A study to determine the effects of certain teaching methods on English proficiency is summarized. Participants were students and teachers of English as a second language at the West New York (New Jersey) Adult Learning Center. A pretest, observation, post test design was used. Data on student background characteristics such as age, sex, previous education and employment, and country of origin were also collected. Results of the statistical analysis indicated that teaching style did affect student gains in proficiency, particularly for certain types of students. The analysis--which considered student background, classroom interaction patterns, pretest performance, and final proficiency--indicated that there were four combinations related to superior classrooms. (1) Students with more education, more previous English instruction, and higher level jobs learned more with a free response mode and a question and answer paradigm. (2) Students who had been in this country longer, had higher level jobs, and had studied English previously demonstrated exceptional gains when instruction included direct-read and/or ask questions along with free response, and question-answer-corrective feedback-prompt-answer interactions. (3) Well-educated, well-employed women with superior English training in their country of origin demonstrated high achievement with the interaction pattern, teacher direct-student read and/or ask question. (4) Younger, less well-educated male students who had recently arrived from Western Europe learned well when interaction patterns were supportive and individualized. (Author/MV)