THESIS 9‘ $3:- «uuzw. ~ 43"." 'I This is to certify that the dissertation entitled WORK MOTIVATION IN ZAMBIA: AN EXPLORATORY STUDY presented by PETER DAVID MACHUNGWA has been accepted towards fulfillment of the requirements for PR‘B . P _\ ‘ r\ degree in 343:, Inx ujtj / \ / ”/ if L;/v/1/AZ /. 51/6 :1 [711$ L w Major professor DateJ/gy /’PZ ’T MS U is an Affirmative Action/Equal Opportunity lrutimtion 042771 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIOIIIIIIIIIIIII — 3 1293 10415 i)V1£3l_] RETURNING MATERIALS: pTace in book drop to LJBRARJES remove this checkout from .-1—. your record. F__I____NES W111 be charged if book is returned after the date stamped below. WORK MOTIVATION IN ZAMBIA: AN EXPLORATORY STUDY By Peter David Machungwa A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Deparmment of Psychology 1981 ABSTRACT WORK MOTIVATION IN ZAMBIA: AN EXPLORATORY STUDY By Peter David Machungwa Behavioral science research conducted mostly in industrially advanced countries has helped develop several theories to explain work motivation. Little of this research was done in develOping countries. However, now that developing countries are trying to industrialize and develop their economies, they face major problems, one of which is the motivation of their work forces. Today, problems of motivation continue to be major and are evident from the astonishingly low productivity of many developing nations. Although cross-cultural researchers have studied motivation across many countries, their research has had severe limitations: cross-cultural research has stressed comparisons across cultures rather than practical applications which developing nations need most. Further, cross-cultural research attempts to validate theories developed on one culture in other cultures. While such theories may do well enough in one culture, they may only be partially, if at all, explanatory in other cultures. Finally, most cross-cultural motivation research has included relatively few developing countries. This study investigated work motivation in the developing country of Zambia to try to: (1) determine the degree to which work motivation in Zambia fit into current motivation theory; (2) develop a theory of motivation if the data so compelled; (3) determine factors that could be manipulated to enhance motivation and productivity in Zambia; and (4) conduct a validity check on the concepts that would emerge from the data. Critical incidents were collected from 341 Zambian employees representing five occupational groups from 11 organizations. Con- tent analysis of the data produced motivational themes which were used to construct a validation questionnaire that was administered to an independent sample of 80 employees. With respect to the validity check, questionnaire responses confirmed the content-analyzed results. Factoring of motivational themes, performed independently by three judges, produced six motivational factors that showed some support for components of some process and content theories, but also included elements that appear unique to Zambia. Five of the six factors showed potential to increase and impair motivation, but each factor could increase motivation more than it could impair motivation or vice versa. A differential positive-negative impact model of work motivation is proposed. Practical and research implications of the findings are discussed. TO MY FATHER ACKNOWLEDGMENTS Throughout my stay at Michigan State University, I have been lucky to have met, known and worked with a number of people who have given me support, encouragement, and needed criticism that has aided in my deve10pment. I am greatly indebted to Dr. Neal Schmitt, who is co—chairperson of my doctoral committee and also served as chairperson of my Masters committee. Dr. Schmitt has especially contributed to my learning and deveIOpment throughout my program at Michigan State. His teaching, guidance and particularly his patience and willingness to answer questions, and offer constructive and practical sug- gestions has proved invaluable to my career thus far and those of many of my colleagues to have worked with him. Special thanks go to Dr. Fred Wickert, the other co-chairperson of my committee. I have greatly profited from Dr. Wickert's wide experience and knowledge of developing countries. His painstaking efforts and patience in reviewing with me and his criticism of various ideas, suggestions, designs, and materials have contributed to not only the quality of this dissertation, but also to my approach to dealing with various ideas. I would like to thank Dr. Larry Foster, who also served as a member of my Masters committee. Dr. Foster, though being geo- graphically isolated from other members of my committee (being in iii another department), has always made very constructive suggestions on my research and has been willing to meet and discuss with me ideas pertaining to my work and when I was looking for teaching appointments in his department. My further thanks go to Dr. Terry Allen, who has made very useful suggestions about my research. Dr. Allen's encouragement, comments, criticisms, and suggestions on various aSpects of my research have definitely helped to improve my work. His ideas on the practical aspects of my research have provided me with en- couragement for future work in this area. I would also like to thank my colleagues, especially Rich Strand and Tom Mitchell, who as fellow graduate students, have shared similar experiences with me and offered professional and social support and willingly helped me factor analyze data for this research. I am grateful, too, to the personnel at the Staff Development Office at the University of Zambia for their help and support while I collected data. I am particularly thankful to the managing directors of Roan and Nchanga Consolidated Copper Mines (Messrs. David Phiri and Francis Kaunda) whose organizations together with the University of Zambia provided funding for this research pro- ject. Special thanks go to my research assistant, Davies Kangwa, tdxo worked so diligently during data collection in various cities in Zambia. Finally, these acknowledgments would not be complete if I omitted to mention the emotional support I have received from my former girlfriend and now my wife, Judi Sue Sweezer-Machungwa. iv And lastly, I feel indebted, an I always will, to my brothers and sisters and especially my mother (late) and father who encouraged me to get to where I am. With all such assistance and support from so many people, all shortcomings in this paper are all mine. TABLE OF CONTENTS page INTRODU“ ION C O O O I O O O O O O O O O O 1 Aim Of This S tudy O O O O O O O O O O O 5 I. THEORIES OF MOTIVATION. . . . . . . . . . . . 7 Content Theories . . . . . . . . . . . . . . 8 Need Hierarchy Theory. . . . . . . . . . 8 NO‘FflCtOI Theory 0 o o o o o o o o o o o 10 Need for Achievement . . . . . . 12 Cognitive Process Theories . . . . . . . . . . . . . 14 Goal setting Theory I O O O O O O O O O O O O O O O O 14 Equity Thea” C C O O O C I O O O O O O O O O O O O 0 l6 Emectancy Thea” U C O O C O O O O I I O O O O O O O 17 A Cognitive Process Theory . . . . . . . . . . . . . . 19 Drive Reinforcement Theory . . . . . . . . . . . . 19 II. CROSS-CULTURAL RESEARCH . . . . . . . . . . . . 22 Leadership 0 O O O C O O O I O O O O O O O O I O O 22 Interpersonal Relations in Organizations . . . . . 24 Exposure to Work Organizations and Worker Attitudes O O O O O O O O O O O O O O O O O O O O O 25 Rank, Education, and Age . . . . . . 27 Pay 0 O O O C O O O O O O O O O O O O O 0 O 28 Job Characteristics. . . . . . 29 M18 cell-means O O O O O O O O O O O O O O O I O O 30 Colonial Heritage. . . . . . . . . . . . . . . . . . 30 Social Change. . . . . . . . . . . . . . . . . 31 corruption 0 O O O O O O O O O O O O O O O O O O 31 Attitude Towards Time. . . . . . . . . . . . . 32 “Dice Of CO‘mtr-y O O O O O O O O O O O O O O O O 32 III 0 2mm. 0 O O O O I O O O O I O O O O O O O O O O O O 33 Geography, History and Economy . . . . . . . . . . . 33 People 0 O O O O I O O O O O O O O O O O O O O O O O 36 Hypotheses . . . . . . . . . . . . . . . . . . . . . . 37 List of Possible Factors . . . . . . . . . . . . 38 IV. V. VI. VII. METHODOLOGICAL PROBLEMS . . . . . . . . . . Possible Techniques. . . . . . . . . . More Useful Techniques . . . . . . The Critical Incident Technique and Wbrk Motivation: The Herzberg Study. . . . . METHOD. . . . . . . . . . . . . . . . . Materials. . . . . . . . . . . . . Subjects . . . . . . . . . . . . . . . . Organizations. . . . . . . . . . . . . . Data Collection. . . . . . . . . . . . . Data Analysis. . . . . . . Content Analysis . . . . . . . . . . . . Validity Check Questionnaire . . . . Scoring. . . . . . . . . . . . . . . . . Formation of Clusters. . . . . . . . . . Correlational Analysis . . . . . RESULTS . . . . . . . . . . . . . . . . . . Structure of Clusters Identified . . . . . Mbtivating Factors (from Good Critical Incidents). Demotivating Factors (from Bad Critical Incidents) Relationships with Biographic and Other Variables. DISCUSSION. . . . . . . . . . . . . . . . . The Factors. . . . . . . . . . . . . . . . Growth and Advancement Opportunity . . . Wbrk Nature and Context. . . . . . . . Material and Physical Provisions . . . . Relations with Others. . . . . . . . . . Fairness in Organizational Practices . . Personal Problems. . . . . . . . A Differential Positive-Negative Impact Model of Wbrk Motivation. . . . . . . . . . . Mbtivation and Other Work-Related Variables. Education. . . . . . . . . . . . . . . . Occupational Level . . . . . . . . . . . Age. . . . . . . . . . . . . . . . . . Years of Paid Employment . . . . . . . . Length of Urban Residence. . . . . . . . Sex. . . . . . . . . . . . . Motivation and Employee's Own Perceptions of mtivates Thw O O O O O O O O I O I O 0 vii page 38 39 4O 41 43 43 44 46 46 48 48 50 51 52 53 54 60 62 63 68 72 73 74 76 81 85 87 90 93 97 98 101 103 104 105 107 109 VIII. CONCLUSIONS AND DIRECTIONS FOR APPENDIX 1. . APPENDIX 2. O O O 0 APPENDIX 3. . . . . LIST OF REFERENCES. viii FUTURE RESEARCH . page 111 116 117 118 120 Table 1. 2. 3a&b LIST OF TABLES Mbtivational Variables Identified through Content Analysis of Data . . . . . . . . . . . . . . . . Mean and Standard Deviation Scores of Mbtivational Factors (Items) Included in the Validity Check Questionnaire. . . . . . . . . . . . . . . . . . Pearson Product Mbment Correlations between Frequency of Mention of Motivational Variables from the Content Analysis and the Means of the Same Variables Included in the Validity Check Questionnaire . ix page 55 57 58 Figure 1. LIST OF FIGURES Page Diagrammatic Representation of the Six Factors Identified Showing the Motivating and Demotivating Items Clustered under Each Factor . . . . . . . . . . . 61 Diagrammatic Representation of the Differential Positive-Negative Impact Model. . . . . . . . . . . . . 96 3': INTRODUCTION The importance of work motivation for employees, organizations and society as a whole cannot be overstated. Productivity of organizations and entire nations depend in part on the motivation of those nations' work force. In developing countries where productivity is comparatively lower than that in developed countries, the issue of motivation assumes added significance. Certainly motivation alone without accompanying technology and employee abilities and aptitudes is not sufficient to boost productivity. But that motivation is so essential an ingredient of productivity is best illustrated in Richman's (1967) study of managers from China, India, Russia and the United States; Richman wrote that "Thus far, Red China has achieved substantial industrial progress more because of managerial motivation and attitudes than because of managerial technical know-how," (p. 70). The importance of work motivation has been long appreciated in the behavioral sciences. Research especially in the field of psychology has led to the develoPment, testing, and refining of theories and constructs of work motivation over the last few decades (Steers and Porter, 1975; Chung, 1977; Campbell and * . Supported by research grants from the University of Zambia and Roan and Nchanga Consolidated Copper Mines. 1 Pritchard, 1976; Lawler, 1973). These theories have been widely applied in work organizations. But the bulk of this research has been conducted in, or on samples from, deveIOped countries. Only little, if any, of the research was done in developing countries. For example, two extensive literature searches conducted inde- pendently by myself and Baklien (1980) found only one reported study pertaining to work motivation in the central-east African region. The result has been that work motivation theory and related constructs developed so far, and even the methods of data collection are more suited to the characteristics (social structures, values and habits, technology and educational levels) of the industrialized countries. But today underdeveIOped nations are attempting to industrialize and develop their economies. These countries can learn and borrow technology from deve10ped countries and can enhance the aptitudes, skills and abilities of their workforce through training. But the vital ingredient of motivation in the development process cannot be easily increased without understanding what will motivate the people being put to work. Unfortunately - as pointed out by Heller (1969) and supported by my own observations - developing nations tend to stress development of infrastructures like roads, railroads, electrical power, etc. as a precondition for rapid economic development without due concerns for factors like moti- vation, values and habits that guide behavior of people in organizations. The implicit assumption of stressing non human factors only appears to be that people in these countries will respond to organizational work situations in the same manner as 3 their counterparts in industrialized nations. This assumption, true only to some extent, ignores the fact that..."the motivation of members to belong, work and advance in the organization may be different from one society to another" (Tannenbaum, 1980, p. 283). And this is clearly evident in the problems of work motivation facing many developing nations today. For example in the young African nation of Zambia, problems of work motivation cannot be more adequately illustrated. Political leaders have on many occasions, complained about the lack of will to work among workers. Leaders have continued to call for "increased productivity," disci- plining of lazy and/or "inefficient" employees. Recently, the President of the country himself highlighted problems of work motivation in the workforce: If by next year all the five million Zambians Choose to be lazy as they are now, President Kenneth D. Kaunda fumed last year, I would willingly step down because I don't want to lead people with lazy bones (Wall Street Journal, February 15, 1979, p. 9). It must be pointed out, however, that cross-cultural researchers in organizational psychology and related disciplines have, for some time now, attempted to learn about work motivation in many countries including some developing nations. (See Barrett and Bass, 1976; Tannenbaum, 1980; for a review.) But the work of these resear- chers have severe limitations. First, cross-cultural research tends to emphasize comparisons across cultures rather than the work motivation factors that apply in a specific culture or country. As a result findings from these studies may not readily lend themselves to practical applications within any specific nation--and deveIOping nations need knowledge on applications. Second, cross-cultural research on work motivation may only provide fragmentary information at best or useless information at worst because most of this re— search is aimed at validating one or two theories (developed on one culture) in a number of cultures. Such theory or theories may do well enough in the one culture but may tap only a small part of the underlying motivational mechanisms or patterns in other cultures or may not be valid at all. Finally, most cross-cultural studies of applicability to work motivation problems have been comparisons among developed nations and the more advanced of the developing nations such as Bulgaria, Israel, India, Mexico (see Barrett and Bass, 1976; Tannenbaum, 1980, for a review), rather than among the less deve10ped nations. Failures and weaknesses in cross- cultural research in the field of organizational psychology led Roberts (1970) to conclude an evaluative review of organizational related research with the statement that "...increment of know- ledge seems minimal and possibly not worth the effort thus far placed in cross-cultural work," (p. 345). Although Roberts' assessment may be too harsh, there is need for investigators wanting to fully understand psychological phenomena like work motivation in different cultures to shift their approach. What is needed is an empirical approach that does not limit itself to some given a priori theoretical framework(s). Such an approach must begin by observing and examining productive and non-productive work behavior in work organizations in regions or cultures thus far not well studied, then proceed to deve10p theoretical concepts around the observations. Such an approach has major advantages: from a theoretical standpoint, analysis of data collected using this approach can allow us to determine the degree to which work motivation in those cultures fits existing theory, and provide material to build new theory for those regions or modify current theory if the data so suggests. Secondly, and more important for the develOpment needs of those regions, this approach would provide information that can be used to increase work motivation in the areas studied. Thus, this approach aims at theory building, practical applications and allows indirect cross-cultural comparisons. Aims of this Study: This research investigated work motivation in Zambia by examining conditions, events and processes surrounding highly productive and highly unproductive work behavior. That is, a critical incident technique (Flanagan, 1954) was used. Content analysis (see Holsti, 1968; Crano and Brewer, 1973) was the method of data analysis. As a check on the construct validity of the findings, motives or factors that were identified in the content analysis were used to develop a questionnaire. This questionnaire was administered to a sample of employees. The employees were asked to indicate-on the basis of their personal work experience--which of the factors (de- rived from content analysis) tended to be associated with situations or incidents when they worked very hard and those with situations when they put very little effort into their work. Although the preposed investigation was not based on any particular theory of work motivation, it is important nevertheless to take a close look at the state of current motivation theory. This is necessary because the aim of this study, among other things, is to determine the degree to which data from a developing nation (Zambia) fits into some existing theory or theories of work moti- vation, or how such data may be used to adapt such theory to the culture and unique conditions of an underdeveloped nation. Thus, we need to examine contemporary theories of work motivation. In the next section, seven such theories are briefly reviewed. Dis— cussion then returns to cross-cultural research where it is noted that despite the weaknesses pointed out earlier, cross-cultural investigations have produced some important findings. This is followed by a presentation of background information on Zambia. Discussion of methodological problems then follows and this leads into the method, results and discussion sections of the paper. I. THEORIES OF MOTIVATION Mbtivation, the process by which behavior is energized, directed, sustained and regulated has been studied for many years. As early as 1789, the philosopher Bentham.coined the phrase "hedonic calculus" to describe the process by which individuals weigh advantages and disadvantages of their behavior. Later, psychologists like James and McDougall (instinct theory), Freud (unconscious theory) and Thorndike (drive theory) proposed more comprehensive theories of motivation at the turn of the century and early this century, (see Cofer and Appley, 1964; Atkinson, 1964 for reviews of early moti- vation models). It is to these early works that today's motivation theories owe their roots. In the summary that follows, discussion concentrates on those formulations which have been used to explain work behavior. Before presenting the theories, a basis for clas- sifying them is first outlined. Contemporary models of motivation can be differentiated on the basis of whether they postulate conscious mental processes (cognitive) or whether they contend that behavior is regulated by past habits or reinforcement history (acognitive). This distinction has some significance for this study. For it can be expected that greater differences across cultures are more likely with theories positing that behavior is determined by conscious decisions of the individual than with theories maintaining that behavior is a function of rein- forcement history. Another basis for distinction is whether the models focus on identifying specific factors within the individual or environment which determine behavior (content) or if they focus on processes by which behavior is controlled (process). In the summaries below, content theories are discussed first, cognitive process theories next, then the single acognitive process theory is presented. Content Theories Need Hierarchy Theory: Maslow (1943, 1954) explained human motivation in terms of a dynamic hierarchy of needs. He postulated a set of five needs - basic physiological, safety, social, esteem and self-actualization - arranged in increasing order or prepotency. According to Maslow, individuals will be motivated to satisfy physiological needs before safety, social and other higher order needs can be fulfilled. Gratification of physiological needs triggers the emergence of safety needs which when satisfied give way to the more psycho- logically oriented social needs. This process continues until the high order needs of esteem and later self-actualization are grati- fied. Thus, relative gratification of a given need submerges it and activates the next higher order need which continues to organize and dominate the individual's personality and capacities until it (need) is satisfied. Maslow also proposed a frustration hypothesis; that non satisfaction of a given need for a long time will cause a fixation for that need. Although Maslow‘s need hierarchy concept is one of the oldest of today's theories of work motivation and has been used to explain and prescribe management practices (McGregor, 1960, 1967; Schein, 1965) and job enrichment (Hackman and Lawler, 1971; Hackman, 1977), it has not received much empirical support. While some studies like those of Aronoff (1967) in the British West Indies and Aronoff and Messé (1969) working with American college students found support for Maslow's formulation, the bulk of research findings have been non supportive of propositions from the theory. Disconfirmatory evidence for propositions of the Maslow model has led some researchers to suggest some modifications. Alderfer (1969, 1972) proposed a three-level (existence, relatedness and growth) need hierarchy theory and this received some empirical support from.Wanous and Zwany (1977). Based on his earlier empirical research (Lawler and Suttle, 1972), Lawler (1973) suggested..."A two-step hierarchy with existence and security needs at the lowest level and all the higher- order needs at the next level," (p. 34). Later, Lawler and Suttle (1975) added a modification that..."prepotency of higher level needs for an individual would be determdned by such things as his child- hood experience, his age..." (p. 45). Some researchers on the other hand found little if any support at all for the need hierarchy model. Rauschenburger, Schmitt and Hunter (1981) found not support whatso- ever for the need hierarchy concept. Cross-cultural finding_: Cross-cultural research on the need concept has met with equally inconsistent findings. If Maslow's formulation was correct, then we could have expected people from underdeveloped economies to be more concerned with satisfying lower order needs than higher order needs. Tannenbaum (1980), succinctly stated this point: ..."'psychological' (quotation marks his) support would be irrelevant in economically disadvantaged societies where lower level needs are not fulfilled; support in such places would have meaning only in terms of actions that contribute directly to physical well being of organizational members," (p. 301). While some studies supported this view, some did not. 10 After studying IBM personnel in poor and rich nations, Greenwood (1971) concluded that job security was more prepotent in economically poor nations while needs for autonomy and challenging work were rela- tively more important in wealthy nations. The study of Singh and Wherry (1963) in India had also come to the same conclusion. But Orpen and Ndlovu (1977) found that black clerical workers in South Africa had stronger higher order needs than whites despite the fact that..."B1acks in South Africa form a disadvantaged caste...(for whom) lower order needs for shelter and for security are only occasionally met," (p. 32). In the same vein, a study by Haire et a1. (1966) involving 3600 managers from fourteen countries did not give strong support to the Maslow position. Despite these weaknesses and perhaps because of its simplicity, the need hierarchy concept has been and continues to be widely used in organizations and has recently been incorporated into modern job design theory (Hackman and Lawler, 1971; Hackman, 1977). For pur- poses of this study, it would be interesting to see, and would pro— vide strong empirical support, if a motivational pattern similar to the need hierarchy formulation emerges from data to be collected. Two-Factor Theory: Another approach to work motivation is that advanced by Herzberg et a1. (1966) following a content analysis of critical incident data collected from.200 American accountants and engineers. This approach, called the two—factor theory states that job satisfaction is quali- tatively different from job dissatisfaction and that these two are caused by two independent sets of factors. The factors that affect satisfaction, the motivators, tend to be more intrinsic and include 11 interesting and challenging work, recognition and advancement, responsibility and achievement. Factors affecting dissatisfaction, the dissatisfiers, are extrinsic and include pay, supervision, interpersonal relations, working conditions and company policies. According to Herzberg, the presence of satisfiers in sufficient amounts will result in employees being satisfied and motivated to work but will not affect dissatisfaction. Similarly presence or absence of dissatisfiers will only affect dissatisfaction and not satisfaction. Support for the two-factor theory has not been strong. Only those studies that used the same methodology as Herzberg have given support to the theory, (Wexley and Yukl, 1977). On the other hand Herzberg and his associates have been accused of equating satisfaction to motivation by asking respondents to talk about good or bad feelings about their jobs and inferring moti- vation from resulting data (Vroom, 1966). Another criticism is that since coding during the content analysis was not entirely de- termined by the coding system but required interpretation by coders, bias could have been introduced into the data during the analysis (Wigdor and House, 1967). Further, Herzberg's claim that satisfaction is affected only by satisfiers and dissatisfaction only by dissatisfiers has been challenged and many researchers now accept that both factors affect satisfaction and dissatisfaction (see Lawler, 1973). Probably the harshest criticism of the two- factor theory is that from Dunnette, Campbell and Hakel (1967) who stated that "It seems that evidence is now sufficient to lay the two factor theory to rest, and we hope that it may be buried peaceably," (p. 173). 12 Despite these criticisms, Herzberg's theory appears to have some validity. The two-factor theory together with Maslow's (1954) need hierarchy concept are cornerstones of modern job design/redesign theory (Hackman and Lawler, 1971; Hackman, 1977) which has received substantial empirical support in field settings. Because the present study will use a methodology similar to the one used by Herzberg, this study may partly be seen as an attempt to replicate Herzberg under a different cultural, social, economic and political environment. Nevertheless this study is different not only in methodological improvements over Herzberg's (see section on method- ological problems) but also in the main assumption that motivation patterns in third world countries might be greatly different from those in the West and consequently might be different from.what Herzberg found. Need for Achievement; Building on Murray's (1938) list of human needs, McClelland (1961), McClelland et al. (1953), and his associates have develOped a theory of human motivation centered around the need for achieve- ment. The achievement motive, as the need for achievement is often referred to, is perhaps the most thoroughly researched individual human motive. According to McClelland, need for achievement is a relatively stable disposition to strive for success. A person with high need for achievement is more intrinsically motivated to perform highly on tasks of medium difficulty and where feedback is available than on extremely difficult or easy tasks with no feedback. Such a person derives more pleasure from accomplishment and more frustration from failure than peOple with low need for achievement. 13 Unlike the two-factor theory just discussed, the need for achieve- ment construct has received wideSpread empirical support. Research findings in both laboratory and field have demonstrated a positive relationship between need for achievement and level of performance in a variety of settings including job success of executives, (Atkinson, 1958; Atkinson and Feather, 1966; McClelland, 1951; McClelland, et a1., 1953; Cumming, 1967; Hundal, 1971; Steers, 1973; Weiner and Rubin, 1969; Weiner and Kukla, 1970). Even though need for achievement is a relatively stable personality variable, McClelland (1965) proposed that a person's need for achievement could be increased through training. He was able to demonstrate this in India though this effort met with limited success (McClelland and Winter, 1969). Cross-cultural finding_; The need for achievement concept has probably been subjected to more cross-cultural investigation than any other formulation of human motivation. In research extending over two decades and covering many develOped and underdeveloped countries, McClelland (1961) and his associates investigated the relationship between need for achievement within a nation and the nation's level of economic growth. McClelland concluded from this research that need for achievement-dmeasured by projective tech- niques-ewes positively related to the level of economic growth measured by amount of electrical output in a given country.‘ Despite the impressive support, the need for achievement frame- work has not been without its critics. Cofer and Appley (1964) cautioned that..."The theory McClelland and his co-workers have developed is neither compelled nor directly derived from their data, 14 but is presumably consistent with the data," (p. 374). McClelland's cross-cultural results linking the achievement motive to economic growth of nations has been questioned by some; Barrett and Franke (1971) raised concern about use of electrical output as an indicator of national economic growth and the assumption implicit in McClelland's research that countries of all sizes, at different levels of economic growth and technology were all affected by need for achievement in the same way. A reanalysis by Barrett and Franke (1971), showed that findings were dependent on time period and measure of economic growth chosen. A finer breakdown of findings revealed that need for achievement related to economic growth only in small already developed countries. Besides these criticisms, McClelland's attempt to link one micro variable (need for achieve- ment) to a macro variable (national growth) is not sufficient to explain the total picture of what motivates productive or non- productive work behavior especially in cultures where work moti- vation has received no systematic investigation. Nevertheless, because of the widespread support the need for achievement construct has received in many developed and developing nation cultures (McClelland, 1961), it is highly probable that achievement need may emerge as a factor of work motivation in this study. Cognitive Process Theories Goal-SettingTTheory: A relatively new formulation of work motivation has been ad- vanced by Locke (1968) based on Lewin's (1935) level of aspiration concept. According to Locke, goals and intentions regulate a person's behavior. Definite goals such as Specific deadlines or is levels of performance lead to higher performance than general goals such as "do your best" or "work as fast as possible." Further, difficult goals result in better performance than easy goals as long as the individual accepts and commits himself to attaining the goals.' Locke further added that though such variables as task characteristics, incentives, supervision and feedback may have some effect on performance, these variables have little or no effect at all when differences in goals and intentions are controlled for. Locke (1968) and his associates have supported the theory with a large number of findings from laboratory studies. Mbre recently, some field studies have also supported goal setting theory propo- . sitions (see Latham and Yukl, 1975). But goal setting theory is not without weaknesses: Firstly, Locke's model contradicts need achievement theory which predicts that achievement-oriented persons (who are also likely to be the high performers) tend to choose and perform well on tasks of medium difficulty. The theory is also in direct disagreement with another cognitive process theory (eXpectancy theory) which would predict greatest motivation on easy tasks where success and therefore rewards are certain-of course, assuming rewards to be contingent on performance. Secondly, Locke's goal- setting theory says little on how goals, expectations or intentions are established in the first place (Howell, 1976); evidence pre- sented by Miner and Dachler (1973) suggest that people prefer goals they expect to be able to reach. Related to this point is perhaps the strongest criticism of goal-setting theory. That is that, when goals are extremely difficult to attain, there may be little commitment on the part of the individual to attain such goals. In a: .3 h. t‘ 21‘ 16 such a situation, goals and intentions may have no effect on per— formance. Unfortunately, none of the studies presented by Locke and his associates varied goal difficulty to near impossible levels to answer this issue (see Dobmeyer, 1970). Equity Theory: A theory formulated with a focus on social interactions in work situations is equity theory (Adams, 1963, 1965). Adams stated that an individual brings into the work situation inputs (such as effort expended, education, training, age, sex, seniority, status, ethnic background) in return for outcomes (such as pay, fringe benefits, job status and rewards intrinsic to the job). The individual com- pares his ratio of outcomes to inputs to that of a "comparison other." If the two ratios match, equity is perceived and the individual is not likely to change his inputs or outcomes. If, however, the ratios are perceived to be unequal, the individual experiences tension and will be motivated to rid himself of the tension and thereby eliminate or reduce the inequity. Adams suggested various modes individuals may use to reduce inequity. These include cognitive distortion of inputs or outcomes, actual alteration of inputs or outcomes or withdrawal from the situation. Thus according to equity theory an employee perceiving himself as underrewarded may reduce work quality or quantity (whichever of the two removes the inequity), demand more pay, be absent or tardy for work or in extreme situations quit the job. On the other hand, overrewarded employees can be eXpected to in- crease work quality or quantity to restore equity. 17 A considerable amount of research has been conducted on equity theory in laboratory, simulated and actual work situations. Recently Welster et a1. (1976) have extended the theory and used it as a general theory of social behavior. They used the equity framework to explain social interactions in dating preferences, helping and exploitative relationships. Although not all studies testing equity theory predictions produced supportive results, the findings from most studies have generally been supportive (see Goodman and Friedman, 1971; Carrell and Dittrich, 1978; Machungwa, 1978; walster, et a1. 1976 for comprehensive reviews). Some of the most frequently observed effects are that persons underpaid under a piece-rate system increase work quantity at the expense of quality to achieve equity. On the other hand, individuals perceiving overpayment inequity under piece- rate increase quality without increasing quantity. Other findings are that overpaid hourly-employees increase quality and/or quantity of their work while those underpaid reduce these aspects of performance to regain equity (see Goodman and Friedman, 1971 for detailed review). The concept of equity is an interesting one to examine in cross- cultural contexts. Because of the political ideology of some countries the equity norm may not be as salient as it might be in other countries. In the proposed study the concept of equity is not directly addressed; however, the approach of this study - starting with no a priori theoretical framework - may very well bring in the equity norm if the idea of equity is salient in the culture to be studied. Expectancy Theory: Drawing heavily on the works of Lewin (1938), Tolman (1959) and Peak (1955), Vroom (1964) proposed a cognitive theory of motivation 5“ 18 to explain work behavior. Like many other versions of expectancy theory (e.g., Campbell, Dunnette, Lawler and Weick, 1970; Graen, 1969; Porter and Lawler, 1968), Vroom's theory posits that a person's behavior results from conscious choices based on a comparative evaluation of possible alternatives. The central concepts in the theory are expectancy (the perceived probability that effort leads to desired performance), valence (the desira- bility of a given event or outcome), outcome (any potential need— related consequence of behavior) and instrumentality (the likelihood that a given level of performance leads to desired outcomes). According to the theory the motivation for an individual to perform a given task is a function of the expectancy that effort leads to required performance, the likelihood that performance leads to outcomes and desirability of the outcomes in question. Mathematically, the theory can be stated as follows: MF 2 EZVI ‘where MF - motivation or force to act, E = expectance that effort leads to performance, V a valence of outcomes and I = the instrumen- tality. Expectancy theory is considered by many to be a leading theory of work motivation. Numerous research studies have been conducted to test prOpositions of the theory. But partly because of the theory's complexity and problems of Operationalizing the key concepts within the model (see Mitchell, 1974) many of the studies had methodological problems and did not provide an adequate test of the theory, or yielded only equivocal findings. Although expectancy theory still seems a promising model for explaining work behavior, l' u h OF A’) v 5‘. . _'§vo’:- r. o' ‘ . .D’?‘ Ab‘ AN: . On- ii I ou$¥e hunt I." v Q! ' . ?-Of' '..ub> § B 5; L.“ ’5 I h 11‘ ‘ r‘M‘ ‘. 19 a number of methodological issues, logical and conceptual problems within the theory itself (Mitchell, 1974, Wahba and House, 1974) need to be resolved before the theory's rather promising explanatory potential can be successfully realized. Another problem is that peOple may not think like the theory suggests and may use a more simplified approach, (Simon, 1965). Acqgnitive Process Theogy Drive-Reinforcement Theggy: Whereas the last three theories discussed treat behavior as having roots in conscious mental processes, drive reinforcement theory holds that behavior exhibited is a result of past reinforce- ment history and stimulus-response relationships. The basis of the theory is Thorndike's law of effect which states that reSponses closely followed by pleasureable events (rewards) are more likely to recur in similar situations while events closely followed by unpleasurable events (punishment) are unlikely to occur in similar situations. Drive-reinforcement theory went through a number of elaborations with the work of Bull (1943 and later Skinner (1963, 1969) who is probably the most well-known and strongest preponent of the theory. According to drive-reinforcement theory, behavior will be strengthened if it is followed closely in time by positive rein— forcement of rewards. Conversely behavior can be weakened or extinguished by withdrawal of rewards and/or presentation of punishment. Rewarding of individuals only part of the time (partial reinforcement) rather than all the time leads to acquisition of behavior more resistant to extinction. However, acquisition of . and" . .» .yd- I a» b 5 '5 "V a: a com .9.» no 0 ch. Pb A $232: 3' ‘1 en fr‘ VsbI 20 behavior is faster when individuals are rewarded each time they emit the required behavioral response. Hundreds of laboratory studies with animal subjects have been conducted. These studies have generally provided impressive support for drive reinforcement theory. Results from studies in work settings have yielded support though this support has not bee as strong as that from the laboratory (see review by Schneier, 1974). Use of pay incentives (a type of positive reinforcement) has been observed to result in higher productivity (Lawler, 1971). Use of drive—rein- forcement principles has also proved effective in reducing absenteeism (Lawler and Hackman, 1969) and tardiness (Hermann et a1. 1973). Other studies such as the one at the Emery airfreight company showed that feedback and praise reinforcement could be used to shape appropriate job behaviors (Anacom, 1973). Drive—reinforcement theory has met with a lot of criticism and resistance. Whyte (1972) has argued that while drive-reinforcement theory has proved explanatory in the laboratory, the theory's pre- dictions are unlikely to hold in the complexities of real life. According to Whyte, drive-reinforcement theory ignores the social comparison process (issues of equity). That even though rewards may be contingent on performance, performance cannot increase and might even decrease if much inequity is perceived. Other researchers have criticized drive-reinforcement theory on the grounds that it advocates an approach to motivation solely based on externally mediated rewards - with total disregard for self-administered rewards (Likert, 1967; Vroom and Deci, 1970; Deci, 1971, 1972). According to Deci (1972), L. L t“: I... .038: I . “av 'I .‘ufiub ' q no .. "86 is 53-. 56115? \I PE??? I o' 21 the drive-reinforcement model of work motivation aims at satisfying lower order needs (Maslow, 1954) and ignores higher order needs. In spite of these criticisms and the opposition this theory has faced from advocates of cognitive models, drive-reinforcement theory can, perhaps, be expected to hold better cross—culturally than cogni- tive and even content theories. Since the reinforcement approach ignores mental processes which are influenced by such things as values, attitudes and beliefs, the effect of culture on this approach is somewhat limited. Thus it can be expected that each individual's behavior (irresPective of culture) will be determined by his unique past reinforcement history. While the design of the proposed study does not permit a direct test of drive-reinforcement theory prOpo— sitions, it is probable that critical incidents from this study will show themes that support the non cognitive theory. Summary: The fact that all theories just reviewed have explained work behavior under certain conditions despite weakness and contradictions illustrates the complexity of human motivation. Behavior is nearly always multi-determined, at times by seemingly conflicting motives. The very nature of this complexity tends to render inadequate an approach relying on a single theory to explain behavior. But with the empirical approach this study pr0poses, the hope is that the data will yield a framework which itself is a product of the complex interactions of motives that determine work behavior. Thus it can, perhaps, be expected that a combination of interaction of variables from the theories reviewed (if these are salient in the country to be studied) plus other work-related and cultural variables may emerge . T“. $5.0 22 from the data. In the next few pages, cultural and work-related variables that influence motivation are discussed. II. CROSS-CULTURAL RESEARCH Employee motivation to work cannot be fully understood without considering other work-related variables that have an impact on employee reactions in work situations. Many of these variables show their influence on work behavior through their effect on work moti- vation. In this section, cross-cultural research that investigated variables known to influence motivation in work settings are re— viewed. Earlier in this paper, some weaknesses of cross-cultural research in organizational psychology (work motivation in particular) were outlined. But despite these weaknesses and failures, cross- cultural research has produced findings that must be noted by any investigator whose work has some cross-cultural aspect, as is the case with the planned study. Studies reviewed here are chosen on the basis that they investigated variables previously shown to have influence on work motivation. Another criterion used is the relevance of variables and questions addressed in the study to the culture and country where the present research will be conducted. This latter criterion is based on the author's own knowledge and familiarity with Zambia. The reader seeking a comprehensive review of work- related cross-cultural research should consult Tannenbaum (1980), and Barrett and Bass (1976). Leadership: Leadership styles and their effectiveness have been extensively investigated in developed countries and a nwmber of theories have ?D 23 been advanced. Behavioral theory of leadership has propounded two leadership dimensions which can be stated as an interpersonal orien- tation, and a task orientation (Bales and Slater, 1955; Fleishman and Harris, 1962; Katz et al., 1950). Interpersonally oriented leaders are considerate, have respect for, and build mutual trust for their group while task oriented leaders tend to structure, define, assign tasks, set deadlines and emphasize performance. Research results concerning the effectiveness of each style or combination of both styles have been as inconsistent across countries as within countries. Thiagarajan and Deep (1970) found a positive relationship between participative supervision and employee satisfaction among workers in Belgium, Italy, United Kingdom and the United States. On the other hand, Meade (1967) observed in India that children led by a democratic leader were absent more often, turned out poorer quality work and liked their leaders less than children in authoritarian—led groups. Similarly, Peruvian workers were found to like their leaders to the extent that they exercised close rather than general supervision (Whyte and Williams, 1963). Another issue pertaining to leadership is the disposition of managers towards participative decision making. Results generally indicate that managers from less developed countries tend to be less participative (Haire et al., 1963; Tannenbaum, et al., 1974; Negandhi, 1973; Bass, 1968; Barrett and Bass, 1970; Thiagarajan and Deep, 1970). Barrett and Ranke (1969) further reported that the preference by managers to use two-way rather than one-way communi— cation correlated .75 with level of economic development of a country. ft“. '1'1‘ A h n O 24 These differences in participative/autocratic styles of leader- ship between developing and industrialized countries have been ex- plained in terms of differences in orientation towards authori- tarianism. Despite some inconsistencies, the general view is that people in developing nations tend to have more authoritarian atti- tudes than their counterparts in developed nations. Hagen (1962) concluded an extensive literature review with the assertion that the harsh and unpredictable conditions in many traditional societies lead people to reliance on the judgment or will of some superior individual or person in authority. If the preference for nonparticipative leadership in developing countries is a universal phenomenon, then it can be expected in this study that good critical incidents will be more associated with non- participative leadership and close supervision. However, since non- participative leadership is not likely to lead to good interpersonal relations (which are stressed in developing nations, as will be seen in the next section), this leadership must be a type of paternalism that blends some consideration with autocracy. Paternalism is a phenomenon which is not strange to many developing countries that have been colonized. In many cases, colonial rulers, when not out— right autocratic, acted paternalistic to the colonial peoples. In the developing nations of Africa, paternalistic leadership had long been practiced by African chiefs and monarchs. Interpersonal Relations in Organizations: Being closely tied to cultural factors, interpersonal relations in work settings can be expected to differ widely across nations. But bureaucracy, which characterizes most work organizations in I'" ‘II vs CCQ 25 nearly all countries, does not consider cultural differences. Emphasis is placed on impersonality of contacts among job occupants (Weber, 1947). Cross-cultural investigators have sought to under- stand the differences in people's interpersonal contacts and reactions to others in organizations across nations. Findings sug- gest that individuals from developed countries tend to be more im- personal in their dealings than persons from less developed countries. For example, Triandis (1967) found that Americans working in Athens were perceived by Greeks to be "inhumanly legal- istic, rigid, cold and overconcerned with efficiency (p. 51)," and gave little weight to interpersonal relations. Similarly, Zurcher (1968) found that Mexican bank employees weighted obligations of friendship in work settings more than American bank workers. In the same vein, Triandis and Vassilou (1972) observed that Greeks attached more importance to recommendations by friends when hiring a new employee than did Americans. These findings are in line with my own observations in Zambia - a developing country. My obser- vations and work experience suggest that people attach greater importance to harmonious social relations with coworkers and superiors than would be the case in a country like the United States, for example. It is plausible, therefore, to expect that in societies like Zambia, interpersonal relations will have a strong impact on work motivation. Exposure to Work Organizations and Worker Attitudes: Populations of most developing nations have not been exposed to industrial organizations for as long as populations in industrialized countries. Some research work done mainly in underdeveloped nations has a u 4- 5y .1 26 has attempted to relate employee work attitudes to degree of exposure to industrial work organizations. It has been found that exposure to, and experience with work organizations appears related to com- patibility with demands of the industrial settings and good psycho- logical adjustment (Inkeles and Smith, 1974). In a longitudinal study, conducted in Argentina, Chile, East Pakistan, India, Israel and Nigeria, Inkeles and Smith found that, compared to peasants, persons who moved from the farm to a factory had better psychological adjustment as measured by feelings of personal efficacy, aspirations, openness to new experiences, belief in calculability of the world and trust in others to meet their obligations. Even more important was the finding that psychological adjustment improved with length of work in industrial settings. Research by Elden and Leviatan (1974) with Israeli agricultural and factory workers in the Kibutzim suggested that Inkeles and Smith findings were not due to mere dif- ference of working on land as opposed to working in the factory. If workers on the lands use modern technology and are exposed to a larger social environment similar to that which factory workers are exposed to, then there would be no difference in psychological ad— justment between land workers and those in the factory. It is difficult to speculate how exposure to and experience with work organizations would relate to work motivation, but one can hazard a guess that there would be a difference in motivation patterns between groups with more and those with less exposure. This study will check for such differences if any exist at all. 27 Rank, Education and Age: Consistent empirical support has been found in the USA for the position that rank is positively related to favorable job attitudes (Herzberg, Mausner, Peterson and Capwell, 1957; Machungwa, 1978). Cross-cultural studies have now confirmed that this finding holds across nations (Tannenbaum et al., 1974; Tannenbaum, et al., in press). A variable closely related to rank is amount of education; that is, people with more education generally hold higher ranks. But unlike rank, education per se, is not positively correlated with favorable job attitudes. 0n the contrary, analyses of research data in Austria, Italy, USA, Yugoslavia and Israel (Tannenbaum et al., 1974), indicated that education is negatively related to job satis— faction and attitudes towards the company. Quoting findings of other investigators, Tannenbaum (1980) states that similar results have been obtained in Brazil, Germany and Hungary. On the other hand, age is positively related to favorable attitudes towards both the job and company. On the basis of these results, it is plausible to expect that rank, education, and age differentials might be reflected in moti- vation levels and patterns. In fact a study by Machungwa (1978) with a U.S. sample found rank to be positively related to level of motivation as measured by amount of perceived equity on the job. The present study will among other things explore any bearing that rank, education and age may have on productive and nonproductive behavior in a developing country. 28 191: Due to the central role pay plays in the interactions in work situations and its relationship to work motivation and other work variables (e.g. job and need satisfaction, performance, etc. Lawler, 1971), even as selective a review as the present one would be in- complete if pay was not discussed. The study of Herzberg et a1. (1966) with a sample of 200 American engineers and accountants con- cluded that pay had a relationship to job attitude, i.e., inadequate pay increased dissatisfaction but had little effect on work moti- vation. On the contrary, there appears some consensus cross- culturally (at least in the few developed countries studied) that pay in itself is not related to job attitudes. Zdravomyslov and Jadov (1964) found with 2550 young Soviet workers that pay level per se had little influence on job satisfaction. Instead factors like job type influenced the pay satisfaction relationship. The two researchers observed that unskilled workers doing monotonous jobs were more dissatisfied with their jobs than skilled workers even though earnings of the unskilled employees were more than those of the skilled workers. These results have received support from Tannenbaum et a1. (1974) and their associates with data from the USA, Austria, Yugoslavia, Germany, Hungary and in Ireland to a limited extent. As earlier noted in our critique of cross-cultural studies, few of the poor, less developed countries were included in the studies referred to above. Because of the relative poverty of many developing countries, pay might be expected to assume greater signi— ficance. In those countries, it can be expected that pay will have 29 a strong link to attitudes towards the job and company. Further, we can speculate that pay will have strong influence on work moti- vation. Job Characteristics: The final work related phenomena of interest in the cross- cultural literature is that of job characteristics. Hackman and Lawler (1971) using an American sample of 200 telephone company employees found that jobs which provided more variety, autonomy, task identity and feedback led to higher employee motivation and job satisfaction both affectively and behaviorally. The effect was stronger for employees with higher order needs. These results have received support from cross-cultural investigations with hardly any conflicting results. It has generally been observed that people doing routinized, monotonous, fractionalized, physically tiring or even dangerous jobs that allow little learning and skill development tend to have more negative attitudes towards their organizations and jobs than employees in jobs providing autonomy, variety, feedback and task identity. These findings hold for the USSR (Zadravomyslov and Jadov, 1964), Israeli Kibbutzin, Italy, USA, Yugoslavia (Tannenbaum, et al., 1974) and Brazil (deSouza and deSouza, in press). According to the Tannenbaum et al. study, negative effects of jobs with poor characteristics persist even when effects of rank, edu- cation and age have been controlled for. The apparent universality of the relationship between job characteristics and job attitudes would suggest that such a relation— ship might exist in developing countries. This study will see if job characteristics are related to work motivation among Zambian workers. 30 Miscellaneous There are many more variables that have influenced and continue to influence motivation in developing countries. Because of the relatively little research conducted in these countries, many of these variables have yet to be identified. Those which have been identified have not been subjected to much empirical research. Con- sequently the nature of the relationship between such variables and task motivation is hardly understood. Most variables that will be examined here fall into this latter category. Although the assump- tion is that these variables are related to motivation, little attempt is made in this section to speculate on the direction and nature of this link. The hope here is that these variables will show up in the content analysis - if they are important determinants of work behavior in the country to be studied. Colonial Heritag_: As indicated earlier, many develOping countries were colonized. And even though they have attained independence, the colonial legacy or heritage (represented by such things as underdevelopment, little education and work skills, rigid organizational structures, authoritarian leadership) still lingers on. Thus organization of the work place and interactions therein continue to be influenced by the colonial history. Whereas the effects on motivation of some aspects of the colonial legacy, e.g., leadership, have been dis- cussed in an earlier section of this paper, effects of some aspects of the colonial experience are not well known. 31 Social Chang : Political independence has brought rapid social and technolo- gical change to many former colonial nations. Chances for higher education and training and resultant status and economic mobility are much greater. Among the more educated, there is a shift towards new values which, more often than not, tend to be those from indus- trialized countries. These values in many cases clash with tradi- tional values. Such conflicts in values can be expected to have an effect on people's reactions to work situations. Not all the effects of social change on motivation are unknown. For example the effects of training, education and exposure to industrial work organizations have been studied cross-culturally and have been dis- cussed earlier (see pages 25-27). However, the influence of con- flicting social values on motivation during this kind of social change is not readily predictable. Corruption: Associated with social change are issues of corruption in de- veloping nations. While corruption is not a monopoly of developing nations or any group of countries, there have been many charges of corruption in developing nations following independence. In work settings, corruption takes the form of bribery, favoritism, nepotism, tribalism or ethnicity in hiring, promotions, firings, work assignment and other personnel decisions. Such actions have a strong impact on the work motivation of the persons involved particularly those perceiving themselves victimized. 32 Attitude Towards Time: An issue that might have possible significance in work settings and which is only occasionally mentioned, is the attitude exhibited towards time by people from developing nations, especially those from Africa. Durand (1960) wrote from French Africa (on the basis of personal observations) that African employees did not consider the time element important in executing their duties as long as they completed the assigned tasks. Hardly any research in Africa has directly addressed this question to support or refute Durand's state- ment. Researchers who inquired into perception of time in Africa focused on such issues as, memory, choice between an immediate reward and a later reward of greater value (see Wober, 1975 for a summary). Since the time element is an important component of work performance in most organizational work situations, attitude toward time is certainly an important issue. While the design of the proposed study does not allow for a direct investigation of Durand's position, it is probable that critical incident data from this study may show the salience of worker attitude toward time if Durand's viewpoint holds in today's modern Africa. Choice of Country Since (as earlier stated) little research on work motivation has been conducted in many developing countries, the study being proposed here would be of theoretical and practical significance in many de- veloping countries. But as it is impossible to conduct a study of this nature in all countries where motivation is relatively unin- vestigated, one has to choose among these countries. A consideration of importance in making a choice is the researcher's familiarity and 33 contact with people and culture to be studied. The importance of this familiarity and contact with the culture a researcher is to work with in cross—cultural psychological research has been aptly stated by Irvine (1968) as follows: "To collect valid data, a psychologist needs adequate and sympathetic training in understanding a culture that is alien, complex and conceptually different," (p. 3). Suggesting that participant observer research would gain greater importance because of the need for contact and familiarity, Irvine added that there is really no substitute for first-hand contact with cultures other than the one the researcher grew up in. This view is also echoed by Wober (1975, p. 215) in his concluding remarks following a descriptive review of psychology research in Africa. Since the intending researcher in this study is Zambian, was born, grew up and lived nearly all his life in Zambia, choice of Zambia as the research site is most appropriate. Another reason for choosing Zambia for this study is that problems of work moti— 'vation have been so clearly evident (see pp. 34-35) that a systematic investigation of motivation is an imperative if this young nation is to make full use of its most valuable resources - human resources. III. ZAMBIA Gepgraphy, History, and Economy: Zambia, located in the Southern central region of Africa, has an area just under 300,000 square miles (approximately five times the size of the State of Michigan). The land-locked country, a former British colony, has a population of five million people. The country became a republic in 1964 and is now ruled by a one- party elected government. 34 The mainstay of Zambia's economy is the giant copper mining in- dustry1 which accounts for about 90% of the country's foreign ex- change earnings. Realizing the susceptibility of copper prices to sharp fluctuations on world markets, the Zambian government em- barked on an ambitious policy to diversify the economy. Manufacturing, processing, chemical, service, agricultural, etc. industries were set up. Such industries were partially (50% or more shares), or in a few cases totally owned by the government. Private interest or multi-nationals held the minority shares. Small private enterprises were also encouraged. Most, if not all, work organizations in Zambia are run bureaucratically. The new industries that had been set up soon ran into trouble. Despite initial success and massive injection of capital, the over- all performance of many of these organizations proved disappointing. Productivity was low and many of these companies continued to rely on government subsidies to meet operating costs. There were many reasons advanced for the mediocre performance of these companies. One among the chief reasons appeared to be problems associated with motivation. It had been noted that productivity in many organi- zations in the country including government ministries and depart- ments was low. Many political leaders including the president began to complain that many Zambians lacked the will to work; that it appeared managers and workers were for some reason unwilling to exert themselves. The leaders called for increased productivity and urged stronger disciplinary measures against "inefficient" and lZambia is the world's third largest exporter of copper. 35 "lazy" employees. Meanwhile the president threatened to quit if the people did not stop being lazy - a threat which he has repeated more recently (see quotation on page 3). In spite of this clear evidence that the work motivation concepts developed in industrialized countries appear to have failed in Zambian work organizations, no systematic inquiry to understand, build and apply newer work moti- vation concepts (if necessary) has been conducted in Zambia to date. In the last few years, copper prices plummeted to an all time low on world markets and forced closure of some copper mines in some countries including the United States. Consequently Zambia was earning little and in some cases incurring considerable loss on copper sales. The loss of copper revenue plus the military and economic demands of being a frontline state2 in the Southern African war of liberation made the situation worse. And with the continuous rise in oil prices and the resultant increase this has on prices of goods that Zambia imports from abroad, the fiscal situation in the country lately has been bordering on chaos. The country now appears to be experiencing a severe recession if not a depression and has had to borrow heavily from the international community. Inflation and unemployment especially among the less educated, unskilled people is fairly high. Under these conditions, it is plausible to expect that many em- ployees in Zambia would be concerned with job security and the pay 2Since its independence in 1964, Zambia has actively supported the liberation struggles of the people of Angola, Mozambique, Zimbabwe (until they gained independence), and Namibia. Consequently, Zambia has had to face and fight off attacks from Portuguese, Rhodesian (before Angola, Mozambique, and Zimbabwe were liberated) and South African troops. 36 they earn. The concern for job security would perhaps be greater for the less educated unskilled workers as these people can be easily replaced. If these suppositions are true, then it can be expected that these concerns about job security and pay might be reflected in the motivation of many Zambian workers. One can also expect, if these assumptions are correct, that concern for job security will be more pronounced in the motivation of the less educated and relatively unskilled workers. People: Although Zambia is a multi—racial society, over 90% of the popu- lation consists of indigeneous African peoples. There are about 72 distinct dialects but these can be grouped into four main linguistic groups. The extended family system is widely practiced throughout the country. However, urbanization and the money economy has con- siderably weakened the larger family unit so that the extended family tends to be strong only in the rural parts of the country. The people are generally friendly and, like most African peoples in the central African region, stress good interpersonal relations with friends, neighbors, workmates and other associates. This stems, I believe, from the way most traditional African societies were organ- ized and this same phenomenon has been noted by anthropologists (Gulliver, 1965; Sangree, 1965; p. 175). In those societies, it would be unheard of, for example, that a man did not know and had not interacted with his neighbors. Hence in today's Zambian society, good interpersonal relations with supervisors, subordinates and co- workers are considered by many to be an essential condition for an effective work atmosphere. Another factor, therefore, that can be 37 expected to affect work motivation - and this is in line with empirical evidence from cross-cultural studies reviewed earlier (see page 24) is interpersonal relationships. Before ending this outline on Zambia, an important point to note for purposes of the research being proposed is the level of education in the workforce. Zambia like most developing nations has a fair degree of illiteracy or semi-illiteracy on its workforce. A very high percentage of workers in low level, manual, unskilled or even some semi-unskilled jobs can be considered functionally non-literate, at least in English even though English is the country's official language. Therefore, a study of work motivation in Zambia that will include these employees cannot use methodologies that require reading or some other school-attained skills. Hypotheses An exploratory study (like the one proposed here) with no a priori theoretical framework(s) and using content analysis is not appropriate for testing hypotheses. This is particularly true when means of validating inferences from the content analyzed data are undetermined or unavailable (Crano and Brewer, 1973, chap. 9). Because of this limitation this study is not intended to pursue rigorous psychometric testing of inferences from the data to be collected. Such testing will be the subject of another follow-up study. Nevertheless, the research studies reviewed earlier and the information presented about Zambia and its people allows specu- lation on some possible factors expected to influence (increase or reduce) work motivation in Zambia. The list of factors presented below is not meant to be exhaustive; it is believed that the a 38 posteriori approach to be used in content analysis (see section on data analysis, p. 48) will enable the data themself to generate the factors determining work motivation. List of Possible Factors: 1. Interpersonal relations with supervisors, codworkers, subordinates 2. Pay (raises, bonuses, etc) 3. Job security or steadiness of employment 4. Type of leadership 5. Amount of education 6. Age 7. Length of employment in work organizations, length of urban residence 8. Occupational level 9. Equity issues 10. Corruption-related issues (nepotism, tribalism, favoritism) IV. METHODOLOGICAL PROBLEMS Researchers of work motivation in the industrialized countries have developed various techniques for assessing task motivation. But as many of these techniques and the mediums they rely on are more suited to conditions in the industrialized countries, they often are unsuitable for use in developing countries for a number of reasons. In this section, I argue that the critical incident technique and content analysis are the most appropriate tools to explore work moti- vation in Zambia. 39 Possible Techniques Questionnaires: Perhaps the most commonly used method to measure motivation has been the written questionnaire. This relatively inexpensive device has a number of problems. Questionnaires being perceptual measures may not tap the construct being investigated. Besides, these instru- ments are subject to faking, and various response sets. When these instruments are to be used in cultures other than those in which they were developed, their validity is highly suspect. Further, question- naires, especially those using abstract scales, may prove cognitively too demanding on subjects. Finally questionnaires assume a literacy level not possessed by considerable segments of populations in de- veloping nations. For these reasons written questionnaires were con- sidered unsuitable for the proposed study. Behavioral Measures: In some laboratory and field studies, motivation has been opera- tionalized as the amount of performance, effort and/or time invested into a given activity. Such behavioral measures while preferable are plagued with difficulties in determining what component of the exhibited behavior is due to motivation, ability or some personality trait. And in any case such techniques of assessing motivation are more useful when we work with some theory to predict or eXplain the behavior. In a study like the one being proposed where the investi- gation is not within the framework of a given motivation theory, the behavioral observation technique is inapprOpriate. 4O Projective Techniques: Another method used in motivational research is the projective technique. For example, the McClelland (1961) study, referred to earlier, examined contents of stories or responses to ambiguous photographs to determine the achievement motive within a culture. A serious problem with projective methods are their doubtful psycho- metric qualities (Anastasi, 1976). Even if this problem were ignored, McClelland's approach is unsuitable because it is theory specific. Besides, using the technique across cultures, as McClelland did, assumes people from various cultures project in similar a manner and to about the same degree - a risky assumption. Mbre Useful Techniques Interview: The interview is another technique that has been used to assess motivation. Unlike the written questionnaires, this method is usable even with pe0p1e with low or no reading skills at all. In the hands of a skilled person, the interview (structured or unstructured) can yield information that questionnaires or mere observation would never be able to get. Problems with interviews include their high cost and the possibility that bad rapport developed in the interview and poor probing would affect the quality and hinder the collecting of information. In addition, unreliability (disagreement among inter- viewers) is a problem with the interview. Critical Incidents: The critical incident technique developed by Flanagan (1954) has, among other things, been used to assess work motivation. The 41 method consists of collecting anecdotes of job behaviors which de- scribe especially good or especially bad job performance. The anecdotes (obtained through observation of job performance, or interview with employees) describe events leading to the incidents, exactly what the individual did that was so effective or ineffective, the perceived consequences of the behavior, and extent to which con- sequences were within the employee's control. The critical incidents approach has major advantages: since respondents are asked about objective events, data collected from this method tends to be more accurate than that from techniques relying on subjective or purely evaluative descriptions. This technique can be used with people with no formal education at all, and, like the interview, it is not cul- ture specific as would a questionnaire develOped in a given culture. In the present study, the critical incident technique is most ideal. It is a less obtrusive way to collect information on what determines productive and non—productive behavior than direct ques- tions like "what motivates you to work very hard?" And because some of the workers to be sampled lack the education and sophistication to deal with the abstract scales of written questionnaires that characterize most research on task.motivation, the critical inci- dent technique (using the interview form) is the compelling alterna- tive for this study. The Critical Incident Technique and Work Motivation: The Herzberg ass-21: A successful use of the critical incident technique was made by Herzberg, Mausner, and Smyderman (1966) to study work motivation. In what has become one of the most cited and controversial studies 42 on work motivation and job satisfaction, Herzberg et al., interviewed 200 accountants and engineers and asked them to: "Think of a time when you felt exceptionally good or exceptionally bad about your job...and tell me what happened," (p. 141). This request to sub- jects was followed with fourteen questions that sought to establish why and how the situation occurred, its consequences and significance to the subject. The process was repeated for both good and bad critical incidents. Using the semi-structured interview procedure, the whats, whys, hows and wherefores of the critical events were collected. Herzberg et a1. then performed a content analysis of their data. The results yielded two groups of factors that constitute the Herzberg two-factor theory (see review of motivation theories, p. 10). Herzberg's methodology and resultant two-factor theory has re- ceived severe criticism as was pointed out earlier. Critics of his methodology accused him of equating a satisfying event to a moti- vating event by asking respondents to talk about good or bad feelings about their job and inferring performance motivation from resulting data (Vroom, 1966). Another criticism is that since coding was not entirely determined by the coding system but required interpretation by raters, bias was introduced into the data during content analysis. Criticisms against the theory include a challenge of the claim that factors causing motivation and those responsible for dissatisfaction are independent. Further, the theory has tended to receive little support from studies using different methodologies - thus it is method bound, (see House and Wigdor, 1967). But as earlier indicated, Herzberg's theory appears to have some validity. The two-factor theory together with need hierarchy are 43 the cornerstone of modern job design/redesign theory (Hackman and Lawler, 1971; Hackman, 1977) whidh has received reasonable support in field settings. If the weaknesses noted in the methodology used in the Herzberg study can be remedied, then the Herzberg methodology is most apprOpriate for the study preposed here. This is because Herzberg's is an empirical approach which is perfect for an explora- tory study and suits conditions in the population to be investigated - no reading skills among some members of the sample. Thus, the critical incident technique (with a semi-structured interview was used to learn about events, conditions and processes surrounding productive and non-productive behavior at work. The methodological modifications to deal with weaknesses in Herzberg's approach will be discussed in the method section, which is presented next . V. METHOD Materials: A semi—structured interview with questionnaire items based on those used by Flanagan (1954, p. 342) and Herzberg et a1. (1966, p. 141-142) was used in this study. Recall that Herzberg et a1. were criticized for equating feelings of satisfaction to motivation in that they asked interviewees to recall "a time when you felt exceptionally good...about your job" and inferred motivation from the data so collected. To deal with this problem, interviewers in the present research asked subjects to "think of a time when you "...a time when worked exceptionally hard" (in the good incident) or you put in exceptionally little effort into your job" (in the bad incident). Such a statement deals with motivation since exceptionally 44 hard work and little effort can be assumed to be associated with high and lomeotivation respectively. Some items used by Herzberg et al., for example the one asking subjects to rate their feelings on an abstract 21 point scale, were drOpped from the scale and new ones considered suitable were added. Further, all questions in the instrument, all of them open-ended type, were subjected to extensive pretesting on a sample of 40 Zambian employees before they were adopted in the final question- naire used in the study. Questions that appeared incomprehensible, vague, or could not induce meaningful responses according to pilot study results were changed, reworded or simplydropped from the final set of questions included in the questionnaire.3 And because a considerable portion of the Zambian workforce (mainly employees in low level jobs) cannot read, write or speak English, the question- naire was translated into two Zambian languages (Bemba and Nyanja) using the back-translation technique. .Bemba and Nyanja were chosen because they are spoken in the major industrialized and urbanized regions of the country where the data were collected. The Bemba and Nyanja (questionnaire) versions were also pretested before the main study began. Subjects: The sample consisted of 341 Zambian employees who included 65 managers/senior administrators, 53 technical personnel, 72 secre- taries, stenographers and typists, 67 clerks and 84-semi—skilled and 3A c0py of the questions finally adOpted for the study (after-. pretesting) are presented in Appendix 1. A short statment used by the researchers to introduce himself to the interviewee is also included. 45 general workers including laborers. The five occupational groups were chosen because they represent categories of employees that occur in many organizations. The employees were drawn from the three main types of organizations existing in Zambia, viz: government, partially government and partially private, and private. Of the 341 employees, 55 were drawn from government departments, 241 from partially government and partially private-owned companies and 45 were from private firms. Ninety-four of the employees were female and the remaining 247 were male. The sample's age ranged from 19 to 67 with a mean of 31.5 and both mode and median of 30 years. Amount of education among the subjects ranged from 0 to 20 years with a mean of 11.3 years. The mode and median were 13 and 12 years, respectively. Years of urban residence in the sample varied from 2 to 50 years with a mean of 21.2. The mode was 23 and the median was 22 years. Inclusion of an employee in the sample was determined by organi- zational officials (the employee's manager or supervisor). Typically the organizational official designated by the chief executive or senior management was asked to provide the researchers with a speci- fied number of employees in each of the five occupational groups included in the sample. The researchers emphasized to this person (manager or supervisor) who actually picked the employees to be as representative as possible. That is, the manager or supervisor was not to pick only the best or worst employees in his or her work unit. Rather, it was to be ensured that employees picked reflected the variations in work behavior existing in the unit or department. 46 Organizations: A number of private, government and quasi-government organizations in the Lusaka and Copperbelt (the most industrialized areas) regions of the country were contacted. The purpose of the study-~to investi- gate motivation and identify factors that could be manipulated to enhance motivation and productivity-dwas explained to the chief executive, senior administrator or person of similar standing in the organization. Participation of the organization in the project was requested; inclusion in the sample depended on whether the organization agreed to participate or not. Of thirteen organizations approached, two--a private company and a government department-declined to participate. The eleven that agreed to participate included two government departments, a city government, three private companies, a university, and four partially government and partially private-owned firms. The eleven organizations represented varying industries or business activities that included brewing, mining, oil refining, heavy engineering, occupational and educational assessment, furniture manufacturing, legal practice, health administration, housing administration, heavy equipment im- porting and distributing, among others. Data Collection: Before the data collection and even pretesting of the question- naires started, a research assistant was trained. The assistant, a senior year public administration student at the University of Zambia, was given background concerning the aim of the study and in- formation on how to conduct a semd-structured interview. He was instructed on establishing rapport, asking questions and ensuring 47 they were understood, active listening, use of the probe to clarify some responses and to keep the conversation on track, and how to encourage subjects through neutral comments like "I see," "Uh-hh," "O.K.," etc. (see the Survey Research Center's Interviewer's Manual for 1976). Procedures for fast note-taking were also discussed and practiced. Role-playing was used to ensure that techniques were correctly and properly learned. After training was completed and questionnaires readied through pretesting and contacts with organizations had been made, data col- lection began. Each interviewer sat in a separate office. When a subject entered, the interviewer introduced himself and went on to establish rapport and explain the purpose of the study. The inter- viewer, who did not ask for the employee's name, emphasized the con- fidential nature of the interview and assured the subject that nothing said in the interview would be told to, or discussed with, anyone other than the researchers. The subject was also told that “participation in the interview was voluntary and could be with- drawn any time. (The introductory statement used in the interviews is shown together with the interview questions in Appendix 1). If subjects expressed unwillingness to participate, they were not interviewed . Only three persons showed reluctance to participate and were subsequently excluded from the study. Other Information: After subjects described and answered questions on critical incidents, the researchers asked them to state conditions, circum- stances, or events that they thought would make them (subjects) work as hard and as little as in the critical incidents narrated. This 48 information was recorded. Following this, subjects were requested to supply some personal information: Their sex, age, job position, education, number of years with company or department, number of years in paid employment and number of years of urban residence. In addition, names of the companies and departments and their location were recorded. Data Analysis Content Analysis: There are two main approaches to content analysis (Lasswell, 1949): One approach is an a priori approach in which classification of content is based on a previously thought out schematic system. The factors proposed earlier (see page 38) would, for example, allow building of a priori classification system or list of factors expected to emerge from the data. The other approach is a posteriori in that its schematic system is based on the data themselves. This approach requires prelimdnary examination of the data and developing categories into which all the data are then classified. In this study, the a posteriori classification was used; that is the category system was based on the data themselves. This is important in light of my earlier criticism of cross-cultural research in motivation which studies cultures with a priori conceptual schema developed on different cultures. The researcher studied all the interview schedules and developed a classification system based on themes or thought units expressed in the interviews. (A theme or thought unit is a "single assertion about some subject," Holsti, 1968, p. 647). The researcher and his assistant then independently examined the 341 interviews and extracted (coded) themes from each interview. 49 Intercoder agreement was found to be 87% over 3025 codings made by each coder. Examination of disagreements between the coders showed that 952 of these disagreements were a result of omission of a theme by one or the other of the coders. The coders then jointly reexamined the interviews where disagreement had occurred. If it was found that one coder had made a mistake of omission in coding a given them, such theme was included in the final list of themes. If, on the other hand, one coder included a theme which on joint examination could not be attributed to the interview being examined, such theme was omitted from the final list of themes. This inclusion or omission of themes on joint examination of interview data did not affect the value of intercoder reliability which had been computed prior to this reexame ination. Each theme was typed on a card. The cards were then sorted in- dependently by four judges (all of whom were Zambian) including the researcher and three other people not connected with the study. The four judges sorted the cards into a classification system earlier developed by the researcher following preliminary examination of all interview schedules. On the basis of the four judges' sorting, 23 factors or motives were identified from the good critical incidents and 27 from the bad incidents (see Table l, p. 55), Since interviewers also collected information on subjects' own subjective impressions of what would make "me work as hard or more than" or "as little or less than" during the incident (good and bad critical incidents respectively), each judge made four sortings: two for factors associated with hard work and two factors inhibiting hard work. 50 Frequency counts were performed for each of the factors extracted from the content analysis. The frequency for each factor represents the number of people within the 341 sample who cited that factor as having encouraged them to work very hard (in the case of motivating factors) or to work very little (in the case of demotivating factors). These frequencies, therefore, serve as a measure of the strength of each factor to encourage or discourage hard work among employees. Validity Check Questionnaire: To check the validity of the factors identified in the content analysis, the motivating and demotivating factors were themselves used as items to construct a questionnaire. To be included in the questionnaire each motivation factor had to have a minimum frequency of 9 (i.e., be mentioned by at least 9 people or approximately 2.6% of the 341 sample). This was done to ensure a reasonably short ques- tionnaire in the hope of avoiding deterioration in response quality often encountered when respondents fatigue and lose interest because of very long questionnaires. Twelve factors encouraging effort at work and 17 discouraging hard work met this criterion and were mixed randomly to form a 29-item three-point scale instrument. The scale simply asked respondents to indicate "basing on your own.work ex- perience (what has happened to you in the past),...which of the listed factors would encourage you to work very hard, which would make you do very little work, and which you are not sure would en- courage or discourage you to work hard" (see Appendix 2). This questionnaire was administered to a sample of 80 employees not in— cluded in the original study. These employees were drawn from eight different organizations representing private, government and 51 quasi-government firms. As in the original sample, the 80 consisted of managers/senior administrators, technical, secretarial, and clerical personnel, and general workers. Scoring: An item which a respondent indicated encouraged hard work was scored three points, one said to discourage work was scored one point, and a "not sure" item received two points. With this scoring, an item that received a mean greater than two after summing and averaging across all respondents is considered motivating. The closer the mean to a value of three, the greater the unanimity among respondents that the item (factor) encourages hard work. Consequently, the size of the mean can be considered a measure of the motivating strength of the items. Another indication of unanimity of subjects' responses to a given time, is the size of the standard deviation of the item. A smaller standard deviation would signify greater unanimity of responses than a larger one. Thus, the standard deviation can also be looked at as an indicator of motivating strength of items. Similarly, items (factors) having a mean score less than two after summing up and averaging across respondents are considered to discourage hard work. The closer the mean value to one, and the smaller the size of the standard deviation, the greater the agreement among respondents that the item discourages working hard and there- fore, the greater the magnitude of such factors or items as a demoti- vator. Items with a mean value of two on summing and averaging across the sample are those where respondents were unsure as to whether such factors encouraged or discouraged them to expend extra work effort; such items may be considered neutral. 52 Convergence of the factors from the content analysis and question- naire study is considered established if items or factors from good critical incidents show mean values greater than two and closer to three, while items or factors from bad critical incidents show mean values lower than two and closer to one. In addition, the strength of the factors as motivators or demotivators as shown by the frequency counts in the content analysis could be confirmed by correlating these frequency counts with the means of the same items following analysis of questionnaire responses. Consequently, the means of questionnaire items were correlated with the frequency counts. Formation of Clusters: The researcher and two other persons (two American advanced doc- toral students of psychology) With a sound background and understan- ding of motivation theories and constructs independently clustered the 23 motivators first and then the 27 demotivators. The three judges separately examined the 23 motivators and then sorted them into categories or clusters. Items were grouped together if (in the judge's opinion) they expressed themes that were related on a content, conceptual and theoretical basis. The judges also defined each cluster of items so formed by giving it a name they thought most descriptive of the factor. This process was repeated for the 27 demotivating items. Mean K coefficient reliabilities [proportion of agreement after chance agreement is removed, Cohen (1960)] across the three judges was found to be .85 for motivators and .82 for de- motivators. 53 Correlational Analysis: For every subject, each motivational variable received a score of one (if that subject mentioned the factor in question) or zero (if the subject did not mention that factor). This was done for both the 23 items derived from content analysis of good critical incidents and the 27 from bad incidents (Table l). Biographic and other work-related variables (e.g., length of employment, length of urban residence, etc.) collected on each subject were correlated with each other and with the dichotomously scored motivational variables. In addition, the frequencies of motivational variables derived from critical incidents were correlated with the frequencies of the same variables derived from subjects' responses when directly asked to state what they thought would motivate or demotivate them. This was separately done for motivating and demotivating items. In order to confirm the content and conceptually-based factoring of motivational variables by the three judges, it would be desirable to statistically factor (Nunnally, 1967; Kim, 1975) the motivating and demotivating items. However, factor analysis is inappropriate for these data for two reasons. First, factor analysis assumes interval data (Stevens, 1946; Kim and Moeller, 1978) which we do not have in this study since the dichotomized variables, scored on the basis of whether a subject mentioned a given variable or not, are nominal. Second, the phi coefficients that would be computed among the variables (and on which the analysis would be performed) become smaller and smaller as p values depart from 0.5 (Nunnally, 1978; Kim and Moeller, 1978). Since, as can be seen from Table 1, the mention 54 frequencies of items vary from 69% to 0.6%, intercorrelations among items would likely be very low and this would affect the factor analysis. VI. RESULTS The motivational variables identified from content analysis of good and bad critical incidents are presented in Table 1 (p. 55). Frequencies for bad critical incidents are somewhat lower than those for good critical incidents. A possible explanation for this might be the one suggested by Vroom and Maier (1961) that incumbents are objective in describing their effective but not their ineffective behavior. Both groups of factors are listed in order of magnitude (measured by how frequently a given item was mentioned as coded from critical incidents). That is, a given motivational variable had a higher frequency than those following it. Since frequency of mention may be seen as an indication of the magnitude of a variable's impact on work motivation, then the effect of the variables on moti- vation increases with proximity to the top of the list. From Table 1, five of the a priori motivational factors are con- firmed by these data. The five are interpersonal relations, pay, job security, equity, and corruption-related issues. (Results re- lating to the other a priori factors are presented on pages 68-72). It can be seen from Table 1 that Herzberg's assertion that motivators and hygienes are independent has been contradicted since some vari- ables (or rather, absence of them) appear as demotivators. For example, work that is interesting, challenging, important and has variety is motivating while work that is boring and unimportant im- pairs motivation. 55 c.: 6.: 9.: a... 9... ~.— N.— r.— r.— u.— _.~ r.m ...p c.~ 6.x 6.:— ~.¢_ ¢.w— c.c— =.=~ c.=¢ c.«e c.6c u .- CCKQWII‘Q‘ en. en. can Nodflmao» L needless ue.aeua acceloc «lease». _oseeuoa ml 0» eeaau- 0;) ouoxuoz|oo .eucuuoeau .ueeeIOu Iauaclueuoaca ecu wee_euu as: — sue: ne-ow n azaleas—e assays .ouo .eus—I—eloo 2:..eess .uo¢.nu has .ueoleu—oae duo: .so-uOIOue ea cassava: Aeooessse use .atec— .uuoeae-uu .c_teamueo .eo—uevallouuev euuueeoa ens—uh Ace—uuequUV aueavowh uueaaoaloo .uaocuoese usuauolfiuun aueoleuaew .u:0>e— sassy ncouuuseou aeo¢e555 census. .un—lnu medueu369cenebquuoaesm Ava—e. useo_u- use .9>uuee—o.uues .e>-ueu~:e:Ouv counvsueesm slo=0ua< condensing .eho_aee:e no :36:- eueweuucoU 1:0 aesuh sou-cumuoaaa as: coo—aeese .auosuDDIOU 5a.) ego—ueyuu .-600ueeueuc— coco .....a.o=oa.o- A._ee~=o m:_>05e see geolo>o_=ue cab-no ueau duosv aeolo>eugu< Auras—ale me_oa assess ouv sausages sea neonates. use no >.—_.a_amoe as Asa—st ~.-ol .u>_u:oue_ .aaeoa .usolbuue_~ he. ¢:_=_euu moan-s. uo\::o as. steam one! shown on oucegu ee.uol6ue no. outage uo Gouge-cum Aces—ueuueexo .ouoaao not! so .auosuo lea. ou—eae .su._.aescu so:- ouv so—u.=nooos ecu-wees gouge—eloU as: sexes. «some: ea «use due: A».o_an> so: .aeouuoth .ue_uco_~ego .uG-ueeuoueuv ages“. duo: Au—=o__u-Q so» .0 goal sou uoev ass: o—aouuuqs so sue) .0 00— < asap—we. anode-nu nose .- .~N .cn .su .0— .c- .nq .N- .N .— 0.: 0.: 6.: 0.: n.~ n.— u.~ o.~ n.n o.n n.m e.n o.m n.o 5.0 c.s e.o— ~._— c.~— 0.~— n.n- 5.:— Q.n— 6.5— u n..9 .n «u PI «a ~ ~— n— c— 0— cm fin nu an en Cc —c 0: No no noeosaamm ace) .6 :o_.:.oe .: Jo:— xuo) :. Ase—uoe=c z: .6 ace—auuoev wosououaoaep 3350:35. 3.5.2.5.: ace—3.850 for 25.2»: mange :- noun—aolloo .e.a:seaa .o eoaeaaogm hour...- Auesoucesv uc_x:.uc sous-sass u: ale-a “name: :59: w.««.— so. so due: :99! och Ant—ueuuoue.:: .uceauosl_== .me—utav u—waaq use: ae~3iow°~5 see Ie.o——oa hassles as: pageaeo-su nu_u:uoe seq hassles a; needle»; ass: as eta—.am ne_:.etu uoauusu uo\1ee eon uses: veal eueou an outage he 4:?— an—aoL—e max-ac no n—wuueae 93:93:: 30.33% duo) aweu.oe s—uoos toque. museu—e—wo.s uo‘psn eo_uslo: A.ouu .oelo— .uuonaceuu .u:_e:o£ no sue—v nuances: cue—um Awh¢au weam¢e— be one—9v c:-e—>~sa:m Aneuduot emu— use. e.goa ass—v son! a: wed—sou muoxuoalou .auo—ueaae ucwuwslcweu noche >u§_ Anusoquelcu no: .we—eua new so...cmoueu no #9:; Abe sue—v CO—aeIOus sq—le. e. sneeze—axsusyc noose—ale u: case‘s) no ale—sous ou ceua.— no he. ones no: so assassinOU .auc_uoe:e .sceelou Ase; :u abuuoafiou no ae_—_eu has $039695 .no>_~:ouc. .uoeueu u-uul .eoeseoa .ueolouo=_ a: Joe. .3:—. use amass—wuoasa .ouoxuoaloo .euouuosaa su‘) eso_ue—eu —eeoausauuuau can .uao .euoce>o_un .eusUIewanne duo) .ewaunu use .osoqualous sq Iouussuuow .lI—ueees .soquea—lcuuaqo peace» .loauquSou .Iouuea—uh Ilsa. gum—ue>—uolovm usuvuoe— new.»«uu we. .A—en I fly counsel .0 oasusozwee 1:- »ueoavouu oe—aeuuuoo no scene a. veucueoua one ue—:e_ul> a cube». .eaea Jou>uoaec at e.gx—eae aeoueoo gasses. vouuuacova ou—ae_uo> ~aco-a>_uoz .- .o~ .eu .nN .NN .—~ .a— .0— .m— .0. .m- .e— .m— .N— .- .c. .c .n .0 .w '3 crit Inc in bar 9 I 9!;- 1.85. be: vat the ite ‘gat 56 Table 2 shows means and standard deviations of items extracted from good and bad critical incidents which were included in the validity check questionnaire. As can be seen from Table 2, all factors extracted from good critical incidents have means greater than two and close to three. In contrast, items from bad critical incidents have means that are less than two and close to one, except for item 17, which had a mean higher than two. The fact that item 17 is double-barrelled perhaps explains this unexpected result. These questionnaire results lend strong support to the factors identified from the content analysis in that all good incident-related factors have been associated with hard work and those from bad incidents have been identified by an independent sample of employees as tending to discourage hard work. Examination of the standard deviations shows the values generally become smaller as item means get closer and closer to three for moti— vating items while the same is true as means get closer and closer to one for demotivating items. This shows response unanimity among the 80 employees who completed the questionnaire, that a particular item is motivating or demotivating and provides an index of moti- vational strength of the items. Another confirmation of the content analyzed - results, particu- larly the ranking of the motivational variables by frequency of mention to indicate strength of their impact on motivation, is shown by the data in Tables 3a and 3b (p. 58). Correlations between frequency counts for the motivational factors (items) identified from the content analysis and the means of the same items computed from the validity check questionnaire are positive 557 nanm.nmen.~ n__e.enmn.~ name..nme.~ ANN—.vnwm.~ nmoe..m~e.~ Anem.e~ee.~ nms~.v~em.~ nmem.vmne.m Amv.v~mm.m nmee.enme.~ Asmp.vm~m.~ neew.neme.~ A.o.mv mzoe xne—em .xno: oo unoameenu meeo— .m:_m:og >eeoeoov mo.eoeoo ooepnn mnonno ooh oz» .m— .mooee¥nox noo» use mommoo .u—omnoox eoozooo mcpoeeomnooeo .meo_oepon ooom one onozh .—p .meonmnooo ones oo ooeezo ace nu'—_o—meoomon no» atop—e now ogp .c. .oofi noo» co oeo5o>o_:oe no» >u__¢o—mmoa m. onog» .o .Aoonoem on o» aponn— uoo one ooxv onooom m. goo .w .o>_u -coocw —e_oeec—n no aeoEonoe— o_nos .moeoo .uooeonoe. non now—venomoe a. onocu oee o—oeeomeon m. xea .n .mewewenu nozonoe non ooeego m. onozo no son ogu ozone onoe eneop oo ooeezo m. onog» .o .=o_ooeono non mooeezo one onozu yes» no ooooeono one :o» .m .one: me_nno3 non oo~_:oooon no ooenena oooevoonooe .oooeoeppoeoo on on ooceso a. onoeh .e oevpoeoo e cage—x oouo_oeoo on own: noxoce aeomn: m. nnox och .n hoonne> we: no menmoo—peeo .oeeonooe. .oe—umonoo=. m_ nnox och .~ .ocoe on an ens: as as. e m. onogn .— exon_ new a z. .xnox Amoo.v~mm.~ opoopp ooo no seas ooo nozono ooeopmme eooo o>eg so» Ammm.vw~e.— Aooxoem on on >—ox.~ one sexy ooeooeonzo m. »o_nooom non .mm0.vmme.— .oox on ones onoz song: momveona noose noon ooe oo enopnooom no aceoeoo one nemn.n~e¢._ .e=_=_enn nogsnsn non ooceeo o_oo.n no son noox so once eneo— ou ooeezo o—oo_— a. onozp .aon on no“ son» oo>oe xpmoooepueoo one so» .oon o_nv none.vmne._ -ooom oe o>ez so» ueno no ooePe—oxo x—noaona no: mp nno: Amm~.v~—o._ .mconuoe snoop—a.omno o» oooooeoom no oooeoomom .ooooEoo one so» .A—uc.vm~c.— .A.ooo .mceop .mcnmoogv mopnocoo om=_nn oe no open“. one onozp fip-.V~—m._ .——e we anc: noox oe unoo— x—one: no Anooe noo» :o mozoeonov a_omo—o momn>noaom nomp>noo:m\mmoo noo» Amme.nmem._ .ecnnno: eon. .oon no: so none so» Amwm.vmom.— .ocooooeooe_ one x~e_ one mommoo no mnonnoaom noo» Ampo.vumm._ .nnoz one: no» cones: -mooon no ueoe__o5oo .eo_oevoonaae .omneno o: m. onogp Ammo.vm~e.— .spon_peo >no> one meowuoeono no» mooeezo no «cowboEOna .meopoona onomoeoo no .eeomnoa nozoo o>eg no» no »p_5em on» on cueoo no mmoenonm m. onozh neee.nnem._ Aewo.vmnm.— .meopoono pecomnoo oo zoom._ no oneu uoe mooo noxoPQEo no >eeoeou Ao—m.VOm~.— .npon.—=: one mo>pocooe_ pepoeeeve noguo one mueoeonoc— u—noe .momoeoo .mueoeonoe. pee :o. m. sea .moernnox noon ooe mommoo no mno.nooow noox .n—omnoo» eooxooo mae'ooeomnooeomwa no meoeue_on men one onogp .mnm.nmn~.. Acmm.vmmc.— .oonaoonnou no am_ooooe .amnoenoo>en .eopuee_swnom¢o .epoen .Em._eo_nu oopuoena mommoo no wno_noa:m n.o.mvmz.ooe ooooon oo me.oeou omofi see new? o5 so one .3532: pee—:no vooa Eon: eoZeZooe snot omeonof 3 952.3 encou— .on_eeeo'umo:o noose havo._e> oco e. woos—oe. Amaoo.v enoooee peeoaue>_ooa eo monoom eonoen>oo oneooeom oee eeoe oe_3ogm opoep N oneen .N— .m— .m— .e— .m— .m— .p— ": . U '6 m U) n) n -i Cc: . .O ;.€ Ite In It: Ere Co: 58 Tables 3 a & b Pearson product moment correlations between frequency of mention of motivational variables from the content analysis and means of the same variables included in the validity questionnaire. Table Ba Correlations for Motivating Items (from good critical incidents) N812 Item Mention Questionnaire Item Mention Frequency Item Means Frequency (Subjective)+ Questionnaire 1.00 Item Means Item Mention .534* 1.00 Frequency Item.Mention . 29 .394 1.00 Frequency (Subjective) 3 *p< .025 Table 3b Correlations for Demotivating Items (from bad critical incidents) N317 Item Mention Questionnaire Item Mention Frequency Item Means Frequency (Subjective) Questionnaire 1.00 Item Means ++Item Mention _ 504* 1 00 Frequency ° ‘ ° ++Item Mention _ ** 0 Frequency (Subjective) '267 ‘841 1’0 *p< .019 **p< .001 + Item Mention Frequency (Subjective) is the number of times an item was mentioned in the case where subjects were asked to state condi- tions, circumstances or events they thought would make them work as hard or as little as they did in the critical incidents. ++' The two negative correlations are positive because questionnaire items were scored such that mean values decreased as demotivating magnitudes increased whereas item mention frequencies increased with demotivating magnitude.’ 59 and highly significant: A positive relationship (r = .534, p‘< .025) has been found between frequency of mention for the 12 motivating items (from good incidents) and means of the same items included in the questionnaire. The correlation between frequencies for the twelve items - in instances where subjects were directly asked what would make them work as hard as they had in the critical incidents - and the mean scores from the questionnaire is positive but nonsignificant (r - .329, p > .149). It is interesting to note here (Table 3a) that the relationship between what motivated employees as coded from good critical incidents (item mention frequency) and what people said would motivate them (item mention frequency - subjective) when asked directly, is positive but nonsignificant) r = .394, p > .103). Support for ordering of demotivating items (from bad incidents) by frequency of mention is also strong - Table 3b - as shown by the significant positive relationship between questionnaire item means and item mention frequency (r B .504, p‘< .019). And, as was the case with the motivating items, a positive but nonsignificant cor- relation (r 8 .267, p > .15) is obtained between validity check questionnaire item means and frequency of mention of items in cases where subjects were directly asked to report conditions or events that would discourage them from working hard. In contrast to the finding with motivating factors, however, a high positive correlation (r - .841, p‘< .001) was found between frequency of mention of an item or factor in the bad critical incidents and the same item's frequency of mention in the more subjective situations where subjects responded to direct questions about what they thought would discourage hard work on their part. These results seem to suggest that employees' 60 ideas and perceptions of what would demotivate them is much more in agreement with what actually discourages them from working hard as arrived at from critical incidents. But their perceptions and thoughts about what encourages them to work hard and what makes them do so as derived from critical incidents is less in agreement. These data show strong convergences between results (factors) from content analysis and those from the questionnaire study. The strong positive correlations between frequencies of items from the content analysis and questionnaire item means (subjects' agreement and unanimity that a given item encouraged or discouraged them to work hard) suggest that ranking items by frequency to roughly indicate the items' effect on motivation is justifiable. Structure of the Clusters Identified Factoring of the motivators and demotivators by the three judges yielded five motivating and six demotivating factors. The judges be- lieved, however (looking at the names they used to describe the factors and from discussion with them after they each had completed factoring), that five of the six demotivating factors were inverses of the five motivating clusters. Each judge had arrived at this decision because five of the six demotivating clusters represented the absence or in- adequacy of conditions defined by the motivating factors. Hence, instead of five motivating and six demotivating clusters, only five factors with potential to both enhance and impair motivation were defined. The sixth of the demotivating factors (Personal Problems) is the only cluster identified which appears to have only a negative impact on motivation. These results, presented in more detail below, are summarized in Figure l. 61 FACTORS Demotivators Motivator! 6? Promotion (lack of) 5. Promotion or chance for promotion 27. Lack of rotation of work | 6. Chance to learn more about job and/or 26. Interference (political or by superior) GROWTH AND further training in work ADVANCEMENT 12. Trust and confidence shown by SPs. CKs 15. lack of chance to learn note about job OPPORTUNITY lo. Responsibility and/or further training 13. Autonomy 21. Proper placement (doing work I was trained for) 18. Feedback (corrective) 20. Too much work or too little work 30“ Yetta: l. A lot of work (sufficient or more but 19. Work itself (boring. unimportant. inn CONTEXT not too much) uninteresting) 2. Work itself (interesting. challenging. 13. Poorly defined work duties important. has variety) 3. Work that is urgent and/or has completion deadline 9. Achievement (work that allows achievement and proving oneself) 3. Pay (low pay, no increment. bonuses. 7. Pay (increment. bonus. incentive, merit nerit raises. incentives. reached pay raise) or possibility of pay increment ceiling or reduction in pay) MATERIAL AND a. Job security (to ensure being employed) 12. Demotion and/or disciplinary action PHYSICAL 16. Physical conditions (work layout. 26. Shortages of essential commodities in shops PROVISIONS equipment) 17. Job security threatened l9. Fringe benefits (accommodation. cafeteria. 23. Physical work conditions (dangerous. transport. laons. pay advances) unfavorable) 11. Fringe benefits (lack of housing. transport. loans. etc.) 7. Lack of recognition (no praise. no 4. Recognition (to show capability. praise compliments) from others. to meetotbers'expectations) 8. Lazy and/or incompetent SPs and Ch. 17. Hardworking SPs. on. 53s 10. Supervision (close. laissea fairs) usurious um 14. Supervision (consultative, participative. 2. Bad interpersonal relations with arms not laissas fairs) SPs. Clo. SPs 11. Good interpersonal relations with CIs. 21. Unjust blame or suspicion SPs. 33s . 15. Supportive-encouraging family. friends 22. Understanding COT. SPs. CIs who listen to my personal problems i. Tripoli... favouricisn, :.¢1.1 PAIRSESS In 20. Fairness (in promotion. work assignment. discrimination. nepotism. corruption ORGANIZATIONAL pay raises. handling complaints. etc.) in promotions. pay raises. work PRACTICES 23. Company keeping promises assignments. grievances. etc. 18. Bad COT policies and procedures 6. COT. SPs. (Is do not care for or listen to problems or welfare of employees 16. Failure to keep promises by company ' S. Death/sickness in family 23. Hunger 9. Feeling or mood (just don't feel like PERSOHAI. working) PROBLEMS 22. Drinking (hangover) 14. Domestic quarrels or other problems Figure l. Diagrammstic representation of the six factors identified showing the motivating and demotivating items *ltem numbers denote the ranking of item by frequency of nention (Table l). SOTI: Abbreviations used in table stand for: Subordinates. C0! - Company or Organisation and Cls - Coworkers or clustered under each factor. colleagues. SPs - Supervisors. She - 62 Motivating;Factors (from Good Critical Incidents): The first factor identified from the data has been defined as Growth and Advancement Opportunity. Existence of growth and advance- ment Opportunity - a very frequently mentioned motivating condition - is ensured when possibilities for promotion are present chance to learn more about the job and/or further training is available, trust and confidence is shown by one's superiors and co-workers, and when job conditions allow responsibility and autonomy. Other items grouped under the factor are proper placement in a job with regard to one's skills and previous training, and corrective feedback received on the job. The second cluster relates to the nature, amount and difficulty of the work and how work is assigned. This factor has been defined as Work Nature and Context. Items that were clustered to form this factor are amount or difficulty level of work assigned (a lot of work - sufficient or more but not too much), urgently required work or one with a completion deadline, work that is interesting, challenging, is perceived important and has variety, and work that allows a sense of achievement and opportunity to prove oneself. Work or assignments that provide these dimensions are highly motivating since these dimensions were very frequently cited by subjects as leading to hard work in the critical incidents. The third factor defined from the data relates to the amount of Material and Physical Provisions present in the job situation. This factor, like the first two, has potential to increase and reduce work motivation. Items making up this material and physical provisions factor to pay (reasonable pay, increments, bonuses, incentives, merit W0? and W0 not PIE no: ha: of Pre 63 raises), job security, fringe benefits, and favorable work conditions. Presence of the dimensions named tends to encourage hard work. Factor four is Relations With Others. Items that judges grouped together to form this factor - those mentioned as tending to encourage hard work - are recognition or praise one gets from superiors, to- workers or subordinates, good supervision, degree to which superiors and coworkers are hard working and competent, good interpersonal rela- tions, supportive family and friends, and degree of understanding and attention to employee problems shown by the company. The fifth factor is Fairness in Organizational Practices. Only two items are grouped under this factor. These are fairness in pro— motions, nwarding of pay raises, work assignment, and other personnel practices, and company keeping promises to employees. These two items that comprise Fairness in Organizational Practices were mentioned as motivators by very few persons in the sample (Table 1). On the other hand, these same items were mentioned by many more people as sources of demotivation. Results from analysis of demotivating factors are presented next. Demotivatinngactors (from Bad Critical Incidents): Clustering of demotivating items by the judges identified six factors. 0f the six factors, five were found to represent an inverse of the five identified from analysis of good critical incidents. The first cluster obtained from this analysis groups together six items. These are unfavorable pay (low pay, lack of increments, bonuses, incentives or merit raises), threatened job security, lack of, or inadequate fringe benefits, and unfavorable or dangerous physical work conditions. In addition, demotion or disciplinary action against 64 an empldyee and shortages of essential food commodities in shops were , grouped under this factor. Since this grouping represents an absence or inadequacy of the conditions, provisions or materials defined under the factor Material and Physical Provisions (identified from analysis of good critical incidents), it is an inverse of that factor. Conse- quently, this cluster of demotivating items represents the demotivating component of material and physical provisions. Thus, the factor Material and Physical Provisions is capable of both increasing and impairing motivation. Two of the items under material and physical provisions (demotion and disciplinary action, and shortages of food commodities in shops) do not appear to have any positive effect on motivation - i.e., none of the subjects indicated that absence of demotions and presence of commodities in shops increase work motivation. The factor Material and Physical Provisions appears to be more associated with deflating than increasing motivation to work since items tending to reduce motivation rank higher (were cited as demo- tivators more often than they were as motivators) than those encour- aging hard work (Table l). The second demotivating cluster identified deals with the nature and difficulty level of the work. Items clustered to form this factor are too much or too little (too difficult or too easy) work assign- ments, poorly defined work, and work that is boring and perceived to be unimportant. As was the case with material and physical provisions, this grouping of items constitutes the inverse of a factor derived from analysis of good critical incident items - namely Work Nature and Context. Hence Work Nature and Context has potential to enhance 65 and also impair motivation. Unlike Material and Physical Provision, however, Work Nature and Context appears to have greater positive than negative association with work behavior since items constituting the positive component of the factor rank much higher than demotivating items. Of the items clustering on Work Nature and Context, two (work that is urgent or has completion deadlines, and work that allows a sense of achievement and chance to prove oneself) were not associated with the demotivating component of the factor. These two items, work that is not urgent or work that does not provide a sense of achievement and opportunity to prove oneself, were not mentioned by subjects as discouraging hard work. Factor three pertains to the degree of fairness or unfairness in organizational practices. Items grouped under this factor are tribalism, racial discrmination, favoritism, nepotism, or corruption in the job situation, failure to keep company promises to employees, had company policies and practices, and unwillingness by the company to listen to employee problems. When these conditions are present, work motivation is very negatively affected. Like the other factors discussed so far, this cluster appears to be the negative component of the factor Fairness in Organizational Practices defined from analysis of motivating items. The degree of Fairness in Organizational Practices appears to have only a minor positive effect on motivation. This is because the two items (fairness in promotions, awarding pay raises, work assign- ments and other personnel activities, and company keeping promises to employees) that constitute the motivating component of the factor 66 were very infrequently mentioned as sources of motivation in the critical incidents. Lack of fairness in organizational practices, however, was quoted much more frequently as a demotivator as is evident from the ranking of the demotivating items composing this factor (Table 1). In fact, one of the items clustered on this factor (tribalism, racial discrimination, favoritism, and corruption at work) is the most frequently cited demotivator. The fourth factor identified from analysis of items from bad critical incidents again seems to represent an inverse of a motiva- ting cluster defined earlier as Growth and Advancement Opportunity. The four items judges clustered to constitute inadequacy or absence of growth and advancement are lack of chance for more learning or training on the job, absence of job rotation, lack of chance for promotion, and interference in one's work by superiors or outsiders ' (8-8-9 politicians). Thus, presence of Growth and Advancement Oppor- ' tunity can encourage hard work while absence of growth and advance- ment chances can discourage hard work. Rankings (Table l) of the items grouped under the positive and negative components of Growth and Advancement Opportunity suggest that subjects associated this factor more with increasing than deflating motivation. This is so because motivating items generally rank higher than demotivating items. Another cluster identified from analysis of demotivating items pertains to an employee's relations with other organizational members. If unfavorable relations are present, motivation to work is negatively affected. Items grouped under this factor are lack of recognition by superiors and coworkers, bad supervision, laziness and incompetence among superiors and coworkers, bad interpersonal relations and unjustly 67 blaming or suspecting an employee for something he or she did not do. This cluster of items, like the others discussed earlier,.is the demotivating component of the factor Relations with Others which was defined from analysis of motivating items. Consequently, Rela- tions with Others can increase and also decrease motivation. Items clustered on the demotivating part of this factor generally rank higher than the corresponding motivating items (Table 1). Thus, it seems that the factor Relations with Others is more often associated with impairing than enhancing work motivation. The sixth and final factor relates to non job-related employee problems. Items clustered around this variable are an employee's mood or feeling of not working, domestic problems or quarrels, hang- overs resulting from alcohol consumption, hunger, and death or sick- ness in the family. This factor has been defined as Personal Problems. Unlike the other five, Personal Problems seems to have power to only impair but not increase work motivation since there were no motivating items to constitute an inverse of this factor. Except for the last cluster (Personal Problems) these results generally contradict Herzberg's position that factors affecting moti- vation positively are independent from those affecting it negatively. What has been observed however, is that most factors have either a greater positive than negative or a greater negative than positive impact on work motivation. There are, however, a few items which were either clustered only on the motivating component or the de- motivating component of a factor. 68 Relationships with Biographic and Other Variables: The final set of results from the analyses deal with relationships among biographic and work related variables (age, education, length of employment, etc.) and between these variables and the various moti- vational items (factors) derived from the content analysis. Only some of these results are presented below; a fuller list of correlations are shown in Appendix 3. All correlations being reported in this section are based on the entire sample that was interviewed, that is N 8 341. One variable which was hypothesized a priori as likely to in- fluence work motivation is an employee's level of education. The data show a meaningful pattern of correlations between level of education and other work related variables and between education and motivational factors: amount of education correlated negatively with age (r - -.27, p‘< .001). This result is to be expected in many de- veloping nations where older persons tend to have less education be— cause of the fewer educational opportunities that existed in those countries until very recently. Education yielded negative correla- tions with both the number of years spent with an organization (r - -.24, p < .001), and number of years in paid employment (r - -.31, p < .001) showing that more educated employees who also tend to be younger have not been long with their organization and have been in paid employment for a shorter time. Of importance are the correlations between education and frequency of mentioning pay, job security and fringe benefits as motivators. More educated employees in the sample mentioned pay as a motivator fewer times than less educated employees (r = -.18, p < .001). This 69 relationship holds true for job security (r = -.18, p'< .001) and fringe benefits (r - -.l3, pi< .008). More educated employees how- ever, mentioned recognition (r = .22, pi< .001), responsibility (r - .19, pi: .001), chance for learning and training (r 8 .15, pi< .003), and interesting, important and challenging work that has variety (r - .30, pi: .001) as motivating them more often than their less educated counterparts said was the case. On the other hand, boring unimportant jobs were mentioned as sources of demotivation more often by more educated than less educated employees (r = .11, pi< .045). These same people (more educated) cited bad physical working conditions and death or sickness in the family as demoti- vating them fewer times than was mentioned by the less educated em- ployees (r = -.15, p'< .01, and r 8 -.l7, p‘< .003, respectively). Another variable that was proposed a priori as likely to have impact on work motivation is an employee's occupational level. Al- though occupational level in this study was operationalized only through a rough ordinal scale (ranging from manager, technician, secretary, clerk to general worker), correlations between this vari- able and others provide meaningful data. A positive relationship between occupational level and education (r 8 .68, p‘< .001) shows that people in higher occupations tend to have more education. Another set of findings are that employees in higher occupations mentioned recognition (r - .23, p‘< .001), chance for learning and training (r 8 .15, p< .003), achievement (r = .11, p< .021), responsibility (r = .17, p<< .001), and interesting and important, challenging work that has variety (r - .25, p<= .001) as sources of motivation more often than employees in lower occupations. Employees 70 in higher occupations also cited lack of chances for learning and training to be a source of demotivation more often than their counterparts in lower occupations (r = .12, p'< .03). Another notable result is that employees in high occupations mentioned hunger as a factor discouraging them from working hard less often than employees in lower occupations (r 8 -.20, p‘< .001). And, finally, as was the case with education level, high occupation level employees referred to death or sickness in the family and bad physical working conditions as sources of demotivation less often than did lower occupational level workers (r 8 -.22, p < .001) and (r = -.11, p < .04) respectively. Age is yet another variable that was proposed to have impact on work motivation. Results show that age is very highly related to number of years in paid employment, (r = .84, p‘< .001) and to number of years with an organization (r - .52, p‘< .001) suggesting, as can be expected, that older persons have been working longer and tend to stay longer with an organization than younger employees. Older employees have also lived in urban areas longer than younger employees (r - .30, p‘< .001). Of more interest perhaps, are the correlations between age and motivational variables. Age correlated negatively with frequency of mentioning chance to learn more or have further training (r - -.l9, pi: .001) and with chance for promotion (r 8 -.ll, pi< .03). This means that older employees in the sample cited chance for more learning or training and promotion opportunities as sources of motivation less often than younger employees. Lastly, the rela— tionship between age and frequency of mentioning urgently required work or work that has a completion deadline is low but negative 71 (r 8 -.ll, p‘< .02) suggesting that older employees in the sample are less motivated by urgent assignments than are younger employees. Closely related to age (as shown by the high correlations between these two variables (r 8 .84, pi: .001) is the number of years of paid employment. Years of paid employment shows a low negative correlation (r 8 -.17, p‘< .001) with frequency of men- tioning chance for more learning or training, and with mentioning urgently required work as motivating (r 8 -.10, p<= ,029). This indicates that employees with more years in paid employment were less likely to cite urgently required tasks and chance for more learning and training as motivating than employees with fewer years in paid work. These results are very likely to be confounded by age. The last of the factors proposed a priori to have influence on motivation in this study was length of urban residence. This vari- able shows low positive but significant correlation with recognition (r 8 .13, p<= .009) and achievement (r-8 .12, p'< .01). This result suggests that people who lived in urban areas larger were more likely to mention achievement and recognition as tending to encourage them to work hard than would individuals who lived shorter periods of time in cities. Urban residence is also positively related to occupational level (r ' .12, pi: .025) indicating people who lived in urban areas longer are more likely to hold higher occupations than those with shorter residence in urban areas. Surprisingly, however, years of urban residence showed a negative relationship with mention of urgently required work as motivating (r 8 -.11, p < .026). But as years of urban residence was positively related to age (r 8 .30, pi: .001) this finding may be confounded by the effect of age. Other relationships 72 are that years of urban residence is associated positively with both number of years with an organization (r 8 .29, p'< .001) and with total years in paid employment (r 8 .32, p‘< .001). One more variable investigated - though not one of the a priori factors - is sex. Sex correlated with years of paid employment (r 8 .29, p'< .001) meaning that women have been in paid employment for fewer years than men. Women cited a lot of work or difficult work as moti- vating more often than men (r 8 -.13, p .10) between conditions employees claimed would en- courage them to expend greater effort and the conditions which, ac— cording to critical incidents, led them to doing so. On the other hand, a strong positive relationship was obtained between conditions employees said would discourage them from working hard (when asked directly) and conditions which, based on critical incidents, dis- couraged them from hard work in real work situations (r 8 .84, p‘< .001). It would appear from this that employees can more accurately describe events, activities or conditions that annoy and lead them to inhibit effort than they can those that encourage them to exert greater effort. This contradicts the view by Vroom and 110 Maier (1961) who suggested that, because of defensive processes, individuals are likely to be more effective describing good critical incidents than bad incidents. It would appear from the present data that the opposite view might in fact be closer to reality than the Vroom and Maier position. This finding could have important implications not only for design of research investigating factors that discourage hard work but also for managers, supervisors, and practitioners interested in identifying and reducing the impact of such factors. While they may have to use unobtrusive and more sophisticated designs to learn about factors likely to increase work motivation, they may simply ask employees to tell them what conditions, factors, or events would depress their motivation to work - a very inexpensive and non-time consuming technique indeed. And since this study has shown that most motiva- tional factors have the ability to both impair and enhance work moti- vation. Further understanding and development of this approach may prove very useful. For once factors claimed by subjects to discourage hard work are obtained, the next step would be to reduce or eliminate such factors and observe if increased motivation would result. Future research should further investigate this phenomenon and look into conditions that may lead to even better agreement between employee assertions about motivation and what actually discourages them from working hard in real life work settings. lll CONCLUSIONS AND DIRECTIONS FOR FURTHER RESEARCH This study found that work motivation in the Zambian workforce is determined by six motivational factors: work nature and context, growth and advancement, material and physical provisions, relations with others, fairness-unfairness in organizational practices and personal problems. Contrary to the position of Herzberg's motivator- hygiene theory, all but one of the six factors can affect work moti- vation both positively and negatively. These factors, moreover, have a differential positive-negative impact on motivation, that is, any given factor tends to have greater potential to increase than to decrease motivation or vice versa. The model of motivation in this paper incorporates within it elements of several (process and content) formulations of work moti- vation: Components of the goal setting, two-factor and need achieve- ment approaches are embodied in the factor work nature and context. The powerful effects of inequity resolution on work behavior have been subsumed under the factor fairness-unfairness in organizational practices. A number of factors from the motivator-hygiene theory have been embodied in factors such as material and physical provi— sions, relations with others, and growth and advancement. And finally, although not directly incorporated in the differential positive- negative impact model, data in this study showed strong support for the existence of some type of hierarchy of needs with potential to moderate the effects of the model. The fact that some components of various theories of motivation have been confirmed and have been incorporated in a more inclusive and possibly more powerful model suggested in this study should be 112 seen as a demonstration of some but incomplete validity of these theories. The lack of support from the data for some components of these theories is an indication of their shortcomings or partial deficiency with respect to validity. This point illustrates and reiterates the position (emphasized in the introduction of this paper) concerning the futility of attempting to explain complex work behavior relying on a single motivation framework or theory since, as has been demonstrated here, no single formulation of motivation appears to sufficiently explain work behavior. However, the theory proposed here has the advantage that it incorporates the more pro- mising parts of several existing theories. Besides being empirically derived, the theory contains elements that may be unique to the social, cultural and economic conditions of many developing countries particularly Zambia. The point that the differential impact model embodies elements that may be unique to Zambia raises the issue of generalizability of the model, especially as it was developed on data from conditions and a sample that was entirely Zambian. While this is acknowledged, the fact that the model has incorporated components of concepts from models of some demonstrated validity in some countries perhaps makes this criticism less damaging. Besides, elements incorporated in the six factors of the theory, that can be said to be unique to Zambia constitute only a small component of the model. The strength of the data presented in this paper and the resulting differential positive- negative impact model accrue from the care taken in the design of this study. First, unlike the Herzberg study which was criticized (Vroom, 1966) for asking employees to remember times when they were 113 very satisfied and very dissatisfied and proceeded to infer motivation from resulting data-(Herzberg, et al., 1966, p. 141), the present study focused on motivation by asking subjects to talk about incidents when they worked very hard and when they put very little effort into their work. Since working very hard or working very little are evidence of motivation, the study clearly addressed the issue of motivation more than the Herzberg study did. Second, questions used in the interview were arrived at after extensive pretesting with a sample of 40 employees drawn from the five occupational categories used in the study. (The pilot study data were not included in the later analyses.) Third, the sample of 341 employees consisting of five occupational groups was care- fully drawn from eleven organizations that represented the three main types of work organizations in Zambia - government, quasi-government and private. Finally, the greatest strength in the design of this research lies in the fact that results from content-analyzed critical incident-interview data were validated using a different method: A questionnaire developed from the data was administered to a new sample of 80 employees drawn from organizations not included in the original study but which again represented government, quasi-government and private organizations. On the practical side, this study may prove to be of immense poten- tial value to Zambia and surrounding regions if the practical recom— mendations presented in this paper were implemented. Consequently, from the point of view of the region studied, this paper goes a long way in helping develop a locally relevant body of knowledge which, 114 with further research and development, may, in the long run, help nations in that area to enhance productivity of their human resources. It must be remembered, however, that this study was only explora- tory and that, despite the very encouraging findings reported here, much work needs to be done to test and refine various aspects of the differential positive-negative impact model of motivation. To begin with, it is now necessary to draw hypotheses from the model and see if actual performance (both hard and soft) data in organizations confirm the predictions of such hypotheses. At the same time, priority must be given to the development of scales based on the six factors. Once the scales are developed, confirmatory factor analyses can be performed on data collected using these scales to empirically confirm the factors identified in this study. Further, the scales, once validated, can prove useful in gauging the level of employee motivation and can also serve as diagnostic instruments which could pinpoint job aspects deficient in certain desirable characteristics. Another concern for research should be an attempt to replicate this study and to validate the differential impact model in other developing countries, particularly those which have social, cultural and economic conditions similar to Zambia. For although this author contends that the differential positive-negative impact theory will hold in many countries with varying social, cultural and economic conditions, the model would probably be more explanatory and pre- dictive in regions that have conditions somewhat similar to those in Zambia. Another issue deserving future research attention is the extent to which expectations of organizational members influence their 115 (members) work behavior in the presence or absence of various factors of the differential positive-negative impact model. An understanding of this issue would make it easier to anticipate and plan for some problems that may be encountered in applying this model to work situations. Finally, more investigation should be made into the findings that there is high agreement between what employees say discourages them from working hard and what actually discourages them. As pointed out earlier, further research on this phenomenon can make it easier to learn about factors that adversely affect motivation to work by making the use of costly non obtrusive approaches unnecessary. APPENDICES H6 >wwnzu~x w Amonmnceoc oconnaonm son Mennoncnnonv mnmnesonnv Hansoncononw mnmnososn" 1% some no ............... neon n70 condoomnnw am unseen. m. nomonson tun: ooo Onaon mmoonnmno.neo nooncnnnom monsoon: on tour sonpcsnuoo one ononconwcnnw. use ououonn so one cOHRpon on sewn noncpm on no canononoon «mononm noon nsmuconoo mooopo.e cant Isnncmnnoo on aropn noon. to one penanceotnnw Announces. owners. moonnnmnpoe. non Hococnonmv moon moconssosn anomnnlomnm. omnsmnmnoue. mnpcmno oolomsnom one are noooon l—nom. «one noooonmnwoo one omnnpopvmnpos no ares mnouoon »m_conw unmounonn on wn new“ rope no canonmnmoa so: to. tone —: weaves. no: uncommon Isnwcmnno: no canon «0 seasons muonconpcnnw on courses. a as moenw no sex woe mono acoonuoom swoon mono Om ween uov nxoonwmonon. H edema were one no unseen are! me onncnmno~< so nauseous. u teen non our non woes none on canoe an noes. aceoenrnom «on spun nowp so an oozmncm2du>r man curr zed an acre HO >z33 5 48! g 332-. Hag—g nan“): san- 48 I>3 s is =3 Pug a )2: ‘3. «Sn i’d «a Gnu a g 3 in!» n. «tons .e s new on tour no es eons. P111149: announces on reason one no: ll nonsense» a. . «on recs cos: seepnseb snares poo Ian: 3 goo :3: (out. o. atone .m e lines no we solo—plasmas. looses-mass. mes-ms.— oo goons—use now adorns: tens. u. “teaspoon on coarse manna-s1 alts—r]: nae—e— ...-51!...ezos. neeoceezel. somonnl on saeemonos. a..- arses ens moon qs-ennose. seasonal-=3 cantons «needs—n. cosmos ll mono deviants. u. «as see seasons.— oe area arses mus sressse nee Cool—3...... a... liars to; no sense» season lee» to unspent suns—s e eel—ass. so. ‘0 rem Vsss wens—ea s—— new. no. anemone on le-oese nose son news on ..eass no monsoon. moor—ass. nu. one no assess-one man arses .e moeenr——.~< use nssnslssn. vases. sen—n nssnslssn on anssssnm— .ssssnnes. nn. itses no so ensues. somessnsanos. salon—Issn on «seems—anon «on teen soar. nu. her no senses Anne ens so» mars—a no we essrsav. —o. «to use o~ns~e nn-aas rsssn—ne asnsosse toes.sm. ~osse. nqsssooqn no soar. esp-nu oneness-v. nm. use use» so mos «sen —.rs seer—om. no. «on ens tales-3. mamas—ls.— ou. morassnl no apes—33.5: son‘s... ~v. 4’s solo-nu on noose—oee no son rssm nae-n meanness capo: tons sons 90 fine. us. use use spaces use senescent—unna as; assess no more ess.n.oss. no. arses pm ensrssnm an seen: n: was «esp—u on one sees eaten I'll masons- oo soleenns moor—sis. no. is as e nil-so no use; lone arose «as one so nvsqs ... assess non neon-.1. one—spam. up. «an be no... In. nssnslssns. tosses... ass—n .ssnslssns ea 3.34 :ssasue— nonssnnese mos son‘s—u. nu. issue one I aspen—one on Inga-assassinate coats: «oases—... v2.8 esmeenoee 3. tonnes In. «one soerlense. nu. «to soar no possession. agape». era—soar! on see eon—e3. up. mega—one on stands. «on season—om one as: saints-v. nu. use moss—:3 no Bloopers.— Awos ens mars-e no to nearly. so. «is; no m moon—0...: non noises!» on «one as... us. «one comexse‘eneneon scmswo-esm shoes: awesssese on Vote ass: on 702:! noore en no.3. to; on e-—. sum. 3 rss no» names; «on s ass news. no. :53 one :3: on so nensas essen—ne 23.325. posse one... 3. is —m =3: stones at asses loss as «one ..ov on :nZs arenas non uses—lo seeps—l. .3. floor .e no» moo-3e: sum—epsi- oq area was ..ses so eossnnns no... 4.! Ins soon-333% lees.— noi as no no... amass see nos—else. on: no ..s—m sin-ass mussels—5 on Deleon—ma so.— asIOn—eszai 31!. no asses nonsense not see so» moo—.1. 118 Zoos—x. u» u. nodesumnpooou >305» conrnwswmnsa onwwnwm om canons-u unspnssnm Anon >H~ nounspmnuose a s uauv >. acnnspmnpono Own: ZOnncmnpsmlaooon moswmsonv aocnn an snnonnno: .oa .oo~ «sons cm coco: «sonnsons .Hu .GN arenas non Hsmnauom nnmpnnom .um .ocu scar «son he wnnsnsmnuom. nsoonnmnn. ornpusowpnw one one cmnpsnw .nu .ocu ”snooomnv»~»nw .uu .OOH xsnomopnnoo .Nu .ocw >n7»s- Isnncmnnosnu con—mouse tons apprenoaocmpw cocoon. u.s.. m confirm—s «snswcsn m moons an ass nu an tom Isnnuonsn so mono um »n ten non. connovHs mum Anonsn mm nouuotmu ls: ~. toss: av «sons am cows: osmpnsoos >aocan on nwmmnscpn pscsp cm tow: use msncnunw «seen an omen saopowssnn bum >aocnn on sncnmnuon «sans can: nos anemone «sons On omen sno~onrns