Difficulty in learning and effectiveness of teaching transitional words

Abstract

The diagnosis of language proficiency for grouping limited English-speaking dents in bilingual education programs has been a problem for three reasons: consistent identification of language proficiency skills for functioning in a and/or bilingual classroom; b) non-specification of appropriate norm-referenced criterion-referenced tests for assessing receptive and expressive English; and availability of criteria for grouping bilingual students. To address these problems, English as a Second Language Assessment Battery (ESLAB), a criterion-measure, was constructed and validated with secondary students. Item analysis based on logical and empirical knowledge included revision rearrangement of items. The data findings were: a) percentile (p) values 89.9% for the Aural Comprehension Test, 1.7% to 64.4% for the Structural Test, and 0.6% to 69.5% for the IRI; and b) the point biserial (RPB) values .67 for the Aural Comprehension Test, -.01 to .47 for the Structural Competency and 0.0 to .70 for the IRI. Reliability involved the computation of internal consistency. The Hoyt estimate of reliability for the Aural Comprehension Test was .81, for the Structural Competency Test .37, and for the IRI .83 and .79 for the total. Cronbach's alpha values were .74 for the IRI and .21 for the total test statistics. The SEm for the three tests was 4.25. Four types of validity were established: a) face validity based on examiner's and examinee's judgements, b) content validity determined by language and reading experts analyzing test items in terms of the specified objectives, c) predictive validity using Kendall's tau (Nie et al., 1975) to correlate each test level results (Beginner I, Begin- ner II, Intermediate I, Intermediate II, and Advanced) with the four Teacher Estimates (T.E.) and ESL report card grades, and d) concurrent validity used only for the IRI. This study examined the Receptive Area (listening and reading) through the Comprehension Test, the Structural Competency Test, and the Informal Reading Inventory (IRI). The Expressive Area (speaking and writing) was analyzed by Rivera (1979) through the Oral Screening Test, the Oral Competency Test, the Dictation Exercise, and the Writing Sample. General guidelines for constructing the battery were specified, then item analysis, reliability, and validity were established. Teacher training workshops were conducted for the six teachers who examined the students: A sample of 59 mostly Hispanic students from an inner city area was tested. This 12 to 16 year old seventh and eighth grade group was composed of 25 boys and 34 girls with varying degrees of English proficiency. A Pearson Coefficient indicated a high correlation between the IRI and Cloze Test .7484 at p<.01; IRI and Stanford Diagnostic Reading Test (1976 Ed.), .6203 at p<.01; Stanford and Cloze .6105 at p<.05; T.E. and Stanford .4248 at p<.05; T.E. and Cloze .3577; and T.E. and IRI -.0456 (Teachers tended to underestimate students' reading levels). It was concluded that the Receptive Area tests are valid measures of language proficiency. The relationships among the four language areas evidenced from the research were: a) listening and reading are related; b) the receptive area is related to the expressive area, and c) oral skills precede literacy skills.