Time Delay Procedures: Constant and Progressive

What Is Time Delay?

Time delay is a prompt fading procedure in which a delay is inserted between the delivery of the instruction (the natural discriminative stimulus) and the delivery of a controlling prompt. Over time, the learner begins to respond before the prompt is delivered, which shows that stimulus control has transferred from the prompt to the natural stimulus. Because the prompt is delivered only after the learner has had a chance to respond, errors are minimized, which makes time delay an errorless learning procedure.

The BCBA exam often asks you to identify time delay from a description, distinguish constant from progressive versions, and decide when it is appropriate.

The Logic Behind the Procedure

At first, the instructor delivers the instruction and the controlling prompt at the same time. This is the zero-second delay. The learner responds correctly because of the prompt and contacts reinforcement. Gradually, a delay is introduced. Because the learner has already experienced the pairing of the instruction with correct responses and reinforcement, the learner begins to anticipate the prompt and respond on their own. The instruction thus acquires stimulus control over the response.

Constant Time Delay

In constant time delay (CTD), the delay interval is fixed after the initial zero-second trials. For example, after several trials with a 0-second delay, the instructor uses a 4-second delay on every trial thereafter.

Procedure

1. Choose the target skill, the instruction, and a controlling prompt that reliably produces a correct response.
2. Select the fixed delay interval, typically between 3 and 5 seconds.
3. Conduct a number of 0-second delay trials, in which the instruction and the prompt are delivered together. Reinforce correct responses.
4. Switch to the fixed delay. Present the instruction, wait the set interval, and then deliver the prompt if the learner has not responded.
5. Record each response type (see below) and reinforce correct responses.
6. Continue until the learner responds independently at the criterion level.

Progressive Time Delay

In progressive time delay (PTD), the delay is increased gradually across trials or sessions. A typical sequence might begin at 0 seconds, then increase by one second every few trials (1, 2, 3, 4 seconds) until it reaches a maximum such as 5 or 6 seconds.

Procedure

1. Begin with several 0-second delay trials.
2. Increase the delay by a set amount (for example, 1 second) after a specified number of trials or sessions.
3. Continue increasing until the learner responds before the prompt or until the maximum delay is reached.
4. Reinforce correct responses before and after the prompt, with higher reinforcement for unprompted responses.

Constant Versus Progressive

• Constant: A single, fixed delay after the zero-second trials. Easier to implement and often equally effective.
• Progressive: Gradual increases give the learner more time to adjust. May be better for learners who would be confused by a sudden jump in delay.
• Research suggests both are effective and that CTD is often somewhat more efficient, but the best choice depends on the learner and the skill.

Recording Responses During Time Delay

Time delay provides rich data because each trial can be categorized by the learner's response relative to the prompt. The four typical categories are:

• Unprompted correct: The learner responds correctly before the prompt. This shows progress toward independence.
• Prompted correct: The learner waits for the prompt and responds correctly.
• Unprompted error: The learner responds incorrectly before the prompt. These are called anticipation errors.
• Prompted error or no response: The learner responds incorrectly after the prompt or does not respond.

Anticipation errors are particularly informative. A few such errors can show that the learner is starting to respond before the prompt but does not yet have the discrimination. If errors are frequent, shorten the delay or revisit the prompt.

Advantages of Time Delay

• Low error rates: Errors are minimized because the prompt is delivered whenever the learner is unlikely to respond correctly.
• Reduced frustration: Learners experience a high rate of reinforcement.
• Simple to implement: The delay is easy for staff and parents to learn.
• Reduced risk of prompt dependence: The built-in delay gives the learner an opportunity to respond without a prompt.
• Versatile: Can be used for many skills, including receptive and expressive language, sight words, daily living skills, and chained tasks.

Limitations

• Learners must be able to wait and attend during the delay. Learners who are highly impulsive may produce anticipation errors.
• If the controlling prompt is not truly controlling, errors will still occur.
• If the delay is lengthened too quickly, the learner may become frustrated or start to wait for the prompt every time.
• Learners who do not benefit from the initial pairing of prompt and instruction may need other approaches.

Comparing Time Delay to Other Prompting Strategies

Most-to-least prompting reduces the intrusiveness of the prompt over time, while time delay keeps the same prompt but delays its delivery. Least-to-most prompting allows the learner to respond independently first and escalates support when needed, which allows more errors. Time delay provides the learner with an opportunity to respond independently on every trial (after the initial zero-second trials), but delivers the prompt quickly enough to prevent most errors. Because it has no prompt hierarchy to manage, it is easy to implement consistently.

Example: Teaching Letter Names

1. Present the card with the letter B and say, "What letter?"
2. During 0-second trials, immediately say "B" as a verbal prompt. The learner echoes "B" and receives praise and a token.
3. After 6 trials, switch to a 4-second constant delay. The instructor says, "What letter?" and waits.
4. If the learner says "B" within 4 seconds, deliver high-value reinforcement.
5. If the learner does not respond, say "B." If the learner echoes, deliver lower-value reinforcement.
6. Track unprompted correct responses and move to the next target when the learner reaches 90 percent over two sessions.

Common Implementation Errors

• Skipping the zero-second trials, which leaves the learner without exposure to the pairing.
• Failing to wait the full delay before prompting.
• Reinforcing prompted and unprompted responses equally.
• Using a prompt that does not reliably evoke the correct response.
• Not recording the type of response, which makes it impossible to evaluate progress.

How This Appears on the BCBA Exam

Look for descriptions such as "the instructor presented the instruction, waited four seconds, and then modeled the response." That is constant time delay. A description of delays that increase from 0 to 1 to 2 to 3 seconds is progressive time delay. You may be asked why time delay is described as errorless, which prompting procedure results in the fewest errors, or how to interpret anticipation errors. You may also be asked what the first trials of the procedure look like (zero-second delay).

Key Takeaways

• Time delay transfers stimulus control by inserting a delay before the prompt.
• Begin with 0-second trials, then use a fixed (constant) or increasing (progressive) delay.
• Track unprompted correct, prompted correct, and error types.
• Reinforce unprompted responses at a higher level than prompted responses.