Review of: |
David F. Larcker and Lessig V. Parker (1983), “An
examination of the linear and retrospective process tracing approaches
to judgment modeling,” Accounting Review, 58, 58-77. |
Judgmental decisions have been 'predicted' from
inferences based on statistical analysis of the data used by the
decision-maker. Alternatively, predictive models have been developed by
asking decision-makers to describe how they made their decisions.
Sometimes the questioning is done as the decisions are made, sometimes
after. The literature on the statistical analysis of judgment (an area
typically referred to as ‘bootstrapping’) has been critical of the
latter approach of questioning after the decisions have been made. The
study compared this post-decision questioning, ‘retrospective process
tracing models,’ with the statistical approach. Buy/no-buy decisions
were made for 45 stocks by 31 subjects. Each stock was described by six
relevant and obvious variables. The statistical inference approach, done
by discriminant analysis, matched the actual decision in 73 per cent of
the cases. The retrospective process tracing model approach (done
immediately after the completion of all stock decisions) was significantly
better (p < 0.05), and it matched the actual for 85 per cent of
the decisions. The gain came at some cost, as the retrospective approach
required l hour with each subject. The authors caution that the results
may not be applicable to more complex problems or to problems where
irrelevant variables are present. They recommend a combined use of
discriminant models, retrospective process tracing models, and concurrent
process tracing (often called protocols). It is a thorough study and the
literature review brings together a number of relevant findings from
accounting, marketing, and psychology. Although this is a tedious and
long-winded paper with much jargon, it is important, and will be rewarding
to those who manage to stick with it.
|