Evaluation of instructional objectives is essential in education in order to make instruction more fruitful by providing feedback to teachers and students. With regard to criterion-referenced evaluation, construct validity has been emphasized more and more in educational testing during the past few decades. This paper is to clarify the relationshipe between the constructs of educational objectives and the test which are intended to measure these objectives; furthermore, the importance of construct validity is emphasized. The differences between psychometric measurements, which focus on individual differences, and edumetric measurements, which focus on withinindividual growth, are also pointed out. The effects of educational instruction are properly measured by edumetric testing. The pretest-posttest method to measure the effectiveness of instruction has been said to have some shortcomings. However, the effectiveness of the method is mentioned.