All family-friendly practices are positively associated with each other, but this
association varies signiWcantly between pairs of variables (e.g. the correlation between
job-sharing and part-time contracts is equal to 0. 51 , while the correlation
between parental leave and childcare subsidies is equal to 0. 10 ). All equal-opportunity
practices are positively associated and the correlation is slightly stronger, on average,
than that among family-friendly practices. Most of the high-involvement practices are
also associated with each other with the exception of the three work-enrichment
measures and motivation as a selection criterion. The three variables concerned with
work enrichment—job variety, method control, and timing control—are weakly
correlated with each other and with most of the other variables in the sample. This
pattern of correlations means that no practices are being used as substitutes for each
other. Even workplace nursery provision and childcare subsidy are not alternatives.
The associations across types of practices vary signiWcantly, but the majority are
statistically signiWcant, though some practices are not associated or negatively
associated. For example, the correlation between adjustments to accommodate
disability and job-sharing is 0. 35 , that between statistics on gender composition
and parental leave is 0. 26 , while that between variable pay and workplace nursery is
an example of a negative correlation (r¼ 0. 10 ), as is that involving internal
recruitment and term-only contracts (r¼ 0. 11 ).^3
- 2 An Integrated Employee-Centered Management?
An integrated approach to management implies more than the fact that practices tend
to coexist; it means that their association reXects an underlying orientation on the part
of management. The correlations that exist between practices are then explained by this
underlying orientation. In other words, if this orientation did not exist, all the practices
would not be associated with each other. It could, for example, be that an association
between appraisal and variable pay simply reXects the fact that appraisal results are fed
into a performance-related pay system. Similarly, an association between job security
and quality circles or suggestion schemesmight exist simply because managements
introduced the employment guarantees to aid idea-capturing. We can assess whether an
orientation underlies the whole set of family-friendly management, equal-opportunity,
and high-involvement practices by using latent variable analysis, the generic name for a
family of statistical models used for testing for common factors (of which factor analysis
is the best known).^4 (Wedo,however,excludefromthisanalysisthosepracticesthatwe
have already seen are unrelated to other practices of a similar type in this study: work
enrichment and motivation as a selection criterion.)
(^3) The non parametric correlation coeYcients and their signiWcance are available at http://www.shef.
ac.uk/iwp/wers.html 4
For an introduction to the latent variable model used see Wood ( 1999 b:411 14).
590 stephen wood and lilian m. de menezes