Calculating Effect Size in Single-Case Research

Abstract

A guide for researchers to calculate effect sizes for single-case research using three nonoverlap methods is provided. These methods are evaluated for their strengths and limitations, applied to a data set, and contrasted with one another. Criteria are presented to determine goodness of fit between nonoverlap procedures and data sets.

Effect sizes as result interpretation aids in single‐subject experimental research: description and application of four nonoverlap methods

“…Second, PND can be used with graphs that have small numbers of data points across phases (that is, n < 20). Third, effect size estimates calculated using PND correlate well with the judgements of visual analysts (Lenz, ; Parker et al., ; Parker, Vannest & Davis, ). One notable disadvantage of PND is that effect sizes using PND are calculated based on a single datum point in the baseline.…”

“…Therefore, PND is vulnerable to an outlier that may exist in the baseline. For example, if one of the baseline datum points approaches the ceiling or floor of the possible score range for the dependent variable, it is possible that PND would show no effect, while an increase in the intervention phase is obvious (Lenz, ).…”

Effectiveness of Dialectical Behavior Therapy for Treating Eating Disorders

“…As noted by Lenz (), the need to demonstrate use of evidence‐supported interventions and accountability practices to third‐party payers is becoming increasingly important for counselors. Given that most of the participants included in studies evaluated within this analysis were Caucasian (89%) women (95%), it is reasonable to surmise that the results of this study may be useful only for counselors treating individuals with similar demographic identities.…”

“…This is a concern, given the contemporary trend for counseling practitioners to demonstrate that interventions with clients are either evidence based or supported by outcome data. Lenz () noted that the promulgation of interventions supported by empirical evidence has positive implications related to the well‐being and economic status of clients and that in an effort to contain costs while providing suitable care to clients, third‐party payers are progressively moving toward business models that exclusively reimburse counselors trained in outcome‐based interventions. Therefore, we regard it as a professional imperative to begin quantitatively consolidating the available outcome evidence related to the treatment of eating disorders using DBT.…”

The Effects of Repeated Exposure to the ARRAY Program With Offenders With Mental Illness

“…We evaluated individual participant ratings on PHQ items using visual and nonoverlap analyses as suggested by Lenz () and to be consistent with our previous evaluations of interventions with OMI (Cox et al, ; Lenz, Speciale, & Aguilar, ). Visual inspection of data included an analysis of data trend and variability as recommended by Vannest and Ninci ().…”