Generalizability of Scaling Gradients on Direct Behavior Ratings

Abstract

Generalizability theory is used to examine the impact of scaling gradients on a single-item Direct Behavior Rating (DBR). A DBR refers to a type of rating scale used to efficiently record target behavior(s) following an observation occasion. Variance components associated with scale gradients are estimated using a random effects design for persons (p) by raters (r) by occasions (o). Data from 106 undergraduate student participants are used in the analysis. Each participant viewed and rated video clips of six elementary-aged students who were engaged in a difficult task. Participant ratings are collected three times for each of two behaviors within three scale gradient conditions (6-, 10-, 14-point scale). Scale gradient does not substantially contribute to the magnitude of observed score variances. In contrast, the largest proportions of variance are attributed to rater and error across all scale gradient conditions. Implications, limitations, and future research considerations are discussed.

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 — 29 citations exist upstream.

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