Organisational commitment was measured by assessing the respondents’ perceived
level of organisational commitment on a 7-point Likert-type scale anchored at both
ends, where 1 = strongly disagree and 7 = strongly agree. The survey instrument
included several items regarding each of the two dimensions of organisational affective
and continuance commitment. The results show that two factors emerged from an
orthogonal Varimax rotated factor analysis. Together, both factors explain 53.95%
of the total variance. The KMO measure of sampling adequacy was 0.828, which is
greater than the required level of 0.6, while Bartlett’s test of sphericity was significant,
confirming the suitability and validity of the data for factor analysis. The measure of
internal reliability for both factors was strong, with a Cronbach Alpha value of 0.818
for affective commitment, and 0.797 for continuance commitment. Z-scores were
constructed for the two factors of organisational commitment to enable regression
analysis of the variables of the study.
Managerial level
The construct of managerial level was categorised into three levels of management.
The three levels that were identified for managers to hold for the current study were
(1) human resource managers, (2) chief financial officers or chief accountants and (3)
chief executive officers. Each participant had to indicate which of these categories they
fell into for their responses to be included in the analyses.
Job satisfaction
The dependent variable, job satisfaction, was measured using a single item measure,
which asked respondents to rate their level of job satisfaction on a Likert-type scale
of 1 to 7. Similar to previous items of the survey instrument, the Likert-type scale was
anchored at both ends, where 1 = strongly disagree and 7 = strongly agree.
Regression diagnostics
Linearity, homoscedasticity, independence of residuals and normality
Homoscedasticity and independence of residuals assumptions can be checked by
observing the standardised residual plot for each of the factors of the independent
variables. The probability plot of the standardised residuals indicated that points are
independent and are not related to each other, and follows the cumulative probabilities
for each factor. Also, there is no fanning out or other systematic behaviour of the
residuals, which means there is no violation of the homoscedasticity assumption
(Pedhazur, 1997). Thus, there is no violation of the assumptions of independence of
residuals or homoscedasticity for the dataset of the study. The assumption of linearity
can be checked by observing the scatterplot of the independent variables of the study.
This plot exhibits a random scatter of points, around the line of best fit, indicating that
the data appear to be normally distributed and linear in nature, showing no violation of
the assumptions for regression analysis.
Normality within the data of the regression model may also be checked by looking at
a histogram or a normal probability plot of the residuals (Pedhazur, 1997). The normal