Time delay procedures have been used for teaching skills such as sight words (Gast, Wolery, Morris, Doyle, & Meyer, 1990), sign language (Browder, Morris, & Snell, 1981), social studies and health facts (Wolery, Cybriwsky, Gast, & Boyle-Gast, 1991), spelling (Coleman-Martin & Heller, 2004), instruction following (Striefel, Bryan, & Aikins, 1974), spontaneous speech (Charlop, Schreibman, & Thibodeau, 1985; Ingenmey & Van Houten, 1991; Taylor & Harris, 1995), vocational assembly tasks (Walls, Haught, & Dowler, 1982), gross motor skills (Zhang, Horvat, & Gast, 1994), word identification (Browder, Hines, McCarthy, & Fees, 1984; Lalli & Browder, 1993) and a variety of other skills (Walker, 2008, Wolery, Ault, & Doyle, 1992). Time delay procedures typically involve the presentation of a discriminative stimulus (e.g., a flashcard with the word LAUNDRY on it) followed by the delivery of an instructional cue (e.g., What word is this?) followed by the provision of the controlling prompt (i.e., stating the correct response; e.g., laundry). In cases where the delays of 0 s (i.e., no delay) are employed the teaching method is often termed as errorless learning (Touchette, 1971, Touchette & Howard, 1984), whereas delays of longer than 0 s are termed as delay. In both procedures the controlling prompt cues the student to engage in the correct response (i.e., a prompted correct response). However, once a delay for providing the controlling prompt is introduced the student has the ability to respond to the salient features of the stimulus independent of the controlling prompt which is the goal of any instructional strategy (i.e., providing independent correct responses). Two typically utilized time delay procedures are constant time delay and progressive time delay procedures (Walker, 2008). Both procedures are similar in implementation: the presentation of the discriminative stimulus, paired with the delivery of an instructional cue with a specified delay to the controlling prompt. The procedures differ on how the delay to when the provision of the controlling prompt is provided. In constant time delay, an arbitrary delay is often selected (e.g., 5 s; Coleman-Martin & Heller, 2004; Lalli, Casey, Goh, & Merlino, 1994), however in progressive time delay, the delay is typically faded in increments (e.g., 1-2 s) based on preset number or percentage of correct responding criteria (e.g., correct responses in 3 consecutive trials; Taylor & Harris, 1995). The effectiveness of both versions of these time delay procedures have been documented through literature reviews regarding time delay procedures (Handen & Zane, 1987; Walker, 2008). However, it has been reported that constant time delay procedures have been associated with slightly more errors and longer delays in the transfer of stimulus control (i.e., correct responding no longer being cued by the controlling prompt; Walker, 2008). Thus, progressive time delay procedures appear to have some advantages over constant time delay procedures. Progressive time delay procedures typically proceed from 0 s to some delay contingent upon a set criterion of correct responding being met. For example, a 0 s delay will be faded to a 2 s delay, then to a 4 s delay and so on until some ceiling limit is reached (e.g., 10 s delay; Taylor & Harris, 1995). In most applications of progressive time delay, delays are typically increased across sessions based on correct responding. However, Touchette's (1971) original procedure consisted of within-session increases to the onset of the controlling prompt for the next trial by 0.5 s following a correct response, and reducing the delay on the next trial by 0.5 s following an incorrect response. Such changes can be burdensome in community based settings, especially when this criterion for increasing/decreasing the delay is applied to stimuli being trained concurrently. …
Read full abstract