Tflflllllflllllllrlliflffli Hilli'lflli 1_293 00634 4182 This is to certify that the thesis entitled Communication Patterns and Upward Job Mobility Among Journalists presented by Julia G. Crystler has been accepted towards fulfillment of the requirements for M . A . degree in Communication [j Major professor Date [1’ X’ 85/ 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES ~.. M RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wi11 be charged if book is returned after the date stamped beiow. i «“4221 snail M193 v ‘ €3er COMMUNICATION PATTERNS AND UPWARD JOB MOBILITY AMONG JOURNALISTS By Julia Gluesing Crystler A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1985 ABSTRACT COMMUNICATION PATTERNS AND UPWARD JOB MOBILITY AMONG JOURNALISTS By Julia G. Crystler This study develops a model of upward job mobility within an expectancy theory framework which includes communication as a significant variable. Mobility can be considered a product of a wide range of contacts, opportunities and relationships that extend through all the activities of organizational members. Achieving and managing this interactive process within an organization is a function of communication. The study examines the relationship of communication to individual job mobility in comparison with other noncommunication factors such as job performance, job satisfaction, age and tenure, and individual perceptions of the work environment. The research was carried out in the specific organizational setting of corporate journalism and also explores the relationship between upward job mobililty characteristics of journalists and their communication patterns. A survey of 1,118 journalists provided the data for this study. Items in the survey were subjected to both exploratory and confirmatory factor analyses which yielded 13 composite indices. These indices were then entered into regression analyses. The results indicate that journalists who have career aspirations and who are upwardly mobile in their profession communicate largely with other journalists, often to the exclusion of contacts with the public. The results also support the conclusion that communication should be included as a variable in any model of upward job mobility. A modified version of the mobility model is suggested for future research. ACKNOWLEDGEMENTS This thesis was completed with the guidance and patience of Dr. Judee Burgoon, Dr. Michael Burgoon, and Dr. Vince Farace. The Drs. Burgoon were kind enough to allow me to participate as a researcher on a Gannett Corporation grant which provided an excellent learning experience and the data base for this study. And Dr. Farace provided much needed encouragement to see me through to completion. I also wish to acknowledge the invaluable contribution made by my husband, Bill and my daughter Carly. Without them, there would be no thesis. ii TABLE OF CONTENTS Chapter Page I OVERVIEW OF THE RESEARCH PROBLEM ....... . ..................... 1 Job Mobility and Communication Among Journalists ............ 1 Organization of the Thesis .................. ..... ........... 5 II REVIEW OF THE JOB MOBILITY LITERATURE ....................... 6 Expectancy Theory of Job Mobility ........................... 17 III METHODS AND PROCEDURES .. .................................... 21 Sample .......................... ....... ....... .............. 21 Procedures ........... . ...................................... 23 Measurement ...... . ....... . ...... . ......... . ...... . ......... . 23 Procedures for Statistical Analysis ........... . ..... . ...... . 36 Hypothesis .......................... .................. ...... 38 IV RESULTS, SUMMARY, AND DISCUSSION ............................ 39 Results .00... O ..... 0.0.0.0.... ...... ......OOOOOOOOOOOOOOO 39 Full Model of Upward Job Mobility ... ............... ........ 39 Partial Model of Upward Job Mobility —— Model I ........ ..... 43 Partial Model of Upward Job Mobility -— Model 2 ............ . 45 Assumptions of Regression Analysis ........... ......... ..... . 47 Discussion of Results ..... . ................................. 49 Limitations .... ............................................. 52 Implications for Future Research ............................ 56 LIST OF REFERENCES .... ............... . ............ . ....... ....... 59 FOOTNOTES ...................... . ................................. 63 APPENDIX A. Descriptive Statistics and Correlations For All Variables In Full Regression Model of Upward Job Mobility ........ 64 iii LIST OF TABLES Table Page 1 Co—Worker vs. Public Communication Practices of Journalists —— Survey II ............... ...... ....... ....... 14 2 Final Factors -— Variables in Regression Equations ......... 32 3 Results of Full Regression Model of Upward Job Mobility .... 41 4 Results of Partial Regression Model of Upward Job Mobility (Model I) ...... ...... ........... .................. ... ...... 45 5 Results of Partial Regression Model of Upward Job Mobility (Model 2) .................................................. 47 iv LIST OF FIGURES Figure Page 1 Items Assessing Communication Practices of Journalists for Survey I -- Grouped by Factors .............................. 12 2 Items Assessing Communication Practices of Journalists for Survey II — Grouped by Factors .. .......... ........ .......... 13 3 Job-related Characteristics of Four Types of Journalist Communicators .... .................. . ........ . ............... 15 4 Theoretical Model of Upward Job Mobility ....... ..... ........ 20 5 Communication Practices of Journalists ....................... 24 6 Work Relations ....................... . ....................... 25 7 Job Satisfaction ................. .... ...... ......... . . 26 8 Perceptions —— Image of the Newspaper TV Station ... ...... .... 27 9 Perceptions -- Outlook Toward the Future ..................... 27 10 Job Performance .............................................. 28 11 Expectations for Upward Job Mobility and Age and Tenure . ..... 29 12 Full Regression Model of Upward Job Mobility ................. 4O 13 Partial Regression of Upward Job Mobility —- Model 1 ......... 44 14 Partial Regression of Upward Job Mobility —— Model 2 ........ . 46 CHAPTER I OVERVIEW OF THE RESEARCH PROBLEM Job Mobility and Communication Among Journalists Individual organizational members are generally concerned at some time in their careers with job mobility. When an individual is promoted upward through the hierarchical levels of an organization, organizational members generally View this upward mobility as a reward (Katz, Kahn, 1978; Jennings, 1967). Vertical mobility typically indicates that the individual's behaviors are perceived as appropriate within the organizational culture and that the individual is worthy of more responsibility and authority (Schneider, 1982). Job mobility can be considered part of the organizational behaviors closely related to work climates. Schneider (1982) suggests that work climates emerge from naturally occurring patterns of the goal—oriented interactions of people with each other and of the changing facets of their work environment. He states that people tend to be attracted to, selected by, and leave from organizations, in a way that yields settings characterized by particular or similar kinds of peOple. It is difficult to separate the nature of the person from the nature of the setting. Mobility can be considered a product of a wide range of contacts, opportunities and relationships that extend through all the activities of organizational members (Jennings, 1967). Achieving and managing this interactive process within an organization is a function of communication (Farace, Monge, & Russell, 1977). Weick (1979) believes that it is possible to obtain a more complete view of the behaviors that occur in organizations if we consider communication as a critical commodity on which organizations function and to which they direct their processes and establish their relations. 2 Job mobility has been studied as an outcome of various organiza— tional processes such as job performance, job satisfaction, age and tenure and individual pereceptions of the work environment (Vardi, 1980). However, communication as a factor related to promotion in organizations has been largely ignored. This paper will examine the questions "What difference does communication make for an individual's job mobility in comparison with other noncommunication factors?" and "What kind of communication patterns positively influence upward job mobility?" Communication is not the sole determinant of job mobility but should be considered along with noncommunication factors to obtain a more complete understanding of this organizational process. In the literature on organizational behavior and career development, job mobility has been broadly defined as organizational—related job movement. Job mobility is usually viewed as a retrospective assessment of a series of events that have occurred over time. Job mobility is conceptualized for this study as the degree of vertical movement in the organizational hierarchy within a specified period of time; mobility is measured by the number of promotions one receives in a given number of years. This definition is consistent with Grusky's (1966) concep— tualization of job mobility as upward organizational movement. It includes mobility from one geographic area to another and from one functional organizational area to another (if the move can be classified as a promotion). The more promotions an individual receives in a given time period, the more mobile he or she is. The research on job mobility conducted for this paper was carried out in the specific organizational setting of corporate journalism. There is a need for communication scientists to study the organizational environment of the modern 3 journalist. In the past 50 years this country has seen a rapid rise in the number of corporate investments in the newspaper industry (Bagdikian, 1971). While journalists remain the keepers of the public information domain, they must also focus attention on the demands of their organization. The professional journalist must be concerned with more than simply improving his or her skills as a reporter. He or she must also look after the progress of a career in the organization that oversees his or her trade. The news stories that fill the papers cannot be evaluated without understanding the organizational context in which they are written. In 1981, the American Society of Newspaper Editors (ASNE) sponsored a study of the attitudes, beliefs, behavior and communication patterns of working journalists as part of the national Newspaper Readership Project. (1) The ASNE findings indicated a potential relationship between upward job mobility in the organization and a restricted flow of information into the newsroom. The ASNE report suggests that some journalists have "lost touch" with their public, contributing to a decline in newspaper readership. Journalists are possibly in danger of becoming a professional elite who communicate primarily among themselves. In a later study, the Gannett Group sponsored a project similar to the ASNE survey, but focusing on its own newsroom personnel. (2) The results of the Gannett study also indicated that journalists are removed from the American public. Subsequently, Burgoon, Burgoon, Buller and Atkin (1983) further assessed journalists' communication practices by examining 1) attitudes toward public involvement, 2) actual contacts with nonjournalists, 3) work—related communication in the newsroom, and 4) social contacts with co—workers. 4 Burgoon et al. (1983) express a concern that journalists are not well integrated into their communities and that they substitute interactions with colleagues for interactions with the public. They suggest that expectations for upward mobility are one of the important variables related to the communication practices of journalists. Newspaper organizations reward those journalists who are "well connected" in the newsroom. They state that: "The young journalists with advancement aspirations may realize that cultivation of contact with other journalists is beneficial and thus spend social time with co—workers at the expense of contacts with nonjournalists." (3) Those journalists who are integrated in their communication practices (i.e., also have contacts outside the newsroom) are less likely to remain with the newspaper or advance into editorial positions, thus increasing the staff's isolation in the newsroom. The concern that journalists are becoming an elite group in American society is shared by professional journalists. Halberstam cautioned his colleagues to avoid this elitist tendency: "...we pretend the press is a body separate of the people, like government, a true Fourth Estate. That is dangerous because the press should protect people's rights. We are people's information—gatherers and the only rights we have are rights people have." (Chien, 1982)) This study develops a model of mobility which includes communication as a significant variable. More specifically, the study is designed to explore the relationship between the upward job mobility characteristics of journalists and their communication patterns. The study addresses the following research question: Is there a positive linear relationship between the amount of communication journalists have with their colleagues and their upward job 5 mobililty in the journalism industry? Organization Qf_the Thesis This paper is organized into four chapters. In Chapter 2, the pertinent literature on job mobility is reviewed and expectancy theory of organizational behavior is presented as a useful framework studying job mobility. Chapter 3 discusses the design of the study, including sample, measures, hypotheses, procedures and analysis. In Chapter 4, the results of the analyses are given. The final chapter is a summary and discussion of the results, including implications for future research. CHAPTER 2 REVIEW OF THE JOB MOBILITY LITERATURE Several authors have discussed the range of theoretical constructs and levels of analyses which comprise the literature on occupational mobility (Vardi, 1980; Zaleznik, Dalton & Barnes, 1970; Glaser, 1968; Schein, 1978). Vardi presents a useful categorization of the various approaches to job mobility by synthesizing the literature into four major models based on differences in paradigms, units of interest, levels of analyses and methodologies. According to Vardi, the four models are psychological, sociological, economic and administrative. The psychological models emphasize the contibution of individual- level factors to mobility. Attitudes, motivation, expectations and behaviors of the individual all contribute to occupational mobility. The sociological models attribute mobility to a process of continuous adjustment to the social and organizational environment. The interaction of organizational and individual characteristics determine the mobility an individual experiences. In the economic models, mobility is regulated by external economic constraints and labor markets which influence the availability of jobs and the potential for mobility of members in the organization. The administrative models attribute mobility to individual merit which is assessed by job performance, seniority, job level and demographic factors such as age and sex. In these models, mobility is conceived of as reward (i.e., promotion). The present study is concerned with the primary variables in the administrative, sociological and psychological models which research findings indicate have an influential impact on vertical job mobility. The variables in the economic models are not considered here since they 7 are related only to influences external to the organization. The important variables are demographic characteristics [age and tenure (both time spent in an industry and time spent in a particular job)], individual motivating factors (such as job satisfaction, work relations, and the opinions one holds of the company where he or she works and of the industry in general), job performance and communication patterns. These variables are all hypothesized to be predictors of actual vertical job mobility. The following sections summarize the research findings which relate to each of these variables: Age and Tenure. Age and tenure are consistent predictors of upward mobility (Michaels & Spector, 1982; Vardi, 1980; Wanous, 1980; Rosenbaum, 1979; Glaser, 1978; Schein, 1978; Spillman, 1972; and Zalenik et al., 1970). From an organizational perspective, these variables become the contingencies on which upward mobility is dependent. Both Schein (1978) and Vardi (1980) discuss age with respect to career opportunities and life stages. They state that age reflects the process of "coming of age" in an organization and is related to increasing seniority or tenure. At a younger age, upward mobility usually occurs at a more rapid rate. Tenure refers to both length of time in the job and length of time in the organization. Rosenbaum (1979) and Michaels and Spector (1982) have researched the changes in status and career mobility related to this variable. Mobility in the earlier portions of one's employment in an organization bears an unequivocal relationship with one's later career, predicting career "ceilings" and career "floors", and successive mobility. In other words, the longer the time one spends in a job the less likely he or she is to be promoted and the fewer the promotions he or she will receive. Length of time in the organization has an inverse relationship to upward mobility. The more time spent in an organization, 8 the fewer the promotions, implying stability in one's later career. Job satisfaction. Job satisfaction is defined as an employee's general overall rating of his/her job as satisfying his/her needs. Several researchers have studied the relationship between job satisfaction and job mobility and have found that an individual's propensity to stay with an organization and to strive for an upwardly mobile career is partially a function of his or her satisfaction with the job. Michaels and Spector (1982), Greenhalgh (1980), Schein (1978), and Hall and Schneider (1973) all state that a person's career mobility is positively related to satisfaction with the job, but that job satisfaction does not lead to upward mobility in and of itself. Other variables such as career expectations and job performance mediate the relationship between satisfaction and vertical job mobility. Job satisfaction is thought to be a motivating variable which can lead an individual to pursue upward mobility in an organization. Work relations. Burgoon et al. (1982) studied work relationships, attitudinal similarity or subordinates with their supervisors, and newsroom conflict. They related these factors to the communication practices of journalists and reported that journalists with frequent interaction in the newsroom were likely to have high quality relationships with other members of the newsroom, perceptions that their organization shared their attitudes, beliefs and values, and perceptions that their supervisors shared their attitudes toward the job. These same journalists were also more likely to have received a promotion in the last two Years. Mowday, Steers and Porter (1979) viewed an individual's work relations as an important part of organizational commitment. They define 9 commitment as the relative strength of an individual's identification with and involvement in a particular organization. Porter and Crampon (1976) and Grusky (1966) studied career mobility and organizational commitment and found that highly mobile managers were generally the most committed. Perceptions. The image an individual holds of the company where he or she is employed and of the industry in general is also related to organizatinal commitment and to vertical job mobility. These perceptions are believed to contribute to an individual's desire for upward mobility. They provide career motivation (Mowday, Steers & Porter, 1979; Grusky, 1966). Bernstein, Burgoon and Burgoon (1984) analyzed attitudes held by a broad cross—section of journalists toward the future of their industry and toward their individual organizations. They factor analyzed 10 attitudinal items related to the newspaper industry's future. The results revealed three independent clusters of attitudes: (1) a pessimistic, gloomy attitude about the future, (2) a confident attitude and commitment to the industry's future, and (3) a present—oriented faith in the industry, both now and in the future. In a regression analysis of these attitudinal clusters with other variables, Bernstein et al. found that those journalists who were pessimistic about the future of their industry were less likely to expect upward mobility and that those who expressed assuredness and commitment to their organization and to the industry intended to make lifelong careers in journalism. Job Performance. Wanous (1980), Michaels and Spector (1982) Spencer and Steers (1981) and Rothman and Perrucci (1970) have all found that job performance can open or close doors to career mobility. When an individual performs poorly on the job, an organization will offer fewer 10 opportunities for promotion. For persons who perform well, of course, the reverse is true. Communication Patterns. Implicit in many of the sociological models of job mobility is the notion of an individual's "fit" in the organization. Pellegrin and Coates (1957) view the vertical mobility of an individual through a corporate hierarchy as dependent on the individual's ability to determine the behaviors appropriate in the organization and to act accordingly. Schein (1978) also proposes that job mobility is dependent on the individual's ability to interact with the key decision—makers in an organization. Interaction with "core managers" is necessary for mobility. To follow an upwardly mobile career path, one must make the proper contacts and maintain positive relationships with coworkers and supervisors. Inclusion in a core group is essential to job mobility because it affords the individual access to and inclusion in information exchange activities which allow him or her to perform a job well. Jennings (1967) states that access to this informal information about appropriate organizational behaviors and expectations can be a powerful determinant of upward mobility. He says that "before the manager is promoted he has talked and behaved as one who belongs at a higher level. He appears so much like his superiors that they formalize the relationship" (p.36). Individuals who are more upwardly mobile tend to belong to cliques of influential organization and industry members (Tichy, 1973). Both Hall (1976) and Janowitz (1968) conducted studies to determine the factors influencing upward mobility. Hall (1976) surveyed 11,000 managers in more than 50 companies and found that in order to move up in the organization and achieve a management position, one must learn to behave like a manager. Janowitz (1968) studied promotion in the military and found that the informal lines of 11 communication were a strong factor influencing promotability. Upward mobility in an organization is also related to an organizational member's communication patterns. Insularity is conceptualized as a communicatin continuum based on the degree to which an individual communicates with others in his or her own and related professions in proportion to the degree of external nonprofessional communication contacts. A highly insular individual engages in exclusive communication within the profession. Time on and off the job is spent communicating with other journalists to cultivate the contacts and glean the information necessary to become a professional, upwardly mobile individual. There is no time left for substantial contacts with people from the "outside world" who represent contacts that are not directly job related. The Burgoon et al. (1982) ASNE monograph led to further exploratory analyses of the relationship of communication patterns and job mobility (Burgoon, Burgoon, Buller & Atkin, 1984). The communication items in Figures 1 and 2 are grouped according to the results of a principal components factor analysis with varimax rotation. FIGURE 1 ITEMS ASSESSING COMMUNICATION PRACTICES OF JOURNALISTS —— SURVEY I GROUPED BY FACTORS Isolation from the Community I have a lot of contact with peOple in walks of life very different from my own. I think journalists should be involved with outside groups and activities. I am involved with a variety of groups and people in the community. Communication with Nonjournalists Regarding News Topics I rarely discuss news events with non—journalists. I feel discussion of news events is best kept in the office. Communication with Journalists I socialize frequently with other members of the newspaper staff. The majority of my friends are associated with the newspaper business in one way or another. My co—workers frequently discuss their stories with me. My editor and I frequently discuss the content of various stories. Isolates Nonjournalists often ask my opinion about things in the news. I regularly seek out nonjournalists' opinions about the news. 12 FIGURE 2 ITEMS ASSESSING COMMUNICATION PRACTICES OF JOURNALISTS -— SURVEY II GROUPED BY FACTORS Isolation from the Community Off the job, I have a lot of contacts with peOple in walks of life very different from my own. The demands of this job leave little time for social contacts with people outside the newsroom. I regularly seek out nonjournalists' Opinions about the news. I think it is vitally important that a journalist be integrated into the local community. I have enough contacts with people outside the newsroom to have a good feel for what is going on in the community. The majority of my friends are associated with the news business in one way or another. I am involved with a variety of groups and people in the community. Communication with Journalists My co—workers and I frequently discuss news stories with one another. I feel well informed about what is going on here at this paper/station. I have a lot of influence on decisions that are made about what stories are covered and what kind of play they receive. 13 14 In Survey I, the ASNE survey, a four factor structure emerged, accounting for 44.7% of the variance. Three factors are related to the insularity issue, with the fourth factor including only one item (it asked about the effectiveness of memos and posted communications). Factor analysis of the communication items in Survey II, the Gannett survey, resulted in a six factor solution accounting for 52.8% of the variance. Two of the factors were related to insular communication practices —— one factor assessing isolation from the community and the other assessing communication with journalists. Burgoon et a1 (1984) further probed the insularity issue among journalists by classifying respondents into four categories from their scores on the two insularity factors found in Survey II. Based on respondents' scores measuring their contact with the public, they were classified as high or low in their amount of work— related and social communication with other journalists. Table 1 indicates the percentage of respondents falling into the four categories of communicator types. TABLE 1 CO—WORKER VS. PUBLIC COMMUNICATION PRACTICES OF JOURNALISTS ——SURVEY II Communication with Other Journalists Low High (1) I I I I Isolates I Fast-trackers I Low I I I I 27% I 25% I Communication I I I With the Public 1 I I I Externals I Talkers I High (2) I I I I 25% I 23% I I I I 15 (1) Based on having a score of 10.38 or above in Survey II (on a scale ranging from 3 to 15 with a midpoint of 9). (2) Based on having a score of 23.25 or higher in Survey II (on a scale ranging from 7 to 35 with a midpoint of 21). The respondents were distributed about equally across the four quadrants. The journalists grouped in each quadrant have been labeled according to their communication patterns. "Isolates" communicate neither with the public nor with their colleagues. "Fast—trackers" fall at the opposite end of the scale. They communicate almost exclusively with their " communicate primarily with the public industry coworkers. "Externals and not with co—workers. Finally, "Talkers" engage in communication with both the public and their fellow coworkers. The four communicator types were also analyzed to see how they differed on job—related characteristics. The results are listed in Figure 3. FIGURE 3 JOB—RELATED CHARACTERISTICS OF FOUR TYPES OF JOURNALIST COMMUNICATORS ISOLATES ... in the industry and on current job slightly longer than average . predominantly in nonsupervisory roles (80% hold jobs as reporters, copy editors, photographers and the like) .. less likely to be promoted (74% had not in last 2 years) ... less likely to receive merit raises (59% had not received one in last 2 years ... read fewer magazines and professional journals than others ... read own newspaper less thoroughly than others ... see their newspaper as nonsensational . hold more negative views toward new technology .. less likely to believe their paper values and uses research ... more likely to see themselves as dissimilar to readers EXTERNALS . in industry longer than average (65% exceed 10 years) ... in current job longer than average .. older than average (45 % are aged 45 or older) ... mostly in nonsupervisory roles (80% are) 16 ... somewhat below average in job satisfaction .. least likely to be promoted (80% had not in last 2 years) .. less likely to receive merit raises (54% had not in last 2 years) .. read more magazines than others but fewer professional journals . consistently rate their paper lower on all facets of image ... hold more negative views toward new technology ... less likely to believe their paper values and uses research ... more likely to see themselves as similar to readers ... hold higher estimates of how many people read the newspaper FAST-TRACKERS ... in industry least time (62% report 10 years or less) . are newcomers to their job (65% have held current position less than 3 years) . are younger and especially likely to fall in 25—34 age bracket (55% are) .. include more supevisors than average (32% are) . are most likely to be promoted (53% have in last 2 years) ... are more likely to receive merit raises (66% have in last 2 years) ... read fewer magazines but more professional journals ... hold consistently more favorable image of their paper on all facets . hold more favorable attitudes toward new technology . more likely to believe their paper uses and values research ... more likely to see themselves as dissimilar to the reader ... hold lower estimates of number of readers in community TALKERS ... in industry less time than average (52% report 10 years or less) ... are newcomers to current job (62% have held it less than 3 years) . are slightly younger than average .. include more supervisors than average (32% are) ... slightly above average on job satisfaction ... more likely to be promoted (44% have in last 2 years) ... more likely to receive merit raises (64% have in last 2 years) . consume more print media than others (magazines, professional journals, own newspaper) ... hold more negative image of paper on sensationalism .. more likely to believe newspaper values and uses research ... more favorable toward new technology . more likely to see themselves as similar to readers Fast—trackers were more mobile than the other types (53% had been promoted in the last two years). Of the talkers, 44% had been promoted past two years. Externals were least likely to be promoted (only 20% had been promoted in the past two years. Externals were least likely to be promoted (only 20% had been promoted in the past two years); among the 17 isolates, only 26% had been promoted. It is expected that the more communication effort one expends to initiate and maintain the necessary contacts and engage in the insular communication patterns that accompany career advancement, the better able one is to perform well in the job and the more upward mobility will be experienced as a result. Expectancy Theory 9f_Job Mobility Little past research on job mobility has been structured by theory. Researchers have built models of job mobility derived from exploratory analyses and conceptualizations based on different units of interest from the individual to the organization (Vardi, 1980). However, Beehr, Taber and Walsh (1980) in a factor analytic study of intraorganizational job mobility, suggest that the expectancy theory of motivation is a significant theoretical domain for the study of mobility. This theory posits that motivation acts as a generalized expectancy about the relationship between one's behavior and various outcomes (Phares, 1976). Expectancy theory has been applied to initial job choice (Wanous, 1972; Vroom, 1964) and in the present study is applied to the expectations of news personnel for obtaining a second position once one already has a job. Since expectancy theory will serve as a framework for this study summary of the theory's basic tenets and the empirical support it has received is presented here. Wanous (1980), Schein (1977), Vroom (1964) and Berlew and Hall (1966) have found that expectations are positively related to job mobility. Expectations for upward mobility lead an individual to behave in consonance with those expectations. The individual might work harder, spend more time on the job and seek the support of coworkers more often in he or she expects that these behaviors might increase the likelihood of obtaining a promotion. Schein (1977) discusses 18 expectations as "career anchors" which influence an individual's career paths. Those who do not expect to be promoted will behave differently and seek different career paths than those who do expect upward job mobility. There are four main assumptions that supporters of expectancy theory make. First, behavior is assumed to be determined by a combination of forces in the individual and forces in the environment. These forces arise from the psychological and social history of the individual and from specific workplace features. For example, individual forces which might influence the behavior of journalists are their tenure in the profession and their job satisfaction, while environmental forces might arise out of the work relations among journalists in the newsroom. Expectancy theory also assumes that people make decisions about their own behavior in organizations. For example, individuals make decisions about membership behavior, e.g. whether to work late hours or to socialize with coworkers. Individuals also make decisions about the amount of effort they devote to their jobs. A third assumption is that different people have different types of needs, desires and goals. Rewards and outcomes should correspond or "match" these needs. Some newspaper reporters may seek to become an editor, whereas others desire to be better reporters but not necessarily rise in the organization. The final assumption of expectancy theory is that people choose among alternative behaviors based on their perceptions (expectancies) of the degree to which a given behavior will lead to desired outcomes. For instance, journalists might feel that it is important to make contact with people in their profession who are well known and well respected and thus further their own careers. 19 Expectancy theory is generally based on a simplified model in which work motivation is treated as the choice between different levels of job performance in light of anticipated outcomes (Connolly, 1976; Reinharth & Wahba, 1975; Mitchell, 1974; Lawler & Suttle, 1973; Mitchell & Albright, 1972; Hackman & Porter, 1968). The original model proposed by Vroom (1964) provides a framework for understanding how behavior may be assessed and directed to meet both individual and organizational needs (Nadler & Lawler, 1983). Expectancy theory has been tested primarily in derivative models of work motivation (Mitchell & Albright, 1972), job satisfaction and performance (Lawler & Suttle, 1973), and needs satisfaction (Salancik & Pfeffer, 1977). In their reviews of the conceptual and methodological issues in expectancy theory, Mitchell (1974) and Connolly (1976) both state that empirical support for the models tested has been uneven. They suggest that part of the variability in empirical support is due both to inaccurate representations of Vroom's basic model and to unreliable measures of the variables in question. However, the evidence for causal relationships between expectancy formulations and behavior is small but positive (Mitchell, 1974). The overall reliability of the constructs is good (coefficient alpha = .74). Only Dachler and Mobley's study (1973) has reported reliability coefficients of less than .40, primarily for test—retests of two months or longer. There is some support for the overall external validity of various expectancy models. Widely different measures, subject populations and criteria produce consistent positive results, although large amounts of variance have not been reliably predicted with expectancy models (Mitchell, 1974). 20 Lawler and Suttle (1973, p. 502) in their review of the theory, suggest "the theory has become so complex that it has exceeded the measures which exist to test it". Behling and Starke (1973) support this statement when they argue that enough questions have been raised to justify a reemphasis in research from extensions of the model to testing of the basic interactive relationship between expectations and behavior. The job mobility model which structures the research in this study is an examination of the theoretical links between motivating variables, expectations, behavior and outcomes. The model reflects the impact of individual forces such as age, tenure in the job, tenure in the profession of journalism and job satisfaction, and environmental forces such as work relations in the newsroom, and expectations for upward mobility. The model also examines the role of expectations in predicting the job performance and communication behaviors of journalists. These behaviors are hypothesized to lead to a specific outcome, vertical job mobility. The basic theoretical model of upward job mobility is illustrated in Figure 4. The proposed research is designed to further explore the relationship between communication behavior and vertical job mobility in an expectancy framework which takes into account the important predictors of mobility. FIGURE 4 THEORETICAL MODEL OF UPWARD JOB MOBILITY I I I I I I I I I I I DEMOGRAPHIC I I PERCEP—I I EXPECTA- I I BEHAVIORAL I I MOBILITY I I CHARACTERIS—I I TIONS I I TIONS I I VARIABLES I I OUTCOMES I I TICS AND I=>I I=>I I=>I I=>I I I WORK I I I I I I I I I I RELATIONS I I I I I I I I I I I I I I I I I I I CHAPTER 3 METHODS AND PROCEDURES This chapter describes the selection of the sample, the questionnaire design and administration, measurement, and the statistical procedures used to analyze the data, and also states the research hypothesis. Sample The sample was drawn from the newsroom staffs of 83 newspapers and seven television stations included in the Gannett Group. The sampled newsrooms mix geographic (regional) location, urban/non—urban locales, publication cycles, circulation sizes, market dominance, degree of competition from other newspapers, presence of a guild (union), presence of a readership committee or ombudsman, size of newspaper staff, history, and presence of a VDT system. Questionnaires were mailed to about 3,500 working journalists in the newsrooms of these organizations. A11 journalists in each newsroom were requested to voluntarily complete the questionnaire. Of the mailed questionnaires, 1,118 (30%) were returned in usable form and provide the sample data on which the present analysis was conducted. A demographic profile of the survey respondents is given below. -- The majority of the respondents are male (67%). —— The average age of the journalists is 35, with half (51%) of them falling within 25 to 34 years of age. The age distribution by category is: under 25 years — 11%; 25 to 34 years — 51%; 35 to 44 years — 21%; 45 to 54 years — 10%; over 55 years - 7%. —- The average newsperson has been in the industry for 12 years, with half (52%) reporting between three and 10 years of experience. The breakdown by years of experience is: under 3 years — 8%; 3 to 5 21 22 years - 22%; 6 to 10 years - 30%; 11 to 20 years — 24%; over 20 years — 16%. Staffers averaged eight years of work for their newspaper or station. The most frequent job title is "reporter/on—the—air newsperson" (32%). The breakdown of job titles is: reporter/on—the air newsperson — 32%; copy editor - 11%; columnist/staff writer — 10%; photographer/graphic artist — 5%; supervising editor (section/ assistant/desk)/assignment editor/news editor/news producer — 26% executive editor/editor/pub1isher/managing editor/news director/ station manager - 3%; other - 13%. Nearly seven out of 10 members of the newsroom staff (69%) hold a bachelor's degree. The distribution of education levels among the newsperson respondents is: high school or less — 4%; some college or A.A. — 14%; B.A. or B.S. - 69%; M.A. or M.S. - 13%; Ph.D. — <1%. Eight out of 10 of the newsroom employees who earned a journalism degree cited reporting, news, and editorial as their major area of interest. The major interest areas of those with journalism degrees are: reporting/news/editorial — 33%; mass communication - 7%; broadcast journalism — 3%; photojournalism — 3%; advertising/public relations — 1%; journalism management - 1%; radio—TV—film — <1%; other — 5% The largest group of respondents with a nonjournalism degree are English majors (35%). The degree backgrounds of nonjournalism majors are: English — 35%; liberal arts/humanities — 20%; social science — 17%; business - 4%; natural science - 2%; technical field — 2%; other — 20%. About half (52%) of the respondents are employed by the smaller newspapers (those with circulatians under 50,000). The distribution 23 of journalists by size of newspapers, and within TV stations, is: less than 50,000 circulation — 52%; 50,000 to 99,999 — 30%; over 100,000 - 15%; TV stations — 3%. Procedures The questions studied in the Gannett research project concerned the influence of naturally occurring phenomena on the interactions of people as they pursue their daily activities. To obtain their reports about these experiences, a survey research design was considered most appropriate for this study. The surveys were distributed to the Gannett newsrooms by mail. Before the surveys were mailed to the newsrooms, the project was announced in the company newsletter and all journalists were encouraged to participate. At each newsroom, a contact person distributed the questionnaires to the journalists and returned the completed surveys to the researchers. Respondents sealed their completed surveys in envelopes which were collected at a central location within the newsroom and then mailed to the researchers for analysis. Measurement In the survey, a series of questions assessing communication practices was included in a larger questionnaire. That questionnaire focused on several issues in the journalists' work environment, including job mobility, work relations, and job satisfaction. [The questionnaire also contained items on variables beyond those of specific interest in this study (these other items served the concerns of other researchers collaborating in the research project).(4)] The discussion below describes the measurement of the independent and dependent variables in the job mobility model. Unless otherwise specified, all variables are measured on a five—point Likert—type scale. 24 All variables were measured on an interval level metric. Communication Practices_gf Journalism. The survey contained 10 items to assess the actual communication practices of the respondents with both journalists and nonjournalists, the nature and frequency of these contacts, and their perceived need to communicate with persons and organizations outside the newsroom. Together, the items were expected to reveal a journalist's insularity from the public —— e.g., the extent to which the person had fewer contacts, perceived less community awareness and/or believed that community integration is undesirable. Figure 5 lists these items. Figure 5 COMMUNICATION PRACTICES OF JOURNALISTS Off the job, I have a lot of contacts with peOple in walks of life very different from my own. The demands of this job leave little time for social contacts with people outside the newsroom. I regularly seek out nonjournalists' opinions about the news. I think it is vitally important that a journalist be integrated into the local community. I have enough contacts with people outside the newsroom to have a good feel for what is going on in the community. The majority of my friends are associated with the news businesss in one way or another. I am involved with a variety of groups and people in the community. My co—workers and I frequently discuss news stories with one another. I feel well informed about what is going on here at this paper/station. I have a lot of influence on decisions that are made about what stories are covered and what kind of play they receive. Work Relations. Also included in the survey were 16 items focusing on work relations. These items addressed issues related to supervisor— 25 subordinate relations, supervisor-subordinate similarity, and conflict within the newsroom. Since the "fit" of an individual with his or her work environment is one determinant of upward mobility, the work relations items were designed to assess the degree to which an individual feels a part of (or at odds with) the organization. The work relations items are listed in Figure 6. FIGURE 6 WORK RELATIONS My supervisor and I frequently discuss my work. Communication among people on this staff is strained. Editors on this staff work to develop the creative potential of the reporters. People in this newsroom are more interested in their own career advancement than in producing a good news product. The information I receive from bulletin boards, memos and other internal communicatin is insufficient. My professional goals are often at odds with the demands placed on me in this organization. There is not enough leadership by top management in this newspaper/station. I feel I receive sufficient guidance and feedback from my supervisors. There is a spirit of cameraderie in this newsroom. There is a difference between the way my supervisor thinks things should be done and the way I think they should be done. My supervisors in this newsroom share my values. My supervisors are different than me. I have a hard time satisfying the conflicting demands of people I work with. There is a definite lack of clear policy and guidelines for doing one's job here. 26 FIGURE 6 WORK RELATIONS (CON'T) This organization really inspires the very best in me in the way of job performance. I find that my values and this paper's/station's values are very similar. Job Satisfaction. Job satisfaction is closely related to work relations, but is not synonymous. Satisfaction with one's job -— with the actual day—to——day work involved in the job -— is separate from the issue of interpersonal relations in the work environment. To assess job satisfaction, a modified version of the Job Description Index (Smith, Kendall & Hulin, 1975) was used to measure employees' satisfaction with the nature of their work. The 18 items which form this scale are included in Figure 7. FIGURE 7 JOB SATISFACTION Listed below are words and phrases that may describe your current job. For each item, either circle Y (for Yes) if it describes your job, N (for No) if it does not describe your job, or 2 if you can't decide. The work that I do is: Fascinating Y 2 N Good Y 2 N Pleasant Y 2 N Routine Y 2 N Creative Y 2 N Useful Y 2 N Satisfying Y 2 N Respected Y 2 N Tiresome Y 2 N Boring Y 2 N High Pressure Y 2 N Healthful Y 2 N Challenging Y 2 N Frustrating Y 2 N Endless Y 2 N Important Y 2 N Simple Y 2 N Gives sense of Z accomplishment Y 2 Perceptions. Since the perceptions an individual holds of the company where a person works and of the industry in general are expected to contribute to an individual's desire for upward mobility, several items were included to analyze journalists' attitudes toward their 27 organization and toward the field of journalism. A five—point semantic differential—type scale was used by respondents to rate their paper/TV on several issues concerning the quality of news reporting and the paper's position in the community. This scale was designed to assess the journalists' current image of their organization. To measure their outlook toward the future of their industry, another 10 items on a five— point Likert—type scale were also incorporated in the questionnaire. Figures 8 and 9 list both the image and the outlook scales. FIGURE 8 PERCEPTIONS —— IMAGE OF THE NEWSPAPER/TV STATION Below are a number of adjectives and phrases that have been used to describe news organizations. For each pair of words or phrases, please rate your newspaper/station as you see it: Inaccurate 1 2 3 4 5 Accurate Doesn't have the latest news 1 2 3 4 5 Has the latest news Not courageous 1 2 3 4 5 Courageous Impersonal 1 2 3 4 5 Personal Can't be trusted 1 2 3 4 5 Can be trusted Not concerned about the Concerned about the community's well-being 1 2 3 4 5 community's well—being Does sensationalize 1 2 3 4 5 Doesn't sensationalize Dull 1 2 3 4 5 Lively Not respected 1 2 3 4 5 Respected Doesn't act as a community Does act as a community watchdog 1 2 3 4 5 watchdog Incompetent 1 2 3 4 5 Competent Doesn't represent the whole Does represent the whole community 1 2 3 4 5 whole community Biased l 2 3 4 5 Unbiased Uninteresting 1 2 3 4 5 Interesting FIGURE 9 PERCEPTIONS —- OUTLOOK TOWARD THE FUTURE I feel that in the future, newspaper readers will turn more and more to broadcast media for news and information. The future for the newspaper industry looks rather gloomy. It is likely our local daily newspaper will lose readers, relative to our 28 population, in the next five years. I feel we have been making progressive changes in content and format in our coverage of news. It is quite likely that I will remain in the field of journalism for as long as I can work. People read our daily newspaper because they have no alternative. The public has more confidence and trust in the local newspaper than in local television news. Our community has a very favorable image of our newspaper/station. Our news presentations have an important impact on readers'/viewers' attitudes toward political and social issues. Advances in new technology will give the newspaper a stronger competitive edge in the future. Job Performance. Two questions were used to assess the journalists' job performance. First, respondents indicated whether they had received a merit raise in the past two years, using three response categories: No; Yes, once; and Yes, more than once. The second question asked if they had been formally evaluated in the past year, and if so, how they were evaluated. These performance items and their scoring are presented in Figure 10. FIGURE 10 JOB PERFORMANCE Have you received a merit raise in the last two years? No Yes, Once Yes, More Than Once (1) (2) (3) Have you received any formal evaluation in the past two years? No Yes —— If Yes, how were you evaluated? (0) Very Negatively 1 2 3 4 5 Very Positively Expectations. Two questions addressed the journalists' expectations for upward mobility in their careers. One question asked if 29 they expected to hold their present position five years from now. If they thought this was somewhat or very unlikely, a second question asked whether they expected to be promoted, change to a lateral position, or to leave the profession. The first item was measured on a five point Likert—type scale and the second contained five response categories. The two items were combined and recoded to measure the degree to which each journalist expected upward job mobility, from low mobility to high. The questions are listed in Figure 11, along with the recoding scheme. FIGURE 11 EXPECTATIONS FOR UPWARD JOB MOBILITY, AND AGE AND TENURE I — EXPECTATIONS FOR UPWARD JOB MOBILITY 1. How likely do you think it is that you will hold the same position five years from now (whether here or in another location)? Very Likely 1 2 3 4 5 Very Unlikely 2. If Somewhat or Very Unlikely —— Do you expect to be promoted, to change to a lateral position here or elsewhere, or to leave the profession? Be Promoted Change Position Leave This Not Other Laterally Profession Sure I 2 3 4 5 Recoding Scheme: If the response to Q1 equals (1) or (2), Mobility Expectations = 0 If the response to 01 equals (3), ..... ...Mobility Expectations = 1 If the response to Q1 equals (4) or (5), and the response to Q2 equals (1), ...... ...... ......... .Mobility Expectations = 2 If the response to Q1 equals (4) or (5), and the response to Q2 equals (2),..... ................... Mobility Expectations = 0 If the response to Q1 equals (4) or (5), and the response to Q2 equals (4), ........................ Mobility Expectations = l Mobility Expectations: Low Expectations = 0 Medium Expectations = 1 High Expectations = 2 30 II — AGE AND TENURE Age What is your age? (In Years) Years i§_Industry How many years have you worked in the newspaper/television industry? (In Years) Years Ip_Present Job How long have you held this present job? (In Years) Have you been promoted in the last two years? No Yes, Once Yes, More Than Once 1 2 3 Age and Tenure. Three demographic items asked journalists to give their age in years and to indicate both the number of years they had spent in the journalism industry and the number of years they had held their present job. Figure 11 includes these questions. Job Mobility. Job mobility, the dependent variable in this analysis, was measured by the number of promotions a journalist received in the last two years. This question is also found in Figure 11. An exploratory factor analysis was run on all the items included in the independent variables described above. Factor analysis was used both as a data reduction tool and to verify the conceptual grouping of the items. Factor analysis is a procedure for evaluating whether a set of variables has a smaller number of underlying dimensions which account for its correlations. Factor analysis reduces a complex set of variables into one or more groups or clusters and can provide insight into the underlying dimension which each cluster represents. Factor loadings indicate the extent of the relationship between the cluster and each variable or item. The researcher must interpret the results and 31 determine whether there are meaningful patterns in the clusters. Factor analysis can provide a clearer understanding of the observed relations in the data by reducing the data to a smaller set of factors. The items which comprised the initial set of variables formed a 74 x 74 correlation matrix. This matrix was subjected to a principal components analysis with communalities in the main diagonal, followed by a varimax rotation of factors having eigenvalues greater than or equal to one (Nie, Hall, Jenkins, Steinbrenner, & Bent, 1975). The resulting 16 factors accounted for 57.1% of the variance in the correlation matrix. The 16 factors corresponded closely to the eight concepts presented above in the discussion of questionnaire design and measurement. Next, the variables in the 16 factors were further analyzed to provide additional confirmation of the factors as underlying indicators of the concepts. First, each of the 16 factors was explored conceptually for patterning in the variables. Variables which grouped together both conceptually and statistically were identified. In some cases, the variables in one of the initial 16 factors were selected out for subsequent analysis, while in other cases, the variables from more than one factor were isolated and factor analyzed. When this step was completed, 11 sets of variables were then analyzed in 11 separate principal components factor analyses (with communalities in the diagonal and varimax rotation of factors having eigenvalues greater than or equal to one). The number of factors expected to result from each analysis was specified a_priori, on the basis of the conceptual and statistical grouping which had just been completed. The confirmatory analyses yielded 14 significant factors having eigenvalues greater than one. Only those analyses which yielded factors that accounted for 35% or more of the variance were selected for the next 32 stage of the analysis. Of the 14 factors from the confirmatory analyses, 10 were selected for the next analysis stage, following the theoretical framework of expectancy theory and the conceptual focus of this research on job mobility. The final 10 factors forming composite variables for regression analysis are listed in Table 2, along with the communalities, eigenvalues, and percentage of total variance accounted for in the set of variables subjected to each confirmatory analysis. TABLE 2 FINAL FACTORS -— VARIABLES IN REGRESSION EQUATIONS Factor 1_:_Demographics (Age, Tenure ig_Industry, Tenure ig_Job2 Factor Items Communality Loading Eigenvalue What is your age? .76 .92 2.34 How many years in industry? .77 .95 How many years in job? .34 .60 Percent of Variance Accounted for: 78.0 Reliability - Coefficient Alpha: .86 Factor 2_:_Job Satisfaction Factor Items Communality Loading Eigenvalue Fascinating .27 .52 4.49 Satisfying .45 .68 Challenging .42 .63 Creative .29 .55 Gives sense of accomplishment .4O .54 Percent of Variance Accounted for: 37.0 Reliability — Coefficient Alpha: .79 Factors 34_4_and 5_:_Work Relations Factor Items in Factor 3 — Sup/Sub Similarity Communality Loading Eigenvalue There is a difference between the way my .36 .58 6.43 supervisor thinks things should be done and the way I think they should be done. My supervisors in this newsroom share my .45 .61 values. My supervisors are different than me. .32 .60 33 TABLE 2 FINAL FACTORS -— VARIABLES IN REGRESSION EQUATIONS (CON'T) .42 .42 Factor Loading Eigenvalue I have a hard time satisfying the con— .32 flicting demands of people I work with. I find that my values and this paper's/ .51 station's values are very similar. My professional goals are often at odds .40 with the demands placed on me in this organization. Percent of Variance Accounted for: 40.2 Reliability — Coefficient Alpha: .80 Items in_Factor 4_:_Conflict Communality Communication among people on this .43 staff is strained. People in this newsroom are more inter— .27 ested in their own career advancement than in producing a good news product. .65 1.18 .50 .65 Factor Loading Eigenvalue There is a spirit of cameraderie in this .43 newsroom. Percent of Variance Accounted for: 7.4 Reliability - Coefficient Alpha: .73 Items ip_Factor 5_:_Leadership Communality My supervisor and I frequently discuss .28 my work. Editors on this staff work to develop the .46 creative potential of the reporters. There is not enough leadership by top .41 management in this newspaper/station. I feel I receive sufficient guidance and .45 feedback from my supervisors. There is a definite lack of clear policy .51 and guidelines for doing one's job here. This organization really inspires the very .58 best in me in the way of job performance. Percent of Variance Accounted for: 7.1 Reliability - Coefficient Alpha: .83 Factors 6 and 7 — Outlook .55 1.13 .58 .47 .64 .46 .49 Factor I feel that in the future, newspaper .21 readers will turn more and more to .55 1.28 34 TABLE 2 FINAL FACTORS —— VARIABLES IN REGRESSION EQUATIONS (CON'T) broadcast media for news and information. The future for the newspaper industry .28 .67 looks rather gloomy. It is likely that our local daily news— .15 .45 paper will lose readers, relative to our population, in the next five years. Advances in new technology will give the .17 .40 newspaper a stronger competitive edge in the future. Percent of Variance Accounted for: 12.8 Reliability - Coefficient Alpha: .61 Factor Items in Factor Z_:_Optimistic Outlook Communality Loading Eigenvalue I feel we have been making progressive .19 .40 2.64 changes in content and format in our coverage of news. People read our daily newspaper because .15 .40 they have no alternative. The public has more confidence and .17 .41 trust in the local newspaper than in local television news. Our community has a very favorable .27 .71 image of our newspaper/station. Percent of Variance Accounted for: 26.4 Reliability — Coefficient Alpha: .54 Factor 8.: Image 9f_Newspaper/Station Factor Items Communality Loading Eigenvalue Inaccurate/Accurate .38 .46 6.64 Can't be Trusted/Can be Trusted .52 .70 Not Concerned About the Community's .44 .63 Well—being/Concerned About the Community's Well—being Dull/Lively .57 .40 Doesn't Care What the Public Thinks/ .38 .59 Does Care What the Public Thinks Incompetent/Competent .57 .60 lDoesn't Represent the Whole Community/ .42 .48 Does Represent the Whole Community IBiased/Unbiased .33 .53 I’ercent of Variance Accounted for: 41.5 Reliability - Coefficient Alpha: .84 35 TABLE 2 FINAL FACTORS —— VARIABLES IN REGRESSION EQUATIONS (CON'T) Factors 9 and 10 — Communication Patterns Factor Items 12 Factor_9‘: Communication Communality Loading Eigenvalue with Public Off the job, I have a lot of contacts .42 .75 2.76 with people in walks of life very different from my own. The demands of this job leave me little .17 .42 time for social contacts with people outside the newsroom. I regularly seek out the nonjournalists' .19 .43 opinions about the news. I have enough contacts with people out— .31 .62 side the newsroom to have a good feel for what is going on in the community. The majority of my friends are associated .28 .54 with the news business in one way or another. I am involved with a variety of groups .33 .66 and people in the community. Percent of Variance Accounted for: 27.6 Reliability — Coefficient Alpha: .74 Factor Items ip_Factor 19_:_Communication Communality Loading Eigenvalue with Journalists My coworkers and I frequently discuss .18 .55 1.76 news stories with one another. I feel well—informed about what is .20 .60 going on here at this paper/station. I have a lot of influence on decisions .15 .49 that are made about what stories are covered and what kind of play they receive. Percent of Variance Accounted for: 17.6 Reliability — Coefficient Alpha: .59 Note: In all factors, negative items have been reverse scored. Two criteria were used to retain a variable within a factor: (1) the variable had a strong conceptual tie with the underlying dimension of the factor, and (2) the variable loaded .40 or greater on the factor. A variable was deleted if it did not discriminate among the factors, i.e., loaded highly on more than one factor. 36 On the basis of the results of the confirmatory analysis, the composite indices were created by summing the unit—weighted scores of the items in each factor which met the above criteria. Reliability analysis was conducted on the 10 indices. Table 11 also reports the coefficient alpha for each index. The indices were labeled by the underlying conceptual dimension they appeared to represent. These labels will be used to refer to these composite variables in the following discussion of statistical analyses and results. Procedures for Statistical Analysis The data were analyzed using the routines of the Statistical Package for the Social Sciences (SPSS). Initial frequencies were run to determine measures of central tendency and distribution for each variable. Scattergrams were run, and indicated that the variables were linearly dependent. Overall, the results of this analysis indicated that most variables approximated a normal distribution. Three variables were skewed somewhat positively; Mobility Expectations and Demographic Characteristics of age and tenure on the job and in the industry (independent variables) and Job Mobility (a dependent variable). The measurement of the expectation and mobility variables allowed for limited variance (a three—point measurement scale) and most respondents did not have high expectations for mobility and had received only one or no promotions in the past two years. Demographic— ally, respondents were in their mid—30's (35 years of age, on the average), and had been in the industry an average of 12 years and at their present jobs an average of four years. Their Job Satisfaction was negatively skewed, indicating that journalists are generally highly satisfied with their work. (See Appendix A for descriptive statistics for all the variables included in the regression analysis.) 37 Regression analyses were conducted to test the main research hypothesis and to evaluate the mobility model as a whole. Hierarchical and stepwise solutions were employed conjointly. Hierarchical regression models include the variables in the regression equation as specified_a priori. These a_priori specifications were established according to theoretical, causal or logical considerations. A hierarchy among sets of variables was specified: First, the demographic characteristics of journalists and the work relations variables are considered to logically precede any variables related to future attitudes or expectations about the job, therefore, this block of variables entered as specified in the expectancy model. The second block of variables included the image and outlook variables which reflected respondents' perceptions about their paper/TV station and their attitudes toward the future of the journalism industry. The variable measuring respondents' expectations for job mobility was entered into the regression equation third, since the preceding variables are thought to heavily influence expectations about job mobility. And finally, job performance and communication patterns, hypothesized to be largely influenced by expectations and to lead to actual mobility, were entered into the equation as the fourth block of variables. In stepwise estimations, variables are typically entered into the analysis beginning with the variable that explains the largest amount of variance in the dependent variable. This is followed by the variable that explains the largest amount of variance not already accounted for by the variable just entered into the regression equation. This procedure is continued until all the variables have been entered. Therefore, variables are entered and removed from the equation based on their unique contribution to the dependent variable. Since no specific causal 38 ordering was designated for the variables within the first block but a notable influence from them was expected, these variables were entered in a stepwise fashion so that each variable could account for its unique contribution to the dependent variable. Three main regression analyses were performed. First, job mobility was regressed on all the variables in the full model. Then each of the two major portions of the model was analyzed, the portion of the model preceding and that following the expectations variable. In the second regression analysis, expectations was regressed on the image and outlook variables, and the demographic and work relations variables. Third, job mobility was regressed on the performance and communication variables and expectations for mobility. The same entry procedures were followed in all three analyses. No direct link between the variables preceding expectations and the dependent variable is hypothesized in the expectancy model in this study. Therefore, a regression analysis specifically testing this relationship was not performed. Hypothesis A research question regarding the relationship between communication patterns among journalists and upward job mobility was stated in Chapter One. The following hypothesis constitutes a testable restatement of this relationship: H: The greater the amount of journalists' communication with their colleagues the greater their upward job mobility. CHAPTER 4 RESULTS, SUMMARY, AND DISCUSSION The results of the regression analysis are presented in Tables 3, 4 and 5. Table 3 lists, by step, the variables entered into the regression equation to test the full model of Upward Job Mobility (Figure 12). Table 4 presents the results of the regression analysis testing Part 1 of the model (Figure 13), and Table 5 indicates the results for Part 2 of the model (Figure 14). An "overall" test for the goodness of fit of the regression equations and the expectancy model of upward job mobility was performed, and the contribution of each variable to the prediction of job mobility was examined. SPSS Regression provides values for the F test of statistical significance. If statistically significant, this test indicates that there is a significant relationship between the set of predictors and the criterion variable. In all the regression analyses, the probability level of .05 was used to establish statistically significant results. The significant beta weights at the p g_.05 level will be denoted by an asterisk (*) in the tables reported with the results of the regression analysis. The analyses are described below, followed by a discussion of their implications for the overall Upward Job Mobility model and the limitations to interpretation of the results. Recommendations for future research are also presented. RESULTS Full Model 9f_Upward Job Mobility. The full model is presented in Figure 12 below, followed by Table 3, in which the results of the regression analysis of this model are presented. In this model, the dependent variable is Upward Job Mobility, and the dependent variables 39 40 are those presented in the four blocks of variables leading to the dependent variable. FIGURE 12 FULL REGRESSION MODEL OF UPWARD JOB MOBILITY IDemographics I IOutlook I IMobilityI IJob Perfor—I IUpward I 1(Age & Tenure)I=>I(Pessi— I=>IExpecta-I=>Imance 1&2 I=>IJob I IJob Satisfac- I Imistic & I Itions I ICommunica- I IMobilityI Ition I IOptimisticI I I Ition with I I I IWork RelationsI IImage I I I IPublic I I I IConflict I I I I I ICommunica— I I I ISupervisor/ I I I I I Ition with I I I I Subordinate I I I I I IJournalistsI I I I Similarity I I I I I I I I I BLOCK ONE: BLOCK TWO: BLOCK THREE: BLOCK FOUR: DEPENDENT VARIABLE: Demographic Perceptions Expectations Behavioral Mobility Characteristics Variables Outcomes And Work Relations The overall accuracy of a prediction equation is given by the R— square statistic, the proportion of variance explained in the dependent variable by the indepdendent variables included in a regression equation. In Table 3, the R-square statistic and the other results of the analysis of the full model are presented. The independent variables in the full model of Upward Job Mobility accounted for 18.6% of the variance in the dependent variable. The overall contribution of the independent variables was significant at the p_g .05 level with an overall F of 14.91. RESULTS OF FULL REGRESSION MODEL OF UPWARD JOB MOBILITY 41 TABLE 3 Step Variables in_the Equation Blockul 1 Demographics (Age, Tenure) 2 Job Satisfaction 3 Work Relations (Leadership) 4 Conflict 5 Supervisor/Sub. Similarity Block_2 6 Pessimistic Outlook Optimistic Outlook Image Block-3 7 Mobility Expectation Block_4 8 Job Performance 1 Communication with Public Job Performance 2 Communication with Journalists Multiple R_ = .431 R—Sguare = .186 Overall F = 14.91 Significance = .002* Significant at p < .05 Beta —.OO6* .022* .000 —.021* .005 —0004 .008 —.005 .117* .096* .010* .005 .048* 95% R- Conf. Interval Square P(—.OO8 I I =====> I I I Subordinate I I Image I I I I Similarity I I I I I I Conflict I I I I I I Job Satisfac— I I I I I I tion I I I I I BLOCK ONE: BLOCK TWO: DEPENDENT VARIABLE: Demographic Perceptions Expectations Characteristics And Work Relations The results of the analysis are shown below in Table 4. The prediction equation has an R—square of (.083) and an overall F of (9.63), which is significant at the p < .05 level. Only two of the eight variables have Beta weights significant at the p < .05 level —- Demographic Characteristics (B=-0.009) and Work Relations/Leadership (B=0.021). Work Relations makes the largest contribution to the prediction of Mobility Expectations compared with the other variables in the equation. 45 An examination of the changes in R—square shows that Demographic Characteristics (age and tenure) and Work Relations/Leadership account for almost all (7.7%) of the explained variance in the dependent variable, Mobility Expectations. Demographic Characteristics, which was entered into the equation first, accounts for the largest percentage of the variance (R—square change = 4.6%). Work Relations/Leadership was entered the equation second, and has an R—square change of 3.1%. The remaining variables in the equation together account for only 0.6% of the total explained variance in Mobility Expectations. TABLE 4 RESULTS OF PARTIAL REGRESSION MODEL OF UPWARD JOB MOBILITY ——MODEL 1 Step Variables R- R—Square ip_the Equation Beta 95% Conf. Interval Square Change Block_1 1 Demographics —.009>-'< P(—.012 1 Communication I === l Mobility I I I I with Public I I I I I I Communication I I I I I I with Journalists I I I BLOCK ONE: BLOCK TWO: DEPENDENT VARIABLE: Expectations Behavioral Mobility Variables Outcomes The results of this analysis are presented below in Table 5. The full regression equation accounts for 13.7% of the variance in the dependent variable, Upward Job Mobility [overall F is 27.21, significant at the p < .05 level]. Four of the five independent variables in the equation are statistically significant. Mobility Expectations is the best predictor of Upward Job Mobility (B=O.146), followed by Job Performance 1/Merit Raise (B=0.099), Journalists'Communication with Colleagues (B=0.049) and Communication With the Public (B=0.017). Job Performance 2/Performance Review was not significant (B=0.016). An incremental partitioning of the variances indicates that Mobility Expectations, the first variable to enter the regression equation, accounts for the largest portion of the explained variance in upward job mobility (R—square change = 0. 053). Journalists' Communication with 47 Colleagues, after controlling for Mobility Expectations, Job Performance 2/Performance Review and Communication with the public, accounts for an additional 4% (R—square change = 0.042) of the explained variance. Job performance 1/Merit Raise, after all the independent variables have been entered into the equation, accounts for 1.5% (R—square change = 0.015) of the variance. Job Performance 2/Performance Review and Communication With the Public, after controlling for Mobility Expectations, contribute about 2% to the total R—square (R—square change = 0.013 and 0.014, respectively). TABLE 5 RESULTS OF PARTIAL REGRESSION MODEL OF UPWARD JOB MOBILITY ——MODEL 2 Step Variables R— R—Square ip_the Eqpation Beta 95% Conf. Interval quare Change Block 1_ 1 Mobility .146* P( .O92