A coherent system of decision making is described that may be incorporated into an instructional sequence to provide a supplement to the experience-based judgment of the classroom teacher. The elements of this decision process incorporate prior information such as a teacher's past experience, experimental results such as a test score, and preferences or values. Beginning with a basic example in which the prior information is the probability that a student having taken a course will have mastered it; the experimental result is a test item score, and the values expressed are the costs of wrong pass-fail decisions. Equations and numerical examples demonstrate the solution to this and to more complex problems. Following these examples is another group of examples that abandons the notion of loss or costs in favor of the notion of the utility functions for various decisions. There are discussions of the formal structure of decision theory and of utility theory. Many of the examples are complex and much of the argument is expressed in integrals or summation equations. Topics include: normal form analysis and extensive form analysis; continuous posterior probabilities; a binomial model; threshold utility; linear utility; quadratic, exponential, and squared exponential utility; and three action examples. (CTM)