Person-Organization Fit and the Attitude of Medical Staff: Professionalism and Work Attitude in a Chinese Hospital

Purposes: Physicians and administrative staffs have different obligations; physicians can have different philosophy and value systems from that of administrative staffs. Therefore, the purposes of this study are: (1) to determine the ways in which the congruence or divergence of values impact upon employee attitudes; and (2) to investigate the moderating role of professionalism in the physicians’ work values. Method: Total of 86 questionnaires was gathered from physicians and staff. This study used previous design developed by Chiu in Taiwan. Polynomial regression was used for multi-dimension value congruence effects. Result: The fit between actual and desired values does not necessarily lead to higher levels of job satisfaction and organizational commitment. The moderating effect of professionalism on fit and attitude only applied to ‘personal respect’ and job satisfaction, plus fit on ‘profit orientation’ and organizational commitment. Conclusion: Any future research in this area should attempt to include other possible intervening variables in the relationship of fit to values and attitude. It is proposed that the hospital should adopt an appropriate competitive strategy that might enable it to achieve a fair compensation system and reasonable profit, and thereby ensure that it can retain the services of its attending physicians.


Introduction
The economy of China has been transformed from a planned economy to an open market system as a result of the country's recent economic reform. In the past, employees of state-owned enterprises (SOEs) were always able to rely upon the so-called 'iron rice bowl', providing them with life-time job security and associated benefits, such as health insurance, retirement plans and accommodation [1]. However, the Communist Party has historically emphasized equalitarianism among all workers and the social status of a physician can be modified in this circumstance. Therefore, it is important to find out how professional staffs such as physicians respond to and perceive the way their organizations operate, compared to administrative staff, under this context.
In the European and North American countries, the essence of professionalism lies in the contract between medicine and society [2]. The contract basically defines the role expectations for physicians given by the public. Physicians take on the heavy obligations of a profession, and they are rewarded with high social status and income [3]. Their education and training programs are focused on expertise and humanitarianism, in order to achieve maximum patient welfare. On the other hand, administrative staffs pursue the goal of maximum profit for the company's key stockholders Bosk [4,] presumes from the perspective of social control theory that surgeons are socialized and taught their identity from role models. The education of physicians to adopt medical practice models would be definitely different from the administrative staff's training process. Assumption has been made that the physicians would hold a different philosophy and values, in contrast to administrative staff, since they both undertake different obligations.
Person-organization fit can be described as the level of compatibility that is effectively established between an individual and an organization when the values, preferences and goals of both parties are appropriately matched. Though person-organization fit researches have been proved in most European and North American countries to enhance better employee attitudes, the idea has not been much applied to socialist countries.
Person-organization (P-O) fit, an area of research which is already well established and which continues to grow, covers three main aspects: antecedent and outcome variables of P-O fit, the content of fit itself, and the fit index.
The related P-O fit studies mostly focus on a discussion of the antecedent and outcome variables. The antecedent variables include human resource management practices [5][6][7], job and career choices [8,9], performance [10], and socialization tactics [11], whilst the outcome variables include job satisfaction, organizational commitment, organizational citizenship behavior, career success, tenure and turnover rates [5,6,[12][13][14][15][16][17][18][19][20][21]. Although the results of the prior studies generally confirm that better P-O fit leads to enhanced organizational commitment and job satisfaction, as well as lower turnover rates, very little research has been undertaken examining P-O fit as a moderator of the relationship between the study variables [1].
Furthermore, the content of fit focuses more on the entry and selection process, such as the 'attraction-selection-attrition' (ASA) three-construct values and job satisfaction as well organizational commitment in a Chinese context. The hypotheses based on the model of organizational commitment model summarized by Mathieu and Zajac [30] and job satisfaction model summarized by Agho et al. [31].
Organizational commitment emphasizes effective linkage [30], i.e. links to 'personal respect' values which are defined as people-oriented and aligned to better physical environments, human resource practices, and interactions with colleagues, and job satisfaction with emphasis on friendly supervisor-subordinate relationships. When employee expectations match how the respondents perceive the organization, the fit is reached and it ultimately enhances positive employee attitudes. Person-organization fit in 'personal respect' in a given level could advance job satisfaction and organizational commitment [27]. Disagreement in the match, regardless of whether it is deficiency (actual less than expected) or excess (actual greater than expected), would result in worse attitudes.
The perception of 'profit orientation' in the pursuit of efficiency and profits is less likely to encourage pleasant and harmonious atmosphere amongst employees. If agreement is reached, it is presumed that fit on 'profit orientation' would result in better job satisfaction and organizational commitment. However, while organizations may demand efficiency, revenue growth, the pursuit of profits, and so on, any circumstances beyond the employees' expectations may lead to a subsequent reversal in their attitudes, if fit is not reached. Agreement in this construct would lead to better attitudes, but misfit would not necessary lead to lower morale.
When a hospital is striving to become a 'leading brand' in the health market, often it may also be accompanied by a requirement for substantially higher standards of performance from its staff members. Nevertheless, employees can also benefit from the prestige of being associated with a leading hospital, thereby fulfilling their sense of self-esteem and potentially enhancing their level of organizational commitment. The organization may focus on the uniqueness of the hospital and the desire to achieve a leading position. As a result, where better fit occurs, employees may accrue greater benefits from their hospital; it is therefore assumed that the fit will have a positive association with job satisfaction and organizational commitment.
Hypothesis 1a: The fit on 'personal respect' has a positive association with job satisfaction and organizational commitment.
Hypothesis 1b: The fit on 'profit orientation' has a positive association with job satisfaction and organizational commitment.
Hypothesis 1c: The fit on 'pride and security' has a positive association with job satisfaction and organizational commitment.

Moderators
The initial predictions regarding the interaction effects on satisfaction and commitment were based on theoretical arguments in the commitment and satisfaction literature. Among antecedents to organizational commitment, situational characteristics account for more variance than do personal characteristics [30][31][32][33]. The combined effect of situational characteristic and personal characteristics based on study of Meyer et al. [27], addresses more than the impact of a single variable. Many of the prior studies have reported a positive relationship between P-O fit, job satisfaction and organizational commitment [34], with little evidence to suggest any moderating role on P-O fit and job satisfaction by professionalism. As compared to administrative staff working within hospitals, it is less likely that physicians will identify with the values of their organization. The prestige of medical theory of Schneider [22], and an emphasis on the process of socialization within organizations which enables employees to learn and adapt to the norms of the organization [11,23]. In addition, the appropriate content describing the shared values of an organization [17,24].
Likewise, there has been general advancement of the use of the fit index to measure congruence. Profile similarity indices (PSI) had been widely used prior to the introduction of polynomial regressions, which reflect multi-dimensional congruence [25]. Research into fit interpreted by polynomial regressions is superior for its lack of ambiguity and its actual and expected information content; however, the coefficients produced by polynomial regressions are quite difficult to interpret [25]. Evidently, although the prior research has succeeded in covering many different aspects from diverse perspectives, there is, nevertheless, a distinct lack of research into the ways in which fit theory applies to specific industries within specific cultural contexts, such as the healthcare industry in China.
The culture in China is quite different from western countries. It is still uncertain that person-organization fit could enhance better employee attitudes in the same way that interactional psychology has been utilized in most western countries to enhance job satisfaction and organizational commitment through reaching better agreement on values. It is important to investigate how person-organization fit is associated with employee attitudes in a given hospital in China, and further explore the moderating role of professionals in relation to P-O fit and attitude. Therefore, the purposes of this study are: (1) to determine the ways in which the congruence or divergence of values impacts upon employee attitudes; and (2) to investigate the moderating role of professionalism. Professional status is used in this study as a proxy for professional identify. This result can ultimately provide advice to the shareholder of this market.

Hypotheses development
Although organizational value is a way of learning how to adapt to one's external environment and gain internal conformity [26], the circumstances and problems faced by every organization are not entirely the same, resulting in the development of very different values. There are many different interpretations on organizational culture with different tools demonstrating various ways of approaching the concept. Most of the studies of person-organization fit have treated value as a onedimensional value construct, but it is not sufficiently comprehensive to illustrate an organization with a singular value, especially since latent constructs of most instruments already exist, such as O'Reilly and Meyer [17,27] modified from Manhardt's work value instrument. This study utilized the 'value questionnaire' proposed by Enz [28], which provides a comprehensive description of basic principles in how the organization handles daily operations, along with the modifications undertaken by Chiu et al. [29] which further identify three factors of major importance; (1) 'personal respect' (the extent to which an individual perceives the level of respect and autonomy provided by the hospital authority); (2) 'profit orientation' (the extent to which an individual perceives the focus of the hospital as being on the pursuit of profit and performance); and (3) the individual's feeling of 'pride and security' (the extent to which an individual perceives the intention of the hospital being to achieve a leading position within the healthcare market). Whilst O'Reilly's [17] framework and scales are widely used in fit research, it was considered not sufficiently specific to this culture and the organization situation. However, Chiu's framework is more suitable for this Chinese sample; hence it was modified according to language specification, and produced meaningful results in Taiwan [29].
There is little literature investigating specific relationships between professionalism gives physicians rewards in higher levels of autonomy, higher reputation, higher income, and the ease in establishing solo practice, while administrative staff emphasize control, productivity, and close supervision to meet the corporate mission [4]. Heterogeneity of values among professionals increases the complexity of human resource management in health care settings. Moreover, physicians are generally less committed and satisfied than non-physicians [29]. Given a high level of P-O fit, physicians will be less likely to be committed or satisfied than non-physicians. Moderator effects are complicated to justify and little justification is available. In order to control for the possible effects of environmental and job characteristics, this study uses professional status as a proxy for professional identity [35] in the relationship between P-O fit and job satisfaction and organizational commitment. Thus this study proposes: Hypothesis 2a: Professional status moderates the relationship between P-O fit and job satisfaction.
Hypothesis 2b: Professional status moderates the relationship between P-O fit and organizational commitment.
The conceptual framework is listed in Figure 1.

Method Sample
The sample hospital, which was located in the north eastern metropolitan area of Shanghai, was founded in 1958 as the very first comprehensive municipal teaching hospital in China, and was in 1994 subsequently appointed a 'Class A, Grade Three' hospital by the Ministry of Public Health of the People's Republic of China (PRC) [36]. The hospital was equipped with 1,318 beds and had more than 3,000 employees. A system of dual authority existed within the sample hospital, with routine hospital administration and daily operation being undertaken by a Superintendent (who was also a physician), and all other matters (mainly involving the promotion of civic values) being undertaken by a Secretary appointed by the Municipal Communist Party Committee.

Procedures
This study issued a survey questionnaire in simplified Chinese during October 2004 and collected 86 completed questionnaires (all of those responding were selected by the vice secretary of the hospital). The composition of the respondents included 42 males (49%), with physicians accounting for 58 (67%) of the respondents, and managers accounting for a further 17 (20%) of the respondents. The average age of the respondents was 35.8 years, and the average organizational tenure within the sample hospital was 13.1 years.

Measures
All of the multi-item scales were measured on a six-point Likert's scale (where 1=strongly disagree/dissatisfied, and 6=strongly agree/ satisfied), with all of the materials being presented in the native language.

Organizational Values
Organizational values were defined in two parts, the 'actual' value (e.g. the level of importance which the employees felt that the hospital actually had) and the 'ideal' value (e.g. the level of importance which the employees expected the hospital should have). The values were measured using 22 items adapted from the 'organizational value congruence scale' [28], with two additional items being included as the means of measuring the level of professional autonomy.
Validation of these items and constructs has been undertaken by various scholars, with a number of adjustments having subsequently been made to the wording of the original items. Two separate answer sheets were designed to collect the responses concerning the perceived level of importance of the items relating to how their hospital 'should be' and how it 'actually was'. All of the responses were measured on a six-point Likert's scale in order to reflect the degree of congruence between the 'ideal' and 'expected' organizational values [12]. In contrast to the adoption of only a single overall construct in Enz (1986), this study used a three-factor construct comprised of 'personal respect', 'profit orientation' and the 'pride and security', in accordance with the constructs outlined in Chiu et al. [29].
A confirmatory factor analysis (CFA) was conducted using LISREL 8.72 to test the three-dimensional structure of the organizational values. The composition of the items was validated under the following three criteria: (1) items with factor loadings greater than 0.5; (2) the Cronbach's alphas for each factor, including a reverse recheck if items were deleted; and (3) key indicators such as Chi-square, Root mean square error or approximation (RMSEA) and comparative fit index (CFI) to ensure the fitness of the model. After removing two items from factor 1, and a further two items from factor 2, it was found that a threefactor model fitted the data quite well (χ 2 =57.73, df=84; RMSEA=0.00; CFI=1.00; GFI=0.96; PNFI=0.86), and very good construct validity.
The 'personal respect' factor included nine items, a sample of these items being 'the willingness to make improvements based upon advice from professionals'. The 'profit orientation' factor was comprised of four items, with a sample item being 'the ability to ensure the survival of the hospital'. Finally, the 'pride and security' factor included seven items, with a sample of these items being 'the desire to be a leading hospital'. The respective Cronbach's alphas for the three factors relating to the 'actual' organizational value were 0.95, 0.91 and 0.85. The respective Cronbach's alphas for the three factors relating to the 'desired' organizational value were 0.90, 0.82 and 0.80. The factor and related items contained within the final model are listed in Table 1.

Job satisfaction
Job satisfaction, which was defined in this study as the overall level of satisfaction, was measured by 20 items [37], with all questions being validated and rephrased by experts from the relevant fields. The responses were measured on a six-point scale, where 1=very dissatisfied and 6=very satisfied. An example of the sample items was 'the chance to do difficult things'. The Cronbach's alpha for job satisfaction was 0.92.

Organizational commitment
Organizational commitment was measured in this study by means of 15 items modified from the 'organizational commitment questionnaire' of Porter et al. [38]. The responses were again measured

Demographic and control variables
The survey provided four demographic attributes for using as controls in this study, with 'gender' taking the value of 1 for male (otherwise 0); 'age' and 'organizational tenure' being measured in years; 'managerial position' assuming the value of 1 for a manager (otherwise 0); and 'professional status' taking the value of 1 for a physician and 0 for an administrator [39].

Analytical strategy
Hierarchical polynomial regression procedures [1,[40][41][42] were used to determine the extent to which value congruence and the possible moderating effects of professional status were related to the outcome variables. The midpoint of each item was subtracted to centralize the scale in order to alleviate the problems of multi-collinearity associated with curvilinear and interaction terms [43]. The outcome variables were then regressed in accordance with the following steps.
The first step involved placing the 'actual' and 'desired' values into the regression model, along with the control variables such as age, professional status, managerial position and gender. In the second step the 'actual value squared' and the 'desired value squared' were entered into the regression model, along with the cross-product of the 'actual value' multiplied by the 'desired value'. The third step involved the inclusion of the hypothesized interaction terms; if any significant R 2 had changed in the second step, we presumed higher-order value congruence and then included (or omitted) the squared term in the third step. Although the equations yielded by the regression are likely to contain coefficients on curvilinear and interactive terms that are difficult to interpret, interpretation is also facilitated by using the regression coefficients to plot the surfaces which relate each pair of entities to the outcome [25].

Results
The respondents reported rather lower scores on actual value items (ranging from 3.3-4.6), and higher scores on desired value items (ranging from 4.8-5.4). As the figures in Table 1 reveal, based upon evaluation by the respondents within the sample hospital, 'profit orientation' and 'the ability to ensure the survival of the hospital' were regarded as the two actual values of primary importance. For employees, the major priorities in terms of desired values were the 'willingness to make improvements based upon advice from professionals' and the 'honesty and integrity of employees'. Three separate actual values can be seen to have positive and significant associations with organizational commitment and job satisfaction (

P-O fit analyses
In order to test the hypothesis, hierarchical moderated regressions were carried out, with the first stage involving the inclusion of age, gender, professionals and positions as the control variables, as well as the actual and desired value variables. The second stage involved the inclusion of the squared term of the values, whilst the third stage involved the inclusion of the hypothesized interaction terms. The final model was determined by the significance of the changes in R 2 . The 'professionals' variable was dummy-coded, with administrative staff being set as the reference group. The squared order and the interaction effect, which followed the procedures proposed by Cohen and Cohen.

Job satisfaction
The fit on three dimensions of value to lead to better attitudes; rather they showed that the actual value of 'personal respect', 'profit orientation', and 'pride and security' contributed significantly to the prediction of job satisfaction (Table 3a). Three-dimensional graphs of the relationships are provided here to assist explanation of the result (Figure 1). The surface plot of desired and actual value on 'personal respect' demonstrates that the fit on 'personal respect' did not predict job satisfaction. Fit on value 'profit orientation' and 'pride and security' were as well not proved to enhance job satisfaction. Moderator effects were only proved in the relationship between fit on the value 'personal respect' and job satisfaction (Figure 2). Moderator effect showed positive interaction among fit on value 'personal respect' and job satisfaction among physicians, but negative interaction among administration staff ( Figure 3A and 3B).

Organizational commitment
Organizational commitment did not result from better fit, and showed that the actual value of 'personal respect', 'profit orientation', and 'pride and security' as well as the desired value of 'personal respect' and 'pride and security' contributed significantly to the prediction of organizational commitment (Table 3b). The surface plot of desired and actual value showed little prediction in relation to organizational commitment ( Figure 2C-2E). The only moderated effects were proved in the relationship between fit on 'profit orientation' and organizational commitment. The surface plot for administrative staff and physicians was illustrated in Figure 3B and 3C.
This result shows that hypotheses 1a, 1b and 1c are not supported; that is, agreement on values did not enhance the satisfaction or commitment of the respondents. Hypotheses 2a and 2b are partially supported.

Discussion
Most P-O fit literature agreed that promoting P-O fit could advance commitment among physicians [44,45], but this study revealed a different outcome.
The findings of this study revealed that the fit between actual and desired values not necessarily leads to higher levels of job satisfaction and organizational commitment. Possible explanations can be: according to our interview with employees, one physician once explained that he totally agreed with the organizational values which emphasized saving lives, assisting the sick and humanitarian culture. But all the same, increasing competition in the Chinese health care market has made the major hospitals suffer in attempting to balance survival and mission. Even though the hospital stressed patient interests rather than hospital profit, the physicians were the ones who took on responsibility and high risk, and received relatively low pay, compared to average workers. One interviewee mentioned that the medical profession held a belief in guarding the benefit of the patient (he totally agreed with the hospital values), however, under the circumstances, he could only    The moderated effect on fit and attitude only proved to fit on 'personal respect' and job satisfaction, as well as on 'profit orientation' and 'organizational commitment'. Value on 'personal respect' was evaluated poorly since poor staff-patient relationships were clearly discernible within the hospital, with some patients even assailing employees, which in turn made the various medical professionals even more remote from some of the patients. Where such attacks did occur, the hospital authority would simply instruct their employees not to retaliate, but instead, to ensure that they quickly left the scene. Employees within this hospital were not protected by the authorities when such conflicts occurred; indeed, there was one simple instruction: 'Do not fight back when you are hit, and do not quarrel when you are being scolded', which may of course explain how the fit on the value of 'personal respect' was related to job satisfaction among physicians   rather than among administrative staff. This might explain why the interaction term of desired and actual value on 'personal respect' is positive among physicians, but remains negative among administrative staff.
Swaak [44] also pointed out the working organization of each danwei (work unit) gives employees a sense of belonging, security, and identity. In some ways, employees might benefit from being a member of a renowned organization; 'face' could be the major determinant of their attitude [45].
The undergoing health care reforms in Shanghai are aimed to expand coverage and pursue financial sustainability started from 2000 until now [46,47]. External environment changes were not limited to the health care reform, but also the booming economic growth in Shanghai and other cities in China. Collected in the year of 2004, this study was cross-sectional survey to observe the transition in the middle points of health care reform and economic boom. From the perspectives of medical professionalism, the physicians suffered poor patient-physician relationship in the year of 2004, and nowadays might suffer worse relationship with patients [48]. Moreover, increasing commercialization of health care is the other issues to break social contract. Therefore, further research is recommended to examine the impact of lasting policy [49].
Future research in this area should attempt to include other possible intervening variables to the relationship of fit on value and attitude. The hospital should adopt an appropriate competitive strategy that might enable it to achieve fair compensation and yield the hospital reasonable profit, and thereby ensure that it can retain the services of its attending physicians.

Limitations
There are four major limitations of this study which should be taken into consideration. Firstly, it was difficult to interpret the polynomial regression coefficients, essentially because this index somehow restricts the limitations that may be caused by other indices, such as combining two corresponding entities into a single score to confound the concept of congruence. Therefore, the polynomial regression was used to interpret the three-dimension equation. Secondly, the potential for generalization was also very limited; clearly, since the sampling was carried out by the hospital authority itself, the problem of selection bias may have been largely unavoidable, and as a result, the representation of the sample was unclear. This sample hospital is public-owned and classified to third grade one in Shanghai metro area. Since the majority of health care service was provided by public hospitals in China until now, we could refer only the research result to similar ranking public hospitals in China. Anyhow, we are reluctant to generalize our result to other ranking public hospitals nor private hospitals in China. Thirdly, the antecedents of P-O fit were multiple, such as job choice, personality, efficiency of socialization, and so on; however, it is clear that some of these antecedents may well interact with P-O fit. Further exploration of these points will be necessary in order to confirm our findings [50]. Finally, common method variance may arise because the variables were all measured by the self-reported instrument. In order to avoid such variance, several processes were included in the design, such as single-blind design, reverse scored questions, and the use of non-rotated factor analysis results to determine whether only a single factor existed [51]. As a result, there was no single factor; therefore, the possibility of common method variance could be eliminated, or at least reduced.

Conclusion
Any future research in this area should attempt to include other possible intervening variables in the relationship of fit to values and attitude. It is proposed that the hospital should adopt an appropriate competitive strategy that might enable it to achieve a fair compensation system and reasonable profit, and thereby ensure that it can retain the services of its attending physicians.