Introspection, Deduction, and Guesswork in Policy Evaluation

There is a need for more use of introspection, deduction, and guesswork in policy/pro gram evaluation to supplement the possible overemphasis on statistical/behavioral analysis. Introspective interviewing of knowledgeable insiders can often determine the magnitude of relations more meaningfully than statistical analysis can in view of reciprocal causation, spurious causation, multicollinearity, and other confounding factors. Deduction from empirically validated premises avoids the problem of lack of an experimental group, a control group, after-data, before-data, bureaucratic inertia, and damages from experimenting. Guesswork can be combined with sensitivity analysis to determine the effects of guesses and missing information on bottom-line conclusions where statistical analysis, introspective authority, or deduction is questionable.