Comprehension of relative clause constructions by children 5 to 10
ABSTRACT
The purpose of this study was to investigate how children between five and ten years of age comprehend sentences containing relative clauses.
Fifteen sentence frames were devised, varying on four syntactic dimensions: voice, relative pronoun deletion, embedding, and function of the common noun. Three nouns (mommy, boy, girl) and two verbs (pushed, chased) were selected for their reversibility and were varied among the sentence frames. Six compound sentences were included among the stimulus items and four training sentences (two compound and two complex) were prepared.
Forty-three subjects in five age groups were selected on the basis of the following criterion: (1) normal intelligence; (2) normal articulation; (3) normal hearing; and (4) social classification of "white".
The experimental task consisted of two parts: (1) sentence imitation and (2) sentence comprehension, each part following a taperecorded presentation of the stimulus sentence. Sentence comprehension was demonstrated by the manipulation of three dolls.
The data consisted of the subjects' responses to the imitation and comprehension tasks. The data were evaluated to determine the effects on performance of: (1) sex; (2) age; (3) sentence difficulty; (4) syntactic variables; (5) semantic variables. In addition, a closer
look at the types of imitation errors was taken. Treatment of the data was accomplished by the use of analyses of variance, Duncan's Multiple Range Test, and Spearman's Rank-Order Correlation.
It was determined that sex was not a factor affecting performance, while chronological age was a significant factor. Age groups one through four showed a progressive increase in correct performance with increases in age, while the fifth group declined in performance.
Sentence difficulty affected performance but the order of difficulty was not completely consistent among the age groups. The rank-order of sentences for imitation did not correlate with the rank-order for comprehension in any age group.
It was found that active sentences were significantly less difficult to comprehend than those of passive or mixed voice; and that sentences in which the common noun served a singular function (subject-subject or object-object) between clauses were less difficult than sentences in which the common noun served a dual function (subject-object or objectsubject).
Children were found to make fewer semantic errors on the imitation task as age increased, while the degree of syntactic errors did not change significantly with age.
The clause "boy pushed mommy" generated significantly more errors than similar clauses of the same form. This occurrence was noted as a possible semantic or expectancy effect--as if the children perceived this action as improbable or anomalous.
Thesis74.S234.pdf
4.01 MB
Adobe PDF
b1141f496fe4f0406037bc317ea87934