International Journal of Nursing Studies 39 (2002) 867878

The inuence of nurses working motivation and job satisfaction on intention to quit: an empirical investigation in Taiwan
Huey-Ming Tzeng*
Department of Nursing, I-Shou University, No. 1, Section 1, Hsueh-Chang Road, Ta-Hsu Hsiang, Kaohsiung County 840, Taiwan Received 21 August 2001; received in revised form 21 January 2002; accepted 22 April 2002

Abstract The present era of cost-containment pressures indicates that nursing executives have to ensure that, their nurses have a work environment with the work characteristics known to be linked to job satisfaction and good outcomes. The research hypothesis investigated here was: the higher nurses levels on general job satisfaction, overall satisfaction with their professional role, and general job happiness, the lower their intention to quit would be. The controlling variables included demographic characteristics, working motivation, and nine job satisfaction subscales. This study was conducted in 3 hospitals located in southern Taiwan. All nurses working for these 3 hospitals were the target subjects. The overall response rate (648 completed questionnaires) was 82%. The nurses Job Satisfaction and the perceptual degree of its importance questionnaire was used. Ordinal logistics regression analyses were utilized. General job satisfaction, general job happiness, satisfaction with salary and promotion, institution, educational background, and age of nurses youngest child were proved to be signicant predictors of nurses intention to quit. Suggestions for future studies and administrative strategies in decreasing nurses intention to quit were discussed. r 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Working motivation; Job satisfaction; Intention to quit; Nurse

1. Introduction A competitive healthcare market is indeed qualityoriented, and improving the quality of healthcare services is a continuing challenge to managers in the healthcare industry. Similar to the western societies, Taiwans healthcare industry became a cost-containing environment after the government implemented the National Health Insurance System in 1995. The present era of cost-containment pressures indicates that nursing executives have to ensure that, their nurses have a work environment with the work characteristics known to be
*Tel.: +886-9261-50912, 886-7-6577711x5752; fax: +886-76577056. E-mail address: tzeng [email protected] (H.-M. Tzeng).

linked to job satisfaction and good outcomes. Research of Leiter et al. (1998) concluded that patients, who stayed on wards where nursing staff felt more exhausted or more frequently expressed their intention to quit, were less satised with their medical care. This study conrmed the importance of understanding nurses job satisfaction and intention to quit as both concepts, which would contribute to patient outcomes. Several recent studies further conrmed that nurses job satisfaction contributed to the perceptual levels of patient satisfaction with nursing care, which is one of the most important clinical outcome indicators (Tzeng and Ketean, 2002; Tzeng et al., 2002; Tzeng et al., 2001). Moreover, both nurses job satisfaction and patient satisfaction have been identied in the Nursing Report Card for Acute Care (American Nurses Association, 1995) as two of the important nursing quality

0020-7489/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 7 4 8 9 ( 0 2 ) 0 0 0 2 7 - 5

868

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

outcome indicators. Nurses intention to quit has a strong effect on their actual action of turnover, which might lead to certain amount of decrease in the quality and increase in the cost for patient care (Alexander et al., 1998). In a local conference for enhancing the communication among hospital nursing executives, nurse educators and government ofcers in Taiwan, it was concluded that nurses turnover rates among hospitals have been decreased from up to 50% down to below 10%. In some public healthcare institutions, the turnover rates were even lower than 5% per year. In other words, Taiwans turnover rates were no longer a severe issue for the nurse market. This trend might be caused by the changes in the macro-economical environment (e.g., cost containment, more supply than demand for nurses, difculties for nurses to change their career path). As for nurses intention to quit was related to their loyalty and work performance (Dolan et al., 1992), nurses intention to quit, instead of turnover, was determined as the target variable of this study. As a result, this exploratory project aimed to investigate the effects of working motivation, job satisfaction, and general perceptual factors as related to jobs, on nurses intention to quit. The purpose of this study was to develop and test a conceptual model of nurses intention to quit, and to contribute to the developing domain of nursing outcome studies.

2. Literature review Previous studies (e.g., McNeese-Smith, 1996; Lee et al., 1999; Tzeng, 1997) report that managerial factors affected employees attitudes, job satisfaction, organizational commitment, and motivation to perform well, and these factors, in turn, inuenced organizational outcomes. Organizational outcomes included, for example, patient satisfaction and employees intention to quit. Study of Bjorvell and Brodins (1992) found that half of the nursing staff wanted to quit their jobs. Those nurses, who had intention to quit, perceived themselves as less satised with several aspects (job dimensions of cooperation, job complexity, help received from superiors, and sufcient time for nursing care delivery), than did those who did not have any intention to quit. Results also showed that the head nurses seemed to have an important supportive function, and suggested that a supportive institution might reduce personnel turnover in hospitals. Dolan et al. (1992) concluded that occupational stressors, lack of professional latitude, and role of problems, predicted nurses intention to quit their working healthcare organizations. For the aims of beneting the quality of patient care and the respective hospitals, possible remedies to address these job

stressors were improving nurses quality of life, mental health, and decreasing the rate of turnover. Keel (1993) indicated that burnout affects mainly nurses, physicians, social workers and teachers. Burnout seems to be caused by stressful working conditions, disproportional-high efforts (time, emotional involvement, and empathy) and dissatisfaction with jobs. For the goals of lowering professional stress and improved satisfaction, social support and improved team cooperation could protect nurses against burnout. Study of Abbott et al. (1994) noted the relationship between the form of practice models (e.g., case management, managed care) and the levels of nurses job satisfaction. They concluded that managed care had a positive impact on nurses job satisfaction. Thus, according to Kovner et al. (1994) research (on the relative impact of various nursing care delivery models and administration interventions on nurses job satisfaction), pay, autonomy, and professional status were important predictors of job satisfaction. Different administration initiatives had impact on the satisfaction levels of interactions and task requirements. Initial dissatisfaction with these administration initiatives was noted. Moreover, study of Kennerly (1996) concluded that initiating shared governance did not inuence the levels of nurses job satisfaction, anticipated turnover, and perceived work effectiveness. Increases in autonomy were not continued over time. Armstrong-Stassen et al. (1994) examined the relationship between nurses job satisfaction dimensions, burnout, and intention to quit. Results demonstrated that career future and burnout (emotional exhaustion) predicted the levels of nurses intention to quit. Kind of work, amount of work, and career future were associated with burnout. Parker and Kulik (1995) demonstrated that work support and job stress were predictors of nurses burnout. Higher turnover rates were associated with poorer self-rated and supervisorrated job performance, more sick leaves, more reported absences for mental health reasons, and higher intention to quit. Lim and Yuens (1998) concluded that demands from patients/relatives, doctors, and perceived job image were correlated with nurses job satisfaction. Demands from patients/relatives and perceived job image were found to be associated with nurses intention to quit. Kivimaki et al. (1995) found that job satisfaction and motivation to perform well were related to the levels of job enrichment. The nurses occupying highly enriched jobs reported signicantly higher job satisfaction and motivation. Professional growth opportunities, workload, dissatisfaction with work hazards, and relationship with coworkers, were found to predict nurses turnover. In nursing care delivery systems, the following aspects (support for autonomy, executives leadership styles, communication, adequate time for patient care, the

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

869

degree of environmental uncertainty, and reported nursing stress) contributed to job satisfaction, motivation to perform well at work, and a work groups effectiveness (Edgar, 1999; Healy and McKay, 1999; Dunham-Taylor, 2000). Kangas and associates (1999) further emphasized that a support environment was most important to nurses job satisfaction. Thus, there were no found differences in nurses job satisfaction or patient satisfaction with nursing care in different organizational structures or nursing care delivery models (Kangas et al., 1999). Adams and Bond (2000) found that in hospitals, nurses interpersonal relationships (cohesive working relationships, relationships with medical staff), their perceptions of workload, and their evaluation of the appropriateness of the system where nursing care are practiced, were inuential factors of nurses job satisfaction. Analyses found that individual nurse characteristics were not related to job satisfaction. Agho (1993) indicated that the major determinants of nurses job satisfaction were routinization, participation, integration, distributive justice, instrumental communication, promotional opportunity, and positive affectivity. Muller-Smith (1999) reported that the potential for growth rather than fear of loss invented energy and excitement, which led to a joyful workplace and a strong sense of accomplishment. Liou et al. (1997) identied similar aspects for nurses job satisfaction, including indirect working environment, direct working environment,

salary and promotion, self-growth, challenging work, interaction and feedback with patients and family members, leadership style, working atmosphere, and family support and religion.

3. The conceptual and tested framework According to the previously reviewed literature, the conceptual framework (see Fig. 1) was developed. Whereas, Adams and Bond (2000) found that individual nurse characteristics were not related to job satisfaction or intention to quit. Nurses demographic characteristics were controlled in the model for exploratory purposes. The explanatory variables of nurses intention to quit were demographic characteristics, nine job satisfaction subscales (indirect working environment, direct working environment, salary and promotion, self-growth, challenge in work, interaction and feedback with patients and family members, leadership style, working atmosphere, and family support and religion), general job satisfaction in the levels of satisfaction (general job satisfaction); overall satisfaction levels with nurses professional role, general feeling when doing the job in the levels of happiness (general job happiness). The demographic characteristics included hospital, nurses age, marital status (single, married, or divorce/separated), years of professional experience, tenure, highest educational degree, going to school for a higher

Demographic Characteristics Hospital Age Marital status (single, married, divorce or separated) Years of professional experience Tenure Motivation (dream job, interest, for money, no other choice, family expectation)

Job Satisfaction Subscales Indirect working environment Direct working environment Salary and promotion Self-growth Challenge in work Interaction with and feedback from patients and family members Leadership style Working atmosphere Family support and religion General Perceptual Factors General job satisfaction Overall satisfaction levels with the professional roles General job happiness

Dependent Variable: Intention to quit

Fig. 1. The conceptual framework of nurses intention to quit. The ordinal dependent variable, having intention to quit, was rated into ve categories: [(1) least agreeable, (2) disagreeable, (3) neutral, (4) agreeable, and (5) most agreeable]. The explanatory variables included demographic characteristics, job satisfaction subscale, and three general perceptual factors.

870

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

educational degree, having a religion, number of children, the age of the youngest child, having administrative responsibilities, working motivation (dream job, interest, for money, no other choice, or family expectation). All nurse subjects in this study were female, and as a result, gender was not included. The dependent variable, intention to quit, was dened as an ordinal variable, and rated in a 5-Likert scale [(1) least agreeable, (2) disagreeable, (3) neutral, (4) agreeable, and (5) most agreeable]. Denitions of included variables were listed in Appendix A: Denitions of the variables. All the scales/items included in this study were operationalized in the nurses job satisfaction and the perceptual degree of its importance questionnaire. The research hypothesis investigated here was: the higher nurses levels on general job satisfaction, overall satisfaction factor with their professional role, and general job happiness, the lower their intention to quit would be, after controlling for demographic characteristics, working motivation, and nine job satisfaction subscales.

participated in this study voluntarily and anonymously. The overall response rate (648 completed questionnaires) was 82%, and the correspondent rate for each individual hospital ranged from 72% to 84%. 4.2. The nurses job satisfaction and the perceptual degree of its importance questionnaire This questionnaire set was written in Mandarin, and only part of the items in this questionnaire was utilized. There were a total of 48 items for measuring the levels of the nine job satisfaction scales. Items were grouped into 9 scales: indirect working environment (6 items); direct working environment (7 items); salary and promotion (7 items); self-growth (3 items); challenge in work (6 items); interaction with and feedback from patients and family members (3 items); leadership style (7 items); working atmosphere (7 items); and family support and religion (2 items). Responses were given on a 5-point Likert scale: ranging from 1 (least satised) to 5 (most satised), and 9 (not applicable). Three general perceptual factors (general job satisfaction, overall satisfaction with professional roles, general job happiness) were rated from very satised or very happy to very dissatised or very unhappy. The ordinal dependent variable, having intention to quit, was rated into ve categories: (1) Least agreeable, (2) disagreeable, (3) neutral, (4) agreeable, and (5) most agreeable. For the consistency of the tool, a test-retest method (2 weeks apart) was used. A total of 46 full-time nurses participated in this test. The test-retest correlation coefcients on items were ranged from 0.81 to 0.96. As for the reliability analyses, alpha coefcients were calculated using all the valid subjects. Except for the satisfaction scale of family support and religion (alpha coefcient=0.64), all the other alpha coefcients were ranged from 0.74 to 0.92. For each scale, the mean value was calculated by averaging the valid values of the grouped items. 4.3. Data analyses Data were entered and processed by using the statistical package for the social sciences (SPSS) software, English version 10.0. Descriptive information for all included variables was presented. Ordinal logistic regression analyses were utilized to develop a model of nurses intention to quit. The demographic characteristics were entered into the regression model rst, followed by nine job satisfaction scales, and three general job-related indicators at last. This procedure allowed us to observe the changes on coefcients when more explanatory variables were added into the equation.

4. Methods 4.1. Subjects This study was conducted in 3 hospitals located in southern Taiwan. These hospitals nursing administrators acted as members of the Human Subject Review committees. Committee members reviewed this research proposal for approval, and assured having appropriate human subject protection. The attached cover sheet indicated that all collected data would be used for research purposes only. Data would be kept condential, and subjects identication will be protected. These 3 hospitals were located in the same demographic area. One of them is a Christian teaching hospital (Hospital A), another one is a private-owned local hospital (Hospital B), and the other one is a local, veterans hospital (Hospital C). Data were collected during the period from December 1999 to March 2000. All the nurses working in these three institutions were the target population (including the ones with managerial responsibilities, and excluding physician assistants), for a total of 786 nurses. The questionnaire packages were sent to the nurse subjects through the hospitals nursing departments. This questionnaire packages (included a letter explaining the study purpose, a demographic information chart, a pre-stamped return envelope, and a gift for appreciating subjects participation) were disseminated to all nurses through subjects managers. The completed questionnaire (in a sealed envelop) was returned to the principal investigator by mail or through their nursing departments. Subjects

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

871

5. Results Table 1 demonstrated the demographic information of nurses. All participants were female, 26 years old with more than 2.5 years of tenure in average. About 29.3% of them were married, divorced or separated. Only

about 4.6% of participants had a baccalaureate, master or higher degree. About 70.6% of nurses worked as a nurse due to their own interest or just for money (35.3% for each). In average, having an intention to quit was rated as 3.02 (3, neutral) (SD=0.99). Except for the aspects of indirect working environment and salary and

Table 1 Descriptive information (n 648) Variables/values Age in years Professional working experience in months Tenure in months Number of children Age of the youngest child in years Indirect working environment Direct working environment Salary and promotion Self growth Challenge in work Interaction with patients and family members Leadership style Working atmosphere Family support and religion General job satisfaction General satisfaction levels toward professional roles General job happiness Intention to quit Frequency (valid percent) Hospital A B C Marital status Single Married Divorced or separated Highest educational degree Diploma Associate degree Baccalaureate degree Master degree or higher Part-time student status No Yes Having a religion No Yes Managerial responsibility No Yes Working motivation (multiple choices) Dream job Interest For money No other choice Family expectation Mean 26.42 66.29 31.24 1.78 6.52 2.94 3.24 2.85 3.08 3.23 3.40 3.21 3.51 3.54 3.26 3.30 3.32 3.02 N 394 174 80 446 182 3 65 511 22 6 509 83 118 530 600 48 78 229 229 114 156 SD 6.65 73.97 52.03 0.69 6.10 0.55 0.55 0.67 0.63 0.55 0.60 0.67 0.57 0.60 0.78 0.64 0.74 0.99 Minimum 19.00 1.00 1.00 1.00 1.00 1.00 1.43 1.00 1.00 1.00 1.00 1.00 1.57 1.50 1.00 1.00 1.00 1.00 % 60.8 26.9 12.3 70.7 28.8 0.5 10.7 84.6 3.6 1.0 86.0 14.0 18.2 81.8 92.4 7.6 12.0 35.3 35.3 17.6 24.1 Maximum 52.00 360.00 357.00 4.00 26.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00

872

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

promotion, most of the nurses felt satised to neutral about their jobs. In assessing overall model t, the goodness of t measure compared the predicted probabilities to the observed probabilities. Smaller values of the 2 log likelihood measure indicated better model t (Hair et al., 1998). Table 2 indicated a good t on all three ordinal regression models. The explanatory variables in the third model were statistically and signicantly associated (Pearson Chi-Square=2677.67, df=1864, p 0:01), however (predictors should be independent to each other.) By observing Model 1 alone (Table 3) (when p was set at 0.10), if a nurse was working at Hospital B (estimate=0.751, Wald w2 5:48; p 0:02), having less

number of children (estimate=0.351, Wald w2 3:39; p 0:07), and her motivation to work was not for money (estimate=0.437, Wald w2 6:66; p 0:01), no other choice (estimate=0.554, Wald w2 6:52; p 0:01), and family members expectation (estimate=0.346, Wald w2 3:19; p 0:07), this nurse would have higher intention to quit (the Nagelkerke Pseudo R2 0:109). For Model 2 (when p was set at 0.10), if a nurse was working at Hospital B (estimate=1.024, Wald w2 7:39; p 0:01) and having lower satisfaction levels with direct working environment (estimate=0.417, Wald w2 2:81; p 0:09), salary and promotion (estimate=0.469, Wald w2 5:81; p 0:02), and challenge in work (estimate=0.697, Wald w2 7:35; p 0:01), and her

Table 2 Model tting information for three ordinal logistic regression models: nurses intention to quit as the dependent variable (the link function is logit) The chi-square test for goodness-of-ta Modelb 1 2 3 2 log likelihood intercept only 1501.89 1361.83 1327.51 Pearsons chi-square testc Model 1 2 3 Model 1 2 3 Chi-square Pearson Deviance Pearson Deviance Pearson Deviance Pseudo R-squared Cox and Snell R2 Nagelkerke Cox and Snell R2 Nagelkerke Cox and Snell R2 Nagelkerke 0.103 0.109 0.266 0.283 0.385 0.410 Degree of freedom 2073.90 1434.85 1895.82 1211.11 2677.67 1096.74 Signicance 2112 2112 1919 1919 1864 1864 0.72 1.00 0.64 1.00 0.00** 1.00 2 log likelihood nal model 1443.17 1211.11 1096.74 Chi-square 58.72 150.72 230.77 Degree of freedom 20 29 32 Signicance 0.00** 0.00** 0.00**

Note: **p 0:01; 2-tailed. a The chi-square test for goodness-of-t compares the results of a study of data to determine how likely the sample results actually are, given the parameter estimates. A good model is one, which results in a high likelihood of the observed values or a small value for 2 Log Likelihood. b The rst model included only demographic variables; the second model was composed of demographic variables and the 8 job satisfaction scales; and the third model included demographic variables, 8 job satisfaction scales, and three general job-related indicators. c The Pearsons chi-square test (also called the chi-square test of independence) examined a set of variables to determine whether they are associated. d The Cox and Snell R2 measure operates in the same manner as the R2 measure in multiple regression, with higher values indicating greater model t (less than 1). Nagelkerke, which has a range of 01, is a modication of the Cox and Snell R2 measure (Hair et al., 1998).

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

873

Table 3 Parameter estimates of the ordinal regression models: nurses intention to quit as the dependent variable (the link function is logit) Model 1 Variable Threshold (intercept) Intention to quita=1 Intention to quit=2 Intention to quit=3 Intention to quit=4 Location (explanatory variables) Nurse age Professional experience Tenure Number of children Age of the youngest child Hospital=A Hospital=B Hospital=C Marital status Single Married Divorced or separated Education Diploma Associate degree Bachelor degree Master degree or higher Part-time student status=no Part-time student status=yes Has a religion=no Has a religion=yes Administrative position=no Administrative position=yes Motivation to work Dream job=no Dream job=yes Interest=no Interest=yes For money=no For money=yes No other choice=no No other choice=yes Family expectation=no Family expectation=yes Threshold (intercept) Intention to quit=1 Intention to quit=2 Intention to quit=3 Intention to quit=4 Location (explanatory variables) Nurse age Professional experience Tenure Number of children Age of the youngest child Estimate Standard error 1.893 1.887 1.887 1.893 0.046 0.004 0.002 0.191 0.033 0.302 0.321 Wald w2 P value 95%CI

2.630 0.315 1.375 3.331 0.010 0.004 0.002 0.351 0.006 0.066 0.751 & 1.554 0.971 & 0.587 0.401 0.777 & 0.204 & 0.116 & 0.145 & 0.074 & 0.234 & 0.437 & 0.554 & 0.346 &

1.93 0.03 0.53 3.10 0.05 0.81 1.06 3.39 0.03 0.05 5.48

0.17 0.88 0.47 0.08* 0.83 0.37 0.30 0.07* 0.86 0.83 0.02**

7.00 4.69 3.04 1.12

to to to to

0.06 2.33 3.98 5.92

0.10 to 0.08 0.004 to 0.01 0.002 to 0.007 0.73 to 0.02 0.06 to 0.07 0.66 to 0.53 0.12 to 1.38

1.144 1.103

1.85 0.77

0.17 0.38

0.69 to 3.80 1.19 to 3.13

0.806 0.770 0.894 0.241 0.194 0.349

0.53 0.27 0.76 0.72 0.36 0.17

0.47 0.60 0.39 0.40 0.55 0.68

2.17 to 0.99 1.91 to 1.11 2.53 to 0.98 0.27 to 0.68 0.27 to 0.50 0.54 to 0.83

0.249 0.184 0.169 0.217 0.194

0.09 1.62 6.66 6.52 3.19

0.77 0.20 0.01*** 0.01*** 0.07*

0.41 to 0.56 0.13 to 0.59 0.77 to 0.11 0.98 to 0.13 0.73 to 0.03

10.294 7.755 5.804 3.664 0.030 0.005 0.003 0.351 0.032

2.413 2.388 2.378 2.375 0.049 0.004 0.003 0.214 0.036

18.20 10.54 5.96 2.38 0.38 1.54 1.19 2.70 0.79

0.00*** 0.00*** 0.02** 0.12 0.54 0.21 0.28 0.10 0.37

15.02 to 5.57 12.44 to 3.07 10.47 to 1.14 8.32 to 0.99 0.13 0.00 0.00 0.77 0.04 to to to to to 0.07 0.01 0.01 0.07 0.10

874 Table 3 (continued) Model Variable

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

Estimate

Standard error 0.241 0.249 0.195 0.180 0.257 0.183 0.193 0.207 0.173 0.362 0.377

Wald w2

P value

95%CI

Job satisfaction indicators Indirect working environment Direct working environment Salary and promotion Self-growth Challenge in work Interaction with patients and family members Leadership style Working atmosphere Family support and religion Hospital=A Hospital=B Hospital=C Marital status Single Married Divorced or separated Education Diploma Associate degree Bachelor degree Master degree or higher Part-time student status=no Part-time student status=yes Has a religion=no Has a religion=yes Administrative position=no Administrative position=yes Motivation to work Dream job=no Dream job=yes Interest=no Interest=yes For money=no For money=yes No other choice=no No other choice=yes Family expectation=no Family expectation=yes Threshold (intercept) Intention to quit=1 Intention to quit=2 Intention to quit=3 Intention to quit=4 Location (explanatory variables) Nurse age Professional experience Tenure Number of children Age of the youngest child

0.047 0.417 0.469 0.211 0.697 0.113 0.315 0.213 0.048 0.325 1.024 & 0.647 0.057 & 1.069 0.983 0.891 & 0.069 & 0.146 & 0.304 & 0.131 & 0.074 & 0.344 & 0.144 & 0.135 & 12.063 9.298 7.179 4.783 0.024 0.005 0.002 0.283 0.068

0.04 2.81 5.81 1.38 7.35 0.39 2.67 1.06 0.08 0.81 7.39

0.85 0.09* 0.02** 0.24 0.01*** 0.54 0.10 0.30 0.78 0.37 0.01***

0.52 0.90 0.85 0.14 1.20 0.47

to to to to to to

0.43 0.07 0.09 0.56 0.19 0.24

0.69 to 0.06 0.62 to 0.19 0.29 to 0.39 0.39 to 1.04 0.29 to 1.76

1.457 1.399

0.20 0.00

0.66 0.97

2.21 to 3.50 2.69 to 2.80

0.841 0.801 0.923 0.267 0.215 0.381

1.62 1.51 0.93 0.07 0.46 0.64

0.20 0.22 0.33 0.80 0.50 0.43

2.72 to 0.58 2.55 to 0.59 2.70 to 0.92 0.46 to 0.59 0.27 to 0.57 0.44 to 1.05

0.272 0.198 0.185 0.238 0.211

0.23 0.14 3.44 0.37 0.41

0.63 0.71 0.06* 0.55 0.52

0.40 to 0.66 0.31 to 0.46 0.71 to 0.02 0.61 to 0.32 0.55 to 0.28

2.498 2.469 2.454 2.448 0.051 0.005 0.003 0.226 0.039

23.32 14.49 8.56 3.82 0.22 1.14 0.66 1.57 2.97

0.00*** 0.00*** 0.00*** 0.05* 0.64 0.29 0.42 0.21 0.09*

16.96 to 7.17 14.24 to 4.56 11.99 to 2.37 9.58 to 0.01 0.12 0.00 0.00 0.73 0.01 to to to to to 0.08 0.01 0.01 0.16 0.15

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878 Table 3 (continued) Model Variable Job satisfaction indicators Indirect working environment Direct working environment Salary and promotion Self-growth Challenge in work Interaction with patients and family members Leadership style Working atmosphere Family support and religion General items General job satisfaction General job happiness Overall satisfaction levels with the professional role Hospital=A Hospital=B Hospital=C Marital status Single Married Divorced or separated Education Diploma Associate degree Bachelor degree Master degree or higher Part-time student status=no Part-time student status=yes Has a religion=no Has a religion=yes Administrative position=no Administrative position=yes Motivation to work Dream job=no Dream job=yes Interest=no Interest=yes For money=no For money=yes No other choice=no No other choice=yes Family expectation=no Family expectation=yes Estimate Standard error 0.252 0.260 0.204 0.186 0.272 0.194 0.200 0.216 0.180 0.188 0.180 0.173 Wald w2 P value 95%CI

875

0.128 0.078 0.430 0.093 0.246 0.085 0.187 0.160 0.177 0.788 0.862 0.150

0.26 0.09 4.41 0.25 0.82 0.19 0.87 0.55 0.96 17.55 22.89 0.75

0.61 0.76 0.04** 0.62 0.37 0.66 0.35 0.46 0.33 0.00*** 0.00*** 385

0.62 0.59 0.83 0.27 0.78 0.47

to to to to to to

0.37 0.43 0.03 0.46 0.29 0.30

0.58 to 0.20 0.58 to.26 0.18 to 0.53 01.16 to 0.42 01.22 to 0.51 0.19 to 0.49

0.564 1.035 & 1.199 0.532 & 1.790 1.639 1.436 & 0.167 & 0.138 & 0.425 & 0.173 & 0.227 & 0.157 & 0.082 & 0.026 &

0.375 0.387

2.26 7.15

0.13 0.01***

0.17 to 1.30 0.28 to 1.79

1.495 1.434

0.64 0.14

0.42 0.71

01.73 to 4.13 02.28 to 3.34

0.866 0.820 0.944 0.280 0.224 0.407

4.28 4.00 2.31 0.36 0.38 1.09

0.04** 0.05** 0.13 0.55 0.54 0.30

3.49 to 0.09 3.25 to 0.03 3.29 to 0.42 0.38 to 0.72 0.30 to 0.58 0.37 to 1.22

0.279 0.208 0.191 0.249 0.218

0.39 1.20 0.68 0.11 0.01

0.54 0.27 0.41 0.74 0.91

0.72 to 0.37 0.64 to 0.18 0.53 to 0.22 0.57 to 0.41 0.40 to 0.45

Notes: *p 0:10; 2-tailed. **p 0:05; 2-tailed. ***p 0:01; 2-tailed. &: This parameter was set to zero because it was redundant. Standard error=estimated precision of the coefcients. 95%CI=95% condence intervals for the coefcients. Wald w2 =the Wald test statistic calculated from the data to be compared with the chi-square distribution with 1 degree of freedom. a The ordinal dependent variable, having intention to quit, was rated into ve categories: (1) least agreeable, (2) disagreeable, (3) neutral, (4) agreeable, and (5) most agreeable.

876

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

motivation to work was NOT for money (estimate=0.344, Wald w2 3:44; p 0:06), this nurse would have higher intention to quit (the Nagelkerke Pseudo R2 0:283). Model 3 was the nal ordinal regression model for testing the research hypothesis. Institution (Hospital B) (estimate=1.035, Wald w2 7:15; p 0:01), age of the youngest child (estimate=0.068, Wald w2 2:97; p 0:09), holding a diploma (estimate=1.790, Wald w2 4:28; p 0:04), holding an associate degree (estimate=1.639, Wald w2 4:00; p 0:05), salary and promotion (estimate=0.430, Wald w2 4:41; p 0:04), general job satisfaction (estimate=0.862, Wald w2 22:89; p 0:00), and general job happiness (estimate=0.788, Wald w2 17:55; p 0:00) were proved to be signicant predictors of nurses intention to quit (when p was set at 0.10). The Nagelkerke Pseudo R2 for Model 3 was 41%.

6. Discussion and conclusion The research hypothesis (the higher nurses levels on general job satisfaction, overall satisfaction with professional roles, and general job happiness, the lower their intention to quit would be, after controlling for demographic characteristics and working motivation rst, and following by nine job satisfaction subscales) was tested in this project. This hypothesis was not fully supported (when p was set at 0.10), however. In these three ordinal regression models, different demographic characteristics (such as working hospital, number of children, age of the youngest child, and motivation to work as a nurse) demonstrated statistically signicantly predicting power toward nurses intention to stay. In Model 2, as compared to direct working environment and salary and promotion, challenge in work was the strongest job satisfaction predictor of intention to quit. However, in Model 3, general job happiness and general job satisfaction had higher values in estimates than the one of salary and promotion. These results showed the sophisticated association and predicting power among included variables and nurses intention to quit. In the nal (third) ordinal regression model, only two primary explanatory variables (general job satisfaction and general job happiness) were statistically signicant predictors of nurses intention to quit. Compared to general job satisfaction, general job happiness was seldom used in predicting outcome variables. Here, job happiness was quantied as a general item to measure nurses overall perceptions towards their jobs when they are doing their jobs in the levels of happiness. Tzeng and Ketean (2002) and Tzeng et al. (2001) studies suggested a signicant relationship between nurses general job happiness and patient satisfaction

with nursing care. In addition, Tzengs (2002) conrmed that nurses job satisfaction and job happiness were two distinct concepts. Muller-Smith (1999) reported that the potential for growth rather than fear of loss invented energy and excitement, which led to a joyful workplace and a strong sense of accomplishment. In conclusion, this study provided strong evidence that job satisfaction and job happiness are two discrete concepts and explained a signicant amount of variance of nurses intention to quit. Future research might investigate whether employees job happiness is a predictor of the other outcome indicators (e.g., work performance, service attitudes). At the mean time, it would be imperative to further investigate the possible differences on employees perceptual levels on job happiness between female and male and employees personality characteristics. Ordinal regression results also demonstrated that working motivation (for money, no other choice, and family expectation) were signicant predicators of intention to quit in Model 1. Three job satisfaction aspects (challenge in work, salary and promotion, and direct working environment) were signicant predicators of nurses intention to quit in Model 2, but the effects of work motivation (only working for money left) were washed out. The variance of nurses satisfaction with challenge in work and direct working environment was washed out after entering general job-related indicators as seen in the nal (third) model, where salary and promotion was still a signicant predictor. Previously reviewed literature also emphasized the importance of building a support environment with appreciated reward systems (Edgar, 1999; Healy and McKay, 1999; Kangas et al., 1999; Adams and Bond, 2000; Dunham-Taylor, 2000). Kivimaki et al. (1995) provided a different insight that the nurses, who are occupying highly enriched jobs, reported signicantly higher job satisfaction and work motivation. Observing the organizational cultures of current working places, building an environment emphasizing afliation and recognition is a norm. However, creating challenging works is still a hard-to-approach matter to nursing executives. Different administration initiatives (such as, implementing case management, shared governance, work redesign, total quality management) had been introduced to the healthcare organizations (Abbott et al., 1994; Kennerly, 1996). Thus, the positive impact of these initiatives on nurses job satisfaction, especially with challenge in work, does not seem to be considerable. For the purpose of decreasing nurses intention to quit, it is suggested to further explore extensively which factors matter to nurses job satisfaction with challenging work, such as achievement, autonomy at work, new challenge, stress levels, and suitable work (workload, context, time, emotional involvement). We believed that for increasing

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878

877

Table 4 Denitions of the variables: operationalized in the Nurses Job Satisfaction and the Perceptual Degree of its Importance Questionnaire Nurses job satisfaction aspects Indirect working environment Direct working environment Salary and promotion Self growth and development Challenging work Denition The following nine scales were described as nurses satisfaction levels with predened job characteristics, which are: A hospitals policies and benets (including leisure activities, housing, parking, and vacation) Medical equipments, working environment, scheduling, rotation to other wards, working hours, workow, and stability of their jobs Clinical ladder, fairness of the performance appraisal systems, bonuses, and paid time off Opportunities and arrangement of on-job training, participating in the research activities, and writing or publication Recognition for their achievements, independent thinking and decisionmaking, independent work, being challenged, competency towards their jobs, and workload The feedback from patients and family members, improvement of patients conditions, and communication with patients and family members How the hospital handles staffs complaints, direct supervisors leadership styles, abilities of coordination, relationship with direct supervisors, collaborative relationship among supervisors, and respect, fair treatment, support and caring from their supervisors The work units atmosphere, communication and the collaborative relationship with physicians and other colleagues working in the same unit, caring and support from their colleagues, and interaction with the colleagues working in the other units Caring and support from relatives regarding their jobs, and the religion of the work place Whether nurses often have intention to leave their currently working hospital (in the scale from most agreeable to least agreeable) Nurses overall perceptions towards the job in the levels of satisfaction Nurses overall perceptions towards the professional roles they played in the levels of satisfaction Nurses overall perceptions towards their jobs, when they are doing their jobs, in the levels of happiness Nurses incentives (including dream job, interest, for money, no other choice, and family expectation) to do their current jobs

Interaction with and feedback from patients and family members Leadership style

Working atmosphere

Family support and religion Intention to quit General job satisfaction Overall satisfaction levels with professional roles General feeling when doing jobs (general job happiness) Working motivation

nurses commitment and loyalty to their working hospital (decreasing their intention to quit), nurse executives should consider various aspects of job characteristics. Thus, some factors, like salary and promotion, might not be as modiable as the other job characteristics. Overall, this project provided an insight on the inuence of nurses working motivation, job satisfaction, and general factors on nurses intention to quit in Taiwan. The healthcare market in Taiwan might not have immediate problems of nursing shortage and high turnover rates. Nursing executives still need to reduce nurses intention to quit and increase their job satisfaction. The reasons are that high intention to quit and low job satisfaction might lead to certain amount of decrease in the quality and patients intention to return for future use, and increase in the cost for patient care under this cost-containment environment.

Appendix A Denitions of the variables: operationalized in the Nurses Job Satisfaction and the Perceptual Degree of its Importance Questionnaire is given in Table 4.

References
Abbott, J., Young, A., Haxton, R., Van Dyke, P., 1994. Collaborative care: a professional model that inuences job satisfaction. Nursing Economics 12(3), 167-9, 174. Adams, A., Bond, S., 2000. Hospital nurses job satisfaction, individual and organizational characteristics. Journal of Advanced Nursing 32 (3), 536543. Agho, A.O., 1993. The moderating effects of dispositional affectivity on relationships between job characteristics and

878

H.-M. Tzeng / International Journal of Nursing Studies 39 (2002) 867878 hospital wards: testing the job characteristics model. Journal of Nursing Management 3 (2), 8791. Kovner, C.T., Hendrickson, G., Knickman, J.R., Finkler, S.A., 1994. Nursing care delivery models and nurse satisfaction. Nursing Administration Quarterly 19 (1), 7485. Lee, J.C., Chang, B.C., Pearson, M.C., Kahn, K.C., Rubenstein, C.V., 1999. Does what nurses do affect clinical outcomes for hospitalized patients? A review of the literature. Health Services Research 34 (5), 10111032. Leiter, M.P., Harvie, P., Frizzell, C., 1998. The correspondence of patient satisfaction and nurse burnout. Social Sciences in Medicine 47 (10), 16111617. Lim, V.K., Yuen, E.C., 1998. Doctors, patients, and perceived job image: an empirical study of stress and nurses in Singapore. Journal of Behavior in Medicine 21 (3), 269283. Liou, C.N., Shi, W.H., Tseng, K.Y., 1997. Job satisfaction of a medical centers nurses and related factors. VGH Nursing 14 (2), 143152. McNeese-Smith, D., 1996. Increasing employee productivity, job satisfaction, and organizational commitment. Hospital & Health Services Administration 41 (2), 160174. Muller-Smith, P., 1999. The energy business. Journal of Perianesthetic Nursing 14 (5), 300302. Parker, P.A., Kulik, J.A., 1995. Burnout, self- and supervisorrated job performance, and absenteeism among nurses. Journal of Behaviors in Medicine 18 (6), 581599. Tzeng, H.M., 1997. The inuence of staff nurses perceptions of organizational culture on inpatient satisfaction with nursing care. Unpublished Doctoral Dissertation, The University of Michigan School of Nursing, Ann Arbor, MI, USA Tzeng, H.M., 2002. Are we satisfying our nurses on the factors they care for? An empirical study in Taiwan. Journal of Nursing Administration 32 (6), forthcoming. Tzeng, H.M., Ketean, S., 2002. The relationship between nurses job satisfaction and inpatient satisfaction: an exploratory study in a Taiwans teaching hospital. Journal of Nursing Care Quality 16 (2), 5060. Tzeng, H.M., Yang, C.S., Lee, S.Y., 2001. Brief report: the relationship between patient satisfaction and health care providers job satisfaction: an exploratory study on the 11 metropolitan public health care institutions in Taiwan. Hospital (Taiwan) 34 (4), 1623. Tzeng, H.M., Ketean, S., Redman, R.W., 2002. Relationship of staff nurses strength of culture, job satisfaction, and inpatient evaluation with nursing care. International Journal of Nursing Studies 39 (1), 7984.

nurses job satisfaction. Research in Nursing and Health 16 (6), 451458. Alexander, J.A., Lichtenstein, R., Oh, H.J., Ullman, E., 1998. A causal model of voluntary turnover among nursing personnel in long-term psychiatric setting. Research in Nursing and Health 21 (5), 415427. American Nurses Association (ANA), 1995. Report to the Board of Directors, 2728. American Nurses Association Congress of Nursing Practice and Congress on Nursing Economics, Agenda item #9. Armstrong-Stassen, M., al-MaAitah, R., Cameron, S., Horsburgh, M., 1994. Determinants and consequences of burnout: a cross-cultural comparison of Canadian and Jordanian nurses. Health Care Women International 15 (5), 413421. Bjorvell, H., Brodin, B., 1992. Hospital staff members are satised with their jobs. Scandium Journal of Caring Sciences 6 (1), 916. Dolan, S.L., Van Ameringen, M.R., Corbin, S., Arsenault, A., 1992. Lack of professional latitude and role problems as correlated or propensity to quit amongst nursing staff. Journal of Advanced Nursing 17 (12), 14551459. Dunham-Taylor, J., 2000. Nurse executive transformational leadership found in participative organizations. Journal of Nursing Administration 30 (5), 241250. Edgar, L., 1999. Nurses motivation and its relationship to the characteristics of nursing care delivery systems: a test of the job characteristics model. Canadian Journal of Nursing Leadership 12 (1), 1422. Hair, J.F., Anderson, R.E., Tatham, R.L., Black, W.C., 1998. Multivariate data analysis, 5th Edition. Prentice-Hall, Upper Saddle River, NJ. Healy, C., McKay, M., 1999. Identifying sources of stress and job satisfaction in the nursing environment. Australian Journal of Advanced Nursing 17 (2), 3035. Kangas, S., Kee, C.C., McKee-Waddle, R., 1999. Organizational factors, nurses job satisfaction, and patient satisfaction with nursing care. Journal of Nursing Administration 29 (1), 3242. Keel, P., 1993. Psychological stress caused by work: burnout syndrome. Sozial und Praeventivmedizin (Article in German) 38 (Suppl 2), S131132. Kennerly, S.M., 1996. Effects of shared governance on perceptions of work and work environment. Nursing Economics 14 (2), 111116. Kivimaki, M., Voutilainen, P., Koskinen, P., 1995. Job enrichment, work motivation, and job satisfaction in