Step 3.1 Determine whether or not PRT has been effective

Determine whether or not PRT has been effective for each meaningful

objective and for the toddler’s overall progress.

As much as possible, the difference between baseline and intervention data should be the effect of the intervention. Otherwise, parents and practitioners may be less confident about whether the intervention was actually effective.

Progress monitoring will result in either proceeding with the next goal of a successful intervention, or making a modification to a less successful intervention. If the intervention is implemented with low fidelity, that is, not as the intervention was intended, modifications or re-training are implemented.

Specific Considerations for Monitoring PRT

Much of PRT’s data collection is based on the procedures commonly used in applied behavior analysis (ABA) interventions. Select the type of data to collect based on:

  • the type of target behavior,
  • the intention of the intervention (e.g., to decrease or increase behavior), and
  • the feasibility associated with collecting the different types of data.

Typical types of data include: frequency or event-recording (e.g., rate and percent occurrence/correct), partial and whole interval, duration, and latency recording. However, there are some additional considerations specific to PRT.

  1. Data collected from language samples during play and daily routines can easily be converted into a variety of behavior data. See Planning PRT for a more in-depth look at communication samples and the types of data that can be extracted. PRT may make more frequent use of language or communication samples.

  2. Data are collected within everyday environments, rather than special purpose assessment contexts.

  3. Monitor whether settings and antecedents are being varied. Given that initiated, spontaneous, and generalized behaviors are key in PRT, it is important to track the varied types of prompts and instructions provided, as well as the varied settings and learning situations. These data on instructional variation assess the association between parent or practitioner instructional behaviors and child behavior and progress.

Sample Questions to Ask When Monitoring Pivotal Response Treatment

Are parents and practitioners providing enough opportunities for spontaneous or initiated child behavior, or instead are they directly initiating most of the interactions through model prompts, choices, and questions? If most of the opportunities were model prompts, this might explain lower levels of a toddler’s spontaneous use of target behaviors.

Are the acquired skills generalizing to diverse situations and are diverse situations being provided? Answering “no” suggests the need to consider teaching with more diverse stimuli, with additional people, and in multiple contexts.

Can the toddler appropriately respond to a variety of relevant antecedents and are you varying stimuli and tasks, such as cues and questions, at appropriate levels? If the toddler cannot respond appropriately, you will need to improve the implementation of PRT.

4. Vary target behaviors in trial-by-trial (opportunity-by-opportunity) data. When collecting percent correct through trial-by-trial data recording, do not provide too many of the same trial or opportunity consecutively. When collecting data on toddler performance during 10 total opportunities, those 10 opportunities are rarely provided all in a row. Avoid “drilling” target behaviors. Data collection systems can support this by providing data recording options for multiple objectives on a single form.

5. Individualize how and when data are collected. This helps data to be as easy as possible to collect. For instance, one mother who was working on question asking with her son started with having her toddler ask where his shoes were before going outside (to preferred activities). In order to make data collection easier, his mom placed an index card on a table by the doors to outside. Mom also appreciated that the card acted as a cue – to remind her to provide the opportunity! Some families put data sheets/cards/post-its on the refrigerator or in the car. A more tech savvy parent might want to set up a free note application on their smart phone to collect data, because it is easy to carry around, they might usually have it with them, and they enjoy technology. Another parent might keep a data sheet by the toddler’s bed, because they are working on teaching their toddler to answer a variety of questions as they read books before bed.

6. Collect representative data probes instead of constant data on every PRT learning opportunity. Data collection is important for monitoring progress, but it may not be necessary to take data on every opportunity provided. A practitioner might choose one or two days a week to collect data across multiple objectives or for certain objectives. You may choose to take data on one objective for a day, however, data collection should be frequent enough to be representative and allow for modifying the intervention in a timely manner, if needed (Koegel & Koegel, 2012).

7. Collect data related to motivation and initiations. Data is collected in order to measure observable characteristics of motivation across learning and social situations. This type of data considers measurable behaviors aside from just accuracy (e.g., correct or incorrect responses) and can include:

  • Responsivity: Is the toddler trying? Percent responding with good attempts
  • Independence: Is the todddler being independent? Percent spontaneous
  • Initiation: Is the toddler starting interactions, engagement, and learning on their own? Rate of initiating
  • Engagement: Is the toddler engaging quickly and for appropriate lengths of time? Latency to and duration of engagement