The use of performance criteria for determining “mastery” in discrete‐trial instruction: A call for research

Abstract

The performance criterion (also known as the mastery criterion) is the operationalized performance standard that behavior must reach before instruction ceases, changes, or becomes focused on other goals, such as stimulus generalization and response maintenance. Although performance criteria are widely used in skill-acquisition research and practice, there has been little experimental research on the topic. Thus, we provide a review of relevant research and offer suggestions for future research based on the different performance-criterion components.

A preliminary comparison of mastery criterion frequency values: Effects on acquisition and maintenance

“…Two recent studies found that both 90% and 100% acquisition‐criterion levels produced adequate maintenance (Longino et al., 2021; Pitts & Hoerger, 2021). The outcomes of these four experiments are similar to those obtained with college students, which have shown that stringent performance criteria produce more durable responses (Fienup & Brodsky, 2017; Fienup & Carr, 2021; Johnston & O'Neill, 1973; Semb, 1974; Richling et al., in press).…”

“…A better understanding of how acquisition‐performance criteria impacts efficiency and effectiveness could have a real‐world positive impact on learners. There is little research on performance criteria within the context of responses taught using discrete trial instruction, despite the ubiquity of this component in discrete instruction (Fienup & Carr, 2021; McDougale et al., 2020). Further research would reveal how various performance criteria values interact with the different components of instruction.…”

“…Regardless, researchers should look at how the number of teaching phases impacts gains in efficiency and the outcomes of this study may not apply to single-phase teaching, such is the case with error correction-only studies (e.g., see Cariveau, La Cruz Montilla, et al, 2019) or trial-based most-to-least prompting whereby fading intrusive prompts is conducted based on consecutive trials rather than session-based criteria (Cengher et al, 2016). Additionally, it is not clear if results would be replicated with different types of target responses (e.g., tacts), a different performance-criterion level (e.g., 80% or 100% accuracy), or learners at different developmental levels (Fienup & Carr, 2021). Indeed, we speculate that children with less developed repertoires, such as those studied by Richling et al (2019), may require longer frequencies of performance criterion in order to maintain target responses after instruction.…”

“…The four published empirical evaluations of acquisition‐performance criterion effects with children with disabilities provide evidence that high levels of performance—across one (Fuller & Fienup, 2018) or three observations (Longino et al., 2021; Pitts & Hoerger, 2021; Richling et al., 2019)—are more likely to maintain over time. A component of performance criteria that has not been studied is the frequency component, or more specifically, the criterion‐level frequency (Fienup & Carr, 2021). The criterion‐level frequency of performance criteria is the number the number of consecutive observations of criterion‐level responding required to terminate a teaching phase (e.g., 90% across one observation vs. 90% accuracy across three observations) (Fuller & Fienup, 2018).…”

Assessing the efficacy of and preference for positive and corrective feedback

“…Because steps contained multiple components (e.g., using the correct number and type of dice on a roll), participants needed to complete all components correctly for the step to count as correct. The mastery criterion was completing three consecutive trials of a game (a trial consisted of completing one entire task analysis one time for one game) with 100% accuracy (Fienup & Carr, 2021; McDougale et al, 2020). No participants were excluded for not meeting the mastery criterion during the acquisition phase.…”