The Effects of Motor Fluency on Dressing Tasks and Decreasing Escape Behaviors

Abstract

Interventions aimed at teaching composite dressing skills assess individuals’ current abilities and improve their performance at more independent and fluent levels. The current case study aims to report on the improvements observed when specific component motor skills are targeted for improved performance using frequency building techniques. We extend previous research by including the measure of escape-maintained behaviors during task completion. Similar to those of previous research, the results of this study reveal a significant increase in speed and accuracy across all targeted composite skills and a reduction in escape-maintained behaviors.

1. A direct assessment of composite skills can be used to identify key component skills for intervention.

2. The frequency building of component motor skills can effectively improve accuracy in completing daily living tasks.

3. Addressing component motor skill deficits can reduce escape-maintained problem behaviors during daily living tasks.

4. Mastery of daily living tasks involves completing tasks accurately, fluently, and in the absence of escape behaviors.

The development of motor skills is crucial for performing and mastering daily activities (e.g., dressing), and it can lead to the acquisition of additional skills. Challenges in acquiring motor skills may impede an individual’s capacity to perform crucial self-care tasks necessary for independent daily functioning (Twarek et al., 2010). Self-care tasks such as brushing teeth, eating, and dressing can be defined as composite skills. A composite skill is a complex task that requires multiple component skills (Vascelli et al., 2020). Composite tasks, such as putting on pants or socks, consist of fine motor movements such as grasping the waistband or pulling on socks, respectively. These foundational motor movements of pulling and grasping are examples of component skills.

Precision teaching (PT) is a measurement system applied to skill acquisition to accelerate behavioral repertoires in any area to a fluent level of mastery (Evans et al., 2021). It involves the use of a standard celeration chart (SCC) to display performance rates, allowing clinicians to make data-based decisions. Building behavior rates to optimal levels is called frequency building, and it involves the repeated timed practice of a behavior up to a specified, and sometimes individualized, accuracy improvement measure that is often conceptualized as an aim (Pennypacker et al., 1972). The fluent performance of the component skills involved in more composite task completion can contribute to behavioral fluency and may result in generative skills that improve other behavior repertoires that are not explicitly targeted (Pallares et al., 2020; Johnson & Layng, 1992; Twarek et al., 2010).

Few studies have examined the effectiveness of component skill interventions in improving composite dressing skills (Twarek et al., 2010; Vascelli et al., 2020). Twarek et al. (2010) targeted the component skills involved in the composite tasks of putting on socks, shirts, and underwear. They recorded the duration of completing the whole-chain task analyses and compared the pre- and postintervention performances. The study reported that the participants displayed 80% accuracy in the steps completed (Twarek et al. 2010). Vascelli et al. (2020) targeted the composite skill of buttoning by using frequency building to improve the performance of this skill.

Neither Twarek et al. (2010) nor Vascelli et al. (2020) measured the effect of fluent component skills on escape-maintained behaviors during composite skill performance. Vascelli et al. (2020) did not report escape-maintained behaviors whereas Twarek et al. (2010) reported that the experimenter terminated the activity when the behaviors occurred during a composite or component skill. Therefore, the current case study aims to replicate component skill interventions to improve the composite task of independently putting on shirts, pants, socks, and shoes. In addition, we extend previous research by evaluating the effect of the intervention on escape-maintained behaviors during dressing tasks.

Method

Participants

Chris, a 4-year-old male; and Harley, a 4-year-old female, participated in the study. Chris and Harley, both diagnosed with autism spectrum disorder, received a comprehensive applied behavior (AB) analysis treatment in a Northern California clinic. The inclusion criterion for participating in this study was the completion of task analyses for putting on socks, shoes, shirts, and pants. Participants had to complete the tasks at or above 80% independence per task across 5 days. Both children took a significant amount of time to complete the tasks cumulatively when all items were presented.

Setting and Materials

The dressing tasks were completed in a bathroom with a primary clinician and a behavior therapist present. The component skills were practiced in the participants’ respective treatment rooms. The materials included a TheraBand, small objects (i.e., wooden blocks), rubber bands, and elastic headbands. A toy shopping cart was included for Harley. The materials for the therapists included a tally counter, timer, and stopwatch.

Target Behaviors

Component Motor Skills

The dependent measures were the component motor skills of reach and pull and the combined skills of push + pull and grasp + release. The left and right hands were used separately. Reach was defined as raising the arm to attempt contact with an object held in front. The items were moved to different positions to ensure that the arm crossed the midline, below the hip, and above the head. Pull (standing) was performed by grasping the TheraBand and pulling it until the arm was bent at a 90-degree angle alongside the body. Each participant performed pull (sitting) by sitting on the floor with the TheraBand around the feet and legs straight. The participant pulled the band back until the elbow made a 90-degree angle alongside the body. For both pull actions, the participant straightened the arm parallel to the floor before engaging in the action again. Push and pull involved sitting in a chair while grasping the shopping cart with one hand. The participant pushed the cart forward until it contacted the wall and then pulled it back. Grasp and release (rubber band) involved sitting on the floor with one leg pulled to the chest and the toes pointing toward the ceiling. The participant then stretched a medium-sized rubber band over the top of the shoe, mirroring the movement of putting on a sock. Grasp and release (headband) involved pulling the headband open and over the top of the head. This action mirrored the movement of putting one’s head through the collar of a shirt. Both the grasp and release component skills were practiced outside frequency building until they were observed to occur accurately and without hesitation.

Escape Behaviors

Escape behaviors for Harley were defined as requesting help without attempts at independence (i.e., asking for help immediately after being given a sock) and crying with or without tears. For Chris, escape behaviors were defined as throwing the clothing item, high-pitched vocalizing, motor stereotypy (i.e., hand flapping and noncontextual gestures), and vocal stereotypy (i.e., repeated sounds or noncontextual babbling).

Design

The current case study used two independent AB designs to evaluate the effects of fluent component skills on the completion of a composite task and on escape behavior. AB designs may not include all necessary elements of single-case experimental designs, such as verification and replication. Nevertheless, they have been proven as valuable sources of insight in research and practice (Brown & Kubina, 2022). The designs are similar to that of Vascelli et al. (2020) as they involved pre- and postapplication checks and the recording of the duration of task completion. Escape behavior frequencies were also recorded during the pre- and postapplication checks.

Procedure

Baseline Composite Assessment

The clinician recorded the duration of completing the dressing tasks of putting shoes, socks, pants, and a shirt. The team recorded the frequency of escape behaviors and the frequency and type of prompts delivered when the participant hesitated to respond for longer than 3 s.

Baseline Component Skills

Information from the composite assessment determined the component skills for the intervention. To test the current performance of the component skill, each participant was asked to engage in the target behavior for 1 min, and the number of movements was recorded. If the participant hesitated to engage in the action, a hand-over-hand guide was used to facilitate the performance of the action once and was then removed. Chris’s pull skill required a full guide. The full guide faded into a partial guide and was completely removed over the course of the intervention. To enable therapists to provide effective guidance and maintain a single record of data, only accurate movements were recorded for both participants. Incorrect movements were not documented.

Intervention

Chris’s intervention included pull (standing), and Harley’s intervention included push + pull. Both participants practiced the actions of reach and pull (sitting) for frequency building. Skills were introduced with a guided prompt until the movement was performed accurately and independently. Doing so allowed the participants to encounter reinforcement and practice the movements at the target frequencies. The aim of achieving fluent movement was to double the individual baseline performance or display approximately 100 movements per minute. Timing began at 6 or 15 s and was increased to 1 min. Specific instructions to perform the action quickly and accurately (i.e., “Let’s see how many times you can reach! Ready? Please begin!”) were provided. The therapists began the timer at the onset of engaging in the target action and used a handheld counter to record the correct movements. The therapists also counted the incorrect movements. Data from previous days were referenced to assess the current performance. The timings were run no more than five times per day. The therapists were instructed to take the best score out of all timings and record this number for the SCC. After obtaining 3 data points on the SCC, the therapists decided to increase, decrease, or continue the current timing interval. Decisions were based on whether the data showed an increasing or decreasing trend, were at or above the target, or displayed no changes.

Composite Skill Reassessment

The application was assessed after individual aims for all component skills were attained by performing the dressing task of independently putting on a shirt, pants, socks, and shoes. The application check also demonstrated the social validity of the intervention.

Data Collection

The daily count per minute chart, which records data across 140 days on the right axis and frequencies between 0.0007 and 1,000 counts per day measured on the left axis (Kubina & Yurich, 2012), was used in this study to record, display, and evaluate the data. Paper datasheets were used to record the timing interval, number of correct, and incorrect component skill movements. The ratio was then converted to obtain the count per minute. Decisions were also recorded on the data sheet and the SCC. The durations for dressing and the frequency of escape behaviors were recorded using web-based data collection software.

Celeration and Bounce

In addition to the visual analysis, the metrics for celeration and bounce were calculated using an online SCC with statistical software. Celeration specifies the accelerating or decelerating direction of a behavior and provides a learning rate (Kubina & Yurich, 2012). A count per minute per day measurement was used to calculate celeration. Bounce refers to the variability in a set of frequencies and was calculated herein as the ratio of the largest to the smallest frequencies (Pennypacker et al., 1972). Smooth and consistent performance suggested improved learning, whereas data points with high bounce and inconsistent or irregular performance suggested less influence on learning (Kubina & Yurich, 2012).

Interobserver Agreement

Interobserver agreement (IOA) was assessed by having a second independent observer record the frequency of trials during baseline dressing assessment, initial treatment, and reassessment. The IOA during dressing was calculated using exact agreement by following the target task analyses, taking the number of intervals in agreement, and dividing the result by the total number of intervals. The IOA values of Harley and Chris during the baseline dressing assessment were 98% (range: 96%–100%) and 97% (range: 94%–100%), respectively. During the dressing reassessment, the IOA values of Harley and Chris were both 100% (range: 99%–100%). The IOA for motor skills was assessed using the count per interval by having a second observer collect data on target actions simultaneously during timings. The mean count per interval was calculated by dividing the smaller count by the larger count in each targeted timing interval (i.e., 6 s, 15 s). Chris’s IOA for motor skills was calculated for 67% of the frequency building trials, with an overall IOA of 84% (range:76%–99%). Harley’s IOA was calculated for 43% of the frequency building trials, with an overall IOA of 98% (range: 95%–100%).

Results

Composite Measurements

The baseline and postintervention data are presented in Table 1. The overall time to complete the composite task of getting dressed and the frequency of escape-maintained behaviors during task completion both decreased for the participants.

Table 1.

Duration of dressing and frequency of escape behaviors

Composite dressing task	Baseline				Post- Intervention
	Duration		Escape behaviors (Frequency)		Duration		Escape behaviors (Frequency)
	C	H	C	H	C	H	C	H
Shirt	4m 0s	2m 12s	6	6	1m 7s	52s	1	3
Pants	4m 51s	57s	9	3	1m 3s	18s	0	0
Socks	4m 37s	2m 26s	5	8	57s	43s	0	2
Shoes	1m 03s	1m 51s	1	10	42s	29s	0	0
Total Duration	14m 52s	7m 43s	21	27	3m 49s	2m 22s	1	5

C Chris, H Harley, m minute, s seconds

The baseline data for Harley indicated that putting on socks took the longest amount of time at 2 min 26 s and that Harley engaged in eight instances of escape behavior. After the intervention, putting on socks took 57 s and two instances of escape behavior were observed. Putting on shoes evoked the highest number of escape behaviors, with 10 instances observed and a total duration of 1 min 51 s. After the intervention, putting on shoes took 29 s and no escape behavior was observed. The baseline duration for Harley to complete all dressing tasks was 7 min 43 s, with 27 observed incidents of escape behavior. After the intervention, the duration decreased to 2 min 22 s and five escape behaviors were observed.

The baseline data for Chris indicated that putting on pants took the longest amount of time at 4 min 51 s and evoked the highest number of escape behaviors in nine observed instances. After the intervention, Chris took 1 min 3 s to put on his pants and did not exhibit escape behavior. The baseline duration to complete all tasks was 14 min 52 s, with a total of 21 escape behaviors observed. After the intervention, Chris decreased task completion to 3 min 49 s and exhibited 1 escape behavior.

Component Skills

Figure 1 displays component skill trends on SCC segments. Celeration values during the intervention are reported with a multiplication sign representing an accelerating trend in Table 2. Celeration for component skills ranged from ×1.01 (1% weekly change rate ) to ×1.67 (67% weekly change rate). Harley’s values increased by an average of 11% during the intervention (×1.11 celeration). Chris’s values increased by an average of 21% during the intervention (×1.21 celeration). Bounce values are presented in Table 2, with scores under ×3.0 representing more control over the intervention (Kubina & Yurich, 2012). Harley’s bounce ranged from ×1.2 to ×1.7, with an average of ×1.43. Chris’s bounce ranged from ×1.2 to ×2.6, with an average of ×1.85.

Fig. 1.

Correct component skills

Table 2.

Fluency component skill outcome

Component motor skill	Harley						Chris
	Baseline		Intervention				Baseline		Intervention
			Bounce		Celeration				Bounce		Celeration
	LH	RH	LH	RH	LH	RH	LH	RH	LH	RH	LH	RH
Pull (sitting)	60	55	×1.4	×1.5	×1.12	×1.09	40	33	×1.2	×1.4	×1.14	×1.13
Reach	52	44	×1.5	×1.7	×1.12	×1.09	35	45	×2	×1.8	×1.07	×1.01
Push and pull	60	60	×1.2	×1.3	×1.12	×1.12
Pull (standing)							20	15	×2.1	×2.6	×1.24	×1.67

LH left hand, RH right hand

Discussion

This case study examined the effects of an intervention on component skills to improve the completion of multiple composite dressing tasks by two children with autism spectrum disorder. In addition, we examined the effects of the intervention on escape-maintained behaviors observed during the targeted dressing skills. The use of the SCC enabled the implementors to make prompt decisions by analyzing response rates, behavior variability, celeration, and frequency patterns over time (Bulla et al., 2021). Frequency building the component skills to optimal rates led to behavioral fluency in the composite task of dressing.

When dressing task analyses were introduced, escape-maintained behaviors were not reflected in the terminal performance criteria. Behaviors were low in intensity and responded to redirection or assistance in the completion of the step in the task. Behaviors were observed to be a barrier to fluent performance when all tasks were presented for completion. The results of the current study suggest that building component skills to fluent levels may be an effective procedure for improving composite task performance and decreasing escape behaviors. During the initial assessment, the participants exhibited escape behavior before receiving assistance. In particular, when Chris displayed motor stereotypy, the therapist provided a positional prompt and guided his hands to accurately complete the task analysis step. By analyzing the task prompts and identifying the necessary component skills, our team developed an effective intervention to address escape behavior.

The results of the current case study align with those of Twarek et al. (2010) and Vascelli et al. (2020) that support the measurement of the accuracy and speed of composite tasks and the use of PT methods during intervention. This study also examined the metrics of bounce and celeration on the basis of the SCC. Instructional control was measured using bounce by analyzing the tightness of the data distribution. A tighter bounce indicated greater stability whereas a wider spread indicated less regularity (Kubina & Yurich, 2012). A bounce of less than ×3.0 indicated little variation in the performance. The participants demonstrated strong control over the component skills, as evidenced by the minimal bounce of less than 2.6 across all movements. This result suggested adherence to the instructions given. Celeration quantifies the learning rate under a given condition. During the intervention, the participants exhibited consistent but low levels of celeration. Harley made steady progress in developing her skills and achieved the desired outcomes for all component skills within three months of the intervention. Likewise, Chris successfully achieved the aim for all component skills within 3 months. However, his progress showed some fluctuation during the frequency building phase, particularly when the timing duration was increased.

Component skills were applied across different types of clothing, such as long-sleeved or short-sleeved shirts, pants, shorts, and long and short socks. We did not directly measure task performance outside the clinical setting and relied on indirect reporting from caregivers. Harley’s family reported a decrease in escape behavior and improved performance with both parents providing the instruction on getting dressed. Chris’ family did not initially observe a decrease in escape behaviors but reported a decrease during the 2-month follow-up. Chris’s father reported a decrease in the duration of getting dressed for school and preparing for bed.

Although the current study produced significant outcomes, it has a few limitations. Clinicians could not control dressing opportunities beyond the clinical settings. Caregivers were informed about the intervention and could have thus reduced their assistance during the dressing tasks at home. Such reduced assistance could explain the decrease in escape behaviors whereas increased practice could have improved task fluency. In addition, the replication design did not demonstrate a functional relationship between fluent component skills and improvement in composite tasks. The reassessment of dressing tasks was conducted once all component skills had reached fluent levels. Future studies should consider conducting weekly evaluations of composite tasks during skill intervention programs to establish the potential correlation between improved component skills and overall task performance.

This study provides valuable insights into PT and the effectiveness of interventions aimed at improving component skills for composite tasks. The findings of this case study suggest that component skill interventions may be effective in reducing escape-maintained behaviors during dressing tasks. In this study, we observed that the fluent performance of component skills resulted in a decrease in escape behaviors. However, we did not investigate the effects of the fluent performance of the targeted component tasks on additional composite tasks. Future studies should examine component skill interventions for decreasing challenging behaviors across additional daily living tasks and the acquisition of novel skills. When analyzing composite tasks, clinicians should evaluate escape-maintained behaviors and identify any deficits in the component skills. Improving the fluency of component skills enables the development of generative skills that increase independence and accelerate the acquisition of new skills.

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Informed Consent

Informed consent was obtained from the participant before the intervention was implemented.

Ethical Approval

All procedures in the context of service delivery were in accordance with the ethical standards of the national research committee.

Conflict of Interest

All authors declare they have no conflict of interest.