Hierarchical Linear Modeling Meta-Analysis of Single-Subject Design Research

被引：24

作者：

Gage, Nicholas A. ^{[1
]}

Lewis, Timothy J. ^{[2
]}

机构：

[1] Univ Connecticut, Storrs, CT 06269 USA

[2] Univ Missouri, Columbia, MO 65211 USA

来源：

JOURNAL OF SPECIAL EDUCATION | 2014年 / 48卷 / 01期

关键词：

single-subject research; meta-analysis; HLM meta-analysis; QUANTITATIVE SYNTHESIS; STATISTICAL-INFERENCE; BEHAVIOR ANALYSIS; EFFECT SIZES; INTERVENTIONS; PERCENTAGE; STUDENTS;

D O I：

10.1177/0022466912443894

中图分类号：

G76 [特殊教育];

学科分类号：

040109 ;

摘要：

The identification of evidence-based practices continues to provoke issues of disagreement across multiple fields. One area of contention is the role of single-subject design (SSD) research in providing scientific evidence. The debate about SSD's utility centers on three issues: sample size, effect size, and serial dependence. One potential method for addressing all three issues is hierarchical linear modeling (HLM) meta-analysis. This study explored the utility of HLM meta-analysis of SSD. A total of 206 functional behavior assessment-based intervention outcome graphs were aggregated to assess whether HLM meta-analysis could identify (a) an overall effect size and statistical significance for mean shift, slope, and variability; (b) how the results mapped to two additional effect size calculations; and (c) whether the procedure met SSD synthesis criteria outlined by Wolery, Busick, Reichow, and Barton.

引用

页码：3 / 16

页数：14

共 64 条

[1] CALCULATING EFFECT SIZES FOR METAANALYSIS - THE CASE OF THE SINGLE-CASE [J].