In a series of studies from 1966 to 1969, Mueller (1971) compared the achievement of students studying audio-lingual materials and cognitive materials. Achievement scores in listening, reading, and writing were much higher for those students working with cognitive materials.
A series of studies, called the GUME project, were conducted in Sweden comparing the "implicit" method, consisting of drills but no analysis or explanation, and the "explicit" method, consisting of analysis, explanations, and practice in the manipulation of forms. For fourteen-year olds in their fourth year of English there was no significant difference between the two methods (Levin, 1969). Another study by Levin and Olsson (1969) with advanced students also found no significant difference between the two approaches. In a later comparison by von Elek and Oskarsson (1972) with adults, the explicit method was superior on class achievement scores and on an oral test, at all age levels, and at all proficiency levels. In addition, the authors concluded that once the students understand the language, structure pattern drills are of only limited value. In general, these studies seemed to concentrate on the acquisition of grammatical concepts rather than language skills.
During the 1966-67 academic year, Chastain and Woerdehoff conducted an investigation comparing the audio-lingual habit theory and the cognitive code-learning theory as applied to teaching first-year Spanish at the college level. The authors defined the audio-lingual approach as having three basic
characteristics: (1) the use of pattern drills, (2) the inductive presentation of new structure, and (3) the introduction of the four language skills in the natural order of listening, speaking, reading, and writing. The cognitive classes were taught (1) using traditional exercises, (2) presenting deductive grammatical explanations, and (3) utilizing all four language skills from the beginning of the course. At the end of the first year, the audio-lingual students received significantly higher scores in repeating sentences after a native speaker; the cognitive students received significantly higher scores in reading. There was no difference in the students' ability to answer questions and describe pictures. Achievement scores in listening and writing favored the cognitive students, although the differences were not significant. The authors concluded that the results favored cognitive teaching procedures.
The most extensive study to date in the general area of methodological comparison of audio-lingual versus traditional teaching procedures was the Pennsylvania Project. 9 This study was conducted at the high-school level and included classes in French and German from various high schools throughout the state. Although careful preparation attempted to control important variables in the project, and although it was well done in this respect, certain weaknesses are inevitable in a project of such vast scope. The Pennsylvania Project was no exception. Nevertheless, the outcomes present important information with regard to conditions as they exist in the schools. In this sense, the study is more correctly viewed as a descriptive study than as a truly experimental one. The important conclusions at the end of the first year were as follows:
1. "Traditional" students exceeded or equalled "functional skills" students in all measures. 10
2. The language laboratory systems (employed twice weekly) had no discernible effect.
3. There was no "optimum" combination of strategy and system.
4. The best combination of predictors of success were the MLA Cooperative Classroom Listening Test, the ML Aptitude Test, and language IQ as measured by the California Test of Mental Maturity (Short Form).
5. Females achieved better than males.
6. Student attitude was independent of the strategy employed.
7. "Functional skills" classes proceeded more slowly than "traditional" classes.
8. There was no relationship between teacher scores on all seven portions of the MLA Teacher Proficiency Tests and the achievement of their students in foreign-language skills.
After another year of investigation and replication, the following conclusions were reached:
1. No significant differences existed among strategies on all skills except reading (TLM<) as measured on contemporary standardized tests after two years.
2. The language laboratory of any type, used twice weekly, had no discernible effect on achievement.
3. The best over-all predictors of success are prior academic success and a modern-language aptitude test.
4. Student opinion of foreign-language study declines throughout the instruction, independent of teaching strategy employed.
5. Published test "norms" and implied in text layout progress were more than most of the experimental population achieved.
6. Within the functional skills strategies students utilizing Holt, Rinehart and Winston materials did significantly better than students using the Audio-Lingual Materials.
7. Neither teacher experience in years and graduate education nor scores on the MLA Teacher Proficiency Tests are related to mean class achievement after one or two years (Smith & Baranyi, 1968).