The impact of locus of control on job stress, job performance and job satisfaction in Taiwan

The Authors

Jui-Chen Chen, Department of Finance and Banking, Cheng Shiu University, Niaosong Township, Taiwan

Colin Silverthorne, Department of Psychology, University of San Francisco, San Francisco, California, USA

Abstract

Purpose – The purpose of this paper is to examine the relationships between locus of control and the work-related behavioral measures of job stress, job satisfaction and job performance in Taiwan.

Design/methodology/approach – Subjects were drawn from a pool of accounting professionals who completed a questionnaire made up of valid and reliable instruments that measured each of the variables studied.

Findings – The findings indicate that one aspect of an accountants' personality, as measured by locus of control, plays an important role in predicting in the level of job satisfaction, stress and performance in CPA firms in Taiwan. Individuals with a higher internal locus of control are more likely to have lower levels of job stress and higher levels of job performance and satisfaction.

Practical implications – The results indicate that locus of control plays an important role in the overall effectiveness of accountants, even in a non-western culture like Taiwan.

Originality/value – This was a study of a non-Western culture and focused on individuals in a profession rather than occupations not requiring professional credentials.

Article Type:

Research paper

Keyword(s):

Stress; Performance management; Job satisfaction; Taiwan; Individual behaviour.

Journal:

Leadership & Organization Development Journal

Volume:

Number:

Year:

2008

pp:

572-582

Copyright ©

Emerald Group Publishing Limited

ISSN:

0143-7739

Introduction

This study examined the relationship between locus of control (LOC) and the work-related behavioral measures of job stress, satisfaction and performance of professional accountants who work for Certificated Public Accountant firms (CPA) in Taiwan. Although previous research results have indicated that an association between personality and behavioral outcomes exists, there has been little attention paid to some of the factors that may impact these relationships. For example, individuals in a profession, such as accounting, may respond differently than individuals in a job that is not associated with a specific profession and the associated training and credentialing required in order to work in that profession.

The primary goal of this study was to assess how LOC impacts individuals and behavior in the workplace in three primary areas: job stress, job satisfaction and job performance. In addition, little or no research has focused on the potential impact of being part of a profession rather than just as an employee. In this study, accountants were used to represent members of a profession that requires specific credentialing for employment. Also, since the majority of research has focused on Western organizations and values, it was thought that an Asian perspective would add to the understanding of both the variables studied and possibly cultural differences. Since Taiwan joined the WTO, a large number of major American accounting companies have opened branches in Taiwan and are facing cultural challenges related to job performance as they develop their businesses.

Literature review and hypotheses

As one aspect of personality, LOC measures an individual's expectancies for either the need for internal or external control of reinforcement (Rotter, 1966). Specifically, it refers “to the extent to which people believe they or external factors, such as chance and powerful others, are in control of the events that influence their lives” (see Firth et al., 2004). Individuals with a low LOC score have an internal LOC (internals) and believe that their own behavior, capacities, or attributes determine the rewards that they obtain. Individuals with a high LOC score have an external LOC (externals) and believe that whether they receive rewards in life is generally outside of their control (Rotter, 1966).

Several studies have been conducted on type of LOC and its impact on such job aspects as stress, satisfaction and organizational commitment (e.g., Martin et al., 2005). Specifically, studies have identified an interaction between locus of control and job stress (Rahim, 1996; Daniels and Guppy, 1994), job satisfaction, and job performance (Judge et al., 2003). Thus, an individual's reaction as measured by job stress, satisfaction, and performance may depend on several individual personality characteristics especially LOC (Martin et al., 2005).

In terms of cultural impacts, Hamid (1994) found that individuals from a collectivist society like the Chinese are more likely to have an external LOC than individuals from an individualistic society like New Zealand. Further, in a comparison study of 24 different nations/territories, researchers found that the Taiwanese people tend to have an external LOC. Specifically, the Taiwan results indicated that they had the sixth most external LOC score of the countries studied (Spector et al., 2001). Collectivism leads a society to place emphasis on the goals of the group rather than those of the individual and defines the self in relation to the group (Triandis, 1995) suggesting that an external locus of control is more likely in such societies.

Since this study involved a consideration of several organizational behavior factors, it is important to consider previous research findings on how they are affected by personality variables, like LOC.

Job stress

Job stress can be defined as an employee's awareness or feeling of personal dysfunction as a result of perceived conditions or happenings in the workplace, and the employee's psychological and physiological reactions caused by these uncomfortable, undesirable, or threats in the employee's immediate workplace environment (Montgomery et al., 1996). Job stress has received substantial attention in past research on accountants like the individuals included in this study (Fisher, 2001; Bernardi, 1997; Patten, 1995; Choo, 1987).

As many of the recently accounting scandals (WorldCom and Enron/Anderson) have demonstrated, because of organizational pressure, “accountants have participated in unethical behavior in the context of a professional framework and of organizational structures, institutional factors have a profound effect on the choices they make” (Gaa and Thorne, 2004). In the practice of accounting, accountants are working at high pressure and often within tight timeframes. Job stress is very much an individual reaction and is different from general stress as it is also organization and job related (Montgomery et al., 1996). Based on these prior findings the following hypothesis was formulated and tested:

H1. Accountants with external LOC would report more job stress than those with internal LOC in Taiwan.

Job satisfaction

Spector (1997) described job satisfaction as simply how people feel about their jobs and different aspects of their jobs. According to Lawler (1990), job satisfaction refers to people's feelings about the rewards they have received on the job. Thus, satisfaction is a consequence of past events and Taiwanese people tend to rank medium/low in their job satisfaction levels compared to other countries studied (Spector et al., 2001). Several determinants of job satisfaction have been established in prior research. They include organizational reward systems, factors such as power distribution and centralization, and individual differences such as self esteem and the need for achievement (Lankau and Scandura, 2002; Lefkowitz, 1994). Many of these studies examined job satisfaction, along with other constructs, since job satisfaction is commonly treated as a critical outcome variable in organizations (e.g., Judge and Watanabe, 1994; Landeweerd and Boumans, 1994). Based on these prior findings the following hypothesis was formulated and tested:

H2. Accountants with external LOC would report less job satisfaction than those with internal LOC in Taiwan.

Job performance

According to Porter and Lawler (1968), there are three types of performance. One is the measure of output rates, amount of sales over a given period of time, the production of a group of employees reporting to manager, and so on. The second type of measure of performance involves ratings of individuals by someone other than the person whose performance is being considered. The third type of performance measures is self-appraisal and self-ratings. As a result, the adoption of self-appraisal and self-rating techniques are useful in encouraging employees to take an active role in setting his or her own goals. Thus, job performance measures the level of achievement of business and social objectives and responsibilities from the perspective of the judging party (Hersey and Blanchard, 1993). Based on these prior findings the following hypothesis was formulated and tested:

H3. Higher performing accountants have more internal LOC and lower performing accounts have more external LOC in Taiwan.

LOC and job factors

The relationship between personality, job stress, job satisfaction and job performance has been investigated. For example, Choo (1986) found that the level of job stress a person experiences is a function of their personality type. Spector and O'Connell (1994) also indicated that some types of personalities may lead to individuals self selecting jobs which typically have higher levels of stress. In the research of the behavior of entrepreneurs and managers, Rahim (1996) concluded that a person with high internal LOC believes that they can cope with stress functionally and more effectively than someone with high external LOC. Kalbers and Fogarty (2005) found those with an internal LOC are less likely to experience a high level of stress but those with an external LOC are more likely to be vulnerable to stress and are more likely to perceive certain events as stressful. In addition, these researchers concluded that external LOC has a significant negative influence on job stress and tends to reduce personal accomplishments and job performance.

Other researchers have suggested that LOC is a significant predictor of job satisfaction and job performance (e.g., Judge et al., 2003). As one might expect, people with internal LOC report higher levels of job satisfaction (Martin et al., 2005). It has long been assumed that higher employee satisfaction leads to an increase in employee performance and productivity (Lucas, 1999). However, while some researchers have found higher job satisfaction may lead to higher job performance others have not (Nerkar et al., 1996). Jamal (1984) found that as job stress increased, the level of job performance decreased. He also found that job stress was significantly related to job satisfaction (Jamal and Baba, 2000).

Accountants' LOC was investigated with specific attention to how it affects several aspects of work-related behavior in Taiwan. If relationships between LOC and the other variables studied exist then LOC might mediate the relationship between job performance, satisfaction and stress, since an individual's belief in their level of ability to control the work environment has an effect on the work-related behaviors that may, in turn, affect job performance. The following additional hypotheses, with specific application to accountants and a Taiwan population, were formulated and tested in this study:

H4. In Taiwan LOC plays a mediating role between job stress and job performance.

H5. In Taiwan LOC plays a mediating role between job satisfaction and job performance.

H6. In Taiwan LOC of control plays a mediating role between job satisfaction and job stress.

Research methodology

Sample

A total of 620 survey instruments were mailed to 310 Certified Public Accountant (CPA) firms in Taiwan with two surveys sent to each firm. The sample was generated from a list of CPA firms on www.roccpa.org.tw/Catalog.php?cndstr= This is a database that includes most of the CPA firms operating in Taiwan. Because not all CPA firms disclose their address, phone number or other available information online, the authors could not select a sample on a completely random basis. Every CPA firm with a complete address listed on the web site was sampled until the desired sample size was reached. To ensure confidentiality, a cover letter explaining the purpose of the study and a preaddressed, stamped envelope were mailed with each questionnaire. All responses were returned anonymously. In total, 209 completed questionnaires were returned for a return rate of 33.7 per cent.

Measures

The following instruments were included in the questionnaire. Locus of control (Rotter, 1966) is a 29-item fixed choice questionnaire where respondents are asked to select the one best of two choices for each item. Six of the items are filler items so the maximum possible score is 23. Generally, scores 13 and above are considered to be indicative of an external locus of control and scores of 12 or less are considered indicative of an internal locus of control. The job stress scale (Price, 2001) had four dimensions: ambiguity, conflict, workload and inadequate resources. There are 22 items each scored on a five-point Likert-like scale. A higher score represents higher the amount of job stress as perceived by the respondent. The job satisfaction scale (Price, 2001) had six items and is also scored using a five-point Likert-type scale with higher scores representing a greater level of job satisfaction. Price suggests averaging rather than summing the scores for comparison purposes. Finally, the job performance scale (Yousef, 2000) uses a seven-point Likert scale and has four items with higher scores representing higher job performance. These surveys have been used previously in research in Taiwan and have been translated and validated for use in a Chinese population. Demographics were also collected on the gender, educational level and managerial versus non-managerial status of the respondents.

Results

The demographic variables are considered first. The final sample consisted of 171 females and 38 males, ranging in age from 21 to 71 with a mean age of 32.34 and a standard deviation of 9.18. The data were analyzed using a MANOVA and the results indicated that female's LOC was significantly greater than the male's, (F (1,205)=7.67, p < 0.01). Accountants in management positions expressed higher job satisfaction levels (F (1,205)=7.06, p < 0.05) and job performance (F (1,205)=10.98, p < .05), and less job stress than accountant not in management positions (F \;(1,205)=4.60, p < 0.05). The significant interaction effect of gender and management position indicated that, for accountants in management positions, males expressed a higher level of job satisfaction than females, while male non-management staff was less satisfied than female staff (F (1,205)=4.19, p < 0.05). Males had lower LOC scores and reported believing they might be able to influence their environment while female possessed higher score of LOC believed that what happened to them was more likely to be based on luck or the actions of others. These findings reflect traditional Chinese cultural values where the male is more likely to be seen as the center of the society and males tend to want power more than females. Educational levels of the respondents are reported in Table I. The results of a one-way ANOVA showed that there were differences based on educational background in reporting LOC score (F (3,205)=6.75, p < 0.01). Also Scheffe post-hoc tests showed that the respondents with a Master's degree reported significantly lower LOC scores than respondents with other educational backgrounds (see Table I). There was no difference in job stress levels based on educational background (F (3,205)=2.24, p > 0.05. The results from a one-way ANOVA also indicated differences in perceived job satisfaction did exist among different educational background, (F (3,205)=4.73, p < 0.01). Scheffe post-hoc tests (see Table II) showed that respondents with a Master's degree reported significantly higher job satisfaction than respondent with either high school or Bachelor's degree levels of education. In addition, respondents with a Master's degree also reported higher job satisfaction than respondent with junior college degrees but this difference was not statistically different at the 0.05 level. Educational background did not affect job performance (F (3,205)=1.56, p > 0.05).

Table III shows the mean scores, reliabilities and the correlations among the variables. All the instrument reliabilities exceeded 0.7. Nunnally (1978) suggested that Cronbach's alpha has to be above 0.5 and greater than 0.7 in order to conclude that the measurement has high reliability.

Pearson correlations, independent t-tests, chi square, and a series of regression analyses were used to test the hypotheses and check for mediating factors. First, Pearson product-moment correlations were computed between LOC, job stress, job satisfaction, and job performance. The correlation values are presented in Table III.

In addition, according to Rotter (1966) LOC scores range from 0 to 23, with a score higher than 13 indicating an external LOC and a score less than 12 indicates an internal LOC. In this study the highest score was 21 and lowest score was 1. Because the assignment of respondents to a high or low LOC group is only separated by a single point using Rotter's approach, this data was split into three groups so that the upper and lower scores associated with approximately 33 per cent of the total sample were considered to be either external or internal LOC. Thus, those with scores between 14 and 21 were considered to have external LOC (34.4 per cent) and those with scores between 1 and 10 (35.9 per cent) were considered to have internal LOC. Categorizing respondents in this way seemed to be a better approach to identifying real differences in LOC behaviors and minimizing regression to the mean effects on found when comparing groups created based on extreme scores. The result of independent t-tests assessing the difference between the internal and external LOC groups are reported in Table IV and show that there was a significant difference between the two groups in their rating of job satisfaction (t=3.30, df=145, p < 0.01), job stress (t=−2.67, df=145, p < 0.01) and job performance (t=3.55, df=145, p < 0.01).

The highest average job performance score was 6.75 and lowest was 3. The sample was again divided into three groups with approximately 33 per cent of total sample in each group for the statistical analysis. The high job performance group had scores between 5.25 and 6.75 (30.1 per cent), the middle level job performance group had scores between 4.50 and 5 (34 per cent) and the low job performance group had scores between 3 and 4.25 (35.9 per cent). The results of a chi squared test indicate that that H3 was supported (χ ²=10.83, df=2, p < 0.01). The high performance accountants reported more internal LOC (70.5 per cent) than external LOC individuals (29.5 per cent), low performance accountants reported more external LOC (62.5 per cent) than internal LOC (37.5 per cent). Once again, LOC appeared to be a significant predictor of job performance.

Finally, a regression analysis was conducted to test for mediating factors and thus evaluate H4, H5, and H6. Baron and Kenny (1986) stated that when testing the mediator, there is no need for hierarchical or stepwise regression or the computation of any correlation. To test for mediation, the three regression equations must be obtained and four conditions met. First the independent variable must affect the mediator (LOC); second, the mediator must affect the dependent variable while controlling for the independent variable; third the independent variable must affect the dependent variable; and fourth, the relationship between independent variable and the dependent variable (condition three) must either reliably reduce or in the perfect situation, become non-significant when the mediator is controlled.

The results of the regression analysis indicated that H4 met the four conditions. First, there was a reliable relationship between job stress and LOC, β=0.251, p < 0.01. Second, LOC scores predicted job performance when controlling for job stress, β=−0.186, p < 0.01. The third and fourth conditions, the negative effect between job stress and job performance β=−0.316, p < 0.01 was reliably reduced after controlling LOC, β=−0.270, p < 0.01 (t=−4.025, p < 0.01). Thus, the results indicate that the fourth hypothesis conditions for mediation were met.

The same steps were used in testing H5. The relationship between job satisfaction and LOC score was significant (β=−0.308, p < 0.01). When controlling for job satisfaction, LOC predicted job performance, β=−0.212, p < 0.01. For conditions 3 and 4, the relationship between job satisfaction and job performance, β=0.201, p < 0.01 was not significant when LOC was controlled (p=0.55). Therefore for H5, the conditions for mediation were met.

The same steps were used in testing H5. The relationship between job satisfaction and LOC score was significant (β=−0.308, p < 0.01). When controlling for job satisfaction, LOC predicted job performance, β=−0.212, p < 0.01. For conditions 3 and 4, the relationship between job satisfaction and job performance, β=0.201, was not significant when LOC was controlled (p > 0.5). Therefore for H5, the conditions for mediation were met.

In testing H6 it was found that the relationship between job satisfaction and LOC was significant (β=−0.308, p < 0.01). When controlling job satisfaction, LOC predicted job stress (β=0.182, p < 0.01). When controlling LOC, the negative effect between job performance, β=−0.277, p < 0.01, was reliable reduced to β=−0.221, p < 0.01 (t=−3.194, p < 0.01). The conditions for mediation were therefore met.

Conclusions and suggestions

The aim of this study was to examine the relationship between locus of control and the work-related behavioral measures of job stress, satisfaction and performance for accountants in Taiwan. Results indicate that an LOC score was a predictor of work related behavior in several areas. Finding these relationships between LOC and the other variables studied is consistent with research results from western cultures (e.g., Choo, 1986; Rahim, 1996; Bernardi, 1997).

Respondents who had an internal LOC perceived lower levels of job stress, reported higher levels of job satisfaction and job performance. Based on individual responses, the results indicate that high performance CPA firms probably have more internal LOC individuals than low performance firms. The mediator function tests also showed that LOC is a mediator, through which job stress influences job performance. In addition, the mediating effects of LOC indicate that job satisfaction affects job performance and job stress. In other words, for an external LOC individual, job stress would have a negative effect on his or her performance while for an internal LOC individual job stress can enhance his or her performance. Compared with external LOC individuals, an internal LOC individual was more easily satisfied with his or her job, thus increasing job performance. Also an internal LOC individual finds it easier to cope with job stress.

There are several potential implications of these results for accounting firms in Taiwan and possibly elsewhere. Since LOC was an important antecedent of accountants' work-related behavior and outcomes, considering this variable may be useful when seeking to hire the right people. As Martin et al. (2005) indicated internal LOC individuals tend to be more adaptive in dealing with task-related problems, along with the results of these findings the authors believed that internal LOC accountants can better cope with stress, develop more job satisfied and higher performance work environment than external LOC accountants. Another interesting aspect of the results is that a traditional male and power conception was still found to exist in the accounting profession in Taiwan. Males in management positions reported higher levels of job satisfaction than female managers while males not in management positions were less satisfied than females in similar positions. These results indicate that LOC plays an important role in the overall effectiveness of individuals, and specifically accountants, even in a non-western culture like Taiwan where traditional gender based cultural values still play a role reflecting gender specific expectations at work.

This study had several limitations. First the research concentrated on accountants working for CPA firms listed in the major database of accounting firms in Taiwan. The use of a specialized sample like this may limit the effective application of these findings to other vocations and countries without further research. A response rate of about one third is good for this type of research, nonetheless it maybe that with only one third responding there is a sample bias created by those willing to respond. However, it is hoped that the research results will help CPA firms in particular to better understand factors that affect accountants' work-related psychological wellbeing. Future research could include personality variables other than LOC and for different professions. In addition, the replication of this study in other countries would help establish the generalizability of these findings for accountants and organizations.

Table IScheffe results for the differences in LOC scores among different educational backgrounds

Table IIScheffe results for the differences in job satisfaction scores among different educational backgrounds

Table IIIMean score, reliabilities, and correlations among all research variables

Table IVIndependent t-test between internal and external control

References

Baron, R.M., Kenny, D.A. (1986), "The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations", Journal of Personality and Social Psychology., Vol. 51 No.6, pp.1173-82.