Planning for a computer-based personnel job opportunities system for the Air Force led to a requirement for a procedure to generate a "payoff" or "value" of the assignment of each person to each possible job. This report discusses three methods of weighting different information to form a single indicator "payoff" or "value," explicit weighting and two implicit weighting methods--policy-capturing and policy-specifying. The two implicit weighting methods are combined into a more comprehensive method referred to as policy-development. The policy-capturing process presents a series of decision situations to one (or more) policy makers and the policy maker assigns to each situation a number which reflects the "value" or "payoff." Then a mathematical model is derived by obtaining the regression equation that best predicts the policy maker's judgments. Policy-specifying, which is the main focus in this report, does not depend on a sample of actual judgments to determine the regression weights, but attempts to translate into mathematical form a policy maker's more global statements about the general properties that the model should have. The mixing of policy-capturing and policy-specifying leads to a process called policy-development. (Author)