Estimating Trend in Progress Monitoring Data: A Comparison of Simple Line-Fitting Methods

Abstract

Four simple line-fitting procedures are presented for practitioners to quickly summarize student time series performance data. Two are in common use — Koenig's “quarter-intersect” and White's “split-middle” adjustment, while two — Tukey I and Tukey II — are less known. Each of the four can be performed on a medium-size classroom dataset in less than 3 minutes. The four procedures were assayed against three criteria: (a) matching line slopes to an ordinary least squares (OLS) standard; (b) “best fit” to the data (minimizing residuals); and (c) prediction of a future reading score at Week 16. Weekly oral reading fluency data were collected on 45 Grade 4 and 5 students with reading disabilities, over a period of 12 weeks. Tukey I and II techniques generally outperformed the Koenig and White line-fitting methods, especially White's “split-middle” adjustment. Performance differences were generally large enough to be educationally meaningful. Given the small database supporting the popular White and Koenig procedures, the authors recommend that practitioners cautiously try out Tukey I and II procedures, comparing results with Koenig's and White's procedures. Of course, further psychometric studies of all four procedures are needed also. The authors discuss three notable study limitations: limited generalizability, use of the future score prediction criterion, and no instructional use of the best-fit lines.

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 — 19 citations exist upstream.

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