ارزش پزشک در مقیاس عمل: ساخت و ساز و اعتبار سنجی اولیه

Measures of values typically appraise the construct globally, across life domains or relative to a broad life domain such as work. We conducted two studies to construct and initially validate an occupation- and context-specific values measure. Study 1, based on a sample of 192 medical students, describes the initial construction and item analysis of the physician values in practice scale (PVIPS), which produced a 15-factor model. Study 2 reports on a further analysis and refinement of the instrument with a national sample of 644 medical students. Results supported the basic psychometric properties of the PVIPS items. Exploratory factor analysis in Study 2 extracted six factors consistent with a theoretical model of values (Dawis, 1991) and accounting for 61% of the variance: Prestige, Service, Autonomy, Lifestyle, Management, and Scholarly Pursuits. The PVIPS shows promise as a values measure for medical students and physicians encountering career specialty and medical practice style decisions. Interested researchers may construct similar scales for other occupations to promote contextualized appraisals of values.

Psychometric scales have long been used to operationally define and appraise values in career counseling and research (Dawis, 1991 and Super, 1983). Measures of values reflect one of two types (Dawis, 1991). The first type of values measure appraises the construct generally across the spectrum of human experience. General, or global values measures include the Rokeach Value Survey (Rokeach, 1975) and the Study of Values (Allport et al., 1970 and Kopelman et al., 2003). The second type of values measure assesses values relative to a specific domain or context; typically the work, vocational, or occupational setting. Domain-specific values measures include the Work Values Inventory (Super, 1970), the Minnesota Importance Questionnaire (Weiss, Dawis, & Lofquist, 1981), and the Values Scale (Super & Nevill, 1985). These measures index values in terms of outcomes or goals individuals seek from work generally in their lives and do not assess values relative to a particular occupational context. Global and work-domain-specific values scales have proven useful for general life design, values clarification, and initial occupational decision making (Zytowski, 1994). Work values assessment in the form of psychometric scales has been criticized, however, for inadequately capturing and conveying the complexity of the construct (Krumboltz, Blando, Kim, & Reikowski, 1994). Such inadequacy may prove especially true when counselors or researchers apply extant work values measures, that appraise values broadly and abstractly, to situations in which individuals confront career specialty decision making. In the case of specialty choice, it would seem more appropriate and useful to consider values relative to the particular work or occupational domain in which those values would be practiced. Recognizing that existing work values measures lack such specificity, Krumboltz et al. (1994) argued for constructing context-specific values scales and suggested that “rather than asking an individual to make value choices outside the context of a particular occupation, assessment instruments might be devised that embed work values in a meaningful context” (p. 57). Further support for constructing context-specific values measures comes from parallel theoretical and empirical work in the arena of personality trait measurement. In this regard, Wright and Mischel (1987) proposed in the theory of conditional dispositions that manifest personality traits are situation-dependent, such that the same person behaves differently under different circumstances (e.g., introverted at work and extroverted at home). Personality measurement research has therefore advanced a line of inquiry that has provided evidence for the incremental validity of using context-specific items in personality appraisal (Bing et al., 2004 and Schmit et al., 1995). Following this line of reasoning and recognizing the similarly situation-dependent nature of values (Ball-Rokeach & Loges, 1996), we innovated a third type of values measure by constructing a scale intended to be both specific to an occupation and distinctive to the context of activities in which persons working in that occupation engage. In the present investigation, physician represented the occupation of interest and medical practice constituted the context relative to which values were assessed. Other occupations certainly could have been chosen, however, we were specifically interested in developing a psychometric tool that could be used to promote medical career development and specialty choice.

The present findings provide preliminary support for the adequacy of the PVIPS’s psychometric properties. A mean Cronbach’s α coefficient of .83 with all αs > .75 for its six scales indicated moderate to high levels of internal consistency. While the α levels indicated item homogeneity among the scales, the factor-analytic data of Study 2 provided direct evidence to support that the PVIPS indeed measures significant value constructs consistent with those identified in the work values literature ( Dawis, 1991, Super, 1970 and Super and Nevill, 1985). The six extracted factors corresponded significantly in item content and seemed parallel conceptually to constructs in the model used to guide development of the PVIPS, thereby covering the range of core values ( Dawis, 1991 and Weiss et al., 1981). Further evidence for construct validity of the PVIPS emerged from results of the inter-scale correlational analyses. Inspection of the correlation matrix presented in Table 2 reveals differences in magnitude and discernible patterns of scale interrelationship. The moderate-to-low strength of the statistically significant intercorrelations suggests that the scales overlap yet seem unique enough to measure distinct value constructs. Positive and statistically significant relationships were found between scales in all but two cases. In one case, the non-significant relationship between Lifestyle and Management suggests that wanting a controllable lifestyle has no bearing on the extent to which one wants to supervise or direct the work of others. A desire for such control may, however, be viewed as incompatible with a career path as a physician academic or researcher as suggested by the significant, albeit weak, inverse correlation between Lifestyle and Scholarly Pursuits (−.10). One interpretation of the many significant positive intercorrelations is that medical students may believe it important to “have it all.” Perhaps the present sample’s endorsement of most of the values reflects their self-selection and socialization into an occupation among the most highly regarded, challenging, service-oriented, and rewarding. Relative to patterns of scale interrelationship, it seems reasonable to consider and interpret the correlational data based on the obtained strength of associations. For example, Prestige correlated most strongly with Management (.50) followed by Autonomy (.36) and Scholarly Pursuits (.34). These findings suggest that individuals highly desirous of recognition and esteem may most likely seek and obtain such status through supervising others, practicing medicine creatively and in their own way, and by engaging in academic, scientific, and scholarly activities. Weaker positive relationships between Prestige and Lifestyle (.17) and Service (.13) seem to indicate that wanting to be held in high regard by others relates only negligibly to a desire for minimizing demands made on one’s time or caring for others without regard for rewards. In the same vein, Service related most highly to Autonomy (.55), suggesting that wanting to practice medicine in a way that helps others without concern for reward involves doing so creatively and in one’s own way. Having few time or other demands, supervising others, and being a scholar/researcher/academician seem much less important relative to Service. The moderate relationship between Autonomy and Lifestyle (.48) points to freedom and independence concomitant to practicing with a sense of stability in terms of schedule and time demands. Management and Scholarly Pursuits related significantly (.39) suggesting that administrating and supervising others’ work may extend to the domain of scientific and academic work. Future research should examine whether or not comparable patterns of scale interrelationship would emerge with different samples of medical students. This might include drawing larger numbers of 4th-year students or residents who have already selected a particular specialty area. Analyses with such groups may yield more differentiated relationships with more distinct patterns of high and low scores and, perhaps, even inverse relationships where they might be theoretically expected (e.g., Prestige and Service, Management and Lifestyle). Analysis of between-groups differences in scores on the six PVIPS factors based on self-reported specialty choice status and preference yielded modest additional evidence for construct validity of the measure. Career-decided students interested in non-primary care oriented specialties scored higher on Prestige and Scholarly Pursuits than did students with expressed interests in primary-care oriented medical practice. These findings proved consistent with research indicating that such values are more strongly endorsed by student aspirants to and physicians working in highly specialized and technical medicine than they are by those inclined to the more people- and service-oriented primary care areas of practice (DeWitt et al., 1998 and Hojat et al., 1998). That students undecided about their specialty choices also scored significantly higher on Scholarly Pursuits than did the primary-care group suggests that they may have an emerging preference for specialized medical practice. Further research with the PVIPS should be conducted with the aim of adding to the rather extensive literature that deals with between-groups differences based on medical specialty choice (Borges & Savickas, 2002). On balance, the PVIPS appears able to measure a significant, relatively broad, and inclusive domain of values pertinent to individuals who have specified and begun to implement physician as an occupational choice. The present investigation effectively produced the first edition of the instrument, which shows promise and appears suitable for further research and development. Future research should address the limitations of the present study, which included non-random and non-representative sampling. Studies that examine the concurrent validity of the measure also represent a logical and needed next step. Administering the PVIPS to practicing physician samples could advance the applicability of the measure by creating specialty group norms. Writing additional items of parallel content toward increasing reliability, especially in the cases of the four-item Lifestyle and three-item Scholarly Pursuits scales should also be pursued. The present research supports the feasibility of developing a physician career values scale. It may prompt interested researchers to develop similar scales for other occupations such as counseling, psychology, law, engineering, and business to provide more contextualized values appraisals.