Advances in curriculum-based measurement: Alternative rate measures for assessing reading skills in pre- and advanced readers.

Abstract

Behavioral researchers have developed Curriculum-Based Measurement (CBM) procedures that yield a rate measure (words read correctly per minute). Investigators have shown that words correct per minute provides a valid, reliable, and sensitive measure of reading proficiency in students. Recently researchers have developed additional rate measures designed to assess pre-reading skills (Dynamic Indicators of Basic Early Literacy Skills) and skill development in advanced readers (reading comprehension rates). The current paper provides a rationale for both procedures and describes an emerging research base related to these measures. Limitations of these procedures along with directions for future research are provided. ********** Curriculum-based assessment (CBA) procedures have emerged as an alternative to traditional norm-referenced standardized achievement measures (Shapiro, 1996). Most CBA procedures use the student's current educational curricula to assess individual skills. Thus, CBA procedures assess what is taught and improve upon the lack of test-text overlap associated with norm-referenced standardized achievement tests (Bell, Lentz, & Graden, 1992; Martens, Steele, Massie, & Diskin, 1995; Shapiro & Derr, 1987). This characteristic of CBA addresses many legal, ethical, and educational concerns related to bias in assessment often associated with standardized tests (e.g., assessment results being overly dependent on extra-school learning). Because of CBA's overlap between what is taught and what is assessed, CBA procedures have clear relevance to instructional planning (Deno, 1989). For example, CBA measures can indicate skills or content students have mastered and areas that require additional work (see Skinner & Schock, 1995 for an applied example). A number of CBA models have been developed (Shapiro, 1990). Some CBA models focus on response accuracy (e.g., Gickling & Havertape, 1981) and others include criterion-referenced measures (e.g., Blankenship, 1985; Idol, Neven, & Paolucci-Witcomb, 1986). However, the model most closely associated with behavioral psychology is known as Curriculum-Based Measurement (CBM). Although CBM procedures have been developed for assessing a variety of skills including mathematics, spelling, and writing (see Deno & Mirkin, 1977 or Shapiro, 1996), the current paper will focus only on CBM reading procedures. After providing a brief overview, analysis, and research summary of traditional CBM reading measures, recent advances in CBM research will be described along with directions for future research. Words Correct Per Minute CBM reading procedures yield measures of oral reading fluency. During CBM assessment procedures, students read aloud for one minute from a passage within their reading text or series while the examiner marks errors (e.g., mispronunciations, substitutions, omissions, skipped lines). The examiner calculates two rate measures: words read correctly per minute and errors per minute (Deno & Mirkin, 1977; Shapiro, 1996). Typically, words correct per minute serves as the primary measure for making educational decisions (e.g., placement in the curricula, evaluation of intervention effectiveness). Researchers have shown that words correct per minute is a sensitive, reliable, and valid measure that can be used to assess students' general reading proficiency and to evaluate progress within students' reading curriculum (Deno, Mirkin, & Chiang, 1982; Fuchs & Deno, 1992; Fuchs & Fuchs, 1992; Fuchs, Fuchs, & Maxwell, 1988; Madelaine & Wheldall, 1999; Marston, 1989; Shinn, Good, Knutson, Tilly, & Collins, 1992). Several characteristics of CBM make words correct per minute ideal for assessing reading skills (Shapiro, 1996; Shinn, 1995). CBM procedures are brief and inexpensive. Multiple measures can be constructed from a student's curriculum. Because students are assessed using probes (i. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 38 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.