ABA ( APPLIED BEHAVIOR ANALYSIS)- antecedent intervention - visual schedule

Vol:.(1234567890)

Journal of Behavioral Education (2021) 30:112–129 https://doi.org/10.1007/s10864-019-09358-1

1 3

ORIGINAL PAPER

Effect of Task Sequence and Preference on On‑Task Behavior

Tiffani Warren1 · Rachel R. Cagliani1  · Erinn Whiteside1 · Kevin M. Ayres1

Published online: 13 November 2019 © Springer Science+Business Media, LLC, part of Springer Nature 2019

Abstract This study compared effects of student choice of task sequence to two variations in teacher-manipulated task sequences on on-task behavior of elementary-aged stu- dents with disabilities. Researchers modified Call et  al.’s (J Appl Behav Anal 42: 723–728, 2009) demand assessment to determine high-, moderate-, and low-prob- ability tasks. Next, researchers applied the results from the demand assessment to inform teacher-manipulated variations in task sequences: a high- to low-proba- bility task sequence and low- to high-probability task sequence. These sequences were then embedded in a visual activity schedule (VAS). Results of task sequence manipulation embedded in a VAS indicated slightly higher median percentages of on-task behavior for the high- to low-probability task sequence. Future directions for research based on these preliminary data are discussed.

Keywords On-task behavior · Off-task behavior · Visual activity schedules · Task sequence · Student choice

Introduction

Classroom success requires active engagement of students; many educators iden- tify on-task behavior as one of the most highly desirable student behaviors (Rich- ards et  al. 2010). Some children with autism and other disabilities may engage in challenging behavior that reduce engagement and interfere with learning (Mira- montez and Schwartz 2016). Disruptive learning environments reduce the amount of instructional time, because teachers have to reallocate teaching time to manag- ing off-task behavior (Santoyo et al. 2017). Therefore, students who engage in high rates of off-task behavior have decreased exposure to learning through meaningful instruction (King et  al. 2014). Kranak et  al. (2017) reported that 40% of teachers

* Rachel R. Cagliani [email protected]

1 Center for Autism and Behavioral Education Research, University of Georgia, 110 Carlton Street, Athens, GA 30602, USA

113

1 3

Journal of Behavioral Education (2021) 30:112–129

say management of students’ challenging behavior is the largest inhibitor of effec- tive instruction. In classroom settings, students may engage in off-task behavior to escape academic demands (May and Howe 2013).

The negative effects that off-task behavior has on teachers and classroom instruction highlight the need for practical and efficient interventions to increase the amount of time that students are on-task (Kranak et al. 2017; Miramontez and Schwartz 2016). Antecedent interventions (AI) are a commonly used approach to address challenging behavior in classroom environments (Einfeld et al. 2013). This type of intervention ideally makes challenging behavior less likely to occur by manipulating the environmental variables that occur prior to an individual engaging in problematic behavior (Kuo and Plavnick 2015). Not only have AIs been shown to be effective for individuals with various disabilities (Kern et al. 2002), they can help teachers avoid lost instructional time and reduce disruptions that occur as a result of reacting to a behavior after the fact (Luke et al. 2014). Further, in Kern et al.’s (2002) review on AIs, the authors reported that 50% of the studies included specifi- cally focused on on-/off-task behavior.

In addition to off-task behavior in the middle of instruction, students may engage in disruptive or off-task behavior during transitions between activities and this can in turn reduce instructional time (Hine et al. 2015). Visual activity schedules (VASs) are an antecedent-based intervention consisting of a series of pictures, images, pho- tographs, or line drawings sometimes coupled with word descriptors designed to illustrate an order of events (Knight et al. 2015). This provides a salient graphical cue which may enhance predictability for students (Knight et al. 2015; Lequia et al. 2012).

Researchers have evaluated VAS for increasing on-task behavior (Bryan and Gast 2000; Carlile et  al. 2013; Cuhadar and Diken 2011; Duttlinger et  al. 2012; Hume and Odom 2007; MacDuff et al. 1993; Morrison et al. 2002) and decreasing chal- lenging behavior (Lequia et al. 2012; Massey and Wheeler 2000; Waters et al. 2009; Zimmerman et  al. 2017). Teachers and therapists have used VAS to teach leisure skills (Betz et al. 2008; Blum-Dimaya et al. 2010; Carlile et al. 2013; Cuhadar and Diken 2011; Dettmer et al. 2000; Hume and Odom 2007; MacDuff et al. 1993), aca- demic skills (Bryan and Gast 2000; Hume and Odom 2007; Waters et  al. 2009), transitions (Cihak 2011; Pierce et al. 2013), and independent living skills (Mechling et  al. 2009; Mechling and Gustafson 2008; Morrison et  al. 2002; Van Laarhoven et al. 2010). While all VASs incorporate images representing a series of tasks, the presentation may vary based on the type of settings, available resources, and user.

VAS can be used flexibly across settings and behaviors, and variations in the sequence of tasks may affect student behavior (e.g., on-task behavior and chal- lenging behavior). Three different behavioral approaches may offer guidance on how task sequence in a VAS may influence responding: behavior momentum theory (BMT), response deprivation hypothesis (Premack principle), and stu- dent choice for a sequence. BMT in this case would involve presenting the task sequence in an order of activities in which the individual is most likely to engage (i.e., high-probability behaviors) followed by a demand in which the individual is less likely to engage in (i.e., low-probability behavior). In a classroom context, researchers have explored BMT with interspersal of mastered content (Ardoin

114 Journal of Behavioral Education (2021) 30:112–129

1 3

et al. 1999; Davis et al. 1994; Jung et al. 2008). For example, a teacher might pre- sent three mastered letters to a student and ask them for the sound before present- ing one new instructional target that the student has not mastered. Thus, structur- ing a schedule with the greater probability of completion activities first might result in more on-task responding which may carry over (i.e., generate momen- tum) into the less probability of completion tasks situated last.

In contrast to BMT, the response deprivation hypothesis or Premack princi- ple operates on the presumption that a high-probability behavior, if restricted, can function as a reinforcer for performance of a low-probability behavior. This presentation of events is sometimes known as “first, then” or “grandma’s law” (Cooper et  al. 2007). Tingstrom and Edwards (1989) recommended teachers to arrange activities so that activities that students are less likely to complete are assigned before activities students are more likely to complete. With a VAS, a teacher would sequence the schedule based on the probability of task completion from low to high probability.

A third option involves allowing students the opportunity to choose the order in which they complete the required activities. Researchers have evaluated stu- dent choice of task sequence and found that incorporating student choice led to increases in on-task behavior and decreased problem behavior for some par- ticipants (Kautz et al. 2018; Kern et al. 2001). Similarly, Smeltzer et al. (2009) also found that students preferred the opportunity to choose the task sequence and time spent on each task decreased when given the opportunity to choose task sequence. Students could sequence tasks in a model reflective of BMT or response deprivation; they could also sequence things randomly (inconsistently) or in other variations. Although teacher-manipulated task sequence and student choice of task sequence have been evaluated independently, there is little compar- ative research addressing the effectiveness of teacher-manipulated task sequences in relation to student choice of task sequence (Lequia et al. 2012).

Further, researchers may have a difficult time determining how students will respond to various demands without a systematic evaluation. Call et  al. (2009, 2016) developed an evaluation to inform which demands to select for functional analysis escape conditions. Traditionally, clinicians and researchers arbitrarily select demands or rely on teacher report. Researchers calculated the latency to respond to various demands and considered demands with the greatest latency “most aversive.” To date, there is not an assessment available to inform the likeli- hood that a student will complete a task, but an assessment similar to that of Call and colleagues may lead to more informed decision-making.

The current study sought to gather preliminary data in order to answer the fol- lowing research question: What are the comparative effects of student choice of task sequence and two variations in teacher-manipulated task sequences on on- task behavior when embedded in a VAS? The purpose of the current study was to use the results of a modified demand assessment to evaluate the comparative effects of high- to low-probability task sequence (HP to LP), low- to high-prob- ability task sequence (LP to HP), and student choice of task sequence (SC) inte- grated on a VAS in relation to the percentage time in which participants were engaged in on-task behavior.

115

1 3

Journal of Behavioral Education (2021) 30:112–129

Method

Participants

Three students receiving special education services in a third- through fifth-grade self-contained classroom participated in the study. Researchers recruited students based on teacher reports of frequent off-task behavior during academic instruc- tion. Two of the three participants also engaged in more severe problem behavior (e.g., aggression). Parents gave permission for each student to participate prior to the study.

Dre was a 10-year-old, African American male whose primary special edu- cation eligibility was moderate intellectual disability (ID). Academically, Dre could expressively identify 20 sight words related to preferred items or activities, expressively identify numbers up to 15, receptively identify letters based on let- ter sound, and write his name. He could also follow one-step directions, use the restroom independently, and perform daily living skills such as brushing his teeth and washing his hands. Dre communicated with students and teachers through complete sentence vocalizations. Occasionally, teachers and staff asked Dre to slow down or repeat vocalizations for clarification. Dre’s classroom teachers indicated that he engaged in frequent off-task behavior, aggression, elopement, disruption, and self-injurious behavior throughout the school day. Dre’s behav- ior intervention plan was implemented throughout sessions of the study and con- sisted of differentially reinforcing any behaviors other than problem behavior on a fixed ratio schedule. Teachers delivered a token for every occurrence of prosocial behavior or compliance. For every five tokens earned, Dre earned a small edible reinforcer, and after receiving 20 tokens Dre was able to trade tokens for a larger reinforcer (e.g., iPad, large edible, toys). Each token was a check mark drawn on a grid with 20 squares using a dry erase marker.

Junior was an 11-year-old, African American male whose primary special educa- tion eligibility was autism spectrum disorder (ASD). Junior could receptively iden- tify 25 high frequency and preferred item sight words, receptively identify numbers up to 30, count with one-on-one correspondence, write letters and numbers, and write sight words from a model. He could also perform daily living tasks indepen- dently, such as using the restroom and washing his hands. Junior occasionally spon- taneously vocalized words, imitated word vocalizations, and utilized a speech-gen- erating device (SGD) to request for preferred items by selecting individual pictures. Junior’s classroom teachers indicated that he engaged in frequent off-task behavior and often required multiple prompts to engage in academic activities during one-on- one instruction. He also engaged in problem behaviors including aggression, disrup- tion, elopement, and touching. Junior’s behavior protocol consisted of differentially reinforcing all behavior other than problem behavior on a fixed ratio schedule of reinforcement. For every two prosocial or compliant behaviors, the teacher delivered a token. Once Junior received three tokens, he was able to trade his tokens in for a backup reinforcer (e.g., edible reinforcer, break from work). Junior chose a rein- forcer by either independently vocalizing or by communicating with his SGD.

116 Journal of Behavioral Education (2021) 30:112–129

1 3

Janet was an 8-year-old, African American girl whose primary special educa- tion eligibility was significant developmental delay (SDD). Janet could expressively identify 50 high frequency and preferred item sight words, solve single-digit addi- tion and subtraction problems without a calculator, pay for items using the dollar plus method, and could write preferred item sight words and CVC words when provided with a corresponding image. She could also perform all daily living skills independently, including using the restroom, washing her hands, brushing her teeth, brushing her hair, and cleaning her environment. Janet communicated by vocalizing complete sentences. Despite a token economy in place to reinforce on-task respond- ing, Janet’s teachers reported that she engaged in frequent off-task behavior during one-on-one academic instruction.

Settings and Materials

Setting

The study took place in a self-contained classroom serving individuals with low incidence disabilities (e.g., ASD), at a Title 1-funded elementary school in the southeastern USA. Many of the students in the classroom engaged in severe problem behavior and experienced difficulties with communication. The classroom was oper- ated by five adults, including university special education and behavior analysis fac- ulty and graduate students. The classroom was led each day by a teacher with mas- ter’s level training in special education and support staff with bachelor’s degrees. All probing, demand assessment sessions, and VAS manipulation sessions occurred in a one-on-one instructional arrangement within the classroom, in the teacher work room, or at a hallway workstation. When working in settings outside of the class- room, no other students were present. During classroom sessions, up to 6 additional students were present but were working with other staff. Students had experience working in these settings prior to the start of the study.

Data Collection Materials

Researchers collected data on handheld devices using the Countee (Gavran and Her- nandez 2015) app during demand assessments. Countee templates were customized for each participant within the app to measure latency to respond and frequency and/ or duration of problem behavior. Researchers used an iPad to video record all VAS manipulation sessions for all participants. The iPad was in a case with a stand on the back so that it could be propped up as needed to record sessions across locations. The Simple Interval Timer (SIT; Kazarova 2011) app was downloaded onto a cell phone and placed next to the iPad during recording so that an audible timer would go off every 10 s when re-watching videos for data collection purposes. Research- ers recorded on-task/off-task behavior and problem behavior on data sheets when scoring videos of all VAS manipulation sessions. Data collection procedures are described later in detail.

117

1 3

Journal of Behavioral Education (2021) 30:112–129

Instructional Materials

Instructional materials varied for each participant depending on the results of their respective demand assessments. Seven sets of instructional materials were included for each student during the demand assessment. Here we only detail the materials ultimately used in instruction (a full description of all 21 tasks can be obtained by contacting the second author). All student materials included a whiteboard, marker, and eraser. The whiteboard was used to write sight words and decoding words. Dre’s materials also included laminated (imitation) dollar bills, notecards with printed sight words, and laminated notecards with pictures of community signs. Junior’s materials included manipulative objects for comparatives (bigger/smaller, longer/ shorter, shorter/taller, and softer/harder) and a calculator or an iPad with a calculator app. Janet also had laminated note cards with sight words in addition to laminated strips of paper with written directions. All targeted demands were drawn from the participants’ existing IEP goals; when participants mastered initial targets, research- ers created new targets addressing the same IEP goal. The basic instructional mate- rials remained the same throughout the study.

Visual Activity Schedules

Each student had three VASs, one for each condition of the study varying in color  (see Fig. 1). Pictures on the VAS represented the different tasks. Each VAS strip without pictures consisted of an 11 cm by 3.8 cm colored piece of cardstock with an affixed 9.5 cm by 2 cm row of three white squares. Each icon, placed into the white squares with Velcro, was roughly 3 cm by 2 cm with actual pictures of the task materials used for teach task.

Fig. 1 Visual activity schedule (VAS). Each VAS condition (HP to LP, LP to HP, and student choice) was represented by a different color VAS. Pictures of the activity materials were placed on the schedule prior to the start of each session

118 Journal of Behavioral Education (2021) 30:112–129

1 3

Dependent Variable and Measurement

Researchers collected live data for the demand assessment conditions and from video recordings for the VAS sessions. Researchers measured latency to respond to demands during demand assessment conditions. Measurement of latency began by starting the session on Countee after a three-second countdown and then immediately placing the target demand. The session ended as soon as the par- ticipant initiated compliance with the demand (e.g., vocally stated a response, putting a marker to a white board to begin writing a response). Therefore, the length of each session was equal to the latency to response initiation. Latency for all sessions of each target demand was then added together and divided by the total number of sessions of that target demand to determine the average latency to respond for each. Researchers electronically recorded and graphed the average latencies for all demand assessment sessions for each participant (see Figs. 2, 3, and 4). Frequency and duration measures were likewise coded in Countee.

The dependent variable was percentage of session engaging in on-task behav- ior. The definition for on-task behavior for Dre and Janet was: eyes oriented toward instructor, task materials, or reinforcement items; engaging in conversa- tion with instructor related to task demands; and complying with task demands. The definition for on-task behavior for Junior was: eyes oriented toward instruc- tor, task materials, reinforcement items, or AAC device; and complying with task demands. Researchers collected data during VAS manipulation sessions by view- ing previously recorded sessions from an iPad. On-task behavior was recorded on data sheets while watching the videos. Researchers used momentary time sam- pling (Cooper et  al. 2007) with 10-s intervals to collect on-task behavior data. They calculated the percentage of on-task behavior by dividing the number of

Fig. 2 Dre’s demand assessment results. Results of modified demand assessment in latency to response to task demand in second for Dre

119

1 3

Journal of Behavioral Education (2021) 30:112–129

intervals in which the participant was on-task by the total number of intervals and multiplying by 100.

Interobserver Agreement (IOA)

Researchers calculated point-by-point IOA for on-task behavior during VAS manip- ulation sessions by dividing the number of agreements by the total number of agree- ments plus disagreements and multiplying by 100 to calculate a percentage. On-task IOA was calculated for 60% of sessions for Dre, 38.8% of sessions for Junior, and

Fig. 3 Junior’s demand assessment results. Results of modified demand assessment in latency to response to task demand in second for Junior

Fig. 4 Janet’s demand assessment results. Results of modified demand assessment in latency to response to task demand in second for Janet

120 Journal of Behavioral Education (2021) 30:112–129

1 3

38.8% of sessions for Janet. For Dre, IOA ranged from 58 to 90.63% with an aver- age agreement of 81.17%. Dre’s IOA dropped below 80% on two occasions, sessions four and fourteen. Junior’s IOA ranged from 76.47 to 91.67% with an average agree- ment of 84.52%. Junior’s IOA dropped below 80% on one occasion, session five. Janet’s IOA ranged from 76.67 to 92.86% with an average agreement of 85.78%. Janet’s IOA dropped below 80% on one occasion, session seven. Researchers noted on the graph the four instances where IOA dropped below 80% with open data points.

Researchers measured procedural fidelity using a checklist of VAS manipula- tion components and procedures. Procedural fidelity was calculated as a percentage of correct implementation by dividing the number of steps implemented correctly by the total number of steps and multiplying by 100. Procedural fidelity for VAS manipulation was assessed in 46.67% of sessions for Dre, 61% of sessions for Jun- ior, and 55.55% of sessions for Janet. Procedural fidelity was 100% across partici- pants for VAS manipulation.

Research Design

The study employed an alternating treatments design to evaluate the effectiveness of the three conditions across time: One condition consisted of the instructor sequenc- ing tasks on the VAS from high probability to low probability (HP to LP). A second condition consisted of the instructor sequencing tasks on the VAS from low prob- ability to high probability (LP to HP). The third condition consisted of the student choosing the sequence of tasks on the VAS (SC). Researchers block randomized the order of conditions for each participant and measured the dependent variable under all conditions in a comparison phase. Alternation of conditions occurred within days with no more than two sessions in a row of the same condition.

Procedures

The independent variable in the current study was the sequence of tasks embed- ded in a VAS. In the HP to LP condition, task sequence consisted of the HP task followed by the moderate probability (MP) task followed by the LP task. The researcher reversed this task sequence in the LP to HP condition, moving from the low- to moderate- to high-probability task. In the SC condition, the participant selected the sequence of tasks. Researchers determined LP, MP, and HP tasks by conducting a modified demand assessment.

Probing

Prior to the start of the demand assessment, researchers probed all academic targets taught through discrete trial training (DTT) and in a one-on-one instructional for- mat. The researcher excluded targets with more than 50% mastery and combined targets with less than 50% mastery into tasks (e.g., sight words) with three targets for each set (e.g., ball, hat, car) to include in the demand assessment.

121

1 3

Journal of Behavioral Education (2021) 30:112–129

Demand Assessment

Researchers used a modified demand assessment to identify low-, moderate-, and high-probability tasks for each participant. The demand assessment differed slightly from the procedures outlined in Call et  al. (2016) in that the session ended when the individual complied with a demand rather than when the individual engaged in problem behavior. The procedural differentiation was made due to severe problem behavior that posed a danger to staff and students and the information related to completion of demands we sought to gather. Further, the demand assessment con- sisted of presenting participants with a demand (i.e., target) and then measuring the latency to response initiation or how long it took the participant to comply with the demand placed. The researcher block randomized the order in which all included tasks were presented as demands.

Based on the data collected in the probe condition, the researcher collected mate- rials associated with non-mastered targets for seven tasks. At onset, the researcher stated, “It’s time do some work” and presented the first target. The session ter- minated upon participant initiation of the first target for each task or after 600  s. Researchers did not prompt the participant to comply with the demand and restated the task demand every 30 s. We identified a hierarchy of high- to low-probability tasks for each participant by first assigning the longest average latency to response initiation as the LP and the shortest average latency as the HP. Then, we identified the moderate probability (MP) task by selecting the task with the median latency when compared to the other tasks.

VAS Manipulation

Researchers used the three tasks from the demand assessment for the VAS manipu- lation conditions. The three conditions were HP to LP, LP to HP, and SC. Each con- dition started with the researcher and participant sitting at a one-on-one academic workstation. During the HP to LP condition and LP to HP condition, the researcher manipulated the VAS prior to the session. The researcher then presented the VAS to the participant and stated, “It’s time to do some work. First, we have to do ‘task A,’ then ‘task B,’ and then ‘task C’.” For each task, the researcher stated the name of the task (e.g., sight words). During the SC condition, the researcher vertically or horizontally placed the three VAS pictures on the workstation and stated, “We have ‘task A,’ ‘task B,’ and ‘task C.’ Which one do you want to do first?” Once the participant selected the first task picture, the researcher (or the participant) put the picture on the farthest left square of the VAS and stated, “Okay. Now we have ‘task B’ and ‘task C.’ Which one do you want to do next?” Once the participant selected the second task picture, the researcher placed it on the middle square on the VAS and then placed the remaining third picture on the farthest right square on the VAS. The researcher then stated, for example, “Okay, it’s time to do some work. First we have to do ‘Sight words,’ then ‘community signs,’ and then ‘dollar plus’” and imme- diately placed the first demand.

Researchers implemented simultaneous prompting procedures typically used in the classroom (i.e., gestural and model prompts for Dre and Junior; gestural, model,

122 Journal of Behavioral Education (2021) 30:112–129

1 3

and physical prompting for Janet). If the student did not respond to the demand within 10 s during the probe, the researcher proceeded with the controlling prompt. The controlling prompt depended on the target task. Researchers implemented the student’s behavior protocols throughout the VAS manipulation condition. For Dre, the researcher presented at least five targets providing Dre with five opportunities to earn tokens (i.e., 3 tasks with 5 targets for each). After the 15 tokens were delivered, 5 additional high-probability demands were placed (e.g., touch your nose, “what color shirt am I wearing?”) for Dre to access a large reinforcer per his behavior plan. For the purpose of the study, data collection ended after the first 15 demands were placed. For Junior, tasks included at least two targets in order to provide Junior an opportunity to access tokens for backup reinforcers based on his behavior protocol. For Janet, the researcher presented five targets for each of the three tasks on the VAS. Janet received a token with every completed target and exchanged the tokens for an edible reinforcer at the completion of each task before resetting the token economy and moving to the next task on the VAS. After completing the final task on the VAS, she was also able to access a leisure activity. For all phases and condi- tions of the current study, off-task behavior and problem behavior were ignored with the exception of elopement in which the researcher returned the participant to their work area. Sessions lasted until students completed the task sequence and earned their selected reinforcer or until 10 min had elapsed without completion of the task sequence.

Results

Dre’s average latencies to respond to each task during the demand assessment are depicted in Fig. 2. Dre’s demand assessment identified expressive identification of functional sight words as his HP task with an average latency to respond of 10 s, paying with dollar plus as his LP task with an average latency to respond of 42 s, and receptive identification of community signs as his MP task with an average latency to respond of 19 s. Results of Dre’s VAS manipulation sessions are shown in Fig. 5. Dre’s median percentages of on-task behavior were 78.95% in the HP to LP condition, 71.43% in the LP to HP condition, and 54% in the SC condition.

Figure  3 displays Junior’s average latencies to respond to each task during the demand assessment. Junior’s demand assessment identified comparatives as his HP task with an average latency to respond of 31  s, single-digit addition on the cal- culator as his LP task with an average latency to respond of 124  s, and receptive identification of sight words as his MP task with an average latency to respond of 94  s. During the study, he mastered all single-digit addition on the calculator tar- gets so double-digit addition on the calculator became his LP task. Figure 6 shows the results of Junior’s VAS manipulation sessions. Junior’s median percentages of on-task behavior were 75.08% for the HP to LP condition, 71.43% in the LP to HP condition, and 68.75% in the SC condition.

Figure  4 displays Janet’s demand assessment results. Janet’s demand assess- ment identified expressive identification of preferred sight words as her HP task with an average latency to respond of 12  s, decoding CVCV words as her LP

123

1 3

Journal of Behavioral Education (2021) 30:112–129

task with an average latency to respond of 42 s, and following written directions as her MP task with an average latency to respond of 19 s. Janet mastered all of her preferred sight word targets during the course of the study so six more pre- ferred sight words were added. Figure 7 displays the results of the VAS condition. Janet’s median percentages for on-task behavior were 86.17% for the HP to LP condition, 71.54% for the LP to HP condition, and 73.505% for the SC condition.

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Pe rc

en ta

ge o

f M om

en ta

ry T

im e

S am

pl es

O n-

Ta sk

Session

Fig. 5 Dre’s VAS manipulation results. Percentage of time in which Dre’s behavior met the definition for on-task during each VAS manipulation condition

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Session

Fig. 6 Junior’s VAS manipulation results. Percentage of time in which Junior’s behavior met the defini- tion for on-task during each VAS manipulation condition

124 Journal of Behavioral Education (2021) 30:112–129

1 3

Discussion

In the current study, researchers compared the effectiveness of three variations in task sequence embedded in a VAS on percentage of on-task behavior. The idiosyn- cratic outcomes across students suggest the need for individualization. Junior’s data resulted in slight differences in median levels of on-task behavior between condi- tions, but with high levels of overlapping data and very minimal differentiation. Janet’s data demonstrated more pronounced differentiation between conditions with moderate levels of overlap. Dre’s on-task behavior data produced the most substan- tial differentiation between conditions, especially between the HP to LP condition and the SC condition. Although not drastic, overall results indicate that there may be some preliminary evidence supporting increased time on-task for the HP to LP task sequence. Additionally, for both Dre and Junior, teachers reported that frequency of problem behavior was lowest in the HP to LP condition. Although slightly higher rates of on-task behavior in the HP to LP condition may not be significant enough alone to warrant suggestion as an effective antecedent intervention, the combination with lower rates of problem behavior may make a case for its potential utility in the classroom. However, future research is warranted to further explore that possibility.

Extending on the current findings, previous research identified student choice as an effective intervention for decreasing problematic behavior (Jolivette et al. 2001; Kern et al. 2001; Watanabe and Sturmey 2003). However, in the current study the SC condition yielded the lowest median level of on-task behavior for two of three participants, Dre and Junior. Although not as significant for Junior, Dre’s median level of on-task behavior was 25% less during the SC condition than it was during the HP to LP condition. According to one theory, perceived incompetence at tasks may decrease the positive effects typically associated with student choice and cause students to perform better when not given a choice (Thompson and Beymer 2015).

low to high

student choice

high to low

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Session

Fig. 7 Janet’s VAS manipulation results. Percentage of time in which Janet’s behavior met the definition for on-task during each VAS manipulation condition

125

1 3

Journal of Behavioral Education (2021) 30:112–129

An additional theory, though not currently supported by research, is that providing a choice may increase the number of demands placed on the student and subsequently lead to more off-task behavior due to the duration of work times.

The response deprivation hypothesis, Premack principle states that a HP activity can serve as a reinforcer to make an individual more likely to engage in a LP activity (Premack 1959) and research has supported the utility of the principle in both ani- mals and people (Hosie et al. 1974). According to this principle, one might assume that the LP to HP condition would present with better results; however, results of the current study did not support this hypothesis. LP to HP task sequence resulted in the lowest percentage of time on-task for one participant, Janet, and modest percentage of time on-task for two participants, Dre and Junior.

Behavioral momentum theory refers to the idea that once a chain of reinforce- ment begins in a certain situation, momentum is gathered for the continuation of the behavior that is occurring in that context (Nevin 1996; Mace and Nevin 2017). In the current study, this theory could provide supporting evidence to the success of the HP to LP condition because it parallels the logic behind the high-probability request sequence (Pitts and Dymond 2012), which is a commonly utilized way of present- ing tasks with individuals with ASD that follows BMT (Pitts and Dymond 2012). For all participants, the HP to LP condition resulted in the highest median levels of on-task behavior. The difference between the percentage of on-task behavior in the HP to LP condition and the other two conditions was most notable for Janet. These results may contribute to the literature on BMT or HP request sequences by provid- ing supporting evidence for its utility with students who engage in off-task behavior. Additionally, the current study provided some support for the use of a variation in a demand assessment to identify a hierarchy of tasks for the purpose of determining task sequence in the context of a visual schedule. Differentiation was evident in the latencies to respond to varying tasks, indicating a higher likelihood that the assess- ment and measures were accurate.

Several limitations exist in the current study. Researchers did not incorporate baseline data. These data would have necessitated the importance of incorporating AI as a means for improving on-task behavior. The effects of the independent vari- ables were inconsistent across participants, and the data within participants were variable and often overlapping across conditions. Through additional analysis of median data, researchers synthesized results, but additional research is warranted. Additionally, although results indicate that HP to LP task sequence may lead to increases in on-task behavior for some students, the implementation of the interven- tion was not measured over an extended period of time so its long-term effective- ness in the classroom cannot be determined. Researchers did not collect data on the choices students made during the students choice condition. Therefore, students may have chosen HP to LP task sequence or LP to HP task sequence. Knowledge of the sequence would have provided further support for the two other conditions. Further, because sessions were conducted in the typical classroom environment, distractions often occurred that may have affected responding. For example, during one session pizza was brought into the classroom for a student’s birthday party, sparking the interest of all the students and affecting on-task behavior. Future research should take careful note of these potential distractors. A specific limitation for Junior is that

126 Journal of Behavioral Education (2021) 30:112–129

1 3

there was an error in block randomization that was undetected until the study’s con- clusion. This resulted in an uneven number of sessions in each condition during the VAS manipulation phase of the study (five in LP to HP, six in HP to LP, and seven in SC), which is problematic in an ATD design. Due to this, his results should be interpreted with extreme caution. Further, a preference assessment of colors was not conducted prior to the current study, so the colors of VAS backgrounds manipulated for discrimination purposes could have affected responding. Four IOA data points dropped below 80% across all participants, potentially due to the complexity of the behavior definition and the frequency of distractions in this particular classroom. Specifically, data collectors found the eye orientation component of the definition, especially difficult without advanced technology. Further research may consider incorporating more sophisticated technology, altering the definition of the depend- ent variable, or allowing for more opportunities to train observers.

Results of the current study have important implications for teachers when attempting to increase on-task behavior and determining the sequence of events within a VAS. In the future, researchers should consider incorporating a base- line phase. Additionally, future research should evaluate the utility of modified demand assessments in creating a hierarchy of demands to inform classroom interventions. Given the variation in the demand assessment, researchers should also evaluate the validity of the procedure change. Finally, researchers should consider evaluating the effects of the independent variable on additional depend- ent variables including challenging behavior, compliance with demands, and accuracy of responses.

Compliance with Ethical Standards

Conflict of interest The authors declare they have no conflict of interest.

Human and Animal Rights All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. This article does not contain any studies with animals performed by any of the authors.

References

Ardoin, S. P., Martens, B. K., & Wolfe, L. A. (1999). Using high-probability instruction sequences with fading to increase student compliance during transitions. Journal of Applied Behavior Analysis, 32, 339–351.

Betz, A., Higbee, T. S., & Reagon, K. A. (2008). Using joint activity schedules to promote peer engage- ment in preschoolers with autism. Journal of Applied Behavior Analysis, 41, 237–241.

Blum-Dimaya, A., Reeve, S. A., Reeve, K. F., & Hoch, H. (2010). Teaching children with autism to play a video game using activity schedules and game-embedded simultaneous video modeling. Educa- tion and Treatment of Children, 33, 351–370.

Bryan, L. C., & Gast, D. L. (2000). Teaching on-task and on-schedule behaviors to high-functioning chil- dren with autism via picture activity schedules. Journal of Autism and Developmental Disorders, 30, 553–567.

127

1 3

Journal of Behavioral Education (2021) 30:112–129

Call, N. A., Miller, S. J., Mintz, J. C., Mevers, J. L., Scheithauer, M. C., Eshelman, J. E., et al. (2016). Use of latency-based demand assessment to identify potential demands for functional analyses. Journal of Applied Behavior Analysis, 49, 900–914.

Call, N. A., Pabico, R. S., & Lomas, J. E. (2009). Use of latency to problem behavior to evaluate demands for inclusion in functional analyses. Journal of Applied Behavior Analysis, 42, 723–728.

Carlile, K. A., Reeve, S. A., Reeve, K. F., & DeBar, R. M. (2013). Using activity schedules on the iPod touch to teach leisure skills to children with autism. Education and Treatment of Children, 36, 33–57.

Cihak, D. F. (2011). Comparing pictorial and video modeling activity schedules during transitions for students with autism spectrum disorders. Research in Autism Spectrum Disorders, 5, 433–441.

Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). Applied behavior analysis. Essex CM202JE: Pear- son Education Limited.

Cuhadar, S., & Diken, I. H. (2011). Effectiveness of instruction performed through activity schedules of leisure skills of children with autism. Education and Training in Autism and Developmental Dis- abilities, 46, 386–398.

Davis, C. A., Brady, M. P., Hamilton, R., McEvoy, M. A., & Williams, R. E. (1994). Effects of high- probability requests on the social interactions of young children with severe disabilities. Journal of Applied Behavior Analysis, 27, 619–637.

Dettmer, S., Simpson, R. L., Smith, M. B., & Ganz, J. B. (2000). The use of visual supports to facili- tate transitions of students with autism. Focus on Autism and Other Developmental Disabilities, 15, 163–169.

Duttlinger, C., Ayres, K. M., Bevill-Davis, A., & Douglas, K. H. (2012). The effects of a picture activity schedule for students with intellectual disability to complete a sequence of tasks following verbal directions. Focus on Autism and Other Developmental Disabilities, 28, 32–43.

Einfeld, S. L., Tonge, B. J., & Clarke, K. S. (2013). Prevention and early intervention for behavior prob- lems in children with developmental disabilities. Current Opinion in Psychiatry, 26, 263–269.

Gavran, P. D. & Hernandez, V. (2015). Countee (1.0.4) [Krushka designs s de rl de vc]. Retrieved from http://itune s.apple .com.

Hine, J. F., Ardoin, S. P., & Foster, T. (2015). Decreasing transition times in elementary school class- rooms: Using computer-assisted instruction to automate intervention components. Journal of Applied Behavior Analysis, 48, 495–510.

Hosie, T. W., Gentile, J. R., & Carroll, J. D. (1974). Pupil preferences and the Premack principle. Ameri- can Educational Research Journal, 11, 241–247.

Hume, K., & Odom, S. (2007). Effects of an individual work system on the independent functioning of students with autism. Journal of Autism and Developmental Disorders, 37, 1166–1180.

Jolivette, K., Wehby, J. H., Canale, J., & Massey, N. G. (2001). Effects of choice-making opportuni- ties on the behavior of students with emotional and behavioral disorders. Behavioral Disorders, 26, 131–145.

Jung, S., Sainato, D. M., & Davis, C. A. (2008). Using high-probability request sequences to increase social interactions in young children with autism. Journal of Early Intervention, 30, 163–187.

Kautz, M. E., DeBar, R. M., Vladescu, J. C., & Graff, R. B. (2018). A further evaluation of choice of task sequence. The Journal of Special Education, 52, 16–28.

Kazarova, A. (2011). Simple Interval Timer (SIT) (2.0.1) [Off]. Retrieved from http://itune s.apple .com. Kern, L., Choutka, C. M., & Sokol, N. G. (2002). Assessment-based antecedent interventions used in

natural settings to reduce challenging behavior: An analysis of the literature. Education and Treat- ment of Children, 25, 113–130.

Kern, L., Mantegna, M. E., Vorndran, C. M., Bailin, D., & Hilt, A. (2001). Choice of task sequence to reduce problem behaviors. Journal of Positive Behavior Interventions, 3, 3–10.

King, B., Radley, K. C., Jenson, W. R., Clark, E., & O’Neill, R. E. (2014). Utilization of video modeling combined with self-monitoring to increase rates of on-task behavior. Behavioral Interventions, 29, 125–144.

Knight, V., Sartini, E., & Spriggs, A. D. (2015). Evaluating visual activity schedules as evidence-based practice for individuals with autism spectrum disorders. Journal of Autism and Developmental Dis- orders, 45, 157–178.

Kranak, M. P., Alber-Morgan, S. R., & Sawyer, M. R. (2017). A parametric analysis of specific praise rates on the on-task behavior of elementary students with autism. Education and Training in Autism and Developmental Disabilities, 52, 453–464.

128 Journal of Behavioral Education (2021) 30:112–129

1 3

Kuo, N., & Plavnick, J. B. (2015). Using an antecedent art intervention to improve the behavior of a child with autism. Art Therapy: Journal of American Art Therapy Association, 32, 54–59.

Lequia, J., Machalicek, W., & Rispoli, M. J. (2012). Effects of activity schedules on challenging behavior exhibited in children with autism spectrum disorders: A systematic review. Research in Autism Spectrum Disorders, 6, 480–492.

Luke, S., Vail, C. O., & Ayres, K. M. (2014). Using antecedent physical activity to increase on-task behavior in young children. Exceptional Children, 80, 489–503.

Macduff, G. S., Krantz, P. J., & McClannahan, L. E. (1993). Teaching children with autism to use photographic activity schedules: Maintenance and generalization of complex response chains. Journal of Applied Behavior Analysis, 26, 89–97.

Mace, F. C., & Nevin, J. A. (2017). Maintenance, generalization, and treatment relapse: A behavioral momentum analysis. Education and Treatment of Children, 40, 27–42.

Massey, N. G., & Wheeler, T. J. (2000). Acquisition and Generalization of activity schedules and their effects on task engagement in a young child with autism in an inclusive pre-school classroom. Division on Autism and Developmental Disabilities, 35, 326–335.

May, M. E., & Howe, A. P. (2013). Evaluating competing reinforcement contingencies on off-task behavior in a preschooler with intellectual disability: A data-based case study. Education and Treatment of Children, 36, 97–109.

Mechling, L. C., Gast, D. L., & Seid, N. H. (2009). Using a personal digital assistant to increase independent task completion by students with autism spectrum disorder. Journal of Autism and Developmental Disorders, 39, 1420–1434.

Mechling, L. C., & Gustagson, M. R. (2008). Comparison of static picture and video prompting on the performance of cooking-related tasks by students with autism. Journal of Special Education Technology, 23, 31–45.

Miramontez, S. K. H., & Schwartz, I. S. (2016). The effects of physical activity on the on-task behav- ior of young children with autism spectrum disorders. International Electronic Journal of Ele- mentary Education, 9, 405–418.

Morrison, R. S., Sainato, D. M., Benchaaban, D., & Endo, S. (2002). Increasing play skills of children with autism using activity schedules and correspondence training. Journal of Early Intervention, 25, 58–72.

Nevin, J. A. (1996). The momentum of compliance. Journal of Applied Behavior Analysis, 29, 535–547.

Pierce, J. M., Spriggs, A. D., Gast, D. L., & Luscre, D. (2013). Effects of visual activity schedules on independent classroom transitions for students with autism. International Journal of Disability, Development and Education, 60, 253–269.

Pitts, L., & Dymond, S. (2012). Increasing compliance of children with autism: Effects of pro- grammed reinforcement for high-probability requests and varied inter-instruction intervals. Research in Autism Spectrum Disorders, 6, 135–143.

Premack, D. (1959). Toward empirical behavior laws I. Positive reinforcement. Psychological Review, 66, 219–233.

Richards, L. C., Heathfield, L. T., & Jenson, W. R. (2010). A classwide peer-modeling intervention package to increase on-task behavior. Psychology in the Schools, 47, 551–566.

Santoyo, C., Jonsson, G. K., Anguera, T., & Lopez-Lopez, J. A. (2017). Observational analysis of the organization of on-task behavior in the classroom using complementary data analysis. Anales de psicologia, 33, 497–514.

Smeltzer, S. S., Graff, R. B., Ahearn, W. H., & Libby, M. E. (2009). Effect of choice of task sequence on responding. Research in Autism Spectrum Disorders, 2009, 734–742.

Thompson, M., & Beymer, P. (2015). The effects of choice in the classroom: Is there too little or too much choice? Support for Learning, 30, 105–120.

Tingstrom, D. H., & Edwards, R. (1989). Eliminating common misconceptions about behavioral psychology: One step toward increased academic productivity. Psychology in the Schools, 26, 194–202.

Van Laarhoven, T., Kraus, E., Karpman, K., Nizzi, R., & Valentino, J. (2010). A comparison of picture and video prompts to teach daily living skills to individuals with autism. Focus on Autism and Other Developmental Disabilities, 25, a195–a208.

Watanabe, M., & Sturmey, P. (2003). The effect of choice-making opportunities during activity schedules on task engagement of adults with autism. Journal of Autism and Developmental Disorders, 33, 535–538.

129

1 3

Journal of Behavioral Education (2021) 30:112–129

Waters, M. B., Lerman, D. C., & Hovanetz, A. (2009). Separate and combined effects of visual schedules and extinction plus differential reinforcement on problem behavior occasioned by transitions. Jour- nal of Applied Behavior Analysis, 42, 309–313.

Zimmerman, K. N., Ledford, J. R., & Barton, E. E. (2017). Using visual activity schedules for young chil- dren with challenging behavior. Journal of Early Intervention, 39, 339–358.

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.