工作满意度指标及其相关因素外文翻译（可编辑） .doc

工作满意度指标及其相关因素外文翻译 外文翻译Job Satisfaction Indicators and Their Correlates Material Source:American Behavioral scientistAuthor: Stanley E. Seashore Conceptions of job satisfaction until very recently have been largely psychological and individualistic in orientation. Empirical studies have been confined to local situations or special populations with interpretive purposes reflecting the values of employed individuals or of their managers. However, if job satisfaction measures are to be useful in monitoring the quality of employment on a societal scale, it will be necessary to enlarge the perspective, to invoke some societal and political values, and to begin to treat job satisfaction in the context of a larger array of associated variablesThe measurement of job satisfaction as a social indicator may have three roles: 1 to represent a valued product of society-a component of the psychological GNP; 2 to provide a monitoring and diagnostic aid for early warning of societal dislocations, policy or program failure, and slowly developing societal changes; and 3 to provide a significant component in the theories and models to be used in the formulation of social policy and programs. Opinions differ on how prominent and how effective job satisfaction measures will be in these three roles. The utility of job satisfaction measures rests on the development of multiple measurement methods that are standardized, suitable for wide use, and capable of detecting population differences and population changes. In addition, the utility rests upon these measures having an agreed conceptual and “real world” reference as well as a known matrix of causal and consequential relationships to other significant variables. Both requirements must be met before convincing proof can be advanced as to the practical utility of job satisfaction measures for anticipating, understanding, and influencing future outcomes of present societal conditionsThese themes provide the structure for this paper. In the next section, we give an overview of the state of the art in the measurement of job satisfaction. The section following that provides an approach to organizing, or modeling, the correlates of job satisfaction. The final section suggests some priorities for further research and development. JOB SATISFACTION INDICATORS This section summarizes considerations that bear upon the choice of approaches and operational methods for measuring job satisfaction. We shall limit the discussion to approaches that rest upon direct inquiry through interview or questionnaire methods to produce data that can be aggregated to provide job satisfaction indicators for variously defined populations. We exclude from discussion: 1 indirect approaches that draw inferences about job satisfaction from presumed causal or consequential phenomena; 2 approaches that are primarily individualistic and diagnostic and, therefore, not usually applicable for generating population indicators; and 3 approaches that have utility primarily for empirical and theoretical discovery rather than for population description purposes. We first review the commonly used forms of primary data, then some commonly used derivative job satisfaction indicators. A scheme is presented to guide the evaluation of these several indicators. These are applied to draw implications for preferred future methods. Throughout the paper, except where noted, we will use the term “job satisfaction” inclusively to refer also to dissatisfaction without intended prejudice whether satisfaction and dissatisfaction are best treated as polar opposites or as two conceptually different variables.PRIMARYDATA By primary data we mean the “raw” responses given by individual respondents to verbal questions or comparable stimuli. There appears to be a fixed roster of basic forms of primary data, even though innumerable variations on these are known. Two kinds of primary data are distinguished: facet-free and facet-specific. Facet-free primary data are obtained when the respondent is asked to indicate his global satisfaction with his job and job environment without specifying in advance the facets to be considered or how they are to be combined. In effect, each respondent provides a net response derived from his own set of facets, weighted or otherwise combined in his own unique fashion, with unstated and unique assumptions not only about the context for evaluation, but also about his own “fit” to the job and its environment, and with the environmental “reality” defined by his own perceptions and cognitions. Normative, cognitive, and unconscious elements in the evaluation are invited. The stimulus questions are usually phrased or nonverbally displayed with an intent to impose the fewest possible constraints upon his perceptual, cognitive, and evaluative processes. Several complementary stimuli may be used to diversify the unavoidable constraints. Facet-specific primary data are obtained when the respondent is asked to represent his satisfaction with respect to some specified facet of his job or job environment. Since the facet specification is never exhaustive or definitive, the difference between a facet-free and a facet-specific inquiry is only one of degree. For example, the query “How satisfied are you with your pay?” elicits a net response that includes consideration of unspecified subfacets amount of pay, certainty of pay, rate of increase, adequacy to need, and so forth, unspecified “reality” last weeks pay, pay after deductions, pay confidently expected next year, and the like, and unknown perceptual, cognitive, and evaluative processes. Nevertheless, facet-specific methods allow the inquirer some control over the range of facets to be included in his data, an added degree of comparability among different respondents, and closer and more confident linkage between the response obtained and the “reality” of the job environment or of the person under investigation. Facet-specific queries, thus, vary in their specificity. In addition, they take the following forms: a direct report of degree of satisfaction with facet satisfaction; b amount or degree of facet provided by job is now; c amount or degree of facet respondent would like to have would like; d amount or degree of facet respondent should be provided should be; e importance of facet to respondent importance. The forms of response exist in great variety, including simple check-list or “yes-no” responses, rank ordering, scalar responses e.g., Likert scales, “faces,” and the like, and the more complex forms such as “self-anchoring” scales. While these alternatives invite useful discussion about their relative reliability, efficiency, simplicity, item utility, and conceptual assumptions, such issues will not be raised here. Each alternative provides primary data permitting aggregation for population comparison or social indicator purposes.DERIVED DATA In the case of primary data that represent the direct or implied expression of job satisfaction, social indicators may be derived by a simple aggregation of primary data for individuals and then an aggregation of individual data for the population. This is often done, for example, with respect to multi-item, facet-free primary data, and with primary data of types a and b above. However, more complex forms of derivative indexes are commonly preferred for various reasons. Procedures for deriving indexes from primary data include: 1 differential weighting of items; 2 clustering of items into factors or dimensions on conceptual or empirical grounds; 3 converting primary data to derived discrepancy scores on theoretical, conceptual, or empirical grounds before aggregation; 4 retaining individual facet item data for differential uses in interpretation or analysis; 5 removing some uncontrolled response variance before aggregation ; and 6 adjusting primary data for known or presumed bias before aggregation. Any of these procedures may be employed singly or in combination with others. The last three procedures are relatively trivial or at least noncontroversial at the present time; the first three are topics of current inquiry and dispute. CORRELATES OF JOB SATISFACTION This section reviews what is known and what should become known with respect to the correlates of job satisfaction. The range of known correlatives is displayed in a way that will aid the assessment of the potential role of job satisfaction as one indicator, among others, of the quality of employment. Some examples of reported empirical correlations will be given for illustrative purposes, but we do not attempt to review and catalog all published reports bearing on the matter, nor to provide evaluation of the various empirical generalizations that have been advanced. We shall ignore for the present the diversity of concept and measurement of job satisfaction treated in the preceding section.SOCIAL INDICATORS AND INTERPRETATIVE MODELS The meaning of any social indicator is found in its assigned role in some conception of how the society “works.” Thus, a measured change in some indicator-infant mortality rate, for example-is uninterpretable apart from some known or assumed dynamic structure of sequential changes that relates the observed change to causes, consequences, and moderating conceptual factors. Ideally, one should have an empirically validated theory, broad in scope, embracing multiple causes and consequences, capable of accommodating additional variables i.e., an open system, and one that treats changes over time i.e., a dynamic theory. Such an interpretive model would permit the evaluation of a change in some social indicator in several useful ways, most importantly in estimating future implications of the observed change and in identifying possible societal actions to forestall or counteract undesirable consequences. With respect to job satisfaction, there does not exist any such comprehensive theoretical model. However, there are micromodels treating limited segments of such a more comprehensive model, and there are known empirical correlations that help to identify classes of variables that must be taken into account and which can guide future work into profitable directions. One example of such a micro-model specifies that more challenging jobs i.e., those with more autonomy, allowing greater use of valued skills, and so on are associated with higher job satisfaction. In a dynamic form with causal specification, it is asserted that “enrichment” with respect to the degree of challenge leads to an increase in job satisfaction. There is ample correlational and experimental evidence that such an association can exist and can be quite strong Lawler, 1969; W. E. Upjohn Institute for Employment Research, 1973: 188-201; but rather little is known about the contextual conditions within which the association holds Hulin and Blood, 1968 and about variables that moderate the strength of the association. The generalization stands as a valid and useful one even though parts of the relevant correlational matrix remain unexplored. Other available micro-models treat job satisfaction in a causal rather than a consequential role. An example is the formulation that occupations that are relatively high in extrinsic job satisfaction will induce relatively high rates of premature death from chronic heart diseases, while occupations relatively high in intrinsic job satisfaction will induce lower death rates. This proposition has been supported in only two correlational tests but with impressively large correlation coefficients House,1972. Two points are illustrated by this example: 1 job satisfaction cannot in all circumstances be treated as a unidimensional construct; and 2 relationships that are trivial and unreliable at the individual level may be highly significant and interpretable when aggregated in this case, aggregated to the occupational level.MACRO-ENVIRONMENTAL FACTORS Although relatively little programmatic inquiry has been made into the role of economic, political, cultural, and similar broad factors as they affect job satisfaction, evidence suggests that this class of variables is indeed relevant to job satisfaction. For example,Hulin and Blood 1968 and also Kendall 1963 found that characteristics of the communities in which workers reside need to be taken into account to understand job satisfaction. Form 1973, comparing auto assembly plant workers in four countries, shows that there are differences in work-related values, motives, and satisfactions associated with degree of industrialization, while other relational regularities appear to be impervious to culture and context. There are speculations, but no adequate evidence, that fluctuations in unemployment rate and general public optimism about future economic conditions impact on job satisfaction.OCCUPATIONAL CHARACTERISTICS That job satisfaction is related to general characteristics of occupations not to be confused with properties of jobs and the occupational structure has been consistently demonstrated from the earliest comparative study of Hoppock 1935 to the more recent studies such as those of Quinn et al. 1973. Numerous studies show significant relationships between job satisfaction and such properties of occupations as status, prestige, power, and control, among others. However, because of defects in study design, not much is known about the degree to which the various occupational characteristics contribute independently to job satisfaction.ORGANIZATIONAL ENVIRONMENT This domain of causal variables is extensively represented in the theoretical and empirical literature. Variables which have shown evidence as significant organizational antecedents to job satisfaction include structural variables such as size and “shape” e.g., Carzo and Yanouzas, 1969, complexity, centralization, and formalization e.g., George and Bishop, 1971; process variables such as prevailing decision-making and conflict management styles, team collaboration and role conflict; and such encompassing variables as “organizational climate”Litwin and Stringer, 1968.THE JOB AND JOB ENVIRONMENT By far the major part of the job satisfaction research has been concerned with the proposition that an individuals job satisfaction is in substantial part a direct product of the objective characteristics of his job and its immediately relevant environment. Many hundreds of reports assert or imply such a proposition and present empirical data bearing upon it. These data are diverse in quality and scope and offer a somewhat bewildering array of correlations and choice of job characteristics for treatment, but they without doubt confirm the general proposition. Smith et al. 1969 report that in a number of replications in different settings, the amount of pay associated with a job correlates positively with degree o