JONA Volume 33, Number 5, pp 293-299 ©2003, Lippincott Williams & Wilkins, Inc.

Factors That Influence Nurses’ Job Satisfaction Chen-Chung Ma, PhD Michael E. Samuels, DrPH Judith W. Alexander, PhD, RN Objective: To examine factors affecting the job satisfaction of registered nurses (RNs). Background: A growing recognition of job dissatisfaction among RNs in South Carolina hospitals has contributed to current problems with recruitment and retention. If administrators identify factors influencing RNs’ job satisfaction in hospitals and implement strategies to address these factors, RN turnover rates will decrease and recruiting and retention rates will increase. Methods: A cross-sectional study of secondary data was designed to identify the individual, work, and geographic factors that impact nursing job satisfaction at the state level.A 27-question self-administered survey was sent to 17,500 RNs in South Carolina with postage-paid envelopes for their responses. Surveys from 3472 nurses were completed anonymously. Univariate statistics were used to describe the study sample. One-way and multivariable Analysis of Variance were used to determine which variables contributed the most to job satisfaction. Results: For about two thirds of the RNs, job satisfaction remained the same or had lessened over the past 2 years. In addition, statistically significant differences were found between job satisfaction and years of service, job position, hospital retirement plan, and geographic area. Conclusions: The findings have implications for nurse managers and hospital administrators for planning and implementing effective health policies that

Author affiliations: Chief (Dr Ma), Division of Planning,Tri-Service General Hospital,Taipei,Taiwan, ROC; Endowed Chair/Distinguished Scholar in Rural Health Policy, Professor of Family Practice (Dr Samuels), Norman J. Arnold School of Public Health, University of Kentucky, Lexington, Ky; Associate Professor (Dr Alexander), College of Nursing, University of South Carolina, Columbia. Corresponding author: Chen-Chung Ma, PhD,Tri-Service General Hospital,Taipei 114,Taiwan, ROC ([email protected] or [email protected]).

JONA • Vol. 33, No. 5 • May 2003

will meet the unique needs of their staffs and organizations. Such research is particularly relevant in this difficult time of nursing shortages throughout the healthcare industry.

Overview Job satisfaction among RNs has long been recognized as a crucial indicator of nurses performance, cost savings, and quality of patient care.1 As RNs’ job satisfaction decreases, the likelihood of their leaving their employment settings increases. If nursing job satisfaction continues to deteriorate, the present nursing shortage will worsen. South Carolina (SC) has already experienced the effects of the nursing shortage. Inadequate staffing of nurses translates into delays in treatment of patients, holding patients in emergency department waiting beds, and shifting of care to patients’ family members. Additionally, increased nurse-to-patient ratios will lead to job dissatisfaction and burnout that intensify the turnover and vacancy rates.2 Therefore, job satisfaction and retention of qualified nurses is currently one of the primary concerns of hospital administrators.With an emerging national and South Carolina nursing shortage and a growing dissatisfaction of RNs, the continued exploration of nursing job satisfaction is important. The aim of this study is to determine which factors affect job satisfaction of nurses in SC hospitals. The findings should raise administrators’ awareness and may inspire them to take steps to improve the level of job satisfaction of nurses.

Literature Review Job Satisfaction Concept Job satisfaction has been widely studied and discussed within several disciplines including nursing, psychology, sociology, and management. The defini-

293

tion of job satisfaction varies from person to person and within any one person from time to time. Locke3 viewed job satisfaction as an evaluation that the employee makes of the job and the environment surrounding the job. Smith et al4 defined job satisfaction as the feelings an employee has about the job in general. Pilkington and Wood5 described job satisfaction as nurses’ degree of positive affective orientation towards their job. Geiger and Davit6 defined job satisfaction as the extent to which a nurse’s felt needs are fulfilled by the job that she/he performs. Generally, job satisfaction can be defined as the difference between the amount of rewards workers receive and the amount they believe they should receive.7 Job Satisfaction Measurement Job satisfaction may be measured in many different ways; therefore, many studies are not measuring the same phenomenon.Two well-established instruments for measuring job satisfaction in nursing are the McCloskey and Mueller Satisfaction Scale (MMSS)8 and the Index of Work Satisfaction Questionnaire (IWS).9 Since this study used secondary data for analyses, neither MMSS nor IWS tools were used. Instead, the SC Nursing Survey was used to measure nursing job satisfaction. Although this survey is not as well-established as the MMSS or the IWS, many items on the survey measure the same phenomenon of job satisfaction among nurses. Job Satisfaction and Nurses The study of the job satisfaction of nurses has a long history. The first major study of job satisfaction in nursing was conducted by Nahm10 at the University of Minnesota in 1940. He concluded that the factors affecting job satisfaction included income, hours of work, relationships with supervisors, interest in work, family and social relationships, and opportunities for advancement. Studies prior to and during the 1970s primarily relied on the theoretical fields of psychology and management.11-13 Then, in the late 1970s and early 1980s, nurses began to emerge as significant theorists regarding the predictors of nursing job satisfaction.14 In this period, organizational characteristics found to contribute to job satisfaction included a participative management model and having a nursing director that reported to the executive level.15 Kangas et al16 explored differences in the relationships among the job satisfaction of registered nurses, nursing care delivery models, and organizational structures.The results indicated elements such as whether nurses perceived the environment as sup-

294

portive and if they worked in a critical care unit were significant predictors of nurses’ job satisfaction. Using research syntheses, Kangas17 examined variables affecting nurses’ job satisfaction from 38 previous studies. The results indicated that job satisfaction for nurses correlated with prestige or job status, independence in decision-making, control over practice, and perceived social support.Additionally, the factors of salary and interpersonal relationships had substantially increased, while job stress continued to correlate negatively with nurses’ job satisfaction. Mills and Blaesing18 found that hospital nurses working in urban areas were more satisfied with nursing as a career than their counterparts in rural and less urban areas. Also important to nurses’ job satisfaction were work schedule, job security, salary, and fringe benefits.

Methods Conceptual Framework Job satisfaction can be considered both as an independent and a dependent variable.As an independent variable, job satisfaction is recognized as the cause of phenomena such as turnover, absenteeism, retention, and productivity.As a dependent variable, job satisfaction is seen as caused by factors such as individual, organizational characteristics, and mechanics of the job.19 In this study, job satisfaction was treated as a dependent variable. The study of factors that affect the job satisfaction among nurses has been extensive over the last 30 years. Job satisfaction is influenced by individual, organization, and environmental characteristics.1 The conceptual framework for this study is based on existing literature and research and is determined by characteristics of the individual, the work, and the geographic area (Figure 1).This conceptual framework is drawn from the theoretical perspectives of Maslow’s11 hierarchy of needs, Herzberg’s12 two-factor theory, expectancy theory,13 and general nursing theory.14 Data Sources and Target Population The data came from a statewide SC nursing survey, the Year 2000 RN relicensure application forms, and a SC Health Alliance (SCHA) internal data file. The SC nursing survey was designed by the Tripp Umbach Synergies research team in cooperation with SCHA and the SC Organization of Nurse Executives (SCONE) in October 2000.The Office of Research and Statistics of the SC State Budget and Control Board provided the names and addresses of each registered nurse who worked in the hospitals in SC.

JONA • Vol. 33, No. 5 • May 2003

Figure 1. Conceptual framework of the study.

The target population for this study was every nurse who was licensed in SC and who was currently working or formerly had worked in a hospital.The settings in which nurses worked included all public and private hospitals in SC, but excluded the federal hospitals of the Veterans Administration and military. Data Collection In October 2000, the survey, accompanied by postage-paid return envelopes, was mailed directly to the home addresses of 17,500 RNs employed in SC hospitals. About 300 of the surveys were returned undelivered, resulting in a total of 17,200 surveys delivered to eligible respondents. Each survey was accompanied by a letter signed by the presidents of SCHA and SCONE explaining the purpose of the research, the voluntary nature of participation, and the confidentiality of the data. Approximately 4,000 surveys were compeleted anonymously. After completing their preliminary analyses, the SCHA provided the survey for secondary data analysis. From the statewide population, 3,931 surveys were found usable for the current study. However, when the SCHA merged these data with additional variables supplies by the Office of Research and Statistics of the State Budget and Control Board, only 3,472 surveys remained.Some respondents tore off or obliterated the identification number, making merger

JONA • Vol. 33, No. 5 • May 2003

of datasets impossible.The final result was a response rate of 20.2%. Representativeness Leslie20 found that when surveys are made of homogeneous populations (a group whose members share a strong group identity [eg, student nurses] concerning attitudes, opinions, and perspectives), significant response-rate bias is probably unlikely. He concluded that if researchers design surveys directly related to homogeneous groups, they should not be overly concerned about the percentage of surveys returned. Representativeness will most likely be excellent. Armstrong and Overton21 proposed a method for estimating nonresponse bias (comparisons with “known” values for the population). They suggested that results from a given survey can be compared with “known” values for the population (eg, age, income). Stratton et al22 compared personal characteristics (eg, age and gender) of staff RNs in their study with similar known population values included in the state Division of Nursing’s 1988 Sample Survey of RNs to investigate the representativeness of the study. The present study compared the variables of gender, age, level of education, and geographic location of the nurses surveyed with those of the state’s RNs in the year 2000.The results showed that the respondent sample does represent the whole population.

295

Study Design This study used a cross-sectional study designed to identify the individual, work, and geographic factors that impact nursing job satisfaction at the state level. The dependent and independent variables were already in place and not subject to manipulation by the researcher. Survey Instruments A two-page double-sided 27-item survey was developed for this study.The instrument consisted of two parts: (1) demographic and work-related questions, and (2) job satisfaction measurements. Based on the survey, the dependent and independent variables used in the study were as follows. Dependent Variable

Overall Measure of Job Satisfaction of Nurses.These 10 items measured job satisfaction. Each item was scored on a 10-point scale ranging from 1 ⫽ very dissatisfied to 10 ⫽ very satisfied. Factor analysis revealed a unidimensional scale of overall job satisfaction containing 10 components.The score for overall job satisfaction was found by totaling the responses to each item to form a quasi-continuous measure that could range from 10 to 100. Independent Variables

The independent variables were age, level of education, years of service, salary, hospital bed size, care setting, job position, hospital retirement plan, teaching hospital status, and geographic area. For analysis purposes, these variables were scored as categorical variables, except for age and years of service, which were treated as both continuous and categorical. Reliability and Validity of the Instrument To test the initial validity and reliability of the job satisfaction instrument, Common Factor Analysis and Internal Consistency Reliability were performed on the 10 items in the survey. The result of the factor loading of all items was positive and ranged from .56 to .78. In addition, only one factor had an Eigen value greater than 1.0 (⫽ 4.02). This analysis suggested that the 10-item instrument sampled one construct, the job satisfaction of nurses. At the same time, the Cronbach coefficient alpha was .867, indicating good internal consistency.23 Data Analysis To maximize the sample size with each variable measured, all usable data were analyzed.The inde-

296

pendent variables in this study represent an array of factors affecting the dependent variable. The data were analyzed using SAS 8.02 for Windows (SAS Institute, Inc., Cary, NC, 1999).

Findings Descriptive Analysis Results Age in the sample (3.472) ranged from 22 to 77 years with a mean age of 43 years and standard deviation of 10.29. The associate degree in nursing was the highest educational preparation of 46.81% (1619) of the participants, whereas 11.74% (406) had completed a diploma program in nursing, 31.74% (1098) had a bachelor’s degree, and 9.71% (336) had a master’s degree or higher. The participants’ years of service ranged from 0.07 to 43 years, with a mean of 8.29 and a standard deviation of 7.45. For the purpose of statistical analysis, the researcher grouped care setting into three categories—intensive care, nonintensive care,and administration—as suggested by Sowell.24 Of the respondents in this study, 50.84% (1,758) reported currently working in intensive care settings, 39.33% (1,360) in non-intensive care settings, and 9.83% (340) in administration. In terms of geographic area, 61.61% (1873) of the nurses worked in an area categorized as larger urban, 20.39% (620) worked in an area categorized as small urban, and 17.99% (547) in an area categorized as rural. The participants reported their overall job satisfaction scores ranging from 10 to 100 (actual range), with a mean score (SD 17.44) of 56.91. For about two thirds of the RNs, job satisfaction remained the same or had lessened over the past 2 years.

Inferential Statistics Results Correlation Among Age, Years of Service, and Overall Job Satisfaction The correlation coefficient (r) for the age and job satisfaction was 0.00178 (P ⫽ .9186); for the years of service and job satisfaction the correlation coefficient was -0.02946 (P ⫽ .0929). Neither of these results was statistically significant at the .05 level. Comparison of Job Satisfaction with Demographics Using ANOVAs Analysis of variance (ANOVA) demonstrated that job satisfaction varied significantly based upon years-ofservice, hospital-bed-size, care-setting, job position, hospital-retirement-plan, and geographic-area factors (Table 1). All tests are based on a 0.05 level of significance.

JONA • Vol. 33, No. 5 • May 2003

Table 1. The Results of One-way ANOVAs Variables

df

P

Age* Years of service* Level of education Salary Hospital bed size Care setting Job position Hospital retirement plan Teaching hospital status Geographic area

3 3 3 4 4 2 4 1 1 2

.9097 ⬍.0001† .4357 .6966 .0002† .0006† .0001† .0045 .7456 .0252†

*For these analyses, age and years of services were coded 4 level categoric variables †Statistically significant at P ⬍ .0

Backward Elimination Procedure To select the best predictor model of job satisfaction, the researcher used the backward elimination procedure, which starts with the largest model of interest and then reduces the model through a sequence of hypothesis tests. The results of this process, which removed the variable with the largest P-value each time, are described in Table 2. After the sixth step, P values of the remaining few variables were all statistically significant. Therefore, the backward elimination procedure was ended, and the selected model consisted of these four variables: years of service group, job position, hospital retirement plan, and geographic area (Table 3). The full model with all 10 variables explained 2.45% of the variance in overall job satisfaction, while the final model with only 4 variables explained 2.0% of the variance. The final model produced essentially the same results as the full model in identifying the variables that explain the greatest amount of variance in overall job satisfaction.

Table 2. The Results of the Backward Elimination Procedure No. of Variables Remaining in the Model 10 9 8 7 6 5

Variable Removed Age group Teaching hospital status Level of education Salary Care setting Hospital bed size

JONA • Vol. 33, No. 5 • May 2003

Table 3. Final Model After the Backward Elimination Procedure No. of Variables Remaining in the Model Variable Removed 4 4 4 4

Years of service Job position Hospital retirement plan Geographic area

P .0003* .0004* .0004* .0026*

*Statistically significant at P ⬍ .05

Discussion and Recommendations Approximately 60% of respondents indicated that their job satisfaction remained the same or that they became less satisfied over the past 2 years.This pattern showed that the job satisfaction among nurses had decreased in the past 2 years.This finding is consistent with the conclusion of the 2000 Nursing Executive Center’s survey25 that most nurses’ satisfaction had decreased in the past 2 years. In 1998, Cumbey and Alexander26 reported the use of the MMSS to measure the job satisfaction of public health nurses in a southeastern state. They found that overall scores on the MMSS ranged from 59 to 155 with a mean score of 114.78. In the present study (data collected in October 2000), the overall job satisfaction scores ranged from 10 to 100, with a mean score of 56.91.The standardized mean score of the MMSS was 74, while the standardized mean score of this study was 51. This comparison also indicated that job satisfaction has decreased in the past several years. The mean scores for each of the 10 components of job satisfaction ranged from 6.61 for competence of clinical staff to 4.88 for recognition, indicating that the nurses were most dissatisfied with their lack of recognition.The results of this study are also consistent with the findings of the 2000 Nursing Executive Center’s survey25 in which 48% of nurses who had held their current job more than one year were very or somewhat dissatisfied with the recognition they received.

P .9420 .8934 .8327 .4015 .2924 .1688

Implications for Nursing Practice This study indicates that about two thirds of nurses expressed that their job satisfaction remained the same or that they became less satisfied over the past 2 years.The result of job dissatisfaction is that nurses leave a place of work or leave the profession entirely.

297

The March 5, 2002 State House press conference,27 which highlighted the nursing shortage in SC, is a good illustration of the outcome of the job dissatisfaction among nurses.This pattern of decreasing satisfaction is a warning sign for nurse administrators who are looking for better strategies to retain their existing and new staff nurses. The implications of this finding are broad and complex because every single factor can affect the job satisfaction of nurses to some degree. Clearly, an implication for nurse managers is to challenge experienced nurses. Nurses who were experienced (over 2 years) had lower levels of job satisfaction with their jobs than those who were inexperienced (0-2 years). According to Decker,28 the longer nurses worked on one particular unit, the greater the likelihood they were dissatisfied with their job and experiencing psychological distress. Nurses with more years of service and commitment to the organization tend to expect more autonomy, recognition, and opportunities. When these factors do not exist, experienced nurses may feel upset and become dissatisfied. The present study demonstrates that job position is a factor that affects the job satisfaction of nurses. Nurses whose job positions were other than staff nurse, charge nurse, clinical nurse specialist, and nurse manager had the highest levels of job satisfaction. Positions within hospitals (eg, utilization review specialist) and those outside hospitals (eg, home health nurse, school nurse, and pharmaceutical company representative) offer more flexible hours and higher compensation and are less stressful than hospital clinical positions.Therefore, nurses who worked in those positions had higher levels of job satisfaction. An important finding of this study is that nurses whose primary place of work was in a small urban area had the highest level of job satisfaction. This finding is different from Mills and Blaesing,18 who found urban nurses to be more satisfied. A reason for this difference may be that nurses in smaller urban hospitals may do different things than their counterparts in larger urban hospitals; time is spent on different tasks, jobs are structured differently, traveling times to work are different, and culture and interpersonal relationships with coworkers and supervisors may be different. Perhaps these differences lead to greater job satisfaction. Furthermore, nurses whose hospitals participated in the SC Retirement System plan had a higher level of job satisfaction than that of nurses whose hospitals did not participate in the retirement sys-

298

tem plan. Surprisingly, about 84% of hospitals did not participate in the SC Retirement System plan. Whether hospitals offer the SC Retirement System plan for their nurses is a factor affecting job satisfaction and also may be a predictor of nurse intention to leave the job. Thus, nurses may voluntarily leave their jobs earlier than hospitals expect.

Areas for Additional Research • This study can serve as a “road-map” for policy makers in SC. A longitudinal study would be useful with data collection taking place at different time intervals (eg, every 2 years). Providing incentives for respondents and conducting telephone sampling or face-to-face interviews may increase the response rate of nurses. Continuing the study over a 6-year period to observe for change in job satisfaction is also recommended. • Future studies of the job satisfaction of nurses need to examine the generalizability of the present findings to more diverse groups such as racial groups, Licensed Practical Nurses (LPNs), or nurses’ aides.

Recommendations and Policy Implications for Nurse and Health Administrators • Hospital and nursing administrators should consider offering a retirement system plan and benefits similar to those in the SC Retirement System plan to improve the work environment. • Administrators should spend more time in communicating with their nursing staffs. Hospital administrators need to talk to the nurses, especially staff nurses and charge nurses, regarding what is happening in their organization and what needs to happen to keep nurses satisfied over the long-term. • Administrators should consider the issues raised in this study. As suggested by other authors,29 special attention should be directed to the concerns of experienced nurses and the issues of benefits and location of working institutions. • Administrators must consider changing the work context, especially concerning salary and promotions policies, in ways that can contribute to job satisfaction and maintain a viable nursing workforce for the future.

JONA • Vol. 33, No. 5 • May 2003

• Administrators should create financial incentives for nurses who agree to work in needy small urban areas and establish educational funds such as tuition reimbursement and scholarships to promote recruitment and retention of nurses. Knowing what incentives will do the most to increase the level of job satisfaction of nurses may help hospital administrators make decisions about the expenditure of scarce resources.

Conclusion The unique findings of this study suggest that administrators should consider revisions in retirement programs to develop institutional loyalty and longterm financial security as a key element of a recruitment and retention strategy. Adding to the growing body of literature about job satisfaction among nurses, this research is particularly relevant to healthcare administrators in facing the possibility of long-term nursing shortages.

References 1. Davidson H, Folcarell P, Crawford S, Duprat L, Clifford J. The effects of health care reforms on job satisfaction and voluntary turnover among hospital-based nurses. Med Care. 1997;35(6):634-645. 2. Loquist R, Pease M. Critical for Care: The SC Nursing Workforce, 2001 and Beyond. Columbia, SC: University of South Carolina; 2001. 3. Locke EA. What is job satisfaction? Organ Behav Hum Performance. 1969;4:309-336. 4. Smith P, Kendall LM, Hulin CL. The Measurement of Satisfaction in Work and Retirement. Chicago, Ill: Rand McNally; 1975. 5. Pilkington W, Wood J. Job satisfaction, role conflict and role ambiguity—a study of hospital nurses. Aust J Adv Nurs. 1986;3(3):3-13. 6. Geiger JW, Davit JS. Self-image and job satisfaction in varied settings. Nurs Manage. 1988;19(12):50-56. 7. Robbins S. Organizational Behavior: Concepts, Controversies, and Applications. 3rd ed. Englewood Cliffs, NJ: Prentice-Hall; 1986:139-143. 8. Muller CW, McCloskey JC. Nurses’ job satisfaction: a proposed measure. Nurs Res.1990;39(2):113-117. 9. Stamps PL. Nurses and Work Satisfaction: An Index for Measurement. 2nd ed. Chicago, Ill: Health Administration Press; 1997. 10. Nahm H. Job satisfaction in nursing. Am J Nurs. 1940;40(12):1389-1392. 11. Maslow A. Motivation and Personality. New York: Harper & Row; 1954. 12. Herzberg F. Work and the Nature of Man. Cleveland, Ohio: World Publishing; 1966. 13. Vroom VH. Work and Motivation. New York: Wiley; 1964. 14. Stamps P, Piedmonte EB, Slavitt DB, Hase AM. Measurement of work satisfaction among health professionals. Med Care. 1978;16:337-352. 15. McClure ML, Poulin MA, Sovie MD, Wandelt MA. Magnet Hospitals: Attraction and Retention of Professional Nurses. Kansas City, Mo: American Nurses Association; 1983.

JONA • Vol. 33, No. 5 • May 2003

16. Kangas S, Kee CC, McKee-Waddle R. Organizational factors, nurses’ job satisfaction, and patient satisfaction with nursing care. J Nurs Adm. 1999;29(1):32-42. 17. Kangas SH. The New Health Care Environment: A Research Synthesis of Impact on Nurses’ Job Satisfaction [dissertation]. Atlanta, Ga: Georgia State University; 1999. 18. Mills A, Blaesing S. A lesson from the last nursing shortage. J Nurs Adm. 2000:30(6):309-315. 19. Hopkins AH. Work and Job Satisfaction in the Public Sector. Totowa, NJ: Rowan & Allanheld; 1983. 20. Leslie L. Are high response rates essential to valid surveys? Soc Sci Res. 1972;1:323-334. 21. Armstrong JS, Overton T. Estimating Nonresponse Bias in Mail Surveys, Reading in Survey Research. Chicago, Ill: American Marketing Association; 1978. 22. Stratton TD, Dunkin JW, Juhl N, Geller JM. Retainment incentives in three rural practice settings: variations in job satisfaction among staff registered nurses. Appl Nurs Res. 1995;8(2):73-80. 23. Devellis, R. F. Scale Development: Theory and Applications. Newbury Park, Calif: Sage; 1991. 24. Sowell RL. A Re-Evaluation of the Relationship Between Role Conflict and Role Ambiguity and Job Satisfaction Among Registered Nurses [master’s thesis]. Augusta, Ga: Medical College of Georgia; 1983. 25. Nursing Executive Center. Reversing the Flight of Talent. Washington, DC: The Advisory Board Company; 2000. 26. Cumbey DA, Alexander JW. The relationship of job satisfaction with organizational variables in public health nursing. J Nurs Adm. 1998;28(5):39-46. 27. Lamb LH. Study recommends ways to ease nursing shortage. The State. March 5, 2002: A1. 28. Decker FH. Occupational and nonoccupational factors in job satisfaction and psychological distress among nurses. Res Nurs Health. 1997;20:453-464. 29. Ingersoll GL, Olsan T, Drew-Cates J, DeVinney BC, Davies J. Nurses’ job satisfaction, organizational commitment, and career intent. J Nurs Adm. 2002;32(5):250-263.

299