The effect of grammatical form class on the comprehension of fast and slow readers

The problem concerns the ability of f as t and slow readers to predict grammatical form classes in context. It has been suggested that the fast reader is better able to make use of redundant sources of information in a passage, and therefore can more accurately predict the alternatives that follow in a sequence of words. 

An experiment was designed in which a prose passage was systematically mutilated and presented in the form of a cloze reading test. The subjects (first year university students) were required to predict the missing word in the gaps or blanks provided in the test. All subjects were tested for speed of reading comprehension and grouped according to fast and slow (the two extreme) groups. 

A 2x5 factorial design was employed in which the main effects were groups (fast and slow readers) and treatments (deletion of five different grammatical form classes). The dependent variables were (1) correct verbatim responses (2) correct form class responses and (3) time taken to complete test. The experimental task was one of reader expectation based on the comprehension of the passage. 

Slow readers did have more difficulty in predicting grammatical form classes than fast readers (p< .05). In particular, they had particular difficulty in predicting adjectives/ad verbs and prepositions/conjunctions. 

Some grammatical form classes were more difficult to predict than others, with adjectives/adverbs being the most difficult of the lexical classes, nouns more difficult than verbs, and prepositions/conjunctions more difficult than function words. These finding proved to be cons is tent with previous studies. There was no interaction groups and the five treatments. 

It is clear from this study that the slow reader's problem is not exclusively linked to speed. For when all time limits were removed from the test situation, the slow readers still made significantly more errors than the fast readers.