In order to test the moderating effect of MCS type on the strategy
performance relationship, moderator regression was used. In the first
part, the entire sample was split into two groups:
1. the group consisting of firms which follow a differentiation strategy
and
2. the group consisting of firms that follow a cost leadership strategy.
The respondent firms in the study were split based on the average
score calculated across the nine strategy items for each firm. Firms
with a strategy value of less than 3 (i.e. the median value) were considered
as firms following a cost leadership strategy and firms
that had an average strategy value of 3 or more were considered as
following a differentiation strategy. A total of 83 (39%) sample
elements belonged to the cost leadership group while the remaining
firms were found to be following the differentiation strategy. Moderator
regression was performed for each group separately. For both
the regressions, performance was the dependent variable.
In the cost leadership group, a firm's strategy value was subtracted
by 3 and the absolute value of the resulting scorewas considered as the
strategy value for the firm. This was because, in the strategy variable,
lower values meant greater reliance on cost leadership strategy, with
value nearing 3, showing a minimal reliance on cost leadership strategy.
The strategy variable was the independent variable and the financial
based MCS variable was the moderator variable for the cost leadership group. For the group of firms that were following a differentiation
strategy, themoderator variablewas specified as the extent towhich the
firms were following non-financial based MCS. The hypotheses were
tested using moderated regression (MAR) analysis. In this analysis,
apart from the independent variable and the moderating variable, a
cross product of the independent variable and moderating variable are
included in the regression equation. According to Aiken and West
(1991), the significance of the moderating effect can be inferred from
the statistical significance of the cross product of the independent
variable and the moderating variable. The degree of moderation could
be plotted using the method suggested by Aiken andWest (1991). The
regression plot gives a good idea about the level and direction of
moderation. The regression equations are given in Table 5.
We also conducted two different sets of regression analysis to
explore: (i) the moderating effect of non-financial MCS on costleadership
strategy and performance relationships and (ii) the
moderating effect of financial MCS on differentiation strategy and
performance relationships. The results are shown in Table 6. However,
in both the regression equations, the regression coefficient of the
interaction term (the cross product of the strategy and MCS variable)
was not found to be significant even at pb0.1 level. A moderating
effect is thus ruled out in the case of non-financial MCS for cost
leadership strategy and financial MCS for differentiation strategy.