number of full quote will be counted and coded. 51~60. Variable 51 to 60 are used to measure the dependent variable of fiequency of news sources used in the whole story. The frequency for each of the 10 types of news sources in the full story will be name/head counted and recorded. 61 ~64. Variables 61~ 64 measure the total number and level of casualty occurring to Israelis and Palestinians, respectively. To measure this concept more precisely, casualty is differentiated between wounded and killed, which are assigned with a different scale. Killed is assigned with a score of 2, and wounded is assigned with a score of 1. That is to say, if there are four Israelis killed and five Israelis wounded, then a total score of 13 will be coded for casualty of Israelis. Similarly, if 2 Palestinians are killed and 7 Palestinian are injured, then a total score of 11 will be entered for casualty of Palestinians. ** News frames (V. 65~ 75): Entman (1993, p, 52) treated news frames as "some aspects of a perceived reality" made more salient in “a communicating text... to promote a 159 particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation.” The news frames or themes to be identified and analyzed are defined as follows. Immediacy of a news frame is defined and measured in terms of the percentile of the type of news frame appearing the first time in the story. It also is represented by the simple formula, l-[(Xth-l)/N], where Xth, as used for measuring placement of the first quote. 65. Israeli sovereignty/security: is defined as inclusion of threats to Israeli national sovereignty or security whether from terrorism, discussions over territory possession that might leave Israel without secure terrains, political entanglement, and so on. 66. Israeli religious/right-wing tenacity: is defined as discussion of actions by firndamental political groups that might be important in policy making, religious fervor or political balance of power. 67. Israeli terrorism/terrorists: is defined as inclusion of acts of terrorism or violence or the threats of terrorism/violence directed at Palestine or other Arabic nations from Israel or Israeli fundamentalists. 68. Palestinian terrorism/terrorists: is defined as inclusion of acts of terrorism or the threats of terrorism directed at Israel from PLA or other Palestinian factions. 69. Palestinian religious/right-wing tenacity: is defined as discussion of actions by Muslim religious groups that might be important in policy making, religious fervor or political balance of power. 70. Palestine independence movement/security: is defined as discussion of a Palestinian homeland, sacred sites, or other political issues that encourage or discourage Palestinian autonomy or religious loyalty, and so on. 71. Peace process/negotiation: is defined as inclusion of present initiatives or setbacks in an effort to promote or achieve peace between Israel and Palestine in particular, and peace in the Middle East in general. 72. World terrorism: is defined as inclusion of acts, reaction or discussion of terrorism or violence or the threats of terrorism/violence. This term is different from and excluding terrorism or violence exchanged between Israel and Palestine. For instance, the War on Terrorism falls into this category. 73. World peace: is defined as inclusion of advancement or setback in peace negotiation between Israel and Palestine that firrther achieves or undermines world peace. 74. American interests: is defined as the U.S. involvement in the Israel/Palestine conflicts that will help promote or curtail American interests in world politics or economy, as well as at home, such as the potential influences on presidential elections. 160 75. Others: Gamson urges that content analysis of news frames should pay attention not only to what is included but also what is omitted. Thus, the “others” category can somewhat serve this function to identify what is less used. 76. The notion and definition of episodic and thematic frame are borrowed from Iyengar’s, who dubs the two terms. He states: “The episodic news frame takes the form of a case study or event-oriented report and depicts public issues in terms of concrete instances. The thematic frame, by contrast, places public issues in some more general or abstract context and takes the form of a ‘takeout,’ or ‘backgrounder,’ report directed at general outcomes or condition.” Iyengar suggest a mixed frame is also common in news coverage, but in most cases, there exists a predominant frame and can be distinguished. Accordingly, episodic frame is more event-oriented, descriptive and focuses on the dramatization of the issue development, whereas thematic frame is more issue-oriented, analytic and pays more attention to the political, cultural, historical or religious background and cause of the news issue. 77 ~ 80. Borrowed from psychology, causal responsibility focuses on the origin of a problem or a conflict (i.e., blames on the intensification of the conflict or delay of the peace talk in the story), whereas treatment responsibility pays attention to who or what has the power to alleviate (or forestall alleviation of) the problem (i.e., stopping the terrorism activity). The responsibility assignments are decided on the overall reading of the story. It is likely that news sources from both sides blame on each other for delaying the peace talk or any violent action, but the decision can be made through reading the headline and the quotes used. Likewise, the decision on the assignment of treatment responsibility can be similarly made. 161 Appendix E: Percentage of agreement of inter-coder reliability for each of the 80 variables, using Scott Pi measure for categorical variables, Spearrnan correlations for ordinal variables, and Pearson correlations for ratio variables. Variable number/name Percentage of agreement 1. Story ID 100 2. Date 100 3. Newspaper 100 4. Total number of paragraphs 100 5. Length of the news story 100 6. Section 100 7. Dateline 100 8. News provider 100 9. Last name of the reporting journalists 100 10. Number of journalist writing for the story 100 11-20 Frequency of sources used in the lead: 1 1. Israeli Prime Minister 100 12. Other Israeli authority 96.8 13. Israeli non-authority 100 14. Yasser Arafat 100 15. Other Palestinian authority 95.9 16. Palestinian non-authority 100 17. American authority 100 18. American non-authority 100 19. Non-Palestinian Arabic nations 100 20. Others 100 21-30 Placement of first quote of sources: 21. Israeli Prime Minister 100 22. Other Israeli authority 100 23. Israeli non-authority 100 24. Yasser Arafat 100 25. Other Palestinian authority 100 26. Palestinian non-authority 100 27. American authority 100 28. American non-authority 100 29. Non-Palestinian Arabic nations 100 30. Others 100 162 31-40 Number of words used in the first quote: 31. Israeli Prime Minister 32. Other Israeli authority 33. Israeli non-authority 34. Yasser Arafat 35. Other Palestinian authority 36. Palestinian non-authority 37. American authority 38. American non-authority 39. Non-Palestinian Arabic nations 40. Others 41-50 Number of full quote for the sources: 41. Israeli Prime Minister 42. Other Israeli authority 43. Israeli non—authority 44. Yasser Arafat 45. Other Palestinian authority 46. Palestinian non-authority 47. American authority 48. American non-authority 49. Non-Palestinian Arabic nations 50. Others 51-60 Frequency of citations for the sources: 51. Israeli Prime Minister 52. Other Israeli authority 53. Israeli non-authority 54. Yasser Arafat 55. Other Palestinian authority 56. Palestinian non-authority 57. American authority 58. American non-authority 59. Non-Palestinian Arabic nations 56. Others 61. Number of Israelis killed 62. Number of Israelis wounded 63. Number of Palestinian killed 64. Number of Palestinian wounded 163 94.5 92.3 93.4 92.6 94.9 95.2 95.4 93.5 94.1 93.8 100 100 100 100 100 100 100 100 100 100 95.7 91.2 94.5 93.7 92.1 90.3 93.8 95.4 93.2 90.6 97.8 92.1 96.4 93.3 65-75 Immediacy of frames mentioned: 65. Israeli sovereignty/security 94.6 66. Israeli religious/right-wing tenacity 92.7 67. Israeli terrorists/terrorism 95.4 68. Palestinian terrorists/terrorism 96.7 69. Palestinian religious/right-wing tenacity 91.8 70. Palestinian independent movement/security 95.2 71. Peace process/negotiation 97.4 72. World terrorism 94.3 73. World peace 93.9 74. American interest 90.6 75. Others 94.1 76. Nature of news content 85.7 77. Causal responsibility assigned to Palestine Israel 100 United States 100 The Arabic world 100 The Western society 92.4 United Nations 94.7 None/NA 93.3 78. Causal responsibility assigned to Israel Palestine 100 United States 97.4 The Arabic world 100 The Western society 95.1 United Nations 93.8 None/NA 94.2 79. Treatment responsibility assigned to Palestine Israel 94.8 United States 92.4 The Arabic world 98.1 The Western society 93.7 United Nations 91.5 None/NA 93.4 164 80. Treatment responsibility assigned to Israel Palestine United States The Arabic world The Western society United Nations None/NA 165 93.6 91.2 95.1 92.5 94.4 93.8 REFERENCES Ahem, T. (1984). Determinants of foreign coverage in U.S. newspapers. In R. L. Stevenson & D. L. Shaw (Eds), Foreign news and news world information (pp. 217-235). Ames, IA: Iowa State University Press. Ashley, L., & Olson, B. (1998). Constructing reality: Print media’s framing of the women’s movement, 1966-1986. Journalism & Mass Communication Quarterly, 75, 263-277. Atwater, T., & Green, N. F. (1988). News sources in network coverage of international terrorism. Journalism Quarterly, 65, 967-971. Belkaoui, J. M. (1979). Images of Arabs and Israelis in the prestige press, 1966-1974. Journalism Quarterly, 55, 732-738. Belle, D. A. (2000). New York Times and network TV news coverage of foreign disaster: The significance of the insignificant variables. Journalism & Mass Communication Quarterly, 77, 50-70. Bennet, J. (2001). (2001, October 18). Far-right leader is slain in Israel; a blow to peace. The New York Times, p. Al. Bennett, W. L. (1990). Toward a theory of press-state relations in the United States. Journal of Communication, 40, 103-125. Bickerton, I. J ., & Klausner, C. L. (2002). A concise history of Arab-Israeli conflict. Upper Saddle River, NJ: Pearson Education, Inc. Bleaney, C. H. (1994). (Ed). World bibliographical series. Oxford: Clio Press. Boulding, K. (1961). The image. Ann Arbor: University of Michigan Press. Boulding, K. E. (197 5). The international system in the Eighties. In G. Sheffer (Ed), Dynamics of a conflict: A re-examination of the Arab-Israeli conflict (pp. 3-18). Atlantic Highlands, NJ: Humanities Press Inc. Cappella, J. A., & Jamieson, K. H. (1997). Spiral of cynicism. New York: Oxford University Press. Chan, J. M., & Lee, C. C. (1984). The journalistic paradigm on civil protests: A case study of Hong Kong. In A. Arno & W. Dissanayke (Eds), The news media in national and international conflict (183—202). Boulder, Co: Westview. 166 Chang, K. K. (1999a). Auto trade policy and the press: Auto elite as a source of the media agenda. Journalism Quarterly & Mass Communication, 76, 312-324. Chang, K. K. (1999b). Foreign policy, ideological exclusion and the media: How the American press shifts its news coverage of Gerry Adams. Paper presented to AEJ MC annual conference, New Orleans, LA. Chang, K. K., & Lee, T. T. (2001). Revisiting the “determinants of the international news coverage: How the U.S. news media cover the world events?” Paper presented to AEJMC annual conference, Washington, DC. Chang, K. K., & Zeldes, G. (2002). How ownership, competition affect papers’ financial performance. Newspaper Research Journal, 23, 101-107. Chang, T. K., Shoemaker, P. J ., & Brendlinger, N. (1987). Determinants of international news coverage in the U.S. media. Communication Research, 14, 396-414. Chang, T. K. (1988). The news and U.S.-China policy: Symbols in newspapers and documents. Journalism Quarterly, 65, 320-327. Chang, T. K., & Lee, J. W. (1992). Factors affecting gatekeepers’ selection of foreign news: A national survey of newspaper editors. Journalism Quarterly, 69, 554-61, Chang, T. K. (1993). The press and China Policy: The illusion of Sino-American relations 1950-1984. Norwood, NJ: Albex Publishing. Chang, T. K. (1998). All countries not created equal to be news: World system and international communication. Communication Research, 25, 528-563. Chang, T. K., Lau, T. Y., & Hao, X. (2000). From the United States with news and more: International flow, television coverage and world system. Gazette, 62, 505-522. Cho, H., & Lacy, S. (2000). International conflict coverage in Japanese local daily newspapers. Journalism & Mass Communication Quarterly, 77, 830-845. Cohen, B. (1963). The press and foreign policy. NJ: Princeton University Press. Cohn-Sherbok, D., & El-Alami, D. (2001). The Palestinian-Israeli conflict. Oxford: Oneworld. Curtius, M., & Wilkinson, T. (2001, December 7). Israel bombs Gaza; Arafat cracks down; Mideast: Jets attack police headquarters. Clashes between Palestinian security forces and Hamas supporters continue. The Los Angels Times, p. A30. Daughterty, D., & Warden, M. (1979). Prestige press editorial treatment of the Mid-East 167 during 11 crisis years. Journalism Quarterly, 56, 776-787. Dimmick, J. (1974). The Gate-Keeper: An Uncertainty Theory. Journalism Monograph. Eban, A. (1975). Some unsystematic thinking about the Arab-Israeli conflict. In G. Sheffer (Ed), Dynamics of a conflict: A re-examination of the Arab-Israeli conflict (pp. 349-366). Atlantic Highlands, NJ: Humanities Press Inc. Eilders, C., & Luter, A. (2000). Germany at war: Competing framing strategies in German public discourse. European Journal of Communication, 15, 425-428. Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43, 51-58. Entman, R. M., & Rojecki, A. (1993). Freezing out the public: Elite and media framing of the U.S. anti-nuclear movement. Political Communications, 10, 155-173. Fico, F. (1984a). The ultimate spokesman revisited: Media visibility of state lawmakers. Journalism Quarterly, 61, 383-391. F ishman, M. (1980). Manufacturing the news. Austin: University of Texas Press. Friedland, L. A., & Zhong, M. (1996). International television coverage of Beijingspring 1989: A comparative approach. Journalism & Mass Communication Monograph. Frisch, H. (1998). Countdown to statehood: Palestinian state formation in the West Bank and Gaza. Albany, NY: State University of New York Press. Gallup poll analyses: Americans show increased support for Israel following terrorist attacks. (Sept. 19, 2001). Gallup poll analyses: Americans against active diplomatic role for United States in Middle East confrontation. (Aug. 16, 2001). Gallup poll analyses: Americans most favorable toward Canada, Australia and Great Britan. (Feb. 16, 2001). Gallup poll analyses: Helping develop peacefirl solution to Middle East crisis is important goal to Americans. (Feb. 12, 2001). Gallup poll analyses: Americans agree that Middle East is vitally important to U.S. interests. (June, 14, 2000). Gallup poll analyses: Americans support active role for U.S. in world affairs. (April 1, 1999). 168 Gamson, W. A. (1989). News as framing. American Behavioral Scientists, 33, 157-161. Gamson, W. A. (1992). Talking politics. Cambridge: Cambridge University Press. Gamson, W. A., & Modigliani, A. (1989). Media discourse and public opinion on unclear power: A constructive approach. American Journal of Sociology, 95, 1-37. Gandy, O. H. (1982). Beyond agenda setting: Information subsidies and public policy. Norwood, NJ: Ablex. Gandy, O. H., Jr., Kopp, K., Hands, T., Frazer, K., & Phillips, D. (1997). Race and risk: Factors affecting the framing of stories about inequality, discrimination, and just plain bad luck. Public Opinion Quarterly, 61, 158-182. Gans, H. (1979). Deciding what ’s news. New York: Pantheon. Galtung, J ., & Ruge, M. H. (1965). The structure of foreign news. Journal of Peace Research, 2, 64-91. Giffard, A. C., & Rivenburgh, N. K. (2000). News agencies, national images, and global media events. Journalism & Mass Communication Quarterly, 77, 8-21. Gitlin, T. (1980). The whole world is watching: Mass media in the making and unmaking of the news left. Berkeley: University of California Press. Goffman, E. (1974). Framing analysis: An essay on the organization of experience. New York: Harper & Row. Goshom, K., & Gandy, O. H. (1995). Race, risk and responsibility: Editorial constraints In the framing of inequality. Journal of Communication, 45, 133-151. Gutierrez-Villalobos, S., Hertog, J. K., & Rush, R. R. (1994). Press support for the U.S. administration during the Panama invasion: Analyses of strategic and tactical critique in the domestic press. Journalism Quarterly, 71, 621-635. Haberman, C. (2001, September 1). Fragile clam holds on Jerusalem’s edge. The New York Times, p. A4. Haberman, C. (2001, August 10). At least 14 dead as suicide bomber strikes Jerusalem. The New York Times, p. Al. Hart, A. (1987). Arafat: Terrorist or peacemaker? London: Sidgwick & Jackson. Hart, J. A. (1961). The flow of international news into Ohio. Journalism Quarterly, 38, 541-543. 169 Hester, A. (1971). An analysis of news flow from developed and developing nations. Gazette, 17, 29-43. Hester, A. (1973). An analysis of news flow from developed and developing nations. Gazette, 19, 239-247. Hockstader, L. (2001, December 20). Emerging Question: If not Arafat, Who? As pressure builds, no clear successor is seen. Washington Post, p. A36. Huang, S. K. (1996). In search of the factors conditioning framing effect: The case of the Hill-Thomas controversy. Paper presented to the AEJMC annual conference, Anaheim, CA. Hur, K. K. (1984). A critical analysis of international news flow research. Critical Studies in Mass Communication, 1, 365-78. Iyengar, S. (1991). Is anyone responsible? Chicago: University of Chicago Press. Iyengar, S., & Simon, A. (1993). News coverage of the Gulf crisis and public opinion: A study of agenda-setting, priming, and framing. Communication Research, 20, 365-3 83. Johnson, M. A. (1997). Predicting news flow from Mexico. Journalism & Mass Communication Quarterly, 74, 315-330. Kapitan, T. (1997). (Ed). Philosophical perspectives on the Israeli-Palestinian conflict. Armonk, NY: M. E. Sharpe. Kempster, N., & Curtius, M. (2001, June 28). U.S. says pace of peace up to 2 Mideast sides. The Los Angeles Times, p. A8. Kim, S. T. (2000). Making a difference: U.S. press coverage of the Kwangju and Tiananmen pro-democracy movement. Journalism & Mass Communication Quarterly, 77, 22-36. Lacy, 8., Chang, T. K., & Lau, T. Y. (1989). Impact of allocation decisions and market factors on foreign news coverage. Newspaper Research Journal, 10, 23-32. Lacy, S., Shaver, M. A., & St. Cyr, C. (1996). Effects of public ownership and newspaper competition on the financial performance of newspaper corporations: A replication and extension. Journalism Quarterly, 73, 332-341. Lawrence, R. G. (2000). Game-frarning the issues: Tracking the strategy frame in public policy news. Political Communication, 17, 93-1 14. Lancaster, J. (2000, July 21). A “dead” summit’s revival; how Clinton persuaded both 170 sides to unpack their bags. Washington Post, p. A1. Lapidoth, R. (1975). Some legal aspects of the Arab-Israeli conflict. In G. Sheffer (Ed), Dynamics of a conflict: A re-examination of the Arab-Israeli conflict (pp. 139- 166). Atlantic Highlands, NJ: Humanities Press Inc. Lee, C. C., & Yang, J. (1995). Foreign news and national interests: Comparing U.S. and Japanese coverage of a Chinese student movement. Gazette, 56, l-18. Liebler, C. M., & Bendix, J. (1996). Old-growth forests on network news: News sources and the framing of an environmental controversy. Journalism & Mass Communication Quarterly, 73, 53-65. Livingston, S., & Eachus, T. (1995). Humanitarian crises and U.S. foreign policy: Somalia and the CNN effect reconsidered. Political Communication, 12, 22-38. Lukacs, Y. (1991). (Ed). The Israeli-Palestinian conflict: A documentary record. Cambridge: Cambridge University Press. Massey, B. L. (2000). How three Southeast-Asian newspapers framed ‘the haze’ of 1997- 98. Asian Journal of Communication, 10, 72-94. McCartney, J. (1994). Rallying around the flag. American Journalism Review, 16, 40-46. McCombs, M. (1992). Explorers and surveyors: Expanding strategies for agenda-setting research. Journalism Quarterly, 69, 813-824. McCombs, M., & Estrada, G. (1997). The news media and the pictures in our head. In S. Iyengar & R. Reeves (Eds), Do the media govern ?: Politicians, voters and reporters in America. Thousand Oaks, CA: Sage. McKenna, E. (1997). Land, property, and occupation: A question of political philosophy. In T. Kapitan (Ed), Philosophical perspectives on the Israeli-Palestinian conflict (pp. 185-204). Armonk, NY: M. E. Sharpe. McLeod, D. M., & Hortog H. K. (1998). Social control and mass media’s role in regulation of protest groups: The communicative acts and perspectives. In D. Demers & K. Viswansth (Eds), Mass media, social control and social change (pp. 305-330) Ames, IA: Iowa State University Press. McLeod, D. M., & Detenber, B. H. (1999). Framing effects of television news coverage of social protest. Journal of Communication, 49, 3-23. McNelly, J ., & Izcaray, F. (1986). International news exposure and images of nations. Journalism Quarterly, 63, 546-663. 171 Merrill, J. C. (1968). The elite press. New York: Pitman Publishing. Miller, M. (2000, October 16). Images are ammunition in battle for world’s sympathy; media: Israelis and Palestinians load cameras, activate beepers and dial cell phones to spread their versions of recent events. The Los Angles Times, p. A14. Mousa, I. S. (1987). The Arab image: The New York Times, 1916-1948. Gazette, 40, 101-20 Neuman, W. R., Just, M. R., & Crigler, A. N. (1992). Common knowledge. Chicago: University of Chicago Press. Norris, P. (1995). The restless search: Network news framing of the post-Cold War world. Political Communication, 12, 357-70. Ostgaard, E. (1965). Factors influencing the flow of news. Journal of Peace Research, 2, 23-31. Pan, 2., & Kosicki, G. M. (1993). Framing analysis: An approach to news discourse. Political Communication, 10, 59-79. Pan, 2., Lee, C. C., Chan, J. M., & So, Y. K. (1999). One event, three stories: Media narratives of the handover of Hong Kong in cultural China. Gazette, 61, 99-112. Peh, D., & Melkote, S. (1991). Bias in newspaper reporting: A content analysis of the coverage of Korean airlines and Iran airbus shooting in the U.S. elite press. Gazette, 47, 59-78. Peretz, D. (1996). The Arab-Israel Dispute. New York: Facts on File, Inc. Quester, G. H. (1975). Miscommunication in the Middle East and its impact on the outbreak of wars. In G. Sheffer (Ed), Dynamics of a conflict: A re-examination of the Arab-Israeli conflict (pp. 279-304). Atlantic Highlands, NJ: Humanities Press Inc. Quandt, W. B. (2001). Peace process: American diplomacy and the Arab-Israeli conflict since 1 96 7. Washington, DC: Brookings Institute Press. Rachlin, Alan. (1988). News as hegemonic reality: American political culture and the framing of news accounts. New York: Praeger. Reese, S. D., Grant, A., and Danielian, L. H. (1994). The structure of news sources on television: A network analysis of ‘CBS,’ ‘Nightline,’ ‘MacNeil/Lehrer,’ and ‘This Week with David Brinkley.’ Journal of Communication, 44, 84-107. Rhee, J. W. (1997). Strategy and issue frames in election campaign coverage: A social cognitive account of fi'aming effects. Journal of Communication, 47, 26-48. 172 Rosengren, K. E. (1972). International news: Intra and extra-media data. Acta Sociologica, 3, 96-109. Rosengren, KB. (1974). International news methods, data and theory. Journal of Peace Research, 11, 35-48. Rusciano, F. L., Fiske-Rusciano, R., & Wang, M. (1997). The impact of world opinion on national identity. Harvard International Journal of Press/Politics, 2, 71-92. Safran, N. (1963). The United States and Israel. Cambridge, MA: Harvard University Press. Segal, J. M. (1997). The state of Palestine. In. Kapitan, T. (1997). (Ed), Philosophical perspectives on the Israeli-Palestinian conflict. Armonk, NY: M. E. Sharpe. Semetko, H., & Valkenburg, P. M. (2000). Framing European politics: A content analysis of press and television news. Journal of Communication, 50, 93-109. Scheufele, D. (1999). Framing as a theory of media effects. Journal of Communication, 49, 103-122. Scott, W. A. (1965). Psychological and social correlations of international images. In H. C. Kelrnan (Ed), International behavior. New York: Holt, Rienhart and Winston. Shamir, S. (1975). Some Arab attitudes toward the conflict with Israel between 1967 and 1973. In G. Sheffer (Ed), Dynamics of a conflict: A re-examination of the Arab- Israeli conflict (pp. 185-200). Atlantic Highlands, NJ: Humanities Press Inc. Shoemaker, P. J., & Reese, S. D. (1991). Mediating the messages: Theories of influences on mass media content. White Plains, NY: Longman. Sigal, L. V. (1973). Reporters and officials: The organization and politics of newsmaking. Lexington, MA: D.C. Heath and Company. Sontag, D. (2000, December 8). As Arafat embraces revolt, his sagging popularity rises. The New York Times, p. A1. Stempel, G. H., III. (1961). The prestige press covers the 1960 presidential campaign. Journalism Quarterly, 38, 157-163. Stempel, G. H., 111. (1965). The prestige press in two presidential elections. Journalism Quarterly, 42, 15-21. Suleiman, M. (1974). National stereotypes as weapons in the Arab-Israeli conflict. Journal of Palestinian Studies, 3, 109-21. 173 Sussman, L. R., & Guida, K. (2000). Freedom Review, 30. Sylvester, J., Wu, H. D., & Hamilton, J. M. (2000). A content analysis of the Philadelphia Inquirer’s Israel/Palestinian coverage. Paper presented to the AEJMC annual conference, Phoenix, AZ. Syrbemy—Mohammadi, A. (1990). U.S. media cover the world. In J. Downing & A. Syrbemy—Moharnmadi (Eds), Questioning the media: A critical introduction (pp. 296-307). Newbury Park, CA: Sage. Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics. Boston: Allyn & Bacon. Takeshita, T. (1997). Exploring the media’s roles in defining reality: From issue-agenda setting to attribute-agenda setting. In M. McCombs, D. L. Shaw, & D. Weaver (Eds), Communication and democracy: Exploring the intellectual frontiers in agenda-setting theory (15-27). Mahwah, NJ: LEA. Terry, J., & Mendanhall, G. (1974). 1973 U.S. press coverage on the Middle East. Journal of Palestinian Studies, 4, 120-133. Trice, R. H. (1979). The American elite press and the Arab-Israeli conflict. The Middle East Journal, 33, 304-325. Tuchman, G. (1978). Making news. New Work: Free Press. Wall, M. A. (1997a). A ‘Pernicious new strain of the old Nazi virus’ and an ‘Orgy of tribal slaughter’: A comparison of U.S. new magazine coverage of the crises in Bosnia and Rwanda. Gazette, 59, 411-428. Wall, M. A. (1997b). The Rwanda crisis: An analysis of news magazine coverage. Gazette, 59, 121-134. Wallerstein, I. (1974). The modern world system. New York: Academic Press. Wallerstein, I. (1979). The capitalist world-economy. Cambridge, MA: Cambridge University Press. Walters, L. M., & Homing, S. (1993). Faces in the news: Network television news coverage of hurricane Hugo and the Lorna Prieta earthquake. Journal of Broadcasting and Electronic Media, 219-32. Wang, Z., & Lowry, D. T. (1999). Wag the press: How changes in U.S. foreign policy toward China were reflected in prestige press coverage of China, 1979 vs. 1997. Paper presented to the AEJMC annual conference, New Orleans, LA. 174 Wanta, W. (1989). Presidential approval rating as a variable in the agenda-setting process. Journalism Quarterly, 68, 672-679. Wanta, W., & Hu, Y. W. (1993). The agenda-setting effects of international news coverage: An examination of differing news frames. International Journal of Public Opinion Research, 5, 250-264. Wanta, W., & Chang, K. K. (2000). Individual variables and the second-level of agenda- setting: Factors affecting the agenda of attributes. Egyptian Journal of Public Opinion, 1, 55-82. Weekly Compilation of Presidential Documents. (2000, July 26). Weekly Compilation of Presidential Documents. (1993, Sept. 13). Wilkinson, T. (2001, June 2). 18 dies, scores hurt suicide blast in Tel Aviv; Terrorism: Anguish and carnage pervade beach area after bomber blows himself up, killing 17 Israelis outside nightclub. Despite Sabbath, Sharon weighs retaliation. The Los Angeles Times, p. A1. Whitney, C. D., Fritzler, M., Jones, 8., Mazzarella, S., & Rakow, L. (1989). Geographic and sources biases in network television news 1982-1984. Journal of Broadcasting and Electronic Media, 33, 159-174 Wilson, T. C. (1996). Cohort and prejudice: Whites’ attitudes toward Blacks, Hispanics, Jews and Asians. Public Opinion Quarterly, 60, 253-274. Woo, J. (1996). Television news discourse in political transition: Framing the 1987 and 1992 Korean presidential elections. Political Communication, 13, 63-80. Wu, H. D. (1996). An enduring schema: The image of the Chinese in American prime time television dramas. Gazette, 58, 69-86. Wu, H. D. (2000). Systemic determinants of international news coverage: A comparison of 38 countries. Journal of Communication, 50, 110-130. Zaller, J ., & Chiu, D. (1996). Government’s little helper: U.S. press coverage of foreign policy crises. Political Communication, 13, 385-405. Zhu, J. H. (1992). Issue competition and attention distraction: A zero-sum theory of agenda-setting. Journalism Quarterly, 69, 825-836. 175