Job strain and determinants in staff working in institutions for people with intellectual disabilities in Taiwan: A test of the Job Demand-Control-Support model

(1)

Job strain and determinants in staff working in institutions for people with intellectual disabilities in Taiwan: A test of the Job

Demand-Control-Support model

Jin-Ding Lin

^a,

* , Tzong-Nan Lee

^a

, Chia-Feng Yen

^b

, Ching-Hui Loh

^c

, Shang-Wei Hsu

^d

, Jia-Ling Wu

^e

, Cordia M. Chu

^f

aSchool of Public Health, National Defense Medical Center, Taipei, Taiwan

bGraduate Institute of Life Sciences, National Defense Medical Center, Taipei, Taiwan

cDepartment of Community and Family Medicine, Tri-Service General Hospital, Taipei, Taiwan

dGraduate Institute of Healthcare Administration, Asia University, Taichung, Taiwan

eResearch Center for Intellectual Disabilities Taiwan, Chung-Hua Foundation for Persons with Intellectual Disabilities, Taipei County, Taiwan

fCenter for Environment and Population Health, Griffith University, Brisbane, Australia Received 15 October 2007; received in revised form 28 January 2008; accepted 12 February 2008

Abstract

Little is known about the job strain of staff working in disability institutions. This study investigated the staff’s job strain profile and its determinants which included the worker characteristics and the psychosocial working environments in Taiwan. A cross-sectional study survey was carried out among 1243 workers by means of a self-answered questionnaire. The outcome variable (high-strain job) was evaluated. The explanatory variables were: worker characteristics and the psychosocial working environment evaluated according to Karasek’s Job Demand-Control-Support model. The results show that many staff characteristics were correlated with job strain, such as staff’s working hours, age, gender, job title, educational level, religion, in-job training, working years in disability institutions and Effort–Reward Imbalance factors.

Organization factors, such as geographical, institutional ownership and accreditation performance and size were also correlated with staff’s job strain. In multiple a logistic regression model of the job strain, we found that the factors of financial reward (high compare to low, OR = 0.95, 95% CI = 0.928–0.975), extrinsic effort (high compare to low, OR = 1.072, 95% CI = 1.072–1.158), perceived job stress (sometimes stressful compare to no stress, OR = 2.305, 95% CI = 1.161–4.575; very stressful compare to no stress, OR = 3.931,

Available online at www.sciencedirect.com

Research in Developmental Disabilities 30 (2009) 146–157

* Corresponding author at: No. 161, Min-Chun E. Road, Section 6, Nei-Hu, Taipei 114, Taiwan.

Tel.: +886 2 8792 3100x18447; fax: +886 2 8792 3147.

E-mail address:a530706@ndmctsgh.edu.tw(J.-D. Lin).

doi:10.1016/j.ridd.2008.02.001

(2)

95% CI = 1.738–8.893) of the staff were significantly correlated to the high job strain of the staff. An important focus of future research should be extending the findings to consider the factors to affect the high job strain to improve the well-being for staff working for people with intellectual disability.

Keywords: Effort–Reward Imbalance model; Job Demand-Control-Support model; Job strain; Staff; Stress

1. Introduction

Staff plays a vital role in the provision and they affect significantly the quality of services for people with intellectual disabilities (ID) (Rose, David, & Jones, 2003), many studies found that the staff is often experience poor morale, high stress, burnout and high turnover rate (Attwood &

Joachim, 1994;Hatton et al., 1999;Larson & Lakin, 1992). However,Skirrow and Hatton (2007) employed a systematic review found that staff in services for adults with ID may not be at increased risk of burnout compared to staff in other services, a significant minority may experience high rates of burnout and this appears to be linked to issues of service delivery and work support. Workplace stress (job strain) clearly is important for staff turnover (Hatton et al., 2001). According toRose and Rose’s (2005)reviews, a number of factors have been implicated in influencing the levels of stress. These include perceived support, staff coping styles and emotional reactions to challenging behavior, the organization and role definition and service user characteristics.Hatton et al. (2001)revealed that work satisfaction, job strain, younger staff age and easier subjective labor conditions were directly associated with intended turnover. Other studies, such asRobertson et al. (2005)found that the greatest perceived sources of stress were lack of resources and lack of staff support. The lowest level of satisfaction was the rate of pay.

They also reported that high levels of intended staff turnover may be more due to job insecurity and lack of support than service user challenging behavior.

Little is known about the working stress profile of staff in disability institutions in Taiwan. An adequate understanding of stress profile and its determinants is important to prevent chronic health problems and absenteeism of staff. TheKarasek’s (1979)model of Job Demand-Control- Support (JDCS) has been adopted in analyzing psychological demands and decision latitude to the jobs. This model has dominated research on working stress for more then two decades (van der Doef & Maes, 1999). Four distinctly different kinds of psychosocial work experience are generated by the interactions of high and low levels of psychological demands and decision latitude: high-strain jobs, active jobs, low-strain jobs, and passive jobs (Karasek’s 1979). High- strain job is that the most adverse reactions of psychological strain (fatigue, anxiety, depression, and physical illness) occur when psychological demands of the job are high and the worker’s decision latitude in the task is low. Active job, typical of professional work, call for the control is high and psychological demand is also high. Such situations, while intensely demanding, involve workers in activities over which they feel a large measure of control, the freedom to use all available skills. Low-strain jobs, situations with few psychological demands and high levels of control. The final kind of the model is a Passive Job represents situations of low demand and low control (Karasek & Theorell, 1990). There were lacking the complete theoretical framework of job strain and determinants in the previous studies on the job strain of the staff for people with ID.

It is difficult to review the general picture of job strain for people working in disability arenas. In addition, the issues of health and well-being of the staff for people with ID are being ignored traditionally in Taiwan society. The study objectives were to describe the stress profile and its

(3)

determinants of staff working in the disability institutions by using Karasek’s JDCS model in Taiwan.

2. Method

2.1. Subjects and methods

The study entire population was composed of 7466 staff who was working in all 244 registered intellectual disability institutions in 2005 in Taiwan. With regard to subject sampling, we used the Marks’s theory (1982) to calculate the minimum meaningful subjects requirement, the least sample size was six times of variables presented in questionnaire of the survey. Due to the nature of low response rate (anticipated 60%) in the field of disability study and mail survey, so we over sampled to obtain meaningful statistical data. In survey research, the method of over sampled is a good alternative to solve the low response rate (Bryman, 2001). From his theory, the present study needs to recruit 1010 subjects at least to achieve meaningful statistical data [(101 variables 6)/60% = 1010]. Subjects were stratified by administrative area (North, South, East and Central Taiwan) selected systematically according to the proportion of the institutions in each area for the study. As choosing the institutions, we mailed the consent letter to the institutions for the survey.

Data were collected by a mail-structured questionnaire that was completed by the institutional staff. The authors mailed out the questionnaires to the participating institutions and the institutional coordinator distributed them to their staff. The survey covered a 3-month period from December 2005 to February 2006. In an attempt to increase the response rate, the questionnaire sent to each respondent was accompanied by a gift of thanks for their cooperation.

In addition, the researchers administering the study also phoned each coordinator in institutions, reminding them to mail back the questionnaire within the data collection period. The questionnaire was specifically designed and, to improve its validity, was reviewed by five experts in the field of family medicine, health policy, epidemiology, nursing and social welfare. The questionnaire concerned demographics, institutional characteristics, Job Content Questionnaire (JCQ), Effort–Reward Imbalance Questionnaire (ERI) and self-perceived heath status. Among the questionnaire, we used the Chinese JCQ and ERI questionnaires which were developed by Professor Cheng Y.W., she and her colleague indicated that Chinese Version of the JCQ and ERI is reliable and valid for assessing psychosocial work conditions among Taiwanese workers (Cheng, Luh, & Guo, 2003). The present study has received the approval to use the Chinese JCQ and ERI questionnaire in 2005. The data were entered into a database and analyzed using SPSS 11.0 software.

3. Results

A total of 1629 questionnaires were mailed to the staff among 24 institutions, 1250 were response, giving a return rate of 76.73%. Of these, seven individuals did not meet eligibility criteria which the response questionnaire were incompletely (less than 70%), and they were excluded from the present study. The final sample consisted of 1243 staff were recruited in the analysis.

The study will use percent or mean (include S.D.) to analyze demographic characteristics working hours and working history of the subjects. The demographic characteristics of the respondents were shown inTables 1 and 2. There were 81.1% female respondents and 18.9%

J.-D. Lin et al. / Research in Developmental Disabilities 30 (2009) 146–157 148

(4)

were male. The average age of staff was 38.61 years (S.D. = 10.18), they work 45 h (S.D. = 8.65) per week and six and half working years on average in institutions. Most of the respondents were first-line workers, such as teaching and nursery workers (37.8%), nearly 60% had college or university educational level and there were 34.8% unmarried staff. In the previous job experience, 81.8% did not have working experience before which means the present job was his/

her first job in their working career. In term of the in-job training, 78.6% felt adequately and 15.7% felt inadequate and 5.7% did not accept any in-job training in institutions.

We used mean, minimum/maximum value and standard deviation to describe the distribution of Demand-Control-Support and Effort–Reward Imbalance conditions of the subjects.

Distribution of Demand-Control-Support and Effort–Reward Imbalance model shown in Table 3. Job control, psychological demand and work-related social support were main domains of Demand-Control-Support model. In-job control, the mean score was 63.1 (S.D. = 8.66), the minimum was 28 and the maximum was 96. The median score (=64) of social support is the cut point of high and low job control. Among the domains of job control, skill discretion and decision authority were the two main indicators. The mean of skill discretion and decision authority were 32.04 (S.D. = 4.21) and 31.08 (S.D. = 5.67) while the mean of psychological demand was 30.38 (S.D. = 4.53). Our median score (=30) was divided into high and low on psychological demands of the job. Work-related social support was 23.99 (S.D. = 3.04), co-worker support was 12.34 (S.D. = 1.58) and supervisor support was 11.66 (S.D. = 1.93). The first 25 percentile (score = 24) was the cut point of work-related social support.

With regard to the Effort–Reward Imbalance model, there were three main indicators:

overcommitment, extrinsic effort and reward. The results showed that the average score of overcommitment was 15.68 (S.D. = 2.57), extrinsic effort was 17.31 (S.D. = 6.27), and reward was 21.98 (S.D. = 8.51). There were three different sub indicators of reward, self-esteem (score = 21.98, S.D. = 3.96), social control (score = 20.33, S.D. = 4.48) and financial rewards (score = 4.26, S.D. = 1.11).

The methods used to analyze the associations of Job Demand-Control-Support/Effort–

Reward Imbalance model and subject characteristics was ANOVA. Associations of Job Demand-Control-Support model and staff characteristics were shown inTables 4 and 5. The job strain was divided into four groups: passive job, low-strain job, active job and high-strain job. Participant characteristics of ‘weekly working hours’, ‘job reward’, ‘extrinsic effort’,

‘overcommitment to the job’, ‘job support’, ‘job category (title)’, ‘gender’, ‘levels of education’, ‘religious status’ and ‘in-job training’ were statistically correlated to the job strain types. We also analyzed the associations of job strain types and organizational characteristics, the factors of geographic location, ownership, accreditation and the size of an institution, health status of the staff, satisfaction to the job, perceived job stress and willing to continue to the present job were statistical significant to the job strain of the staff (Tables 6 and 7). The job strain types also related significantly to the factors of the Effort–Reward Imbalance model (Table 8).

Table 1

Age, working hours and working history of the subjects

Category (N = 1243) N Min Max Mean S.D.

Age 1197 19 64 38.61 10.18

Working hours (per week) 1116 20 104 45.02 8.65

Work history of the present job (month) 1194 1 415 79.89 75.78

(5)

In the multiple logistic regression model of the high-strain job, we found that the factors of financial reward (high reward compare to low reward, OR = 0.95, 95% CI = 0.928–0.975), extrinsic effort (high effort compare to low effort, OR = 1.072, 95% CI = 1.072–1.158), perceived job stress (felt sometimes stressful compare to no stress at all, OR = 2.305, 95%

Table 2

Demographic characteristics of the subjects

Category N Percent

Gender (N = 1238)

Male 234 18.9

Female 1004 81.1

Job title (N = 1203)

Admin workers 149 12.4

Social worker 89 7.4

Nurses 45 3.7

Teaching and nursery 455 37.8

Counselor 147 12.2

Living carers 150 12.5

Substitute military 46 3.8

Other 122 10.1

Levels of education (N = 1229)

Less than primary school 72 5.9

Junior high school 73 5.9

Senior high school 354 28.8

College 306 24.9

University 382 31.1

Masters 42 3.4

Marital status (N = 1225)

Unmarried 426 34.8

Married 705 57.5

Divorced 48 3.9

Separated 11 0.9

Widow/-er 35 2.9

Religion (N = 1204)

Buddhism 378 31.4

Tao 243 20.2

I-Kuan Tao 40 3.3

Christian 111 9.2

Catholic 47 3.9

Muslim 3 0.2

Non-specific 370 30.7

Other 12 1.0

In-job training (N = 1213)

No 69 5.7

Yes, adequate 954 78.6

Yes, inadequate 190 15.7

Previous job experience (N = 1222)

No 1000 81.8

Yes, similarity 147 12.0

Yes, different 75 6.1

(6)

CI = 1.161–4.575; very stressful compare to no stress at all, OR = 3.931, 95% CI = 1.738–8.893) of the staff were significantly correlated to the job strain of the staff (Table 9).

4. Discussions

The present study attempts to describe the job strain and its determinants of staff working in disability institutions. Job strain clearly is important for staff turnover. Adverse job conditions may lead to the development of anxiety and depressive symptoms (Siegrist, 1996; Turner, Wheaton, & Lloyd, 1995). Therefore, the assessment of psychosocial work environment may identify workers at risk, and serve as a basis for job-redesign (Sanne, Mykletun, Dahl, Moen, &

Tell, 2005). Results of the present study showed that most staff working in disability institutions was female (81.1%) and the present job was their first job in their working career (81.8%).

Escriba-Aguir and Tenias-Burillo (2004)state that gender role has a negative influence on the psychological well-being of hospital staff. Hatton et al. (1999)also revealed that the wishful thinking, stress linked to work-home conflict and role ambiguity will affect the general distress of the staff. Support is the most strongly associated with anxiety and depression in women (Sanne et al., 2005). Therefore, the working environments in the disability institutions should review the supportive health and welfare services to protect the majority workforce–women staff to improve their psychological well-being in disability services. There were over one-fifth staff felt no or inadequately toward the in-job training in the present study. The staff’s training in appropriate methods for dealing with challenging behavior of people with intellectual disabilities are necessary in institutions (Robertson et al., 2005). Due to most staff in disability institutions were their first career job, the authorities should focus on and examine the issue of in-job training to satisfy the needs of staff working in institutions.

Many staff characteristics were correlated with job strain in the study, such as staff’s working hours, age, gender, job title, educational level, religion, in-job training, working years in disability institutions and Effort–Reward Imbalance factors. Organization factors, such as geographic location, institutional ownership and accreditation performance and size were also correlated with staff’s job strain. AsHatton et al. (2001)stated that organizations can attempt to

Table 3

Distribution of Demand-Control-Support and Effort–Reward Imbalance model

Category (N = 1243) N Min Max Mean S.D.

Demand-Control-Support model

Job control 1216 28 96 63.13 8.66

Skill discretion 1223 14 48 32.04 4.21

Decision authority 1227 12 48 31.08 5.67

Psychological demand 1220 18 48 30.38 4.53

Work-related social support 1232 8 32 23.99 3.04

Co-worker support 1236 4 16 12.34 1.58

Supervisor support 1232 4 16 11.66 1.93

Effort–Reward Imbalance model

Overcommitment 1227 7 24 15.68 2.57

Extrinsic effort 1214 7 35 17.31 6.27

Reward 1211 11 55 46.54 8.51

Self-esteem 1220 5 25 21.98 3.96

Social control 1225 5 25 20.33 4.48

Financial rewards 1237 1 5 4.26 1.11

(7)

J.-D.Linetal./ResearchinDevelopmentalDisabilities30(2009)146–157152

Table 4

Associations of Job Demand-Control-Support model and staff characteristics (I) Variables Job strain types (N = 1243)

Passive job Low-strain job Active job High-strain job F P Scheffe

N Mean S.D. N Mean S.D. N Mean S.D. N Mean S.D.

Work hours/p.w. 310 44.68 9.33 338 43.99 6.86 181 45.57 8.02 278 46.27 9.94 3.987 .008 4 > 2

Age 323 39.26 10.77 367 39.63 10.60 278 37.36 9.30 302 37.46 9.38 3.967 .008 N.S.

Working years 321 83.45 79.78 366 83.03 78.42 195 66.20 76.86 304 81.94 76.86 2.639 .048 N.S.

Job experiences 319 6.2 25.51 371 8.16 29.16 195 9.69 30.69 298 8.11 25.95 .684 .562

Job reward 336 47.17 7.43 366 49.72 6.84 193 48.13 7.3 303 41.02 9.52 73.892 <0.001 4 < 123

Extrinsic effort 335 14.95 5.18 370 14.51 5.12 200 19.9 5.68 308 21.56 5.96 130.280 <0.001 4 > 123 Overcommitment 337 15.02 2.35 376 15.14 2.35 194 16.44 2.59 309 16.59 2.66 33.958 <0.001 3 > 12,4 > 12

Job support 341 23.71 2.74 373 24.91 3.00 197 24.76 2.95 312 22.71 2.98 37.860 <0.001 4 < 123

(8)

reduce job strain by: fostering appropriate practical coping skills of staff, encouraging a positive commitment to the organization, making staff roles clear, increasing the support available to staff, considering streamlining bureaucratic procedures where this is consistent with service quality, and increasing the control which staff have. The health professionals should be aware of perceived adverse psychosocial work environment as a potential risk factor for anxiety and depression of staff (Sanne et al., 2005).

Finally, multiple logistic regression model was used to evaluate the factors associated with high job strain in the study. This model illustrated that the factor of reward, extrinsic effort and self-perceived job stress were correlated with staff’s high-strain job.Robertson et al. (2005) suggested that employers seeking to reduce turnover or stress of the staff should pay attention to basic pay and conditions. High demands, low control and low support individually, but particularly combined, are risk factors for anxiety and depression. In addition, social relationship

Table 5

Associations of Job Demand-Control-Support model and staff characteristics (II)

Variable Job strain model (N = 1243)

Passive Low strain Active High strain x² P-value

N (%) N (%) N (%) N (%)

Job title (N = 1151) 13.173 0.04

Admin staff 34 2.8 52 4.4 27 2.3 36 149

Front line worker 239 20 250 20.9 150 12.6 239 878

Other 53 4.4 62 5.2 22 1.8 30 167

Gender (N = 1227) 9.539 0.023

Male 76 6.2 81 6.6 32 2.6 45 3.7

Female 263 21.4 295 24 168 13.7 267 21.8

Levels of education (N = 1219) 48.410 <0.001

Less than junior high 63 5.2 39 3.2 14 1.1 27 2.2

Senior high 120 9.8 109 8.9 40 3.3 82 6.7

College/university 156 12.8 226 18.5 144 11.8 199 16.3

Marital status (N = 1214) 11.155 0.084

Unmarried 106 8.7 116 9.6 78 6.4 124 10.2

Married 198 16.3 231 19 107 8.8 22 13.3

Other 31 2.6 27 2.2 13 1.1 307 1.8

Religion (N = 1193) 33.195 0.001

Buddhism 112 9.4 122 10.2 45 3.8 97 8.1

Dao 79 6.6 76 6.4 27 2.3 61 5.1

Christian/Catholic 36 3 40 3.4 38 3.2 38 3.2

No-specific 86 7.2 112 9.4 71 6 99 8.3

Other 14 1.2 16 1.3 16 1.3 8 0.7

Job experience (N = 1212) 9.084 0.169

No 280 23.1 314 25.9 153 12.6 246 20.3

Yes, same type 31 2.6 43 3.5 28 2.3 43 3.5

Yes, different type 22 1.8 16 1.3 17 1.4 19 1.6

In-job training (N = 1205) 14.322 0.026

No 25 2.1 23 1.9 8 0.7 13 1.1

Yes, adequate 259 21.5 303 25.1 152 12.6 234 19.4

Yes, inadequate 48 4 42 3.5 38 3.2 60 5

(9)

in the workplace accounted for a great variability of job satisfaction in Taiwanese societies (Cheng et al., 2003). Staff support was significantly correlated with psychological outcomes, with low levels of support being associated with high levels of stress (Harris & Thomson, 1993;

Rose et al., 2003). Positive changes in perceived social support, decision latitude, and

Table 6

Associations of Job Demand-Control-Support model and organizational characteristics

Variable Job strain types (N = 1243)

N (%) N (%) N (%) N (%)

Location (N = 1232) 18.122 0.006

North Taiwan 104 8.4 117 9.5 61 5 94 7.6

Central Taiwan 127 10.3 155 12.6 52 4.2 113 9.2

South and East Taiwan 111 9 105 8.5 87 7.1 106 8.6

Ownership (N = 1232) 46.395 <0.001

Public 166 13.5 130 10.6 52 4.2 160 13

Private 176 14.3 247 20 148 12 153 12.4

Accreditation (N = 1171) 19.492 <0.001

Excellent (Grade A) 125 10.7 104 8.9 63 5.4 134 11.4

Good (Grade B) and the less (Grades C and D)

197 16.8 255 21.8 128 10.9 165 14.1

Size (N = 1232) 16.452 0.001

Large (3100 persons) 297 24.1 321 26.1 157 12.7 285 23.1

Other (<100 persons) 45 3.7 56 4.5 43 3.5 28 2.3

Table 7

Associations of Job Demand-Control-Support model and job characteristics

N (%) N (%) N (%) N (%)

Health status 36.504 <0.001

Better than before 48 3.9 75 6.1 34 2.8 51 4.1

Similarity 187 15.2 214 17.4 87 7.1 130 10.6

Worse than before 107 8.7 87 7.1 79 6.4 131 10.7

Satisfaction to the job 104.86 <0.001

Satisfactory 223 18.4 296 24.4 131 10.8 136 11.2

Neutral 92 7.6 63 5.2 46 3.8 117 9.7

Unsatisfactory 22 1.8 14 1.2 16 1.3 56 4.6

Willing to Continue the present job 81.254 <0.001

Continue 207 17.1 278 22.9 121 10 135 11.1

Neutral 89 7.3 71 5.9 61 5 109 9

Discontinue 41 3.4 24 2 12 1 65 5.4

Perceived job stress 161.4 <0.001

No stress 93 7.7 103 8.5 18 1.5 17 1.4

Sometimes 217 17.9 230 18.9 118 9.7 180 14.8

Very stressful 26 2.1 41 3.4 59 4.9 112 9.2

(10)

psychological job demands went together with a decrease in fatigue, emotional exhaustion and psychological distress (Janssen & Nijhuis, 2004).

This study has many limitations that should be acknowledged. First, healthy worker effect will possible affect the results of the study. Moreover, our data were collected during the season of Chinese New Year and the Sixth Disability Welfare Institution Accreditation of Taiwan, these events might increase the job strain. However, the present paper was the first study to describe the job strain and determinants of staff working in disability institutions in Taiwan. It gives the health welfare professionals a general picture of job strain toward staff working for people with ID. An important focus of future research should be extending the findings to consider the factors to affect the high job strain to improve the well-being for staff working for people with ID.

Table 9

Factors associated with high job strain in multiple logistic regression model

Variable Reference b OR 95% CI

Working hours per week .001 .999 .979 1.019

Age .008 1.008 .981 1.036

Working years .001 .999 .996 1.002

BMI .022 .978 .923 1.037

Financial reward .050 .951 .928 .975

Extrinsic effort .108 1.114 1.072 1.158

Overcommitment .000 1.000 .919 1.088

Job support .170 .844 .783 .910

Job title Admin worker

First-line workers .026 1.026 .586 1.797

Other .273 .761 .350 1.655

Gender Male

Female .007 .993 .576 1.714

Table 8

Associations of Job Demand-Control-Support and Effort–Reward Imbalance model

N (%) N (%) N (%) N (%)

Effort status (N = 1207) 220.15 <0.001

Low effort 224 18.6 262 21.7 64 5.3 68 5.6

High effort 111 9.2 108 8.9 130 10.8 240 19.9

Job commitment (N = 1219) 64.029 <0.001

Non-overcommitment 244 20 263 21.6 98 8 148 12.1

Overcommitment 93 7.6 113 9.3 99 8.1 161 13.2

Reward status (N = 1205) 113.06 <0.001

Low reward 151 12.5 110 9.1 79 6.6 213 17.7

High reward 185 15.4 256 21.2 121 10 90 7.5

Job support (N = 1224) 48.979 <0.001

Low support 293 23.9 264 21.6 137 11.2 272 22.2

High support 48 3.9 109 8.9 61 5 40 3.3

(11)

Acknowledgements

Authors would like to acknowledge the in-part grant support from the Chung-Hua Foundation for Persons with Intellectual Disabilities, Taipei County, Taiwan for this research and those disability institutional staff who participated in the study. In addition, the authors wants to acknowledge the kindness of professor Cheng Y.W., agreed to use the Chinese Version of the JCQ and ERI scales in the study.

References

Attwood, T., & Joachim, R. (1994). The prevention and management of seriously disruptive behaviour in Australia. In N.

Bouras (Ed.), Mental health in mental retardation: Recent advances and practices (pp. 365–374). Cambridge:

Cambridge University Press.

Table 9 (Continued )

Variable Reference b OR 95% CI

Level of education Less than junior high

Senior high school .152 .859 .410 1.803

College/university .334 .716 .333 1.541

Religion Non-specific

Buddhism .008 1.008 .608 1.671

Dao .191 .826 .477 1.430

Christian/Catholic .131 1.140 .609 2.134

Other .991 .371 .122 1.129

In-job training No

Yes, adequate .142 .868 .378 1.991

Yes, inadequate .330 .719 .288 1.796

Institution location North

Central .057 1.058 .637 1.759

East and south .048 .953 .531 1.711

Institution ownership Public

Private .591 .554 .348 .883

Institution accreditation A Grade

B Grade and the less .098 .907 .569 1.445

Size of the institution Large

Non-large .196 .822 .416 1.623

Health status Better than before

Similarity .357 .699 .404 1.212

Worse than before .454 .635 .356 1.134

Job satisfaction Satisfactory

Neutral .126 1.135 .705 1.826

Unsatisfactory .379 .685 .328 1.431

Job continuity Discontinue

Neutral .034 1.034 .654 1.636

Continuity .049 1.051 .552 1.998

Perceived job stress No stress

Sometimes .835 2.305 1.161 4.575

Very stressful 1.369 3.931 1.738 8.893

(12)

Bryman, A. (2001). Social research methods. New York: Oxford University Press.

Cheng, Y., Luh, W. M., & Guo, Y. L. (2003). Reliability and validity of the Chinese Version of the Job Content Questionnaire in Taiwanese Workers. International Journal of Behavioral Medicine, 10, 15–30.

Escriba-Aguir, V., & Tenias-Burillo, J. M. (2004). Psychological well-being among hospital personnel: The role of family demands and psychosocial work environment. International Achieves of Occupational and Environmental Health, 77, 401–408.

Harris, P., & Thomson, G. (1993). The staff support questionnaire—a means of measuring support among staff working with people with challenging behavior. Mental Handicap, 21, 122–127.

Hatton, C., Emerson, E., Rivers, M., Mason, H., Swarbrick, R., Mason, L., et al. (2001). Factors associated with intended staff turnover and job search behavior in services for people with intellectual disability. Journal of Intellectual Disability Research, 45, 258–270.

Hatton, C., Rivers, M., Mason, H., Mason, L., Kiernan, C., Emerson, E., et al. (1999). Staff stressors and staff outcomes in services for adults with intellectual disabilities: The staff stressor questionnaire. Research in Developmental Disabilities, 20, 269–285.

Janssen, N., & Nijhuis, E. J. (2004). Associations between positive changes in perceived work characteristics and changes in fatigue. Journal of Occupational and Environmental Medicine, 46, 866–875.

Karasek, R. A. (1979). Job demands, job decision latitude, and mental strains: Implications for job redesign. Admin- istrative Science Quarterly, 24, 285–307.

Karasek, R., & Theorell, T. (1990). Health work: Stress, productivity, and the reconstruction of working life. New York:

Basic Book.

Larson, S. A., & Lakin, K. C. (1992). Direct-care staff stability in a national sample of small group homes. Mental Retardation, 30, 13–22.

Marks, R. G. (1982). Designing a research project. Belmont: Wadsworth.

Robertson, J., Hatton, C., Felce, D., Meek, A., Carr, D., Knapp, M., et al. (2005). Staff stress and morale in community- based settings for people with intellectual disabilities and challenging behavior: A brief report. Journal of Applied Research in Intellectual Disabilities, 18, 271–277.

Rose, R., David, G., & Jones, C. (2003). Staff who work with people who have intellectual disabilities: The importance of personality. Journal of Applied Research in Intellectual Disabilities, 16, 267–277.

Rose, D., & Rose, J. (2005). Staff in service for people with intellectual disabilities: The impact of stress on attributions of challenging behavior. Journal of Intellectual Disability Research, 49, 827–838.

Sanne, B., Mykletun, A., Dahl, A. A., Moen, B. E., & Tell, G. S. (2005). Testing the Job Demand-Control-Support model with anxiety and depression as outcomes: The Hordaland Health Study. Occupational Medicine, 55, 463–473.

Siegrist, J. (1996). Adverse health effects of high-effort/low-reward conditions. Journal of Occupational Health Psychology, 1, 27–41.

Skirrow, P., & Hatton, C. (2007). ‘Burnout’ among director care workers in services for adults with intellectual disabilities: A systematic review of research findings and initial normative data. Journal of Applied Research in Intellectual Disabilities, 20, 131–144.

Turner, R. J., Wheaton, B., & Lloyd, D. A. (1995). The epidemiology of social stress. American Sociological Review, 60, 104–125.

van der Doef, M., & Maes, S. (1999). The Job Demand-Control (-Support) Model and psychological well-being: A review of 20 years of empirical research. Work Stress, 13, 87–114.