A Study of Some Factors Affecting Generalization of Language Training

Abstract

Two related experiments examined generalization across contexts involving different sentence forms when selected grammatical features were trained within a single syntactic context. Results obtained within a single-subject multiple baseline design showed that training the verbal/auxiliary (contracted with pronoun he) was sufficient to induce its generalization across a variety of verbs, object noun phrases, and a different subjective case pronoun, but not across an objective case pronoun. Similar results were obtained for the uncontractible auxiliary in the past tense. Generalization of the uncontractible auxiliary in the present tense was noted across different verbs; so also was the generalization of contractible copula to different adjectives qualifying the subjective case pronouns he and she (but not in qualifying an objective case pronoun it). Finally, generalization of possessive s inflection across male, female, and animal categories was evident. Trained behavior generalized to home situation in one of the experiments. Some unexpected results also suggest that training on certain forms of either verbal auxiliary or copula may be sufficient to generate correct production of both of them.

Individual Differences in Language Delayed Children's Responses to Direct and Interactive Preschool Instruction

“…Efficacy research on DI for language instruction has clearly documented the value of DI for teaching information in a very specific setting and set of stimuli (e.g., Gottsleben, Tyack, &C Buschini, 1974;Gray & Ryan, 1973;Hegde, 1980;Hegde, Noll, & Pecora, 1979;Weller, 1979). There is also substantial evidence that DI does not teach language in a manner that allows the skills to be generalized outside the training setting (Jeffree, Wheldall, &C Mittler, 1973;Mulac & Tomlinson, 1977;Zwitman &c Sonderman, 1979).…”

Can language acquisition be facilitated in cochlear implanted children? Comparison of cognitive and behavioral psychologists’ viewpoints

“…A number of single subject studies which have specifically focused on the language responses of language impaired children have demonstrated diverse language behaviors in this group of children[12,20]. …”

“…However, it should be noted that generalization occurs with different number of newly trained items in the case of a pronoun for example, than in the case of a verb. So, training in each modality is not influenced by training in other modalities[20]. …”

Short-term syntax learning and generalization effects

“…One of the treatments was a highly structured, repetitive drill procedure which included many features of small-step behavioural teaching programmes used in earlier studies (Gray and Ryan, 1973;Hegde, Noll and Pecora, 1979). The other was a more loosely structured adaptation of the Interactive Language Development Teaching (ILDT) procedure by Lee, Koenigsknecht and Mulhern (1975).…”

“…18-27% of utterances) in children's speech (Ingram, 1974;Shields, 1974); therefore spontaneous language samples should include sufficient opportunities to observe possible generalization of the target. Copular be includes two morphological forms, is and are, which may be viewed as independent, in that treatment on one is not expected to affect the other (Hegde, Noll and Pecora, 1979;Hegde and McConn, 1981). Therefore, these two forms of the same target can be used to keep separate the effects of two treatment programmes for purposes of measuring generalization.…”