Information Technology Reference
In-Depth Information
of financial nature like the improvement of profitability measures are usually influenced by long term
objectives of non-financial type. Therefore Kaplan and Norton (1992) postulate a balanced selection of
strategic goals out of (at least) four distinct dimensions: Financial, customer-, process- and, develop-
ment-specific perspectives should be included in the strategic process. Between the goals and measures
of these dimensions the decision maker has to make hypotheses about the underlying cause-and-effect
relations. Subsequently they can be used to disaggregate main strategic goals into tactical objectives
and measures.
Similar principles are encompassed by the French Tableau de Board methodology (Mendoza et al.,
2002) as well as by cybernetic management principles as proposed for example by Vester (1988) in his
Biocybernetic Approach whose main concepts are reused in the St. Gallen Management Model (Gomez
& Probst, 1999; Schwaninger, 2001; Spickers, 2003).
Although these managerial approaches for strategic decision support provide some practical ap-
proaches for the reduction of complexity coming with a sense-making process the implementations of
these ideas in the form of DSS are rather weak: It can be observed that software tools supporting such
approaches are focused on techniques out of the method base in order to draw conclusions from a hypo-
thetically assumed cause-and-effect model as outlined before (Hillbrand & Karagiannis, 2002a, p. 368).
Therefore most DSS of this type provide simulation techniques as well as how-to-achieve- and what-
if-analyses. However, the model base usually remains unproven with respect to the empirical evidence
of the hypothetically assumed cause-and-effect relations between the business variables. As a logical
consequence, the overall quality of the decision support provided by such a system is directly related to
the completeness and soundness of the underlying causal hypotheses. Moreover these techniques are not
able to provide quantitative forecasts for future impacts of an analyzed strategic scenario. The reason
for this lack of approaches for causal proof and quantitative techniques for managerial cause-and-effect
models can be traced back to a proposition of Kaplan and Norton (1996): In their topic they recommend
“correlational studies” in order to infer causal knowledge from time series of business variables in the
course of double loop learning. This postulate caused an intense discussion about the admissibility of
such techniques with respect to the purpose as mentioned before: As for example Weber and Schäffer
(2000) conjecture, the “basic problem of an analytic derivation of the `correct` cause-and-effect relations
cannot be solved in this way” (p. 8). Horvàth & Partner (2001) come to a similar conclusion:
Cause-and-effect relations [...] do not describe arithmetical logics e.g. in the form of the known ROI-
scheme [...]. The goals and for this reason also the measures are causally associated in a logical but
not in a calculative sense. If the value of one goal changes, the impacts on another variable of the same
system cannot be predicted exactly.
(p. 44f.)
Other authors only refer to the need for further research in this area (Probst, 2001, p. 81). However,
this discussion and its limitation to the construct of correlation leads to a rather dogmatic conception
that managerial cause-and-effect models must not be evaluated in a quantitative way which is typical for
the relevant managerial literature. According to Schneiderman (1999) this is one of the major reasons
for the failure of Balanced Scorecard projects:
We all know that correlation does not mean causality. But try explaining these data to someone who
has been only reluctantly convinced that the non-financial scorecard metrics are a leading indicator of
future financial success.
(p. 10)
Search WWH ::
Custom Search