A study of the ability of young children to discriminate among such letters as p and q, d and b, using four different methods of discrimination (matching, copying, naming, and writing to dictation). Results show that the accuracy of discrimination is dependent upon the method of assessment employed. (Author/WY)

Discusses study which looked at the variance from target language norms in second-language behavior which arises in semantic information in individual words. Results suggest that study of grammar as a knowledge of basic lexical classifications requires little attention because work class identifications learned in L1 are smoothly transferred to…

Four purposes of testing (screening, classification, program planning, and program evaluation) are identified, and tests are futher broken down into type, methods of scoring, and methods of administration. The author emphasizes that tests should be selected to accomplish the purposes of testing. (CL)

Thirty five-year-olds were given number conservation tasks within their subitising and their counting ranges. For the experimental group, counters were joined by threads. These children's superior performance suggests that, when such one-to-one correspondence is a salient clue, young children can effectively apply the principle of invariance.…

The author reviews and expands the work of J. Boyd (EC 062 111) on the use of the Slosson Intelligence Test scored according to Wechsler Intelligence Scale for Children categories for children who may need special education. (CL)

Florida has adopted and implemented a variety of ways to test handicapped students without harming them, including exemptions, special standards, and modification of test techniques. (Author/CL)

Twenty-one male and 21 female adolescents were tested individually for conservation of mass, weight, and volume. (CM)

Tests are described that will objectively measure playing ability and discriminate between successful and unsuccessful basketball players. (JD)

Investigating use of maximum v averaged scores in physical and motor multiple-trial tests as indicators of performance, this article concludes use of mean scores is still most appropriate in terms of scientific estimations of true performance given multiple fallible empirical measures. (JC)