The effects of varying levels of reading complexity on measures of comprehension in reading Spanish expository prose

Mary Sue Barry, Purdue University

Abstract

A native English-speaking adult confronted with an authentic Spanish text may have a variety of problems which interact to impede the reading process. Some difficulties are reader based, while others are created by the specific characteristics of the text. One of the reader-based problems, predicted to decrease readability of a text, is the lack of familiarity with the content/topic. Another source of reading difficulty related to text characteristics is embedded phrases or clauses inserted as elaborations and/or digressions. Several researchers have found a higher incidence of this characteristic in written Spanish discourse than in similar English texts. Two experiments were conducted with Spanish language texts. The passages were written as expository discourse and organized as a collection of descriptions. Experiment I investigated the effect of increasing the number of additive clauses on proportional recall scores and timed recall scores for subjects familiar with the content of the reading selection. Each subject was presented with three different topics at three different levels of reading complexity defined as the number of additional embedded clauses per sentence. For each topic, there was a kernel reading defined as level-one reading complexity that presented the information essential to the learning task. The reading selections defined as second and third level reading complexity contained the kernel reading of each topic plus one and then two additional embedded clauses. These additions were either elaborations or digressions which did not present information essential to the learning task. Results show that increasing the number of embedded clauses decreases the average proportional recall score as well as the average time per proposition recalled. Experiment II investigated the effects of the same text characteristic, i.e., three levels of reading complexity, repeated for the same three topics, on the proportional recall scores of subjects unfamiliar with the content of the reading selections. The proportional recall scores followed the same pattern as for Experiment I, the with-schemata group recalled considerably more from the kernel text than the without-schemata group. However, this advantage disappeared by the third level of reading complexity.

Degree

Ph.D.

Advisors

Garfinkel, Purdue University.

Subject Area

Language|Literacy|Reading instruction|Educational psychology

COinS