Mobile Edition
Print
Get the Mag
Weekly UpdatesDescriptive Statistics
The mean and standard deviation for variables and frequency distribution for attributes are shown in Table 1-A and 1-B. The mean scores for organizational commitment, job and organizational characteristics range from 4.79 (slightly agree) to 5.84 (moderately agree). There are more females (61.7%) and more public institutions (77.8%) than their counterparts. Returns are almost evenly distributed by age (less than 44 years old vs. 44 years old or older), number of years of education (graduate degrees vs. others) and size of student body (less than 10,000 vs. 10,000 or more).
Linear Regression Model - Full Model
The relationship of the job and organizational characteristics (mean values) along with the demographic factors (independent variables) and organizational commitment (dependent variable) were examined using a multiple linear regression model. Demographic variables in Table 1 - B were recoded into dummy variables to capture the effects of these variables on organizational commitment. The mean of the fifteen questions that measured organizational commitment was used as the dependent variable. This value (dependent variable) was squared to stabilize nonconstant variance in the residual value. The results are listed in Table 2. Table 2 also includes standardized regression coefficients and a 95% confidence estimate of regression coefficients.
The model is reliable (F = 32.467, p [less than] 0.000) and it has good explanatory power (R-square = 0.692). No apparent colinearity problem among independent variables exists as the variance inflation factors (VIF) for all independent variables are smaller than 3 (not shown here) (Neter et al, 1997). Among the demographic variables, organizational tenure appears to be the only variable that has a positive significant relationship (p [less than] 0.05) with organizational commitment. Skill variety appears to have a statistically significant positive influence over organizational commitment (p [less than]0.01). The feedback variable, although statistically significant, has an unexpected negative relationship (p [less than] 0.05). Other variables in the job characteristics fail to generate any significant relationship with the dependent variable.
As to the organizational characteristics, every subconstruct variable appears to have a significant relationship with organizational commitment. The standardized regression coefficients indicate that among several independent variables, organizational dependability and instrumental communication seem to have the largest impact on the dependent variable.
Factor Analysis Model
In light of multiple questions in each characteristic (12 questions under job characteristics and 13 questions under organizational characteristics), a factor analysis was attempted to simplify the number of possible independent variables in each area to a manageable number, and to provide decision makers with a manageable set of decision variables. The results are shown in Table 3 - A (job characteristics) and Table 3 - B (organizational characteristics).
The Kaiser-Meyer-Olkin sampling adequacy indicates factor analysis is appropriate for both constructs as the coefficients exceed 0.8 for both characteristics. The Bartlett's test of sphericity appears to reject the hypothesis that the correlation matrix is an identity matrix.
Three factor loadings emerged for the job characteristic variables with the eigen value exceeding one. The first factor loading (eigen value = 5.020 and 41.5% variance) is named "Job Meaningfulness" as it encompasses most of the factors related to the experienced meaningfulness of the work. The second factor loading (eigen value = 1.514 and 12.6% variance) is clearly grouped "Feedback" that employees receive either directly or indirectly from their colleagues and management. Factor Loading 3 was named "Task Autonomy" as it includes factors related to task autonomy (eigen value = 1.049 and 8.7% variance).
As to the organizational characteristics (Table 3-B), only two factor loadings emerged with total variance of 52.9%. It appears that the employees' perceptions toward the organization play a significant role in Factor Loading 1 (42.6% variance), whereas the other colleagues attitudes toward the organization seem to play a crucial role in Factor Loading 2 (10.3% variance).
Linear Regression Model - Reduced Model
Three factor scores from job characteristics and two factor scores from organizational characteristics, along with dichotomous demographic variables, were entered as independent variables in a multiple regression model. An average of the fifteen organizational commitment questions was used as the dependent variable. This value (dependent variable) was also squared to stabilize nonconstant variance in the residual value. The results are reported in Table 4. The model as a whole was reliable (F = 48.929, p [less than] .001) with R2 = 0.695. Table 4 also includes the standardized regression coefficients as well as a 95% confidence for regression coefficients.
It appears that employees with a longer tenure have a higher degree of organizational commitment than their counterparts (p [less than] 0.05). In this reduced model, female respondents appear to have a higher level of commitment than male respondents (p [less than] 0.05). Other demographic variables (AGE, EDUCATION, SIZE, TYPE) failed to show any reliable relationship with the organizational commitment variable.
As to the job characteristics variables, only one factor loading (JOB MEANINGFULNESS) seemed to have a significant relationship with the organizational commitment variable. Employees tend to value a challenging task utilizing a variety of skills to produce an identifiable product. Job feedback and task autonomy factors have negative relationships with the dependent variable, which were unexpected. However, the relationship could go in either direction as the range of regression coefficients for these two factors included negative as well as positive relationships.
Both factor loadings of organizational characteristics variables were found to be highly related to organizational commitment. A good perception of the organization by employees and a positive attitude toward the organization by other employees in the organization appear to create strong organizational commitment.
Discussion
While some of the findings in this study confirm the relationships found in other studies, other findings show that the variables that impact the organizational commitment of internal auditors in colleges and universities differ from the findings for other professions.
Gender and organizational tenure were the only demographic variables that were significantly related to organizational commitment in this study. Relationships found in past studies between organizational commitment and age (Sommer et al., 1996; Mathieu and Zajac, 1990), level of education (Steers, 1977; Mathieu and Zajac, 1990), organizational size (Sommer et al., 1996), and organizational type (Zeffane, 1994; Bourantas and Papalexandris, 1992) were not found to be significant for internal auditors. The insignificant relationship between organization size and organizational commitment did, however, confirm the finding from Mathieu and Zajac's (1990) meta-analysis.
A significant relationship was found between gender and organizational commitment. Based on the results of the reduced multiple regression model (Table 4), the organizational commitment of females was higher than that of males. This result may be explained by differences in perceived alternate employment opportunities. In a study of the organizational commitment of nurses, the employee's perception of alternative employment opportunities was negatively related to organizational commitment (Curry et al., 1986). If males perceive that they have more employment opportunities than females, the fact that females are more committed than males is at least tenable. Two meta-analyses have examined this relationship. One confirms the relationship found in our study (Mathieu and Zajac, 1990), while the other found no relationship between gender and attitudinal commitment (Aven et al., 1993).
Organizational tenure was found to have a statistically significant positive relationship with organizational commitment. This finding supports the study results by Curry et al. (1986) and Mathieu and Zajac (1990) which show that tenure has a significant impact on organizational commitment.
Organizational characteristics had a significant positive relationship with organizational commitment both in the expanded model (Table 2) and the reduced model (Table 4), which confirms previous findings. The strength and direction of these relationships is to be expected when the structure of the internal audit position is considered. Without authority, internal auditors must rely on management support if their work is to affect organizational operations. Because of this situation, internal auditors' perceptions of organizational characteristics are crucial for effective discharge of their responsibilities as internal auditors; therefore, such a strong relationship is not unexpected.
Inconsistent relationships between job characteristics and organizational commitment were found in this study. Only two of the five job constructs (Table 2) were significantly related to organizational commitment. Skill variety had a significant positive relationship with organizational commitment (p [less than] 0.01), which supports the relationship found in other studies (Dunham et al, 1994; Bhuian et at., 1996). Another job characteristic, feedback, had a significant negative relationship with organizational commitment (p [less than] 0.05). This result contradicts the hypothesized relationship and the result found by Dunham et al. (1994). One possible explanation is that the feedback received may be more negative in nature than positive. The questions measuring feedback in the survey instrument measured only the level, not the type, of feedback.
FEED