Curriculum-based measurement of oral reading: A preliminary investigation of confidence interval overlap to detect reliable growth.

SCHOOL PSYCHOLOGY QUARTERLY Pub Date : 2016-02-11 DOI:10.1037/spq0000146

Ethan R. Van Norman

{"title":"Curriculum-based measurement of oral reading: A preliminary investigation of confidence interval overlap to detect reliable growth.","authors":"Ethan R. Van Norman","doi":"10.1037/spq0000146","DOIUrl":null,"url":null,"abstract":"Curriculum-based measurement of oral reading (CBM-R) progress monitoring data is used to measure student response to instruction. Federal legislation permits educators to use CBM-R progress monitoring data as a basis for determining the presence of specific learning disabilities. However, decision making frameworks originally developed for CBM-R progress monitoring data were not intended for such high stakes assessments. Numerous documented issues with trend line estimation undermine the validity of using slope estimates to infer progress. One proposed recommendation is to use confidence interval overlap as a means of judging reliable growth. This project explored the degree to which confidence interval overlap was related to true growth magnitude using simulation methodology. True and observed CBM-R scores were generated across 7 durations of data collection (range 6-18 weeks), 3 levels of dataset quality or residual variance (5, 10, and 15 words read correct per minute) and 2 types of data collection schedules. Descriptive and inferential analyses were conducted to explore interactions between overlap status, progress monitoring scenarios, and true growth magnitude. A small but statistically significant interaction was observed between overlap status, duration, and dataset quality, b = -0.004, t(20992) =-7.96, p < .001. In general, confidence interval overlap does not appear to meaningfully account for variance in true growth across many progress monitoring conditions. Implications for research and practice are discussed. Limitations and directions for future research are addressed. (PsycINFO Database Record","PeriodicalId":48005,"journal":{"name":"SCHOOL PSYCHOLOGY QUARTERLY","volume":"31 3 1","pages":"405-418"},"PeriodicalIF":0.0000,"publicationDate":"2016-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"SCHOOL PSYCHOLOGY QUARTERLY","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1037/spq0000146","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 1

Abstract

Curriculum-based measurement of oral reading (CBM-R) progress monitoring data is used to measure student response to instruction. Federal legislation permits educators to use CBM-R progress monitoring data as a basis for determining the presence of specific learning disabilities. However, decision making frameworks originally developed for CBM-R progress monitoring data were not intended for such high stakes assessments. Numerous documented issues with trend line estimation undermine the validity of using slope estimates to infer progress. One proposed recommendation is to use confidence interval overlap as a means of judging reliable growth. This project explored the degree to which confidence interval overlap was related to true growth magnitude using simulation methodology. True and observed CBM-R scores were generated across 7 durations of data collection (range 6-18 weeks), 3 levels of dataset quality or residual variance (5, 10, and 15 words read correct per minute) and 2 types of data collection schedules. Descriptive and inferential analyses were conducted to explore interactions between overlap status, progress monitoring scenarios, and true growth magnitude. A small but statistically significant interaction was observed between overlap status, duration, and dataset quality, b = -0.004, t(20992) =-7.96, p < .001. In general, confidence interval overlap does not appear to meaningfully account for variance in true growth across many progress monitoring conditions. Implications for research and practice are discussed. Limitations and directions for future research are addressed. (PsycINFO Database Record

查看原文本刊更多论文

基于课程的口语阅读测量:可信区间重叠的初步调查，以检测可靠的增长。

基于课程的口语阅读测量(CBM-R)进度监测数据用于测量学生对教学的反应。联邦立法允许教育工作者使用CBM-R进度监测数据作为确定特定学习障碍是否存在的基础。然而，最初为CBM-R进度监测数据开发的决策框架并不打算用于这种高风险评估。许多关于趋势线估计的文件问题削弱了使用斜率估计来推断进展的有效性。一项建议是使用置信区间重叠作为判断可靠增长的手段。该项目探索了置信区间重叠与真实增长幅度相关的程度，使用模拟方法。真实的和观察到的CBM-R评分是在7个数据收集持续时间(范围6-18周)、3个数据集质量或剩余方差水平(每分钟正确阅读5、10和15个单词)和2种数据收集计划中生成的。进行描述性和推断性分析以探索重叠状态、进度监测情景和真实增长幅度之间的相互作用。重叠状态、持续时间和数据集质量之间存在较小但具有统计学意义的交互作用，b = -0.004, t(20992) =-7.96, p < 0.001。一般来说，置信区间重叠似乎不能有效地解释在许多进度监测条件下真实增长的差异。讨论了对研究和实践的启示。指出了未来研究的局限性和方向。(PsycINFO数据库记录

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

SCHOOL PSYCHOLOGY QUARTERLY PSYCHOLOGY, EDUCATIONAL-

自引率

0.00%

发文量

期刊介绍： The flagship scholarly journal in the field of school psychology, the journal publishes empirical studies, theoretical analyses, and literature reviews encompassing a full range of methodologies and orientations, including educational, cognitive, social, cognitive behavioral, preventive, dynamic, multicultural, and organizational psychology. Focusing primarily on children, youth, and the adults who serve them, School Psychology Quarterly publishes information pertaining to populations across the life span.