English proficiency of beginning elementary Spanish-speaking students



Journal Title

Journal ISSN

Volume Title



The purpose of this study was to provide more extensive normative data, reliability data and data which would allow a judgement to be made about the minimum level of English proficiency required for predicting adequate performance in kindergarten and the first grade using the Children's English Proficiency Test (CEPT) (Webster, 1971). Two hundred and nineteen Mexican-American beginning elementary students were tested with CEPT. The sample included forty-two kindergarten students and one-hundred and seventy-seven first grade students from upper, middle, and lower socio-economic status. The age range of the students was sixty-five months to eighty-seven months. The testing was done by two experienced Mexican-American examiners in ten elementary schools in the Corpus Christi Independent School District during January, 1973. The procedures used to provide more extensive normative data included an attempt to create an age scale and to identify difference in English proficiency among different educational groups. When an attempt was made to form an age scale with CEPT scores the data failed to produce a scale that differentiated between age groups. Two possible factors which may have contributed to this failure were discussed. The data for three educational groups, kindergarten students (kg), first grade students without kindergarten experience (1-nk) and first grade students with kindergarten experience (1-kg) were analyzed. The 1-kg was found to score higher on CEPT than either kg or 1-nkg. The latter two educational groups did not differ in mean performance on CEPT. Original statistical analysis of test-retest data resulted in a very low correlation coefficient. Statistical analysis of test-retest data using a formula for correction of curtailed distributions yielded a high correlation. Three measures of achievement including the Metropolitan Readiness Test (MRT-A), Stanford Early School Achievement Test, Level I (SE-SAT) and the Science Research Associates Assessment Survey, Primary I, Form E (SRA-PI) were used in determining the usefulness of CEPT in predicting adequate performance in kindergarten and the first grade. A criterion score of stanine three or below was selected to represent inadequate achievement. CEPT scores of thirty-two and thirty-three were examined as possible cutting points on CEPT. The results indicated a CEPT score of thirty-two was a reasonably effective cutting point. Statistical analysis of the relationship between performance on CEPT and performance on the MRT-A, SE-SAT, and SRA-PI resulted in thirteen correlation coefficients, twelve of which were significant beyond the .05 level. The major conclusion was that CEPT is valid in relation to achievement for kindergarten and first-grade Mexican-American students. Several recommendations were made for the use of CEPT in schools and use in future research.



English language, Study and teaching, Spanish speakers, Ability testing