The Effects of the Audio Lingual Method vs. the Total Physical Response Method on Initial Vocabulary Acquisition in Adult Learners of a Model Language
Method
Subjects

The sample for this study was selected from the total student population of Christ For The Nations Institute. The population is ethnically and chronologically diverse. From this population ten volunteers were chosen and randomly assigned to one of two groups of five subjects each, using a table of random numbers.

The Audio-Lingual group (group A) concisted of four volunteers which ranged in age from 18 to 48 with an average age of 27.5 years. All members of this group were Anglo-American. Two were first time language learners one has studied American Sign Language and the final member of group A was a bilingual Spanish speaker. Three members of this group were female and one male. There was origionally a fifth member of this group which dropped out of the experiment after the first lesson.

The composition of the Total Physical Response group (group B) varied significantly from that of group A. The ages of the group B participants varied from 18 to 56 with an average age of 31 years. The ethnic composition of this group broke down as follows: two Anglo-Americans, one African-American, one Asian-American and one Indian national with English as the primary home language. Two members of this group identified themselves as monolingual English speakers; two identified themselves as bilingals and the fifth member identified herself as a quadrilingual speaker. All members of group B had previously studied a second language. One member of this group attended only two lessons and was not tested.

Materials Design

The researcher felt that one of the most important factors to controll for in this experiment would be prior exposure to the target language. Therefore, since one of the researcher's hobbies is glossopoeia (the construction of model languages), the researcher decided to construct a language for this experiment. Lrahran was a project already under way at the time this experiment was designed.

It has a very non-English phonology consisting solely of liquids, nasals and simi-vowels with a four vowel system concisting of /a/, /O/, /u/ and /E/.

The grammar is of the inflecting type with nouns inflecting for case and number and verbs inflecting for tense, aspect and mood. Adjectives do not exist in Lrahran. Instead adjectival concepts are expressed as verbs.

Syntax is of the SVO type with modifier following modified.

In preparing the materials for this study several sources were consulted. Among those sources which were especially helpful were Daine Larsen-Freeman (1986) whose book outlines and demonstrates numerous instructional techniques including the Audio-Lingual Method and Total Physical Response, the two methods which were the subject of the current study. The sample lessons in Rashed (pp. 208-226) were also highly useful.

Experimental Design

The design applied in this study will be a posttest-only design. This design was chosen because random assignment to groups is possible due to the impossibility of subjects having been previously exposed to the target language. The combination of random assignment and the establishment of two groups, each serving as control to the other, eliminates the majority of threats to both internal and external validity of the study. However, mortality could have become a major threat to the internal validity. An attempt at controlling for mortality was made in the design by having the subjects sign a contract which entitled them to a reward of $20.00 for completing the experiment. One member of group A left the experiment after the first session and one member of group B attended only two sessions and was not tested. Other than these losses the compesition remained steady throughout the study.


Group      Assignment      n       Treatment      Posttest
  A        Random          4       Audio-Lingual  researcher's
                                   Method         impliment
  B        Random          5       Total Physical researcher's
                                   Response       impliment
Instrument

A non-standardized instrument created by the researcher was used to measure the acquisition of the vocabulary taught. Since no writing was taught during the course of this experiment a written test would not have been feasable. The test consisted of two sections of equal length. One section consisted of twenty-five drawings and/or photographs of nouns which the subjects were to identify in Lrahran. The second section of the test consisted of a list of twenty-five verbs which the researcher read in English and the subjects attempted to correctly identify using Lrahran. The vocabulary items on the test were chosen using a table of random numbers to randomly select from a list of the vocabulary items covered by both methodologies.

No validity, reliability or norms data will be available due to the nature of the instrument selected. Despite this fact, this method of testing was selected due to the fact that the experimental treatment will be very brief and the number of vocabulary items taught correspondingly small making standardized tests impractical.

Proceedure

In the first weeks of the Spring '98 semester ten volunteers were acquired through monetary encouragement which proved necessary. From this group of ten volunteers, the subjects were randomly assigned to one of two treatement groups. Each of these two groups were instructed in the Lrahran language. One of these groups (group A) was instructed using the Audio-Lingual method. The second (group B) was instructed using the Total Physical Response method. Each of these groups was instructed by the researcher.

The study lasted for the four weeks of the month of February, chosen because it avoided such historical factors as spring break, mid-terms, and crunch times at the researcher's place of employment.

The Audio-Lingual group was instructed using traditional A-L techniques calling for immediate student response to teacher prompts in the form of correct production of the language forms modeled, and immediate correction of "errors."

Likewise the Total Physical Response group was instructed using traditional TPR techniques which call for student response to imperatives given by the instructor and later by the students.

After four weeks of instruction identical tests were administered to both groups and the results compared using a t-test.

Results

A t test for independent samples (alpha =.05) was used to compare the test results of the two experimental groups. This statistical technique was employed because it seemed to be the one most suited to the analysis of the data collected. When the data was analyzed it was found that there was a significant difference in the means of the two groups (see table two). Therefore, the original hypothesis that there there would be no significant difference in the acquisition of vocabulary between those students using the Audio-Lingual methodology versus those using the Total Physical Response methodology must be rejected.

TABLE 2
Test Means, Standard Deviations and t for the Experimental Groups _________________________________________________________________

Group
Audio-Linguala Total Physical Responsea t*
M
           2.75                       15.63             3.44
SD
            1.77                        6.39
_________________________________________________________________
Note. Maximum score = 50
an= 4
*p < .05

Bibliography