The 73- and 23-item versions are too lengthy for this research and no theory-
based rationale for the choice of the eight-from-eighteen items used by
Borchgrevink and Boster (either in Borchgrevink and Boster (1994) or in their
source (Miller et al. 1990 )). Similarly, no theory-based rationale was provided by
Chiang and Jang (2008) for their choice of three items from the JDI (one related
to general job satisfaction, one to empowerment satisfaction and one to
happiness with the organisation). Furthermore, it is difficult to see how Chiang
and Jang’s happiness with the organisation measure relates to any of the original
JDI items.
The challenge, then, is to create a more succinct method of measuring job
satisfaction to keep the questionnaire reasonably short and which relates to Smith
et al.’s JDI. Ironson et al. (1989) developed a method for truncating the JDI to
form a unidimensional (rather than the original five dimensions) measure of job
satisfaction which measured respondents’ evaluation of general rather than
specific job characteristics. Following a similar approach it was decided to use
each of the five sub-scale concepts as an individual item in an attempt to
generate a ‘general job satisfaction’ measure for this study.
This is a somewhat exploratory approach – not least because there is no
guarantee that the five items representing the original sub-scales will covary
sufficiently strongly to demonstrate convergent validity. This concern is
underlined by the fact only moderate correlations (0.25 to 0.45) between the five
sub-scales was oberved by Smith et al. (1969; cited in Ironson et al. 1989: 194).
Different studies using different samples in different contexts do produce different
correlations, however, and Tutuncu and Kozak (2007) reported inter-factor
correlations as high as 0.644 (with many in the 0.300 to 0.600 range) in their
application of a truncated JDI in a hospitality sample.
Nevertheless, because the approach being used here is somewhat tentative and
exploratory, an additional item (JS6 - All things considered, how satisfied are you
with your job?) was added to provide a ‘back-up’ global measure of job
satisfaction. The successful use of such global measures is discussed, specifically
in a job satisfaction context, by Yiing and Ahmad (2009: 62).
A further safety net is provided by the combined inclusion of the items JS1 and
JS6 which, together, correspond directly with two of the three items used by
Susskind et al. (2000a; 2000b). Susskind et al. use three items, one for