|
Some common pitfalls when developing alternatives include:
-
Anchoring and Tweaking - Starting from the
status quo or the first proposed alternative or what's going
on in a neighboring jurisdiction and making only minor
incremental changes
-
Recommending - Developing a single
recommended alternative and trying to get everyone to buy
into it
-
Accepting constraints - when they could be
removed or softened
-
Considering sunk costs - Choosing
alternatives that justify past decisions
-
Avoiding trade-offs - Not presenting an
alternative because it would be controversial and/or
glossing over difficult value-based choices or trying to
eliminate them too soon -
-
Not knowing when to quit - You can
introduce new alternatives at any time, but at some point
(when the incremental gains are not worth the incremental
effort) you need to satisfice.
Refining Alternatives and Criteria
You'll never get it all right the first time.
The key to good alternatives is iteration. Eliminate clear
losers ""dominated alternatives" and look for
win-wins or joint gains. You are developing the best possible
alternatives for decision makers to consider. Their job is to
make value-based judgments, not technical refinements.
Lay out the alternatives in a consequence table. Characterize
their impacts on each evaluation criterion, either
quantitatively or qualitatively. In the early stages you may
have many criteria and many alternatives. Often they can be
reduced in number by testing the alternatives for dominance, and
testing the criteria for sensitivity.
One alternative is dominated by another if it performs worse on
at least one evaluation criterion and the same or worse on every
other criterion. It can be removed from further consideration,
since there is no constituency that would prefer it.
A criterion is insensitive to the alternatives when its value
does not change across the range of alternatives. When a
criterion is insensitive to the alternatives - regardless of how
important it is in a general sense, if a criterion is
insensitive to the alternatives, then it is not useful in
selecting among them and can be eliminated for this decision,
and for this set of alternatives.
Next:
Estimating Consequences>>
|
There are a number of pitfalls to watch out
for when generating alternatives