The Complexity of a First-Grade Core Reading Program Over Seven Decades: The Early-Grades Lexile Text-Complexity System Analysis

• In Event: 70.021 - Multiple Perspectives on the Complexity of First-Grade Texts Over the Past 50 Years

Abstract

Objectives/Purpose: This study was designed to address the question of whether first-grade texts have decelerated in difficulty from 1960 to the present.

Perspective: The unique perspective on text complexity in the present work involves the following basic assumptions/principles: (a) Understanding early-grades text complexity requires theorization about the child cognitions that are acquired in the initial phase of learning to read. (b) Complexity depends on more than the number and variety of a text’s constituent elements. It also depends on the elaborateness of the elements’ interrelational structure (Rescher, 1998, p. 1). (c) Early-grades texts are complex systems consisting of many mutually interacting characteristics and ensembles of characteristics to impact text complexity. Exploration of interplay among text characteristics is essential to successful explanation of complexity.

Sources/Methods: We used an innovative text-characteristic measurement system designed specifically for early-grades texts. First, using computer-based operationalizations, nine text characteristics were summarized for the each text in the corpus: for word-structure linguistic level—decoding demand, number of syllables in words; word meaning—age of acquisition, abstractness, and word rareness; and sentence and discourse level—intersentential complexity, phrase diversity, text density/information load, and non-compressibility. Second, for each of the five core-reading units within year and for each overall program year, Lexile measures were derived (by applying a previously determined algorithm from a random forest regression model involving the nine text characteristics). Third, to address differences in within-year progression across the seven years and differences across years on the whole, a hierarchical linear model was accomplished with the five units nested in the seven years. Fourth, for each program year, a series of visual analyses (e.g., CART scans, Nason, Emerson, & LeBlanc, 2004) was accomplished to examine text-complexity characteristics such as structure and variable interplay.

Results: (a) Overall Lexile level shifted from a median of 250L in 1963 to approximately 400L in 2007/2013. (b) The progression of complexity Lexile level within year was different across the seven years. Year 1993 displayed the most challenging progression, starting at a higher Lexile level (150L) than the other years and ending at a much higher level (810L). (c) The within-year text-complexity-level variation was somewhat different across years, with 1993 displaying the most within-year variability and the year 2000 the least. (d) Word structure and sentence/discourse-level variable medians remained reasonably consistent across years, but again, variability for both linguistic levels changed somewhat across years, with no general pattern of shift. However, a relatively major shift was witnessed for word meaning variables—age of acquisition and word rareness increased in challenge level over time, while abstractness tended to decrease over time. (d) Visual analyses revealed little to no differential interplay of selected text characteristics across years.

Significance: As one of the first studies of first-grade core-reading program complexity, the text-complexity descriptions by year are newsworthy. The first-grade results do not support the Common Core contention that the texts have become easier over time.

Authors

• Jill Fitzgerald, UNC at Chapel Hill and MetaMetrics

• Jeffrey Elmore, MetaMetrics

• Heather Hughes Koons, MetaMetrics