The proposed framework assumes six developmental stages according to mean length of utterance in morphemes. Within each stage, syntactic forms and semantic relations interact. In assessment, each utterance is analyzed for features which describe current language abilities. Language goals are developed based on existing and expected features.…

Examines the issue of giving English-language oral proficiency examinations in overseas English-medium universities as part of student evaluations. An oral test given at Jordan's Yarmouk University is reviewed showing an attempt to administer biannual, oral-communicative-proficiency tests to 1,000 students at a time. (24 references) (GLR)

Examines two proficiency testing models, the ACTFL/ETS Oral Proficiency Interview and another model developed at the University of Wisconsin, to determine which is more appropriate for use in academic settings and whether it is possible to combine the best of both models. (GLR)

Comparison of language test performance of 26 learning disabled and 26 non-learning disabled Black elementary-aged children is discussed in terms of (1) relative performance of the two groups, (2) the most useful assessment measures for this population, and (3) effects of socioeconomic factors on language test performance. (Author/DB)

The advantages and disadvantages of the use of the oral interview and role-playing to assess communicative competence are outlined. The interview is criticized and deemed an inappropriate form of testing, and it is proposed that role-playing allows better assessment of sociolinguistic competence. Group interaction is suggested as providing another…

The Denver Developmental Screening Test (DDST) and an expressive language measure were administered to 114 children (aged 24 to 74 months) suspected of developmental difficulties. The DDST did not identify the majority of children who failed the expressive language screening, even after cutoff scores were made more rigorous. (Author/JDD)

The results of a study are presented in which 40 preschool children were given 5 tasks including a test of story comprehension (listening and responding to questions) and a test of pragmatic awareness divided into 4 parts (listening to a story, retelling it, judging what was strange in it, and explaining the consequences of the "strange" event).…

Suggests that language testing should be more performance-based and realigned with a current and richer understanding of the language teaching and learning processes. This kind of testing has questionable validity until a wide range of variables are successfully defined and delimited. (CB)

Comparison of results obtained for 62 normally developing preschool children on the Bracken Basic Concept Scale, the Preschool Language Scale, and the Slosson Intelligence Test revealed low to moderate correlations between the three tests. Results suggest the tests do not measure the same abilities and thus cannot be used interchangeably. (DB)

A comparison of two well-known test formats for evaluating and characterizing foreign language learners' interlanguages revealed that open-ended production test questions were more difficult than closed-item recognition test questions. (CB)

Violations of item unidimensionality on language tests produced distorted estimates of person ability, and violations of person unidimensionality produced distorted estimates of item difficulty. The Bejar Method was sensitive to such distortions. (Author)

Third semester college students of French viewed part of a film, read part, and then were given a surprise vocabulary test with colloquial words from the text. Their performance, compared to a control group, suggests that incidental vocabulary acquisition is possible in a foreign language situation. The test is appended. (Contains eight…

Computer adaptive language testing (CALT) offers a variety of advantages; however, since CALT cannot test the multidimensional nature of language, it does not assess communicative/functional language. This article proposes to replace multiple choice and cloze formats and to apply CALT to live-action simulations. (18 references) (LB)

Argues that teachers need to become "informed consumers" of standardized tests in order to influence decisions about test use and about ways to help students perform at their best. Six strategies for considering the content of a test form are presented. (LR)

Narratives from three studies were analyzed using seven variables hypothesized to measure a variety of language abilities used in narrative production. Results indicated that the variables represented global organization of content, and within- and across-sentence structure. Variables representing sentence structure were most effective as…