Control variables
Six individual demographic variables were controlled: age, tenure,
education, gender and position in the company, as well as the job held by
the individual. It is important to control these variables because they may
influence the level of organizational commitment (e.g. Meyer & Allen, 1991;
Chang, 1999). For instance, as employees get older, stay in the current
company longer, and are promoted to higher positions, they may perceive
higher justice in the current system and become more committed to the
company. Gender was dummy-coded, with 0 for female employees and 1
for male employees. Position was measured by the position’s level in the
company, starting from clerk, coded 1, up to general manager, coded 5. An
individual job was effect-coded into four dummy-type variables. Effect
coding has been regarded as more appropriate than dummy coding when
the criterion is randomly selected as is the case of the current research,
because each job is compared with an entire set of jobs rather than a
randomly selected one (Cohen, 2001).
In examining the effect of a company’s HR practices on employees’
overall perception, company size was controlled. Company size was
measured by categorizing the number of employees into five scales.