An English proficiency test for pre-school Spanish-speaking children

This study was conducted to build an instrument for measuring the English proficiency of pre-school children from Spanish-speaking homes and to evaluate this instrument. The need for this test was indicated from the fact that many Mexican-American children have been placed in special education classes on the basis of tests administered in English when English was not clearly understood or spoken by the child. The Children's English Proficiency Test (CEPT) was built from items teachers expect children to comprehend on entrance into first grade. A pilot test was designed and tested on Anglo, Negro, and Mexican-American pre-school children. Revisions of the items were then made for a final version of the test. An alternate form in Spanish was made by members of the Mexican-American community to parallel the English version as closely as possible. In its final form CEPT is an individually administered test requiring approximately seven (7) minutes for complete administration. It is scored on a 'rights' only basis. Sixty-seven (67) children (10 Anglo, 10 Negro, 10 Mexican-American with kindergarten experience, and 37 Mexican-American children without kindergarten experience) entering first grade in Houston in the fall of 1971, were tested with CEPT and with a Test of General Ability from the InterAmerican Test Series using the Spanish or English form according to the child's primary language. Children who were unable to respond to CEPT in English were given CEPT in Spanish to prevent a possible failure experience. All of the children were from lower socio-economic areas in the city. The Mexican-American sample was retested with CEPT within three (3) weeks of the original testing. Statistical analysis indicated that CEPT was internally consistent and reliable over time. CEPT also appeared to be construct valid as all of the hypotheses were confirmed: proficiency in English increased with age; Anglo and Negro children were significantly more proficient in English than Mexican-American children; general ability level accounted for only a small part of the variance on CEPT; and Mexican-American children who scored low on CEPT scored as high on the Spanish translation of CEPT as did Anglos and Negros on the English version. The major conclusion was that CEPT is a valid and reliable instrument for screening pre-school Spanish speaking children on English proficiency. Several recommendations for the use of CEPT in schools and in pre-school projects were made along with suggestions for further research with this instrument.