Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/unpaywall/cognitive-elements-of-empowerment-an-interpretive-model-of-intrinsic-dw98nSD05A
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- Unpaywall
- ISSN
- 0363-7425
- DOI
- 10.5465/amr.1990.4310926
- Publisher site
- See Article on Publisher Site
Abstract
Calhoun: The NPS Institutional Archive DSpace Repository NPS Scholarship Publications Cognitive Elements of Empowerment: An "Interpretive" Model of Intrinsic Task Motivation Thomas, Kenneth W.; Velthouse, Betty A. Academy of Management Review, 1900, Vol. 15, No. 4, pp. 666-681. https://hdl.handle.net/10945/45984 This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States. Downloaded from NPS Archive: Calhoun o Academy 01 Management Review . 1900. Vol. IS. No. 4. 666 - 68 1. Cognitive Elements of Empowerment: An "Interpretive" Model of Intrinsic Task Motivation KENNETH W. THOMAS Naval Postgraduate School BETTY A. VELTHOUSE University of Michigan-Flint This article presents a cognitive model of empowerment. Here, em powerment is defined as increased intrinsic task motivation, and our subsequent model identifies four cognitions (task assessments) as the basis for worker empowerment: sense of impact, competence, mean ingfulness, and choice. Adopting an interpretive perspective, we have used the model also to describe cognitive processes through which workers reach these conclusions. Central to the processes we describe are workers ' interpretive styles and global beliefs . Both pre liminary evidence for the model and general implications for re search are discussed . Empowerment has become a Widely used ventions, while being more explicit about what word within the organizational sciences (e.g., those effects are. More specifically, Conger and Bennis & Nanus, 1985; Block, 1987; Burke, 1986; Kanungo (1988) proposed that empowerment be Harrison, 1983 ; House, 1988 ; Kanter, 1983; defined as increases in workers' effort-perfor Neilsen, 1986). At this early stage of its usage, mance expectancies (p. 475) or, using Bandura's however, empowerment has no agreed-upon (1977 , 1986, p. 474) term , feelings of self-efficacy. definition. Rather, the term has been used , often Empowering interventions, therefore, enable loosely, to capture a family of somewhat related workers to feel they can perform their work com meanings . For example, the word has been petently. This empowering experience, in turn, is asserted by Conger and Kanungo (1988, p. used to describe a variety of specific interven tions, as well as the presumed effects of those 476) to increase "both initiation and persistence interventions on workers (Conger & Kanungo, of subordinates' task behavior." Building on this 1988). definition, those authors constructed a model of In 1988 Conger and Kanungo took an impor organizational conditions, managerial strate tant step toward clarifying this concept. These gies, and types of information that produce em powerment and its behavioral effects. authors recommended that empowerment be defined in terms of motivational processes in This article further develops the general ap workers. This approach allows researchers to proach taken by Conger and Kanungo. That is, study the empowering effects of different inter- empowerment is conceptualized here in terms of 666 changes in cognitive variables (called task as forced a search for alternative forms of manage sessments), which determine motivation in ment that encourage commitment, risk-taking, workers. However, a more complex cognitive and innovation (e .g., Harrison, 1983; Kanter, model is developed. Our model attempts to im 1983; Walton, 1985). This trend has been espe prove upon Conger and Kanungo's in three cially apparent in the fields of leadership and ways. First, the concept of empowerment as mo organizational culture, where research has tivation is made more precise by identifying em shown how transformational and charismatic powerment with a type of motivation, referred to leaders can energize workers by tapping ideal here as intrinsic task motivation. Second, we at ism and building faith in the ability to accom tempt to specify a more nearly complete or suf plish meaningful goals (e.g., Bennis & Nanus, ficient set of task assessments that produce this 1985; Burns, 1978; Deal & Kennedy, 1982; House, motivation. Although self-efficacy is included in 1977; Schein , 1985). Recent conceptualizations of this set, it is not regarded as sufficient, and it is organizations also have demonstrated how con supplemented by three additional task assess trols can be reduced under conditions of strong ments. Third, the model attempts to capture the goal alignment to allow initiative without sacri interpretive processes through which workers ficing coordination (Berlew , 1986; Harrison, arrive at those task assessments. In Conger and 1983; Kanter , 1983; Walton, 1985). Finally, stud Kanungo's model. individuals' judgments re ies of successful firms and leaders have given garding personal efficacy are assumed to reflect credibility to these approaches by providing objective conditions/events and information that compelling examples of organizations that oper flows from those events. In this way, they made ate in this manner (Bennis & Nanus , 1985; stimulus-response assumptions about the causal Kanter, 1983; O'Toole , 1985; Peters & Austin, effects of those external stimuli upon individu 1985; Peters & Waterman, 1982). als. In contrast, we view such judgments as sub A common theme in this literature has been jective interpretations (constructions) of reality, the limitation of the classicallbureaucratic para so that task assessments are also influenced by digm of strict controls combined with contingent individual differences in interpretive processes. rewards and punishments-a paradigm in which work tasks are presumed to have only instrumental value to workers, and in which the Empowerment and worker's role is primarily to comply (e.g., Block, Intrinsic Task Motivation 1987). In contrast, the newer paradigm involves To empower means to give power to. Power, relaxed (or broad) controls and an emphasis on however , has several meanings . In a legal internalized commitment to the task itself. Thus, sense , power means authority, so that empow authors have emphasized the importance of the "pull" of the task rather than the "push" of man erment can mean authorization . Power also agement (Berlew , 1986), making work meaning may be used to describe capacity, as in the self efficacy definition of Conger and Kanungo. ful (Block, 1987; Schein, 1985), identifying with However, power also means energy. Thus, to the task (Bennis & Nanus, 1985), or finding ex pressive value in the task (Shamir, House, & Ar empower also can mean to energize. This latter thur, 1989). In this way, recent developments in meaning best captures the present motivational the management literature converge with the usage of the term. Our perception is that the word empowerment motivational assumptions of the job design liter ature (e.g., Hackman & Oldham, 1980). has become popular because it provides a label We use the word empowerment to refer to for a nontraditional paradigm of motivation . the motivational content of this newer paradigm Widespread use of the term has come at a time of management. This use of the term is consis- when foreign competition and change have 667 tent with usage by Bennis and Nanus (1985) and is the major purpose. That model contrasts Block (1987). It is also broad enough to include sharply with models .derived from work settings, the self-efficacy definition as one factor in pro where purposes are considered less trivial. For ducing this motivation (as detailed in the follow example, the Hackman and Oldham (1980) ing sections). model includes job "meaningfulness" (based in part on task Significance) as a necessary psy Intrinsic Task Motivation chological component of intrinsic work motiva tion. Likewise, Shamir and his colleagues (1989) Because the focus of this new paradigm is on concluded that the most important motivational commitment to the task itself, the present model aspect of charismatic/transformational leader operationalizes empowerment in terms of intrin ship is the heightened intrinsic value of goal ac sic task motivation. In general terms, intrinsic complishment produced by the articulation of a task motivation refers to what Brief and Aldag meaningful vision or mission (see also Bennis & (1977) referred to as intrinsic work motivation , Nanus, 1985, and Block, 1987, on the role of lead but at the level of analysis of individual tasks or ers' visions in subordinates' empowerment). projects. The terms intrinsic and intrinsic moti vation have been used inconsistently (Dyer & An "Interpretive" Approach Parker, 1975; Lee, 1987), so we shall be precise in spelling out our definitions and assumptions. Existing models relating to empowerment and Essentially, intrinsic task motivation involves intrinsic task motivation have focused predomi positively valued experiences that individuals nantly on the role of objective external condi derive directly from a task. In the present cogni tions/events as independent variables-for ex tive modeL intrinsic task motivation involves ample, job characteristics (Hackman & Oldham, 1980) , leaders' behavior (Bennis & Nanus, 1985), those generic conditions by an individual, per taining directly to the task, that produce motiva and other variables outside the individual (see tion and satisfaction. The core of the modeL Conger & Kanungo , 1988). In a critique of re then, involves identifying these cognitions , search regarding job characteristics, Roberts called task assessments. Task assessments are and Glick (1981. p. 193) identified the need for presumed to be the proximal cause of intrinsic "alternative theoretical perspectives that distin task motivation and satisfaction. They occur guish between situational attributes ... and in within the person and refer to the task itself, cumbent cognitions about those attributes" (see rather than to the context of the task or to re also Glick, Jenkins, & Gupta, 1986). The same wards/punishments mediated by others . Fi critique could be extended to the more extensive nally, they are assessments along dimensions literature regarding empowerment and intrinsic that are generic enough to be applicable to any task motivation. task. The present model attempts to provide the sort The construct of task is central to our defini of "alternative theoretical perspective" recom tion. Here, task refers to a set of activities di mended by Roberts and Glick (1981). Rather rected toward a purpose. (A task can be as than simply capturing individuals' cognitions re signed or chosen.) Thus, a task includes both garding task characteristics, however, this activities and a purpose (e.g., Staw, 1976). This model attempts to identify key interpretive pro distinction is crucial. The Deci model of intrinsic cesses that introduce diversity into individuals' motivation (Deci, 1975; Dad & Ryan, 1985), for cognitions. To do this, it adds a fundamentally instance, refers only to activities. Thus, the Deci new set of independent variables to organiza tional behavior, "interpretive styles." These in model has been researched most often in the context of games, where enjoyment of activities terpretive styles, along with "objective" vari- 668 abIes in the individual's environment, are pre The Cognitive Model sumed to have an additive effect upon the individual's task assessments and, hence, on The proposed model is shown in Figure 1. In their empowerment. general form, the model resembles the social The interpretive approach taken in this model learning sequence of stimulus , organism , be diverges from existing cognitive models of orga havior, and consequences (S-O-B-C) (Davis & nizational behavior by explicating a form of soft Luthans, 1980). However, its focus is on intra social constructionism (Burrell & Morgan, 1979). personal cognitive processes. Thus, conse Although cognitive models are now common quences and stimuli are simplified into a single place in organizational behavior , their design external element , environmental events, ers have been overwhelmingly objecti vist in whereas intrapersonal cognition (the organism their assumptions. It is assumed in these models, part of the S-O-B-C paradigm) is expanded into then , that individuals' cognitions are about a several elements. verifiable , external reality. Therefore, the domi An overview of the model's operation will help nant criterion in evaluating cognition (percep clarify subsequent discussion of its elements. tion or judgment) is its objectivity or accuracy , The model's core is the ongoing cycle of envi that is, the degree to which it matches external ronmental events, task assessments, and be reality. Deviations from objectivity are consid havior (the loop composed of Elements 1, 2, 3 in ered dysfunctional errors. (For a summary of Figure 1). Environmental events provide data to such research, see Bazerman, 1986.) the individual about the consequences of ongo In contrast, our model applies a soft construc ing task behavior and about conditions and tionist perspective to intrinsic task motivation. events relevant to future behavior. This data is Observable external events and conditions are seen as shaping the individual's task assess regarded as verifiable (1. e. , as factual or objec ments regarding impact, competence, mean tive) . However, individuals' judgments and be ingfulness, and choice. These task assessments, havior regarding tasks also are shaped by cog in turn , energize and sustain the individual's be nitions that go beyond verifiable reality. Such havior. This behavior then impacts environmen interpretive cognitions go beyond the percep tal events, and so on. tion of facts to provide additionaL needed mean Consistent with the earlier discussion, the task ing for an individual. In particular, the model assessments (Element 2) are viewed as interpre identifies three interpretive processes through tations or constructions of reality, rather than which individuals add meaning to factual per simple recordings of objective facts . Thus, intrin ceptions about tasks: evaluation, a ttribution , sic task motivation (and subsequent behavior) is and envisioning. These processes provide task asserted to be influenced not solely by external events, but also by the way those events are related cognitions about how well things are go ing, about what may have caused past events, construed. Accordingly, two intrapersonal ele and about what could happen in the future . ments (Elements 4 and 5 in Figure 1) are added Styles of performing each of these processes are to the model to help explain variations in the construction of task assessments. identified. These styles lead individuals to con Global assessments (Element 4) are an indi strue events in ways that may be equally con vidual's generalized beliefs about impact, com sistent with facts, but which have markedly dif ferent effects on task assessments (and, hence , petence, meaningfulness, and choice. These are more abstract beliefs in contrast to the spe motivation). Rather than being a matter of accu cific assessments (Element 2) that motivate a per racy these interpretive styles can be more prop son's behavior in a given task situation. As erly evaluated in terms of their usefulness for shown in Figure 1, global and situational as- individuals in empowering themselves. 669 5 Int.rpretiv. Styl •• - Attributing Global - Evaluating A ..... m.nt. - Envisioning -Impact -Competence -Meaningfulness -Choice Ta.k A •• e •• m.nt. - Impact Environm.ntal 6 Interv.ntion. 2 -Competence Ev.nt. -Meaningfulness -Choice B.havior - Activity - Concentration - Initiative - Resiliency - Flexibility Figure 1. Cognitive model of empowerment. dividual's assessments of his or her impact on sessments are assumed to shape each other. assessments are assumed to be induc specific tasks. In tum, task assessments are Global tive generalizations from past task assessments, shaped deductively, in part, by a person's glob and, thus, they represent an individual's cumu al assessments. That is, when available infor lative learning about these factors. For exam mation leaves room for ambiguity when inter ple, a person's assessment of global impact is preting events, that individual tends to make analogous to Rotter's (1966) notion of locus of task assessments that are consistent with his or her past experience. control, that is, it is formed over time by the in- 670 Interpretive styles (Element 5) are tendencies ities-they must be interpreted. Event z must be regarding an individual's interpretive process evaluated as progress or setback, causal attri ing of events. This processing adds subjective butions for that outcome must be made, and information regarding evaluation, attribution , probable effects must be anticipated. The objec and envisioning. Specific styles of performing tive "accuracy " of these interpretations may each process are asserted to have direct effects only be clear long afterward (if at all), viewed in on an individual's task assessments . For exam the context of subsequent events. ple, an absolutistic style of evaluating implies In addition to individuals' direct experience of that outcomes will more likely be construed as the outcomes of their behavior, a variety of other failures, thus reducing the individual's task as external events provides data on which to base sessment of impact (and , in turn , diminishing task assessments (see Bandura, 1977, and Con that person's continued striving and the likeli ger & Kanungo, 1988 , on alternative sources of hood of further successful outcomes). Because of information regarding self -efficacy). These their key role in the motivational cycle , as noted, events include inputs from superiors, staff , these interpretive styles identify a significant peers, and subordinates, for example, perfor way in which individuals may empower or dis mance evaluations, charismatic appeals, train empower themselves, setting up self-enhancing ing sessions , m e ntoring advice , and general or self-debilitating cycles. discussions of ongoing projects. Such inputs Finally, empowering interventions (Element 6) also must be interpreted and reconciled with the provide ways of influencing the variables in the individuals' more direct experiences. model to increase an individual's task assess Task Assessments ments (as well as producing gradual increases in global assessments) . Two general interven tion strategies are identified: changing the envi Processing this somewhat ambiguous data, ronmental events on which the individual bases individuals make a number of assessments or his or her task assessments and changing the judgments with respect to specific tasks. Four individual's styles of interpreting those events. dimensions of assessment are included as cog The following is a more detailed discussion of nitive components of intrinsic motivation: im the model's elements . pact, competence, meaningfulness, and choice. These four task assessments are seen as having Environmental Events additive motivational effects. As detailed below, these four assessments are offered as a synthesis As we noted , environmental events are of the cognitive motivational content of a variety sources of data about the consequences of the of theorists , including Deci (1975), Hackman and Oldham (1980) , Bandura (1977, 1986), and oth individual's ongoing behavior and about condi tions relevant to that person's future behavior. ers. These assessments also capture motiva For many tasks , especially novel or complex tional aspects of current leadership models that ones, the consequences of behavior are inher apply at the individual level of analysis. (Lead ently ambiguous. Descriptions of these out ership-oriented models include some uniquely "group" content, such as a sense of community, comes may be verifiably true or false with re spect to their observable components: "I per which is not included here. See, for example, formed actions x and y, which were followed by Bennis and Nanus, 1985.) event z." To have motivational implications for Consistent with the eclecticism of the social learning approach (Kreitner & Luthans, 1984), the individual, however, these "objective" out these four variables combine both expectancy comes must be given meaning or significance and reinforcement dynamics. Thus, from an ex- with respect to the individual's goals and activ- 671 pectancy theory perspective (Lawler, 1973), im nents-universal helplessness and personal pact represents a performance-outcome expec helplessness. Universal helplessness (similar to tancy, competence an effort-performance ex external locus of control) occurs when impact is pectancy, and meaningfulness an anticipated seen as unlikely, regardless of performance. In outcome valence (for intrinsic motivation), contrast, personal helplessness occurs when a whereas choice represents the perceived oppor person perceives that impact would be possible, tunity for a decision based on these variables . but the competence to perform is lacking. During the course of an activity, however, each Using the distinction between task and global of the four assessments also can be viewed as assessments in the present model, research an intrinsic reinforcement. That is, each assess about learned helplessness has been focused on ment is also a reward that individuals can give assessments of a person's impact on specific themselves during the course of an activity, a tasks. Experimental work has demonstrated the reward that reinforces their continued striving. following negative effects (deficits) produced by Each of these variables, in fact, has been treated universal helplessness: dampened ability to rec by earlier theorists as a separate intrinsic need ognize opportunities, reduced motivation , and or reinforcer. depressed affect (Abramson et aI. , 1978). (Re The following is a more detailed description of search on locus of control. which has been used each situational assessment, along with a brief to examine global assessments of impact, will discussion of behavioral effects. be discussed in a later section of this article.) Impact. This assessment refers to the degree Competence. This assessment refers to the de to which behavior is seen as " making a gree to which a person can perform task activi difference" in terms of accomplishing the pur ties skillfully when he or she tries. In clinical psy pose of the task, that is, producing intended ef chological literature , this variable has been fects in one's task environment. The general no studied by Bandura (1977, 1986) under the terms tion of perceived impact has been studied under self-efficacy or personal mastery. Here we have a variety of labels, including locus of control (Rot used White's (1959) simpler term, competence. ter, 1966) and learned helplessness (Abramson, Bandura (1977) observed that low self-efficacy Seligman, & Teasdale, 1978). Within the Hack leads people to avoid situations that require the man and Oldham (1980) model of job redesign, relevant skills. This avoidance behavior, in turn, impact is analogous to knowledge of results. tends to prevent an individual from confronting We have labeled this variable impact in the fears , building competencies, and improving light of conceptual refinements that help to clar perceived competence. Furthermore, Abram ify the variable and to distinguish it from com son and his associates (1978) concluded that petence. Briefly, the constructs of control and people who experience personal helplessness, in addition to suffering deficits from universal helplessness are somewhat ambiguous because perceived control over the environment (or lack helplessness (low impact), experience deficits in of helplessness) involves both the belief that a self-esteem. In contrast, high self-efficacy tends person's behavior could have an impact and the to result in initiating behaviors, high effort, and belief that one could perform the relevant be persistence in the face of obstacles (Bandura, 1977). havior competently. Accordingly , Bandura (1977) noted that research on locus of control Meaningfulness . This assessment concerns usually involves the instrumentality of behavior the value of the task goal or purpose, judged in (the performance-outcome relationship, rather relation to the individual's own ideals or stan than the effort-performance relationship). Like dards. In other words, it involves the individual's intrinsic caring about a given task. This use of wise, Abramson and his colleagues (1978) di meaningfulness is analogous to Hackman and vided learned helplessness into two compo- 672 Oldham's (1980) term, although it is at the level dimensions of behavior shown in Figure 1. Two of specific tasks or projects. In psychoanalytic of these dimensions are also characteristic of ex terms, meaningfulness represents a kind of trinsic motivation: activity (as opposed to passiv cathexis (or investment of psychic energy) with ity) and concentration of energy upon task. respect to a task. These two dimensions translate roughly into fo Low degrees of meaningfulness are believed cused effort or "working hard." What is distinc to result in apathy, feeling detached and unre tive about intrinsically motivated behavior is lated to significant events (May, 1969). Higher that this effort is not dependent upon the super levels of meaningfulness, in contrast, are be vision of others nor upon rewards mediated by lieved to result in commitment, involvement, others. Thus , individuals may not only work in and concentration of energy (e.g., Kanter, 1968; the absence of close supervision, but also they Sjoberg, Olsson, & Salay, 1983). may demonstrate flexibility in controlling their Choice. This last assessment involves causal own task accomplishment, initiation of new responsibility for a person's actions ; it is what tasks as problems or opportunities arise, and re deC harms (1968) termed locus of causality. Deci siliency to obstacles, sustaining motivation in and Ryan (1985 , p. 152) noted that locus of cau the face of problems or ambiguity. sality is quite different from Rotter's (1966) locus Other things being equal, the behaviors in El ement 3 of the model increase the likelihood that of control: Locus of control involves outcome contingencies, whereas locus of causality in individuals will achieve outcomes that will, in volves the issue of whether a person's behavior tum, proVide further evidence of competence, is perceived as self-determined. deCharms choice, and impact on meaningful goals. Thus, the task assessments, through their effects upon (1968) argued that perceiving oneself as the lo cus of causality for one's behavior (as origin behavior and subsequent outcomes, have some rather than pawn) is the fundamental require tendency to initiate self-reinforcing cycles. The ment for intrinsic motivation. Deci and Ryan task assessments, in other words, have some of the qualities of self-fulfilling prophecies. Lowas (1985) also included experienced self-determi along with competence, as an essential sessments may initiate debilitating cycles of in nation, ingredient of intrinsic motivation . Hackman and activity, low initiative, and so on, which pro Oldham's (1980) model, likewise, included an duce further evidence of low impact, lack of experienced sense of responsibility that stems competence, and so forth . In contrast, high sit uational assessments may lead to self-enhanc from autonomy. Deci and Ryan (1985, p . 105) noted that the central issue in self-determination ing cycles that strengthen or confirm those as is the experience of choice. We have used the sessments. word choice here , rather than the more abstract The discussion now turns to the two intraper sonal elements of the model that help to explain or philosophical term, self-determination . Deci and Ryan (1985) observed that perceived variations in task assessments-global assess choice (self-determination) produces greater ments and interpretive styles (Elements 4 and 5 in Figure 1). flexibility, creativity, initiative, resiliency, and self-regulation. In contrast, the sense that a per Global Assessments son is controlled by events leads to tension, a more negative emotional tone, and decreased As previously noted, global assessments are generalized beliefs about the four assessment self-esteem. dimensions, aggregated across tasks and over Behavior time. They represent cumulative learnings from past task assessments and are used to help "fill In the previous section we have summarized in the gaps" in assessing novel situations. If a research that links the task assessments to the 673 person defines faith as belief held in the ab noted that experiences of efficacy (competence) sence of complete evidence, it follows that the most predictably increase the individual's ex pectancies of efficacy in similar activities. How motivation to engage in any new activity re quires some faith. Global assessments, then, ever, many experiences "instill a more general provide a potential reservoir of faith for under ized sense of efficacy that extends well beyond taking and sustaining new activities. If an indi the specific treatment situation" (Bandura, 1977, vidual's global assessments are relatively high, p. 194); "once established, enhanced self-effi the model implies that he or she will be more cacy tends to generalize to other situations in optimistic than others in undertaking activities in which performance was self-debilitated" (Ban dura, 1977, p. 195). which information about task assessments is ambiguous. In contrast, individuals who have Global Meaningfulness. This variable reflects lower global assessments will tend to be more the aggregate extent to which individuals psy pessimistic. At a given point in time, the global chologically invest in the tasks in which they are assessments can be considered individual dif involved; it can be described as their general ference variables. However, global assessments level of caring or commitment. The notion that can change over time in response to evidence people differ in overall cathexis (global mean ingfulness) is well established in the psycholog that causes new patterns of task assessments. ical literature (e.g., Solomon & Patch, 1971; Global Impact. Research on locus of control provides strong support for this part of the May, 1969). Low levels of global meaningfulness model. As noted, conceptual refinements by correspond to alienation (e.g., Brickman, 1978), Bandura (1977) have equated internal locus of and they imply a decreased anticipation that control with the perceived impact of a person's new tasks will be meaningful. behavior. Rotter's (1966) measure thus appears Global Choice. This global assessment is roughly equivalent to deCharms's (1968) notion to be a global assessment of one's impact. Con of locus of causality, that is, a person's general sistent with the model, individuals with internal locus of control (high global impact) usually ized tendency to experience him- or herself as have higher expectancies of impact on specific origin or pawn. deC harms found this variable to tasks (Gregory, 1981). Accordingly, in such a sit change over time in response to life experi uation, locus of control research provides exten ences. In the mid-198Gs, Deci and Ryan (1985) sive evidence of a link to proactive behavior, operationalized this construct as the strength of an "autonomy orientation," which leads individ resiliency to setbacks, and measures of emo uals to interpret situations in terms of greater or tional adjustment. (For reviews, see Joe, 1971; Lefcourt, 1973; Rotter, 1966.) Moreover, although lesser degrees of self-determination. locus of control has been treated as a personal Interpretive Styles ity variable, research also has shown that it changes over time to reflect new experiences Interpretive styles play a key role in the (Hoffman, 1978; Simmons & Parsons, 1983). model. They add interpretive information to Global Competence. This global assessment data from external events to produce additional information for task assessments. Global assess is equivalent to self-confidence, and it is one ments, because they are generalizations from way in which self-esteem has frequently been operationalized (Wells & Marwell, 1976). Global task assessments, also reflect the influence of competence represents a generalized sense of a these styles. person's ability to perform adequately in new Attributional styles have been used in a num ber of SOCial-learning models of motivation (e.g., situations. The existence of a global assessment Luthans & Kreitner, 1975; Martinko & Gardner, of competence has been supported by Bandu ra's (1977) research on self-efficacy. Bandura 1982). However, the present model provides a 674 more comprehensive set, adding styles of eval of competence) or external (general, permanent uating and envisioning as well. In effect, the in nonimpactfulness of the environment). Likewise, corporation of interpretive styles enables the re any style which underutilizes stable, global at searcher to use the model to delineate ways in tributions for successes is predicted to Similarly which individuals contribute to their own em handicap an individual by reducing assess powerment or disempowerment. They do this by ments of competence and/or impact. construing events in ways that set up the self Evaluating . Dysfunctional styles of evaluating enhancing or self-debilitating cycles discussed are the focus of Ellis's rational emotive therapy earlier. Such dynamics have received empirical (RET) (Ellis, 1980; Morris & Kanitz, 1975). The support in personality and clinical research starting point for the RET paradigm is a setback (which will be discussed in the following sec or frustration of the individual's expectations. tion), in addition to their prominent role in the The effects of these setbacks are mediated by self-help literature (e.g., Peale, 1954). Note that "belief systems," which appear more precisely although we draw upon clinical research, our to be evaluative standards. These standards, as primary interest is in extrapolating these dy applied to the setback, produce dysfunctional namics to nonclinical populations, that is, to the emotional and behavioral consequences. Indi broad range of style variation nearer the center viduals, however , typically are unaware of the of the distribution. effects of their standards, believing that the set backs directly cause their reactions . Thus, RET Attributing. Causal attributions are central to Abramson and his associates' (1978) reformula focuses on helping individuals to identify their tion of learned helplessness theory. The theory use of dysfunctional standards and then to mod centers on attributions made to account for ify them. (For an overview of RET, see Ellis, 1980. "failures"-typically not reaching performance For a summary of research support, see Ellis goals . Causal attributions are distinguished 1977.) along three dimensions: internal versus exter Ellis concluded that dysfunctional standards nal. stable versus unstable (or transient), and take the form of unqualified and absolutistic "musts. " In this article, for example, such a stan global versus specific. Laboratory investigations dard might be, "I must achieve perfection on have shown that depressed individuals are more likely to make internal, stable, global at nearly all dimensions of a task." Such standards tributions for such outcomes (e.g. , "I am incom are difficult to satisfy and tend to label any out come short of near-total perfection as negative. petent") than nondepressed individuals, who are more likely to cite task difficulty (external, Thus, Ellis (1980, p. 8) noted that such standards stable, specific) , lack of effort (internal, unstable, tend to produce "awfulizing" about outcomes and specific), and so on. Furthermore, depressed in dissatisfaction with the current state of one's life. In the present model, absolutistic standards dividuals appear less likely to see successes as clearly tend to reduce assessments of impact, evidence of competence. These attributional styles, in turn, have predictable negative effects because anything short of total, continuous suc upon expectancies of success at subsequent cess may be labeled failure. Absolutistic stan dards, as applied to an individual's purposes, tasks and upon actual performance. (For re are also likely to reduce meaningfulness. In con views, see Abramson et a1., 1978; Kammer, trast to the purity of ideals, concrete purposes 1983.) In the present model, any attributional style tend to involve compromises and imperfections. Absolutistic standards would tend to produce that favors stable, global explanations for set low assessments of the meaningfulness of such backs is seen as negatively influencing motiva "flawed" purposes. tion by overgeneralizing the existence of obsta Envisioning. This third process involves cog- cles, whether internal (general. permanent lack 675 displacement is expected to occur as cognitions nitive imagery of future events-visualizations or anticipations of what can happen. Harrison shift from attaining their purposes toward cop (1983) noted that interview studies of high ing with possible obstacles. performing individuals in various fields have re Interventions ported consistently that these individuals work at creating vivid mental images of successes This last element of the model (Element 6 in Figure 1) refers to deliberate attempts to produce and avoiding images of setbacks or failures. This conclusion has also been drawn by Bennis empowerment (increased intrinsic task motiva and Nanus (1985) concerning successful leaders tion) through changes in the environmental and by Garfield (1984) concerning peak per events that impinge upon individua ls, or formers in various fields. Earlier, Ford (1969) had through changes in these individuals' manner of made the same observation at the organiza interpreting those events . Both intervention tional level of analysis. Although these studies routes seek to produce increases in individuals' are impressionistic , the strong consensus task assessments (and, in turn, gradual in among them is notable . Briefly, then, high creases in their global assessments). In the performing individuals seem to work at antici model. the effects upon task assessments of en pating the positive and worrying less about the vironmental variables (see Table 1) and the in negative. terpretive styles are regarded as additive. In our model. this form of envisioning in Environmental Changes. Judgments regard creases motivation through effects upon task as ing the task assessments (and thus empower ment) are shaped in part by "objective" vari sessments of impact, competence, and mean ables in the individual's environment. Thus, the ingfulness . As an individual's phenomenal field becomes filled with images of successes, as op conventional approach to empowermen t has in posed to visions of obstacles and setbacks, per volved interventions that target such variables. ceptions and expectancies of competence and Table 1 contains a selected list of environmental impact are expected to increase. In addition , en variables that have been asserted to shape the visioning successes allows the individual to con task assessments, together with representative centrate upon his or her purpose and its associ authors. The Xs in the table show the primary impact of each intervention upon the task as- ated meaningfulness . In contrast, a form of goal Table 1 Selected Environmental Variables That Shape the Task Assessments Task Anessments Meaningfulne •• Variable. Impact Competence Choice Leadership Charismatic (House, 1977) x x Transformative (BenniS & Nanus , 1985) x x x Delegation (Leana, 1987) x Job De.ign x x (Hackman & Oldham, 1980) x Reward System. (Oed. 1975) x x 676 sessments, as interpreted from discussions in the tive styles and task assessments. Study I (Lee, cited works. Due to space limitations, this list is 1987) measured task assessments as part of an not intended to be comprehensive. Rather, it experimental laboratory investigation of the ef emphasizes research areas in which theorists fects of positive feedback upon intrinsic task mo have been relatively explicit about motivational tivation. In this study, self-report ratings of the assumptions and the psychological targets of in task assessments, summed across assessment terventions. More interventions will need to be 'dimensions, correlated very strongly with two added to this list as more explicit theory/ existing pencil-and-paper measures of intrinsic research becomes available that links them to motivation. Furthermore, path analysis con intrinsic task motivation. For example, a person firmed that the task assessments were mediating would expect task assessments also to be af variables in the causal relationship between fected by aspects of mentoring, changes in or positive feedback (the experimental manipula ganizational culture, team building, and orga tion) and ratings of intrinsic motivation on those nizational redesign. measures. Interpretive Intervention Strategies. The A more extensive test of the model. Study II present model also suggests the importance of (Tymon, 1988), involved a questionnaire survey interpretive interventions that more directly ad of 164 managers in three organizations, using dress how individuals construe environmental multiple-item measures of the task assessments events. The interpretive styles in the model are and interpretive styles. Factor analYSis of the viewed as developed habits rather than innate task assessment items demonstrated four sepa abilities. These habits are relatively isolated rate factors, corresponding to the task assess from day-to-day learning activities. Generaliz ments in the present model. Task assessments ing from Ellis's (1980) observation, individuals were strongly related to job satisfaction and typically are unaware of the role of such styles in stress, and they were modestly related to perfor shaping their interpretations of events, so that mance (as rated by superiors). Factor analysis of the styles themselves are not likely to be altered interpretive style items yielded three factors that when unfavorable outcomes are produced. This were clearly related to the styles in the model. explanation is analogous to Argyris's (1982) no Different subsets of styles strongly influenced the tion of single-loop learning. Consistent with Ar four task assessments. Furthermore, the task as gyris's thinking, however, researchers/clini sessments intervened in the relationship be tween interpretive styles and both job satisfac cians have demonstrated that such styles can be tion and job stress. changed by making the individual aware of as sumptions that are inherent in a style and by Implications for Research teaching individuals to consciously monitor This article has identified empowerment with those ongoing interpretations and their conse an emerging, nontraditional paradigm of man quences (e.g., Abramson et aI., 1978; Ellis, agement. We have argued that the motivational 1980). Thus, according to the model. "se1£ content of this paradigm involves the fostering of empowerment" programs are feasible solutions intrinsic task motivation among workers. The to help individuals identify and practice styles of core of the article has explicated a relatively attributing, evaluating, and envisioning, which comprehensive, cognitive model of intrinsic task would enhance their task assessments. motivation to describe the empowerment pro Empirical Support cess in individuals. Although it draws from pre vious models, this newer model has several Thus far, two studies have provided support comparative advantages. Those advantages, for the parts of the model that involve interpre- 677 and their implications for research, are elabo viding a sense ' of meaningfulness, as well as rated in this final section. competence and/or impact (e.g., Bass, 1985; Task Assessments . The model provides a Bennis & Nanus, 1985; House, 1977). Therefore, more comprehensive set of task assessments what are the consequences of omitting choice as than other models related to intrinsic motivation. a variable? Writers have criticized this literature As noted previously, such assessments are the for failing to discriminate between charismatic key cognitions (called critical psychological leaders who foster independence and ego de states by Hackman & Oldham, 1980), which are velopment in followers versus those who foster presumed to be the proximal cause of intrinsic dependency and the subordination of will (e .g., task motivation. Earlier models have been fo Musser , 1986; Rutan & Rice, 1981). This distinc cused on subsets of these variables that have tion centers on subordinate choice-allowing been asserted to be influenced by particular subordinates to be the origins of their actions , in types of empowering interventions (as detailed deCharms's (1968) terms, rather than the pawns. in the intervention section above). Thus , this Adding choice to existing models of charismatic model provides a common framewo rk which leadership would help theorists to discriminate makes it possible for researchers to compare the between empowering leaders and the dynam separate and combined effects of different types ics underlying the Jonestown tragedy. of interventions. It makes it possible , for exam Finally, the four task assessments also provide ple, for theorists to research the comparative or a framework for evaluating the probable effec additive effects of job redesign and charismatic tiveness of empowerment interventions that leadership on work groups . In the past , such have not been based upon explicit motivational phenomena have been treated as qualitatively assumptions. Conger and Kanungo (1988), for different in motivational effect. example , noted that empowerment has often Because the current model provides a more been equated with participative management comprehensive set of task assessments than pre techniques. In the present model, participative vious models , this also suggests research ques techniques are aimed primarily at providing tions about the consequences of variables that choice to subordinates. (See, for example, the classification of delegation in Table 1.) The have been omitted from those models. For ex ample, according to the Deci (1975) model, feed model implies that the amount of committed par back and reward systems will influence intrinsic ticipation (intrinsically motivated task behavior) motivation through perceptions of self-determi generated by these techniques will depend nation (choice) and competence. What effect heavily on the state of the other three task as does the omission of meaningfulness and im sessments, that is , whether the task objectives pact have on that model's accuracy? The current under discussion are seen as meaningful rather than trivial, whether individuals see themselves model implies that the predictions of Deci 's model will hold most reliably in contexts where as having sufficient competencies (including meaningfulness and impact are irrelevant (e .g . , technical expertise and participative problem in recreational games where no longer-term solving skills) to participate effectively, and task purpose is involved, that is, the dominant whether suggestions/decisions are followed up or supported by top management (impact). experimental setting for studies of the Deci Thus, the present model provides a concise way model). Intrinsic motivation will be predicted of explaining failures in power sharing or other less reliably in settings where task meaningful ness and impact are relevant and variable. participative management techniques (see As another example, theorists of charismatic! Sashkin, 1982). inspirational leadership have focused on pro- Global Assessments. The model provides a 678 systematic way of explaining the nature, source, features of the currenf model is its explication of and role of glol;xIl assessments in the context of the interpretive dynamics that underlie task as intrinsic task motivation . The model adds preci sessments. The present model adds interpretive sion to distinctions between the individual differ styles as a new class of causal variables that ence variables of locus of control. locus of cau shape individual empowerment at both the task sality, self-esteemlself-confidence, and overall level and the global level. The model asserts cathexis (alienation/involvement) by reconcep that interpretive styles will increase the amount tualizing them as generalized versions of the of explained variance in task assessments (and, four task assessments . They are viewed as in thus, intrinsic task motivation) when added to ductive generalizations over time-as cumula existing models of job redesign, charismatic tive abstractions from the individual's ongoing leadership , and so forth . task assessments . Moreover, they are viewed as The current model operationalizes interpre shaping a person's current task assessments de tive styles in terms of discrete cognitive habits ductively (Le., as applications of general beliefs t!lOt can be monitored by an individual. We to specific instances) when situational cues are have tried to identify variables that can be ambiguous. changed through cognitive behavior modifica The distinction between task assessments and tion techniques (Meichenbaum , 1977). Thus , one global assessments clarifies the difference be set of research questions raised by the model tween two different levels of empowerment. Al involves the effectiveness of such techniques in though most interventions in the literature have changing interpre tive styles and in producing been directed at empowering workers with re "downstream" effects on empowerment. spect to a given task, a few (e.g . , Simmons & Finally, although this article has developed Parsons, 1983) have focused upon more global the construct of interpretive styles in the context of intrinsic task motivation, it is clear that such or developmental changes in individuals. This amounts to adopting an individual's global styles also play a causal role regarding other (rather than task) assessments as the change phenomena . For example , the styles in the target. In such interventions, the objective is to present model also would influence extrinsic produce lasting increases in individual empow motivational dynamics and would thus appear erment that generalize across tasks . The issue of to be a promising addition to expectancy theory how to produce such changes has been rela models in general. helping to explain between tively unexamined in the organizational em individual variance in expectancies and out come valences. Likewise, the Tymon (1988) powerment literature. The model implies that one approach to global empowerment is study mentioned previously also found strong through the alignment of organizational pro relationships between interpretive styles and cesses and structures to consistently enhance in the affective phenomena of stress and job satis faction. Further research regarding the role that dividuals' task assessments. (Isolated effects of charismatic appeals, for example, seem un interpretive styles play regarding an individu likely to produce lasting or generalized in al's stress is underway. Also underway is an investigation of the effects of leaders' interpre creases in intrinsic task motivation.) Another ap tive styles on the intrinsic task motivation and proach is through interventions that provide un stress of subordinates. Thus, interpretive styles usually dramatic, memorable examples of high may prove to be an important aspect of leader task assessments and, thus, are more likely to ship behavior, providing a framework for cap shift a person's global assessments. This ap turing the ways in which leaders affect subordi proach is analogous to the one taken in such nates' interpretation of (and motivational/affec programs as Upward Bound. tive responses to) ongoing events. Interpretive Styles. One of the most innovative 679 References Abramson , L. Y. , Seligman, M. E. P ., & Teasdale, J. D. extrinsic dichotomy. fournal of Applied Psychology, 60, (1978) Learned helplessness in humans: Critique and re 455-458. formulation . Journal of Abnormal Psychology, 87, 19- 74. Ellis, A. (1977) Rational-emotive therapy: Research data that supports the clinical and personality hypotheses of RET Argyris, C. (1982) Reasoning , learning , and action . San and other modes of cognitive-behavior therapy. Counsel Francisco: Jossey-Bass. ing Psychologist, 7, 2-42. Bandura, A. (1977) Self-efficacy: Toward a unifying theory of Ellis, A. (1980) An overview of the clinical theory of rational behavioral change. Psychological Review, 84, 191-215. emotive therapy. In R. Grieger & J. Boyd (Eds.), Rational Bandura, A. (1986) Social foundations of thought and action : emotive therapy (pp. 1-31). New York: Van Nostrand A social-cognitive view. Englewood Cliffs , NJ: Prentice Reinhold. Ford, C. H. (1969) If you 're problem-oriented. you're in trou Hall. ble. Business Management , 35(5), 24-28. Bass, B. M. (1985) Leadership and performance beyond ex Garfield, C . (1984) Peak performance. Los Angeles: J. P. pectations. New York: Free Press. Tarcher. Bazerman, M. H. (1986) Judgment in managerial decision Glick. W. H. , Jenkins, G. D., Jr., & Gupta, N. (1986) Method making . New York: Wiley. versus substance: How strong are underlying relation Bennis, W., & Nanus , B. (1985) Leaders . New York: Harper & ships between job characteristics and attitudinal out Row . comes? Academy of Management fournal, 29, 441-464. Berlew, D. E. (1986) Managing human energy: Pushing ver Gregory, W. L. (1981) Expectancies for controllability. per sus pulling. In S. Srivastva (Ed. ), Executive power (pp. formance attributions , and behavior. In H. M. Lefcourt 33 - 50). San Francisco: Jossey-Bass. (Ed.), Research with the locus of control construct (Vol. I, pp. 67-124) . New York: Academic Press. Block, P. (1987) The empowered manager. San Francisco: Jossey-Bass. Hackman, J. R., & Oldham, G. R. (1980) Work redesign. Reading , MA: Addison-Wesley. Brickman. P. (1978) Is it real? In J. H. Harvey, W. Ickes, & Harrison , R. (1983) Strategies for a new age. Human Re R. F. Kidd (Eds.), New directions in attributional research source Management . 22, 209-235. (Vol. 2, pp. 5--34). Hillsdale, NJ: Erlbaum. Hoffman. C. (1978) Empowerment movements and mental Brief, A. B., & Aldag, R. J. (1977) The intrinsic-extrinsic di health: Locus of control and commitment to the United chotomy: Toward conceptual clarity. Academy of Man Farm Workers. Journal of Community Psychology, 6, 216- agement Review , 2, 496-500. Burke, W. (1986) Leadership as empowering others. In S. House. R. J. (1977) A 1976 theory of charismatic leadership. Srivastva (Ed. ), Executive power (pp. 51-77). San Francis In J. G . Hunt & L. L. Larson (Eds.), Leadership: The cutting co: Jossey-Bass. edge (pp. 189-207). Carbondale: Southern illinois Univer Bums, J. M. (1978) Leadership. New York: Harper & Row . sity Press. Burrell, G., & Morgan, G. (1979) Sociological paradigms and House. R. J. (1988) Power and personality in complex orga organizational analysis . Exeter, NH: Heinemann. nizations. In L. L. Cummings & B. M. Staw (Eds.), Re Conger, J. A. , & Kanungo , R. N. (1988) The empowerment search in organirotional behavior (Vol. 10, pp. 305-357). process: Integrating theory and practice . Academy of Greenwich, Cf: JAl Press. Management Review, 13, 471-482. Joe. V. C. (1971) Review of the internal-external control con Davis, T. R. V., & Luthans, F. (1980) A social learning ap struct as a personality variable. Psychological Reports, 28 , proach to organizational behavior. Academy of Manage 619-640. ment Review, 5, 281-290. Kammer, D. (1983) Depression, attrlbutional style, and fail Deal. T. E., & Kennedy, A. A. (1982) Corporate cultures. ure generalization. Cognitive Therapy and Research, 7, Reading , MA: Addison-Wesley. 413-424. deCharms, R. (1968) Personal causation: The internal affec Kanter, R. M. (1968) Commitment and social organization: A tive determinants of behavior. New York: Academic Press. study of commitment mechanisms in utopian communi Deci, E. L. (1975) Intrinsic motivation. New York: Plenum. ties. American Sociological Review, 33, 499-517. Kanter, R. M. (1983) The change masters. New York: Simon Deci, E. L., & Ryan, R. M. (1985) Intrinsic motivation and & Schuster. self-determination in human behavior. New York: Ple num. Kreitner. R., & Luthans, F. (1984 , Autumn) A social learning Dyer, L. , & Parker. D. F. (1975) Classifying outcomes in work approach to behavioral management: Radical behavior motivation research: An examination of the intrinsic- ists "mellowing out." Organiwtional Dynamics, 12, 47 -65. 680 Lawler. E. E .. m. (1973) Motivation in work organizations. Roberts . K. H .. & Glick. W. (1981) The job characteristics ap Monterey. CA: Brooks/Cole. proach to task design : A critical review. Journal of ApplJed Psychology. 66. 193 - 217. Leana. C. R. (1987) Power relinquishment versus power sharing: Theoretical clarification and empirical compari Rotter. I. B. (1966) Generalized expectancies for internal ver son of delegation and participation. Journal of Applied sus external control of reinforcement. Psychological Mono Psychology. 72. 228-233 . graphs 80 (Whole No. 609). Lee. Y. D. (1987) The enhancement of intrinsic motivation Rutan. I .. & Rice. C. (1981) The charismatic leader: Asset or through the mechanism of feedback. Unpublished doc liability. Psychotherapy: Theory. Research . and Practice. toral dissertation. University of Pittsburgh. 18. 487 - 492. Lefcourt. H. M. (1973) The function of the illusions of control Sashldn . M. (1982) A manager's guide to participative man and freedom . American Psychologist. 5. 417-425. agement . New York: American Management Association. Luthans. F .. & Kreitner. R. (1975) Organizational behavior Schein. E. H. (1985) Organizational culture and leadership: modification. Glenview. IL: Scott. Foresman. A dynamic view. San Francisco: Iossey-Bass. Martinko. M. I .. & Gardner. W. L. (1982) Learned helpless Shamir. B .. House. R. I .. & Arthur. M. B. (1989) The transfor ness: An alternative explanation for performance deficits. mational effects of charismatic leadership: A motivational Academy of Management Review. 7. 195 - 204 . theory. Working paper. Reginald Iones Center for Strate gic Management. The Wharton School of Management. May. R. (1969) Love and will . New York: Dell. Philadelphia. PA. Meichenbaum. D. (1977) Cognitive behavior modification : Simmons. C . H .. & Parsons. R. I. (1983) Developing internal An integrative approach. New York: Plenum . ity and perceived competence : The empowerment of ad Morris. K. T .. & Kanitz. H. M. (1975) Rational-emotive ther olescent girls. Adolescence. 18.917 - 922. apy. Boston: Houghton Mifflin. Sjoberg. L. G .. Olsson . & Salay. F. (1983) Cathectic orienta Musser. S. I. (1986. August) The determination of positive tion . goal setting and mood. Journal of Personality Assess and negative charismatic leadership. Paper presented at ment. 47. 307 - 313. the Annual Meeting of the Academy of Management . Chi Solomon. P .. & Patch. V. D. (1971) Handbook of psychiatry cago. (2nd ed. ). Los Angeles: Lange Medical Publications. Neilsen . E. (1986) Empowerment strategies: Balancing au Staw. B. M. (1976) Intrinsic and extrinsic motivation . Morris thority and responsibility. In S. Srivastva (Ed .). Executive town . N): General Learning Press. power (pp. 78-110) . San Francisco: Iossey-Bass. Tymon. W. G .. Ir. (1988) An empirical investigation of a cog O'Toole. I. (1985) Vanguard management. New York: Dou nitive model of empowerment. Unpublished doctoral dis bleday. sertation. Temple Un iversity. Philadelphia. Peale. N. V. (1954) The power of positive thinking. New York: Walton. R. E. (1985) From control to commitment in the work Prentice-Hall . pla ce. Harvard Business Review. 63(2) . 77 - 84. Peters. T. I .. & Austin. N. (1985) A passion for excellence . Wells. L. E .. & Marwell . G. (1976) Self-esteem : Its conceptu New York: Random House. alJzation and measurement. Beverly Hills. CA : Sage. White. R. W. (1959) Motivation reconsidered: The concept of Peters. T. I .. & Waterman. R. H. (1982) In search of excel competence. Psychological Review. 66. 297 - 333 . lence. New York: Harper & Row . Part of the material reported here was presented at the annual meeting of the Academy of Management. San Diego. 1985. The authors are indebted to Larry W . Boone. Amy Fried. Young Doo Lee. and Walter G. Tymon . Jr. . for comments on earlier drafts of this paper and in particular to Walter G . Tymon . Jr .. and Young Doo Lee regarding their substantive discus sions about the content of the paper. This report was prepared in conjunction with research conducted un der the Naval Postgraduate School Foundation Re search Program . Requests for reprints or further information should be sent to Professor Kenneth W. Thomas . Department of Administrative Sciences. Naval Postgraduate School. Monterey. CA 93943.
Journal
Academy of Management Review – Unpaywall
Published: Oct 1, 1990