range of indicator, or manifest, variables to ‘construct’ these latent variables.
Structural equation modelling (SEM) has been chosen as the data analytical
method for this research owing to its capacity to both (i) measure latent
constructs using a range of indicator variables and (ii) evaluate the relationships
between latent constructs. In SEM parlance, these two processes are referred to
as (i) the measurement model and (ii) the structural model.
To proceed with the measurement model it is necessary to identify and select a
range of indicator variables for each latent construct. This process is described in
Chapter 5 where a range of relevant organisational research studies (both
hospitality-specific and more general) are examined to identify precedents for
measuring these constructs.
Regarding the structural relations modelling dimension: prior to measuring the
relationships between latent constructs, it is necessary to specify, based on
theory, the hypothesised linkages between the various constructs. While it is
possible to look to the generic and hospitality-specific leadership literatures to
develop some hypotheses (e.g. motivational leadership may be a predictor of
both work meaning and extra effort/motivation) many of the constructs of
interest are not specific to leadership. All of the identified constructs are,
however, related to employee motivation and, accordingly, attention is now
turned to the generic and applied (hospitality) work motivation literature to
ensure that the research takes account of, and can be informed by, those bodies
of research knowledge.
This chapter firstly provides (Section 4.1) an overview of the generic work
motivation research to identify a theory-based organising framework to inform
the development of research hypotheses linking the latent variables. This review
finds that the field of work motivation does not operate under a unifying
theoretical framework (see e.g. Ambrose and Kulik 1999: 280; Mitchell and
Daniels 2002: 227), but that there are a variety of contrasting theories and
research approaches. As a consequence, the chapter goes on (in Section 4.2) to
briefly examine the overarching field of industrial and organisational (I/O)
psychology. At this level it is possible to identify an organising framework which
is capable of both (a) satisfactorily accommodating the range of identified latent
variables and (b) providing insights regarding the likely causal relations between
them.