The study took place at one elementary school in the School District of Indian River County, Florida.
The study included 42 English language learners in grades 3–5. (The study began with 50 students. Minor attrition occurred, with eight students moving out of the district during the implementation of the study. Of the eight students, three left the bilingual tutor group, four left the English-only tutor group, and one left the comparison group.) These students were native Spanish-speaking and were children of Mexican and Mexican-American migrant workers who seasonally reside in Florida to pick citrus fruits. English language learners were administered a pretest, the IDEA Oral Language Proficiency Test I (K-6) (Ballard, Tighe, & Dalton, 1982, as cited by Serrano, 1987) and were divided into two levels of English language proficiency. Students at each level were randomly assigned to one of three groups. Overall, 12 students were assigned to the bilingual tutoring group, 13 students were assigned to the English-only tutoring group, and 17 students were assigned to the comparison group. The analytic sample for the first and second interventions is 29 and 30 students respectively.
Students participated in a three-month tutoring program. Two versions of the program were examined: a tutoring group where the ELL tutee worked with a bilingual (somewhat
proficient in both English and Spanish) student tutor and a tutoring group where the ELL tutee worked with an English-speaking tutor who did not speak Spanish. Students
were assigned to their tutors based on age, sex, and grade level criteria. Tutoring included daily 20-minute sessions. A total of 37 sessions were implemented in the study
for a total of 12.3 hours of tutoring. Tutoring focused on English language instruction and included lessons on life skills and every day tasks. For example, tutors introduced
vocabulary, played a cassette tape that asked tutees to respond to directions and commands, and used a set of pictures to help ask comprehension questions. Each tutoring
lesson focused on a life skill task (such as caring for a cut).
Students in the comparison condition did not receive tutoring. The control group consisted of whole-group second language instruction led by the teacher.
The primary outcome was oral language proficiency as measured by the IDEA Oral Language Proficiency Test I (K-6) (Ballard, Tighe, & Dalton, 1982, as cited by Serrano, 1987). The test assesses syntax, comprehension, vocabulary, and verbal expression.
Support for implementation
Student tutors participated in a series of 20-minute training sessions before tutoring began. Training content included explanations and demonstrations of effective second language teaching, modeling instructions, prompting, asking questions, and managing time and behavior. Role-playing was also included in training where the trainer played the role of the learner to help tutors practice tutoring skills.