It is better to prepare for the worst
and judge the complexity dimension
to be worse (highly complex) than it
may be in reality (low complexity).
2. For a high penalty setting, the time-based approach (SS1) performs
well compared with the cost-based
The preferred solution strategy
depends on the (actual and perceived) complexity of the judgment
error. Even though the time-based
strategy (SS1) performs slightly better than the cost-based strategy (SS2)
when the complexity is high but
judged to be low, the difference was
found to be statistically insignificant. When the complexity is low but
judged to be high, the cost-based
strategy comes out on top.
When it comes to the uncertainty
dimension, SS1 performs slightly better than SS2 for both judgment error
positions. Interestingly, the activity
costs are lower for SS2 but the higher
penalty costs push the global cost
deviation of SS2 higher than that of
3. For a low penalty setting, the cost-based strategy (SS2) clearly outperforms the time-based strategy (SS1).
Influence of the level of effort. In order
to ensure that no large differences in the
level of effort materialize for the generated
projects, the level of effort was controlled
significance results of the main experi-
ment, whereas the judgment error results
can be found in Table 10. The third para-
graph looks at how a higher level of effort
impacts the cost performance of both
Performance. In this paragraph, we
limit ourselves to situations in which
the decision maker judged the complexity and uncertainty correctly. The following five observations regarding the
performance of both solution strategies
can be made:
1. The penalty costs for the time-based
strategy (SS1) are lower compared
with those of the cost-based strategy
(SS2) across all complexity, uncertainty, and penalty levels.
This implies that the project duration attained by SS1 lies closer to the
deadline than that for SS2, resulting
in a lower amount of incurred penalty costs.
2. For a high penalty setting, a larger
deadline deviation leads to a steep
It is no surprise that due to this
increased importance of the timing
aspect SS1 thrives in a high penalty
3. Even though SS1 has a smaller share
of penalty costs, the activity costs of
SS2 are much lower, indicating that a
better trade-off selection takes place.
The timing aspect does not have a
substantial impact when the penalty
is low; therefore, SS2 almost always
returns better results than SS1. The
difference between both strategies is
more pronounced for a high degree of
4. The complex search process for better
trade-offs proves advantageous for a
cost-based approach (SS2).
When the complexity is high, SS2 is
slightly but not significantly better
even when the penalty is high. In
that case, the proportion of penalty
costs is larger than for SS1 but the
activity costs are much lower.
5. When there is little uncertainty, SS2
performs better or there is only a
small difference compared with SS1.
Clearly, a low degree of uncertainty
has only a minor impact on a project’s duration.
Judgment error. A central topic in
this article is the discrepancy between
the real complexity or uncertainty and
how it is judged; therefore, judgment
errors can be made in which a dimension is low but judged to be high or
vice versa. The results of these judgment errors lead to the following three
1. A general conclusion for both strategies is that safety is the best policy.
Dimension Actual Perceived Penalty
Global Cost Deviation Penalty Share
SS1 SS2 Sign. SS1 SS2 Sign.
Complexity Low High Low
High Low Low
Uncertainty Low High Low
High Low Low
Table 10: Results of the main experiment (judgment error). An asterisk denotes a significant difference: p , 0.05.