Skip to main content
NCBI home page
Behavior Analysis in Practice logoLink to Behavior Analysis in Practice
. 2019 Sep 6;13(1):63–80. doi: 10.1007/s40617-019-00388-9

The Effects of a Self-Instruction Package and Group Training on Trial-Based Functional Analysis Administration

Kristin R Griffith 1,2,✉, Jenifer N Price 1, Becky Penrod 1
PMCID: PMC7070140  PMID: 32231968

Abstract

This study examined the effects of a self-instruction package comprised of a task analysis data sheet, detailed written instructions, and a small group performance feedback training on the training of 12 undergraduate students to conduct trial-based functional analyses (TBFAs). In contrast to previous research, training omitted technical language and did not include didactic instruction on the principles of applied behavior analysis or the underlying rationale for functional analysis methodology. Although the self-instruction package produced significant increases in performance and reliable data collection on the targeted problem behavior, small group performance feedback training was necessary to achieve mastery across all trial types. Results offer a socially valid training package to teach paraprofessionals to assist Board Certified Behavior Analysts® in conducting TBFAs in applied settings.

Keywords: behavioral skills training, self-instruction, staff training, trial-based functional analysis, task analysis

Although the standard functional analysis (FA) described by Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994) is the gold standard for assessment of problem behavior, surveys of practitioners have repeatedly ranked experimental manipulation as less informative and more difficult to use than other functional assessment methods (Desrochers, Hile, & Williams-Moseley, 1997; Ellingson, Miltenberger, & Long, 1999; Roscoe, Phillips, Kelly, Farber, & Dube, 2015; Van Acker, Boreson, Gable, & Potterton, 2005). Trial-based functional analysis (TBFA) is a variation of FA methodology that addresses many practitioner concerns and has been demonstrated to accurately identify the function of problem behavior (Bloom, Iwata, Fritz, Roscoe, & Carreau, 2011; LaRue et al., 2010; Rispoli, Ninci, Neely, & Zaini, 2014). Further, the procedures, data collection, and analysis may be easier for entry-level staff to implement over a traditional FA (Rispoli, Neely, Healy, & Gregori, 2016).

As opposed to exposing participants to experimental sessions up to 15 min in duration in a highly controlled analog environment, TBFA trials are brief (generally under 4 min) and interspersed throughout the day in the natural environment (Austin, Groves, Reynish, & Francis, 2015; Bloom, Lambert, Dayton, & Samaha, 2013; Kodak, Fisher, Paden, & Dickes, 2013; Lambert, Bloom, & Irvin, 2012; Sigafoos & Saggers, 1995). These modifications eliminate the need for time and space outside educational or vocational settings required to conduct the assessment. In addition, because the trial ends after one occurrence of targeted problem behavior, the procedure limits the frequency of potentially dangerous behavior during the assessment. Data are collected on either the presence or absence of problem behavior during the test and control segments and are often analyzed with simple bar graphs (Austin et al., 2015; Bloom et al., 2013; Lambert et al., 2012).

Prior research also shows TBFAs can be effectively implemented by caregivers and teachers (Bloom et al., 2013; Flynn & Lo, 2016; Kunnavanta, Bloom, Samaha, & Dayton, 2013; Lambert, Bloom, Clay, Kunnavanta, & Collins, 2014; Lambert, Bloom, Kunnavatana, Collins, & Clay, 2013; Lloyd, Weaver, & Staubitz, 2016; Rispoli et al., 2015, 2016). Implementation by familiar individuals ensures relevant variables that maintain problem behavior in the natural environment are captured (English & Anderson, 2004), while reducing the need for additional staff to conduct the assessment. Combined, these benefits of TBFAs lend themselves to strong social validity ratings (Rispoli et al., 2016) and provide a rationale for dissemination of the methodology to natural settings where it may be useful.

Because the TBFA has the potential to increase the use of FA methodology in settings where practitioners previously noted barriers to implementation, research regarding the most efficacious methods of training people how to conduct TBFAs is warranted. Previous research has demonstrated that behavioral skills training (BST; Sarokoff & Sturmey, 2004) is an effective way to teach a wide variety of skills, including behavior-analytic assessments and interventions. BST generally involves a didactic training component (e.g., reading written instructions, attending a lecture), modeling of the desired behavior, and opportunities to rehearse and receive feedback on performance. Findings from the FA and TBFA training literature show that even individuals with little to no knowledge of applied behavior analysis (ABA) can learn to correctly implement FA and TBFA procedures in brief periods using BST (Bloom et al., 2013; Flynn & Lo, 2016; Iwata et al., 2000; Kunnavanta et al., 2013a; Lambert, Bloom, et al., 2014; Lambert et al., 2013; Lloyd et al., 2015; Moore et al., 2002; Phillips & Mudford, 2008; Rispoli et al., 2015, 2016; Wallace, Doney, Mintz-Resudek, & Tarbox, 2004; Ward-Horner & Sturmey, 2012).

To further reduce training time by identifying the BST components responsible for skill acquisition, Ward-Horner and Sturmey (2012) examined the effects of various training components. While training three teachers to conduct traditional FAs, the authors found that written instructions and rehearsal without feedback were insufficient to produce FA procedural fidelity. It was not until video modeling and rehearsal with feedback were introduced that teachers learned to implement procedures correctly. Although these findings were limited by the failure to isolate the didactic (i.e., written instructions), modeling, and feedback components, they suggest that modeling and rehearsal with feedback may have been responsible for behavior change. This notion is bolstered by findings from Lambert, Lloyd, Staubitz, Weaver, and Jennings (2014), who examined whether an automated didactic presentation with embedded video models was sufficient to train 10 ABA master’s students to implement TBFAs. Investigators found that didactic instruction and modeling alone were ineffective in producing procedural fidelity or accurate data collection, providing further evidence that additional training components (e.g., rehearsal, feedback) may be necessary to produce desired performances. Authors also highlighted the need for additional research to standardize and validate didactic trainings, stating that this component of BST “may vary considerably across groups of trainers or researchers” (p. 345).

Although didactic instructions provided in FA and TBFA trainings vary in duration, delivery method, and content, the bulk of training research provides participants with relevant journal articles (Iwata et al., 1982/1994; Bloom et al., 2011, 2013) to review. Researchers noted that journal articles were used because it is unlikely that participants would be able to implement procedures without some orientation to FA background and rationale (Lambert, Bloom, et al., 2014; Phillips & Mudford, 2008). Despite the need to orient participants to the subject matter, providing participants with these materials alone has repeatedly been shown to be ineffective in increasing performance—so much so that providing a journal article to review is a procedure commonly integrated in baseline conditions of FA training studies (Flynn & Lo, 2016; Kunnavatana, Bloom, Samaha, & Dayton, 2013; Kunnavatana, Bloom, Samaha, Lignugaris-Kraft, et al., 2013; Lambert et al., 2013; Lambert, Bloom, et al., 2014; Moore et al., 2002; Phillips & Mudford, 2008; Rispoli et al., 2015, 2016). These results call into question the necessity of the didactic training component, and if nothing else, the content of this training component deserves further investigation.

Staff training research has produced valuable information that may guide the development of more effective didactic trainings. Graff and Karsten (2012) examined the use of self-instruction packages to train 11 teachers to conduct stimulus preference assessments. When provided with written instructions based on previously published literature alone or in combination with a data sheet, performance was low to moderate across participants. Accuracy greatly improved with all participants achieving mastery when they were provided with enhanced written instructions that included pictures, diagrams, step-by-step examples, and minimal use of technical jargon. A replication and extension of this study were conducted by Shapiro, Kazemi, Pogosjana, Rios, and Mendoza (2016). Results of this study indicated that the enhanced written instructions were sufficient to train five of seven undergraduate students and four of five in-home behavior therapists to mastery. This replication demonstrates that even those with little to no ABA background can effectively be trained using only written instructions in conventional language.

This finding is echoed in the organizational behavior management (OBM) literature, where task clarification is considered an essential component of effective training (Gravina, VanWagner, & Austin, 2008). Task clarification involves providing trainees with a “precise specification of behavioral components of a job” (Crowell, Anderson, Abel, & Sergio, 1988, p. 65), often in the form of a written checklist (Mirman, 1982). Although OBM training studies have produced less robust results with task clarification alone (Anderson, Crowell, Hantula, & Siroky, 1988; Squires et al., 2007; Wilson, Boni, & Hogg, 1997) than self-instruction package studies (Graff & Karsten, 2012; Shapiro et al., 2016), the call for use of conventional language and step-by-step instructions is a common theme across investigations, providing direction for future training research. Because many of the individuals who may be trained to assist behavioral professionals in conducting TBFAs are entry-level therapists, parents, caregivers, and teachers, it may be critical to adapt written materials for use with these intended audiences.

Thus, the purpose of the current investigation was to systematically evaluate the use of a self-instruction package comprised of a task analysis (TA) data sheet and detailed written instructions, both using only conventional language, on the fidelity of TBFA administration by 12 participants with little to no behavior-analytic background.

Method

Participants and Setting

Two male and ten female college students with no background or formal education in behavior analysis participated in this study. In addition, participants had no prior training in FAs and less than 6 months’ experience providing behavior-analytic intervention services. Two participants (Alexa and Brianna) had previous training on the collection of antecedent-behavior-consequence data as part of their work history. TBFA training sessions were conducted in a university classroom, whereas assessment sessions were completed in a laboratory space on the campus. All sessions were videotaped for the purposes of procedural fidelity and interobserver agreement (IOA) data collection.

Dependent Measures and IOA

This study examined the procedural fidelity of TBFA administration and reliability of problem behavior data collection during assessment sessions to evaluate the effectiveness of training procedures. All data were collected via paper and pencil.

TBFA procedural fidelity

Similar to procedures used by Lambert et al. (2013) and Lambert, Lloyd, et al. (2014), a TA data sheet composed of a checklist of procedural steps specific to each TBFA trial type was used to assess procedural fidelity (see Appendix). A trained research assistant recorded whether the participant completed the following TBFA steps correctly or incorrectly (yes/no): (a) the presence of the putative reinforcer during the control segment and the absence of said reinforcer during the test segment, (b) correct responses to the targeted problem behavior (and not any other problem behavior) during the control and test segments, (c) the absence of demands (except for during escape trials), and (d) the presence of materials necessary to conduct each trial type. Procedural fidelity was calculated on 100% of trials by dividing the number of correctly implemented steps by the total number of steps for each specific trial type, then multiplying by 100.

IOA data on procedural fidelity were collected via video by a secondary observer for 62% of trials across participants for all experimental phases and trial types. The IOA data collector used the same TA data sheet described previously to record participants’ procedural fidelity for each TBFA trial type. When both the primary and secondary data collectors scored the same response (e.g., both scored “yes,” or both scored “no”) on each step, an agreement was recorded. Alternatively, if the primary and secondary data collectors scored a specific checklist step differently (e.g., one scored a “yes,” whereas the other scored a “no”), a disagreement was recorded. IOA on procedural fidelity was then calculated using the point-by-point method (Kazdin, 2011). Procedural fidelity IOA averaged 92.5% across participants and trial types (range 78%–100%).

Reliability of problem behavior data collection

Participants collected data on the occurrence of the targeted problem behavior during both the control and test segments of all trial types. For trials in the naturalistic and article review baseline conditions (described in what follows), participants recorded this information on the standard TBFA data sheet used by Lambert, Lloyd, et al. (2014), whereas the TA data sheet was used to record this information in all other conditions. A trained independent observer also recorded data on the occurrence of targeted problem behavior on 100% of trials using the TA data sheet. Reliability of targeted problem behavior data collection was calculated by dividing the number of agreements by the total number of trials, and then multiplying by 100. In this case, an agreement was defined as the participant and the independent observer scoring the same response in both the control and test segments of the trial. If scores were congruent on only one of the trial segments (e.g., only agreement on the test segment but not the control), a disagreement was recorded for that trial. A secondary independent observer recorded IOA data on the occurrence of the targeted problem behavior on 30% of trials using the same TA data sheet. Agreement definitions and IOA calculations were the same as those used for reliability calculations. IOA on the occurrence of the targeted problem behavior across participants and conditions was 94% (range 91%–100%).

Experimental Design

Due to logistical constraints, we used a nonconcurrent multiple-baseline across-participants design (Kazdin, 2011) to evaluate the effects of training components on TBFA procedural fidelity and accurate data collection. The design was similar to that used in previous research (Kunnavatana, Bloom, Samaha, & Dayton, 2013; Lambert et al., 2013; Lambert, Bloom, et al., 2014; Phillips & Mudford, 2008; Rispoli et al., 2015, 2016); however, participants were divided into four groups of three participants. Each group of participants experienced a different combination of experimental conditions (described in what follows) to isolate and compare the effects of various didactic training components. The mastery criterion during all training conditions was defined as 100% procedural fidelity and 100% reliability on data collection for all four trial types.

Procedures

To familiarize participants with the targeted problem behavior they were to assess, the experimenter provided each participant with an operational definition accompanied by a 1-min video clip of an adult confederate engaging in the targeted problem behavior. Nine 5 cm × 8 cm pictures of tangible items to be used in the assessment were also provided (e.g., three highly preferred items, three moderately preferred items, and three low-preferred items). During the naturalistic baseline, the article review baseline, the TA data sheet baseline, and the self-instruction package conditions, participants reviewed materials alone at a desk in the university clinic space.

Various baseline, training, and assessment phases required 4–8 hr over two experimental visits to complete. Brief breaks (10–15 min) were scheduled between each experimental phase, and we provided the opportunity for additional breaks away from the confederate researcher posing as a client whenever participants requested. During these breaks, all assessment materials were removed, but participants often elected to remain in the assessment setting.

Naturalistic Baseline

During this baseline condition, researchers provided Allison, Alexa, and Andrew (Group 1) with the informed consent outlining the duration of each experimental phase and the overall training schedule. To assess how those unfamiliar with ABA might attempt to determine the function of problem behavior, we did not provide any specific procedures on how to implement TBFA trials, but participants did receive the following written instructions:

You will be assessing the reason why the client is engaging in the targeted problem behavior. You will have the opportunity to make multiple brief attempts to determine the reason. Each attempt (or trial) during the assessment session will be 2–4 min in length. If you are taking longer, you may want to stop and try again in another trial. You will have 15 min to look over the data sheet and the materials and develop any ideas of how to assess why the targeted problem behavior is occurring. Since this is a training study and not an actual assessment, the client will be a confederate researcher enacting scripted behavior. This means that your different attempts to determine the reason why—all the things you do or try in these assessments—will NOT influence the client’s behavior.

Participants also received the standard TBFA data sheet used by Lambert, Lloyd, et al. (2014), and researchers instructed participants to record yes/no data on the occurrence of the targeted problem behavior during the control and test segments of each trial type. Participants also had access to an iPhone timer and the pictures of materials to be used in the assessment and were told to “use them as they saw fit.” Researchers gave participants 15 min to review these materials and devise any strategies on how to administer TBFA trials. Because no prior TBFA or FA training research has been conducted using the self-instruction materials described in what follows, the duration of this review period was selected based on a sampling of three undergraduate research assistants with no prior background in FA requiring 5–15 min to review these materials.

Article Review Baseline

Because the baseline condition in many previous FA training studies involved providing participants with a relevant journal article to review, participants in this condition (Group 2: Ben, Brianna, and Belle) were provided the Bloom et al. (2011) article with the Method section highlighted. Participants also received a standard TBFA data sheet (Lambert et al., 2013; Lambert, Lloyd, et al., 2014) and a brief review on how to use it. We instructed participants to review the highlighted section of the article and to use the data sheet to record yes/no data for the occurrence of the targeted problem behavior during the control and test segments of each trial. The experimenter gave participants 45 min to review these materials to yoke this review period to that of the self-instruction package condition described in what follows.

TA Data Sheet Baseline

Participants in this condition (Group 3: Candice, Carol, and Claire) reviewed the same TA data sheets used by researchers to record procedural fidelity (see Appendix), including a checklist of steps required to implement each trial segment correctly and a section to record the occurrence of the targeted problem behavior for each trial type. The TA data sheets included a specific section to write in the targeted problem behavior, the highly preferred items used in tangible trials, the moderately preferred items used in attention trials, and the task demand used in escape trials. The TA data sheet was written in common language and included no behavioral terminology, background, or rationale. The experimenter gave participants 15 min to review the TA data sheets based on a sampling of the time required for the same three undergraduate research assistants discussed previously to review these materials.

Self-Instruction Package Condition

In this condition, participants received the same TA data sheet described previously along with detailed written instructions that included procedural information and suggested practice activities for each trial type (available from the first author). Participants in Group 4 (Deborah, Danielle, and Dawn) experienced this condition as a baseline, whereas all remaining participants experienced this condition following their respective baselines. The 45-min review period duration was determined again by sampling the time required for three undergraduate research assistants with no background in FA to read all materials and complete the suggested practice activities.

Small Group Training

Prior to the start of the study, a graduate researcher was trained to implement the small group training according to a training rubric. The same TA data sheet and written instructions provided to participants were used to train the trainer, ensuring that she could accurately identify any participant errors that may occur, describe the correct behavior by referring to the self-instruction materials, model correct trial administration, and provide participants with an opportunity to practice and receive feedback. A trained independent observer collected data on the occurrence of each training task listed on a training rubric to assess the trainer’s implementation of training skills. Once the trainer met 100% procedural fidelity with the small group training rubric, she went on to conduct the small group training with participants. During group training, trainer fidelity to the training script averaged 98.75% across trial types and participants (attention: 97%; tangible: 99%: escape 100%; ignore: 99%). A second independent observer collected IOA data on the trainer’s performance on 60% of small group training trials across participants using a simplified yes/no data sheet like that described previously. Using the same agreement definitions and point-by-point calculations described previously, IOA on the trainer’s performance of the script during small group training was 95%.

Participants completed the small group training following the self-instruction package condition. Procedurally, this training was similar to that used by Bloom et al. (2013), except that this training was done in a small group format with two participants trained at a time. Because each group of three participants experienced the same order of experimental conditions, this meant that some participants completed the small group training with participants who experienced different conditions prior to training. The small group training, however, was the same for all participants regardless of which group participants were assigned to. Also, the training only included procedural rehearsal and performance feedback without any information on ABA or FA rationale. Each participant was instructed to follow the directions provided in the self-instruction package and to rehearse each trial type with the group trainer acting as the client while the other participant observed the performance and feedback.

To mimic TBFA administration in clinical settings where the topography of the targeted problem behavior varies, the targeted problem behavior enacted by the trainer differed from the targeted problem behavior enacted during assessment trials. Additionally, the trainer engaged in different topographies of nontargeted problem behavior and used a different task demand during escape trials from those participants encountered in assessment trials. The trainer exposed participants to various scenarios likely to occur during TBFA administration in the natural environment, including identifying the presence or absence of targeted problem behavior during the control and test segments, identifying failed trials, and ignoring nontargeted problem behavior.

Participants administered each TBFA trial type, and the trainer provided feedback on their performance immediately following completion of the trial. The trainer either made a brief statement acknowledging that the participant had completed all the steps in the self-instruction package correctly or provided performance feedback for each error made. Performance feedback included identification of the error, a description of the correct procedure as described in the self-instruction package, modeling the correct procedure with the participant acting as the client (a script was provided), and an opportunity for the participant to rehearse the trial again with the same script. The experimenter randomly selected which participant would rehearse first and then alternated between participants for all trial types that followed. For example, when one participant rehearsed the attention trial first with a second participant observing, the second participant would rehearse the tangible trial first while the first participant observed. The small group training continued until participants demonstrated 100% procedural fidelity with all four trial types. Participants completed small group training in an average of 2 hr 20 min (range 1 hr 30 min to 3 hr).

TBFA Assessment Trials

A graduate researcher posed as a confederate client during all assessment trials. Prior to the start of the study, in a fashion similar to that used by Lambert, Lloyd, et al. (2014), this researcher was trained to follow a script that required enacting a series of behaviors relevant to each trial type. Specifically, the confederate was to engage in the following: the targeted problem behavior, a nontargeted problem behavior, and appropriate play and communication behaviors, as well as compliance and noncompliance with demands during escape trials. When the confederate enacted two trials of each trial type (eight trials total) with 100% accuracy to the script, confederate training was complete.

Because the confederate engaged in all the relevant behaviors as scripted for each trial type, each assessment trial allowed for the evaluation of each component step from the self-instruction package required for correct TBFA administration. A trained observer collected yes/no data on whether the confederate enacted each part of the assessment script during 100% of assessment trials. Confederate fidelity to the script averaged 81% across participants, trial types, and conditions (range 50%–100%). The lowest confederate fidelity to the script occurred during the naturalistic (67%) and article review (50%) baseline conditions, where participants’ performance was also more variable, a finding that mirrors that of Lambert, Bloom, et al. (2014). This result is readily explained by the incorrect implementation of TBFA trials during these baseline conditions, where participants often transitioned from the control to test segments of the trial prematurely, effectively eliminating the opportunity for the confederate to engage in all behaviors outlined in the script. A second independent observer rated confederate fidelity to the script on 28% of trials across participants and conditions. IOA on confederate client fidelity to the script was calculated using the agreement definitions and the point-by-point method previously described and was 92%.

TBFA assessment trials were conducted in the same order as materials were presented in the self-instruction package and small group training conditions—namely, attention, tangible, escape, and ignore. All trial types were assessed, even if participants demonstrated mastery on particular trial types during prior conditions. Participants administered 4 to 12 TBFA trials in accordance with the multiple-baseline design. An additional 4 to 8 assessment trials were conducted after review of the self-instruction package and another four trials after small group training. Additional assessment trials were conducted on specific trial types where participants failed to demonstrate mastery after group training, thus requiring one-on-one feedback. During assessment trials, the confederate researcher and participant sat at a child-sized table in chairs during the assessment, with a rolling set of plastic drawers with necessary materials positioned to their right. Participants were permitted to keep and use any training materials, notes, and writing utensils as they saw fit, but only the data sheet relevant to the current experimental phase and trial type was provided.

One-on-One Feedback

If a participant did not score 100% on procedural fidelity checks during any of the assessment trials following small group training, the group trainer provided one-on-one feedback in the assessment setting. Just as in the small group training, the trainer identified any errors that occurred, described the correct procedure per the self-instruction package, modeled correct implementation of the trial type, and provided an opportunity for the participant to practice the trial with the same script before conducting a new assessment trial with the confederate and a new script. Participants received one-on-one feedback until they achieved 100% procedural fidelity for every trial type. During the one-on-one feedback condition, trainer fidelity to the training script was 100% for all trial types. A second independent observer collected IOA data on the trainer’s performance on 60% of one-on-one feedback trials across participants using the same yes/no checklist described in the small group training previously. Using the agreement definitions and point-by-point method previously described, IOA on the trainer’s performance of the script during the one-on-one feedback condition was 94%.

Social Validity

Researchers collected participant responses regarding the social acceptability of training procedures via a modified version of the Training Acceptability Rating Scale (TARS; Davis, Ramana, & Capponi, 1989). The TARS is a self-report questionnaire that consists of 14 items intended to assess the social acceptability of training procedures (e.g., satisfaction, appropriateness, social validity). Participants provided quantitative data on 11 items via a 4-point Likert scale (ranging from 1 for not at all to 4 for a great deal) and qualitative feedback on the remaining three open-ended questions once at the conclusion of the study. Individual items were modified to include the relevant training topic (TBFAs), but the original format was otherwise preserved. The TARS has acceptable test-retest reliability (r = .83), as well as internal consistency reliability (α = .99).

Results

Baseline and Self-Instruction Package Performances

TBFA procedural fidelity scores for Group 1 (Allison, Alexa, and Andrew) in the naturalistic baseline condition averaged 17.7% (range 0%–54%; Figure 1). After exposure to the self-instruction package, procedural fidelity scores averaged 87.7% (range 53%–100%). All participants mastered the ignore trials, Allison mastered the attention and tangible trial types, and Alexa mastered the escape trial procedure and approached mastery on the tangible and attention trials.

Fig. 1.

Fig. 1

Open in a new tab

TBFA procedural fidelity scores for Group 1 (Allison, Alexa, and Andrew) across the naturalistic baseline and subsequent training phases

For participants in Group 2 (Ben, Brianna, and Belle), procedural fidelity scores following review of the Bloom et al. (2011) journal article were initially low with a mix of increasing and decreasing trends (M = 43.8%; range 0%–100%; Figure 2). Following the introduction of the self-instruction package, procedural fidelity scores averaged 77.8% (range 31%–100%). Brianna mastered all but the escape trial, and Ben and Belle each mastered one additional trial type.

Fig. 2.

Fig. 2

Open in a new tab

TBFA procedural fidelity scores for Group 2 (Ben, Brianna, and Belle) across the article review baseline and subsequent training phases

During assessment trials for Group 3 (Candice, Carol, and Claire) following the TA data sheet baseline, performances were slightly higher than those observed for participants in the naturalistic or article review baselines. Procedural fidelity averaged 58.7% (range 22%–100%). Following the introduction of the complete self-instruction package, average performance increased to 88.2% (range 57%–100%), and two participants (Carol and Claire) mastered the ignore trial condition. After exposure to the self-instruction package baseline, Group 4 participants (Deborah, Danielle, and Dawn) performed appreciably better than participants in all other baseline conditions, with procedural fidelity averaging 88.5% (range 58%–100%).

Small Group Training Performances

During the small group training, participants in all groups demonstrated 100% procedural fidelity on all four trial types within an average of 6.75 trials (range 4–10 trials). There were no notable differences in the average number of small group training trials needed to obtain mastery based on previous exposure to various baseline conditions (Group 1: 5 trials, range 6–9; Group 2: 6.33 trials, range 4–10; Group 3: 6.33 trials, range 6–7; Group 4: 6.66 trials, range 5–10). There were also notable differences in the number of small group training trials to obtain mastery for participants who were the first to be assessed for a trial type versus those who were able to observe their peer rehearse and obtain feedback for a trial type prior to their own assessment of that same trial type.

Fig. 3.

Fig. 3

Open in a new tab

TBFA procedural fidelity scores for Group 3 (Candice, Carol, and Claire) across the TA data sheet baseline and subsequent training phases

Fig. 4.

Fig. 4

Open in a new tab

TBFA procedural fidelity scores for Group 4 (Deborah, Danielle, and Dawn) across the self-instruction package baseline and subsequent training phases

In assessment sessions following the small group training for Group 1, Allison maintained procedural fidelity during ignore and tangible trials but demonstrated a decrease in procedural fidelity on the attention trial type. Alexa and Andrew demonstrated mastery on two additional trial types (attention, tangible), but Alexa’s performance on the escape trial type was less accurate than that observed following the previous self-instruction package condition. Additionally, although Andrew’s performance on the escape assessment trial did not indicate mastery, procedural fidelity improved by 39%. Following small group training, all three participants in Group 2 maintained procedural fidelity on the ignore trial type, Ben mastered the tangible trial type, and Belle mastered the escape trial type. Although Ben demonstrated improved performance on the attention and escape trials, performance did not meet the mastery criterion. Additionally, Brianna’s performance on the attention trial type decreased. Participants in Group 3 all demonstrated improved performance on at least two additional trial types following small group training. Candice mastered all four trial types, Carol mastered the attention and escape trial types, and Claire mastered the tangible and escape trial types. In Group 4, Deborah and Danielle mastered the attention and escape trial types following group training, Danielle also mastered the tangible trial type, and Dawn performed all trial types correctly.

One-on-One Feedback Performances

2Across groups, participants required an average of 1.2 one-on-one performance feedback trials to demonstrate mastery during assessment trials following small group training (range 1–3). From Group 1, all participants required one-on-one feedback to implement escape trials correctly, and Allison required one additional one-on-one feedback trial to demonstrate 100% procedural fidelity on the attention trial, despite mastering this trial in the prior self-instruction package condition. From Group 2, Ben and Brianna required one-on-one feedback for the attention and escape trials, and Belle required individualized feedback for the tangible trial type. Only Carol and Claire from Group 3 required one-on-one feedback to master the tangible and attention trials, respectively, whereas Candice required no additional feedback. In Group 4, only Deborah required any additional one-on-one feedback trial for the tangible trial type. No participants in any group required additional training for the ignore trial type following group training. Participants in Group 4 required the fewest number of one-on-one feedback trials to master all trial types (average of 0.33 trials across participants, range 0–1), whereas participants in Group 2 required the greatest number of feedback trials (average of 2 trials across participants, range 1–3). The need for one-on-one feedback was similar for participants in Groups 1 and 3, with those groups requiring an average of 1.33 (range 1–2) and 1 (range 0–2) individual training trials, respectively.

Reliability of Problem Behavior Data Collection

2General increases in participants’ reliable collection of problem behavior data were noted as interventions were introduced (see Table 1). Groups 1–3 reliably collected data on the occurrence and nonoccurrence of targeted problem behavior on an average of 65% of trials (range 58%–71% across groups). Once the self-instruction package was introduced during the baseline condition for Group 4 and following the baseline conditions for the remaining groups, reliable data collection occurred on 90% of trials (79 of 88). Participants in all groups reliably collected data on 98% of trials (47 of 48) following small group training and 100% of one-on-one feedback trials (14 of 14). The largest gains in reliability of problem behavior data collection were obtained after the introduction of the self-instruction package and small group training, with average increases across participants and trial types of 25% and 8%, respectively.

Table 1.

Percentage of trials with reliable problem behavior data collection by experimental group

Experimental Condition Experimental Groups Average Across Groups
1 2 3 4
Naturalistic baseline 71 — — —
Article review baseline — 58 — — 65
TA data sheet baseline — — 67 —
Self-instruction package 100 85 85 88 90
Small group training 100 92 100 100 98
One-on-one feedback 100 100 100 100 100
Open in a new tab

Note. TA = task analysis.

Social Validity

All participants scored 9 of 11 items on the modified version of the TARS (Davis et al., 1989) with either a 3 (quite a lot) or 4 (a great deal), indicating high levels of acceptability for the training packages evaluated. Although the majority of participants rated the effectiveness of the self-instruction package as either quite a lot or a great deal, two participants rated the effectiveness of this training component as not at all or a little (scores of 1 or 2, respectively). Almost all participants stated that their perception of the effectiveness of the small group training was a great deal, although one participant rated group training effectiveness as a little (scale score of 2).

The following common themes were identified in qualitative feedback when participants were asked to list the one or two most helpful aspects of training: the group training in general (n = 5), receiving feedback during the group training (n = 8), and watching another participant administer TBFA trials and receive feedback during the group training (n = 7). Additionally, one participant stated that the TA data sheet was helpful, writing “The listed steps to follow after I tried it myself [was most helpful].” When asked what changes should be made to improve the training, one participant aptly suggested that future trainings should require each participant to perform each trial type correctly twice before moving on to the next trial.

Discussion

Participants in Group 1 had relatively low procedural fidelity in the naturalistic baseline condition when they were asked to devise their own method to determine why the targeted problem behavior was occurring. Similarly, and in accordance with previous research (e.g., Lambert et al., 2013; Lambert, Bloom, et al., 2014; Rispoli et al., 2015, 2016), didactic instruction with behavior-analytic terminology and a standard TBFA data sheet used in the article review baseline (Group 2) was generally ineffective in training TBFA assessment procedures. Predominantly low to moderate fidelity scores, idiosyncratic trends across participants, and inconsistent scores across trial types replicate previous findings and suggest that the written materials used in the article review baseline do not lead to accurate assessment administration. In contrast, participants initially exposed to the self-instruction package baseline (Group 4) mastered one or more trial types, and even for those trial types that did not meet mastery criterion, fidelity was relatively high and stable. The high and stable baseline data for Group 4 participants contradicts previous findings that suggest written information in isolation is an ineffective training procedure (Flynn & Lo, 2016; Kunnavantana et al. (2013) Kunnavantana et al. (2013) Lambert et al., 2013; Lambert, Bloom, et al., 2014; Moore et al., 2002; Phillips & Mudford, 2008; Rispoli et al., 2015, 2016).

Once Group 1 and 2 participants received the self-instruction package, performance increased substantially, with all participants mastering at least one trial type. Again, even for trial types that were not mastered, procedural fidelity was moderate to high, with substantial increases in fidelity for one or more trial types for all participants. For participants in Group 3, the percentage of trials with fidelity during the TA data sheet baseline was slightly greater overall in comparison to Group 2 participants supplied with the journal article during baseline. Once the written instruction component was introduced in combination with the TA data sheet, fidelity further increased for almost all trial types across participants in this group, suggesting that the written self-instruction package provided more complete instructions needed for accurate implementation. In other words, it is possible that written instructions provided the participants with sufficient information to respond to the prompts provided by the TA data sheet. Although the use of the TA data sheet in isolation was less effective in increasing TBFA procedural fidelity than when used in combination with more complete written instructions, more research is needed to determine the minimal content of written materials required to produce optimal outcomes.

In addition to gains in procedural fidelity, all participants could reliably collect data on the targeted problem behavior on 90% of trials after being provided only the self-instruction package. These increases in performance are again contrary to findings of previous research that indicates the use of written materials in traditional FA and TBFA trainings is ineffective (Flynn & Lo, 2016; Kunnavatana, Bloom, Samaha, & Dayton, 2013; Kunnavatana, Bloom, Samaha, Lignugaris-Kraft, et al., 2013; Lambert et al., 2013; Lambert, Bloom, et al., 2014; Moore et al., 2002; Phillips & Mudford, 2008; Rispoli et al., 2015, 2016; Ward-Horner & Sturmey, 2012). This discrepancy may be attributed to the fact that the content of written materials used in this study differed significantly from that used in previous research. As opposed to providing participants with a journal article to review, or lecture on ABA or FA rationale, the self-instruction package used in this study was written in everyday language, omitting technical behavioral terms, background information on ABA, and FA rationale. The written instructions also included examples of correct and incorrect responses and suggested practice activities using assessment materials. Like the participants included in this study, paraprofessionals who might assist in the implementation of a TBFA likely have limited knowledge of ABA or FA methodology.

Allen and Warzak (2002) suggested that when parents and paraprofessionals are unfamiliar with ABA, the use of behavioral terminology during training could impede performance. These researchers point to the possibility that the learner’s limited understanding of behavioral terms may conflict with their conventional understanding of the same words. This may lead to the development of ineffective rules, which in turn leads to poor procedural fidelity. Written materials used in this study may have been more effective in increasing procedural fidelity because any rules generated after reading them were not likely to conflict with participants’ existing understanding of the terms they encountered. Instead, the self-instruction package clearly specified what to do and what not to do, as opposed to clouding instructions with background and rationale information unlikely to influence performance.

Although educating others on ABA background and rationale is a necessary and commendable task, including this information along with procedural instructions may actually impede performance. Although TBFAs should always be conducted with appropriate oversight and supervision by a behavioral professional with a more comprehensive conceptual understanding of FA methodology, revising written materials for use by nonbehavioral audiences may prove to be a valuable and cost-effective training strategy that facilitates dissemination. Modifying written materials used in clinical settings by eliminating technical jargon and providing concise, complete rules presents a time- and cost-efficient training option and aligns with our ethical obligation to communicate with others using “language and graphic displays of data that are reasonably understandable” (Behavior Analyst Certification Board, 2014, p. 11). Thus, further evaluations of such modifications deserve investigation.

Although improvements in procedural fidelity achieved with the use of the self-instruction package were substantial and demonstrated a valuable contribution to the TBFA training literature, the need for small group training and even one-on-one feedback was not entirely eliminated. This need was underscored by the observation that participants who had previously mastered specific trial types in the self-instruction package condition made errors during the small group training and assessment trials that followed. Given that the small group training scripts included a wider variety of scenarios likely to occur in the natural environment, this result is understandable. Despite this limitation, the self-instruction package provides a reasonable starting place for training that is economical and requires little effort to implement. The small group training model then provides a viable way to “bridge the gap” between training and implementation. In applied settings where training time and qualified staff are limited, valuable resources can be saved by only providing group training for trial types not mastered through self-instruction alone. Last, participants who received the self-instruction package in baseline (Group 4) required the fewest number of one-on-one feedback trials, and the limited need for individualized feedback across groups indicates that the training packages evaluated align with the demand to develop effective and efficient clinical training procedures.

This study is also limited by the failure to isolate the effects of the written instructions without the use of the TA data sheet. Although we examined the use of the TA data sheet without the written instructions with Group 3, the role of the written instructions alone has yet to be evaluated. Future research should examine both components of the self-instruction package, in the context of TBFA training and in training a variety of other behavioral assessments and interventions. In addition, the order of trial types during training and assessment phases in this study did not vary as it would when implementing a TBFA in the natural environment. Although the order of trial administration may be a limitation, it provided participants with predictability and allowed them to build on skills learned in previous trial types. While the order of trial administration may complicate the interpretation of results, the low number of training trials required in small group training may indicate that this fixed order facilitated skill acquisition. Future studies should examine the order of TBFA training trial administration to determine whether participants who learn to conduct TBFA trials in a specific order can generalize skills to administration settings in the natural environment where trials are conducted in an unspecified order.

2Incorporation of generalization and maintenance measures, as done in previous studies (e.g., Kunnavatana, Bloom, Samaha, & Dayton, 2013; Moore et al., 2002; Rispoli et al., 2016), would also provide useful information about whether training actually transfers to natural contexts where skills should be used. Because training environments and rehearsal with adult confederates likely differs significantly from the wide variety of clients and settings encountered when actually conducting a TBFA, generalization and maintenance information would provide important insight into the true efficacy of the training procedures evaluated in this study. Finally, future research should seek to apply the most effective and efficient TBFA training package on a larger scale to increase the external validity of results. Because many behavioral service providers conduct trainings for groups of employees, replications with a greater number of participants would allow for evaluation of the feasibility of procedures. Specifically, because many people whom we wish to train to conduct TBFAs have little to no ABA or FA background, future investigators should continue to examine this population to ensure training reaches those who would benefit most.

Acknowledgements

We thank Gaby Chagolla, Stephanie Hanley, and Kendall Rankin for their assistance in conducting this project. The research presented in this article was completed in partial fulfillment of thesis requirements for the master’s degree by the first and second authors at California State University, Sacramento.

Appendix

graphic file with name 40617_2019_388_Figaa_HTML.jpg

graphic file with name 40617_2019_388_Figab_HTML.jpg

graphic file with name 40617_2019_388_Figac_HTML.jpg

graphic file with name 40617_2019_388_Figad_HTML.jpg

Funding

No funding was received for this study.

Compliance with Ethical Standards

Conflict of Interest

All authors declare that there are no conflicts of interest.

Ethical Approval

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.

Informed Consent

Informed consent was obtained from all individual participants included in the study.

Footnotes

Implications for Practice

Researchers and practitioners should:

• design training packages adapted for use with audiences with less behavioral experience to effectively disseminate TBFA technology;

• modify didactic training materials to eliminate technical jargon to achieve increases in TBFA procedural fidelity; and

• consider additional training components for learners to master TBFA administration.

Publisher’s Note

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

References

  1. Allen KD, Warzak WJ. The problem of parental non-adherence in clinical behavior analysis: Effective treatment is not enough. Journal of Applied Behavior Analysis. 2002;33:373–391. doi: 10.1901/jaba.2000.33-373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anderson DC, Crowell CR, Hantula DA, Siroky LM. Task clarification and individual performance posting for improving cleaning in a student-managed university bar. Journal of Organizational Behavior Management. 1988;9:523–548. doi: 10.1300/j075v09n02_06. [DOI] [Google Scholar]
  3. Austin JL, Groves EA, Reynish LC, Francis LL. Validating trial-based functional analysis in mainstream primary school classrooms. Journal of Applied Behavior Analysis. 2015;48:274–288. doi: 10.1002/jaba.208. [DOI] [PubMed] [Google Scholar]
  4. Behavior Analyst Certification Board . Professional and ethical compliance code for behavior analysts. Littleton, CO: Author; 2014. [Google Scholar]
  5. Bloom SE, Iwata BA, Fritz JN, Roscoe EM, Carreau AB. Classroom application of a trial-based functional analysis. Journal of Applied Behavior Analysis. 2011;44:19–32. doi: 10.1901/jaba.2011.44-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bloom SE, Lambert JM, Dayton E, Samaha AL. Teacher-conducted trial-based functional analyses as the basis for intervention. Journal of Applied Behavior Analysis. 2013;46:208–218. doi: 10.1002/jaba.21. [DOI] [PubMed] [Google Scholar]
  7. Crowell CR, Anderson DC, Abel DM, Sergio JP. Task clarification, performance feedback, and social praise: Procedures for improving the customer service of bank tellers. Journal of Applied Behavior Analysis. 1988;21:65–71. doi: 10.1901/jaba.1988.21-65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Davis JR, Ramana EP, Capponi DR. Acceptability of behavioral staff management techniques. Behavioral Residential Treatment. 1989;4:23–44. doi: 10.1002/bin.2360040104. [DOI] [Google Scholar]
  9. Desrochers MN, Hile MG, Williams-Moseley TL. Survey of functional assessment procedures used with individuals who display mental retardation and severe problem behaviors. American Journal on Mental Retardation. 1997;101:535–546. [PubMed] [Google Scholar]
  10. Ellingson SA, Miltenberger RG, Long ES. A survey of the use of functional assessment procedures in agencies serving individuals with developmental disabilities. Behavioral Interventions. 1999;14:187–198. doi: 10.1002/(SICI)1099-078X(199910/12)14:4<187::AID-BIN38>3.0.CO;2-A. [DOI] [Google Scholar]
  11. English CL, Anderson CM. Effects of familiar versus unfamiliar therapists on responding in the analog functional analysis. Research in Developmental Disabilities. 2004;25:39–55. doi: 10.1016/j.ridd.2003.04.002. [DOI] [PubMed] [Google Scholar]
  12. Flynn SD, Lo Y. Teacher implementation of trial-based functional analysis and differential reinforcement of alternative behavior for students with challenging behavior. Journal of Behavioral Education. 2016;25:1–31. doi: 10.1007/s10864-015-9231-2. [DOI] [Google Scholar]
  13. Graff RB, Karsten AM. Evaluation of a self-instruction package for conducting stimulus preference assessments. Journal of Applied Behavior Analysis. 2012;45:69–82. doi: 10.1901/jaba.2012.45-69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gravina N, VanWagner M, Austin J. Increasing physical therapy equipment preparation using task clarification, feedback and environmental manipulations. Journal of Organizational Behavior Management. 2008;28:110–122. doi: 10.1080/01608060802100931. [DOI] [Google Scholar]
  15. Iwata BA, Dorsey MF, Slifer KJ, Bauman KE, Richman GS. Toward a functional analysis of self-injury. Journal of Applied Behavior Analysis. 1994;27:197–209. doi: 10.1901/jaba.1994.27-197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Iwata BA, Wallace MD, Kahng S, Lindberg JS, Roscoe EM, Connors J, et al. Skill acquisition in the implementation of functional analysis methodology. Journal of Applied Behavior Analysis. 2000;33:181–194. doi: 10.1901/jaba.2000.33-181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kazdin AE. Single-case research designs: Methods for clinical and applied settings. 2. New York, NY: Oxford University Press; 2011. [Google Scholar]
  18. Kodak T, Fisher WW, Paden A, Dickes N. Evaluation of the utility of a discrete-trial functional analysis in early intervention classrooms. Journal of Applied Behavior Analysis. 2013;46:301–306. doi: 10.1002/jaba.2. [DOI] [PubMed] [Google Scholar]
  19. Kunnavatana SS, Bloom SE, Samaha AL, Dayton E. Training teachers to conduct trial-based functional analyses. Behavior Modification. 2013;37:707–722. doi: 10.1177/0145445513490950. [DOI] [PubMed] [Google Scholar]
  20. Kunnavatana SS, Bloom SE, Samaha AL, Lignugaris-Kraft B, Dayton E, Harris SK. Using a modified pyramidal training model to teach special educators to conduct trial-based functional analyses. Teacher Education and Special Education: The Journal of the Teacher Education Division of the Council for Exceptional Children. 2013;36:267–285. doi: 10.1177/0888406413500152. [DOI] [Google Scholar]
  21. Lambert JM, Bloom SE, Clay CJ, Kunnavatana SS, Collins SD. Training residential staff and supervisors to conduct traditional functional analyses. Research in Developmental Disabilities. 2014;35:1757–1765. doi: 10.1016/j.ridd.2014.02.014. [DOI] [PubMed] [Google Scholar]
  22. Lambert JM, Bloom SE, Irvin J. Trial-based functional analysis and functional communication training in an early childhood setting. Journal of Applied Behavior Analysis. 2012;45:579–584. doi: 10.1901/jaba.2012.45-579. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lambert JM, Bloom SE, Kunnavatana SS, Collins SD, Clay CJ. Training residential treatment staff to conduct trial-based functional analyses. Journal of Applied Behavior Analysis. 2013;46:296–300. doi: 10.1002/jaba.17. [DOI] [PubMed] [Google Scholar]
  24. Lambert JM, Lloyd BP, Staubitz JL, Weaver ES, Jennings CM. Effect of an automated training presentation on pre-service behavior analysts’ implementation of trial-based functional analysis. Journal of Behavioral Education. 2014;23:344–367. doi: 10.1007/s10864-014-9197-5. [DOI] [Google Scholar]
  25. LaRue RH, Lenard K, Weiss MJ, Bamond M, Palmieri M, Kelley ME. Comparison of traditional and trial-based methodologies for conducting functional analyses. Research in Developmental Disabilities. 2010;31:480–487. doi: 10.1016/j.ridd.2009.10.020. [DOI] [PubMed] [Google Scholar]
  26. Lloyd BP, Weaver ES, Staubitz JL. A review of functional analysis methods conducted in public school classrooms settings. Journal of Behavioral Education. 2016;25:324–356. doi: 10.1007/s10864-015-9243-y. [DOI] [Google Scholar]
  27. Lloyd BP, Wehby JH, Weaver ES, Goldman SE, Harvey MN, Sherlock DR. Implementation and validation of trial-based functional analyses in public elementary school settings. Journal of Behavioral Education. 2015;24:167–195. doi: 10.1007/s10864-014-9217-5. [DOI] [Google Scholar]
  28. Mirman R. Sales management: An effective performance system. In: O’Brien RM, Dickinson AM, Rosow MP, editors. Individual behavior modification. New York, NY: Pergamon Press; 1982. pp. 159–182. [Google Scholar]
  29. Moore JW, Edwards RP, Sterling-Turner HE, Riley J, DuBard M, McGeorge A. Teacher acquisition of functional analysis methodology. Journal of Applied Behavior Analysis. 2002;35:73–77. doi: 10.1901/jaba.2002.35-73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Phillips KJ, Mudford OC. Functional analysis skills training for residential caregivers. Behavioral Interventions. 2008;23:1–12. doi: 10.1002/bin.252. [DOI] [Google Scholar]
  31. Rispoli M, Burke MD, Hatton H, Ninci J, Zaini S, Sanchez L. Training Head Start teachers to conduct trial-based functional analysis of challenging behavior. Journal of Positive Behavior Interventions. 2015;17:235–244. doi: 10.1177/1098300715577428. [DOI] [Google Scholar]
  32. Rispoli M, Neely L, Healy O, Gregori E. Training public school special educators to implement two functional analysis models. Journal of Behavioral Education. 2016;25:249–274. doi: 10.1007/s10864-016-9247-2. [DOI] [Google Scholar]
  33. Rispoli M, Ninci J, Neely L, Zaini S. A systematic literature review of trial-based functional analysis of challenging behavior. Journal of Developmental and Physical Disabilities. 2014;26:271–282. doi: 10.1007/s10882-013-9363-z. [DOI] [Google Scholar]
  34. Roscoe EM, Phillips KM, Kelly MA, Farber R, Dube WV. A statewide survey assessing practitioners’ use and perceived utility of functional assessment. Journal of Applied Behavior Analysis. 2015;48(4):830–844. doi: 10.1002/jaba.259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Sarokoff RA, Sturmey P. The effects of behavioral skills training on staff implementation of discrete-trial teaching. Journal of Applied Behavior Analysis. 2004;37(4):535–538. doi: 10.1901/jaba.2004.37-535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Shapiro M, Kazemi E, Pogosjana M, Rios D, Mendoza M. Preference assessment training via self-instruction: A replication and extension. Journal of Applied Behavior Analysis. 2016;49:1–15. doi: 10.1002/jaba.339. [DOI] [PubMed] [Google Scholar]
  37. Sigafoos J, Saggers E. A discrete-trials approach to the functional analysis of aggressive behavior in two boys with autism. Australia & New Zealand Journal of Developmental Disabilities. 1995;20:287–297. doi: 10.1080/07263869500035621. [DOI] [Google Scholar]
  38. Squires J, Wilder DA, Fixsen A, Hess E, Rost K, Curran R, Zonneveld K. The effects of task clarification, visual prompts, and graphic feedback on customer greeting and up-selling in a restaurant. Journal of Organizational Behavior Management. 2007;27:1–13. doi: 10.1300/j075v27n03_01. [DOI] [Google Scholar]
  39. Van Acker R, Boreson L, Gable R, Potterton T. Are we on the right course? Lessons learned about current FBA/BIP practices in schools. Journal of Behavioral Education. 2005;14:35–56. doi: 10.1007/s10864-005-0960-5. [DOI] [Google Scholar]
  40. Wallace MD, Doney JK, Mintz-Resudek CM, Tarbox RE. Training educators to implement functional analyses. Journal of Applied Behavior Analysis. 2004;37:89–92. doi: 10.1901/jaba.2004.37-89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Ward-Horner J, Sturmey P. Component analyses of behavior skills training in functional analysis. Behavior Interventions. 2012;27:75–92. doi: 10.1002/bin.1339. [DOI] [Google Scholar]
  42. Wilson C, Boni N, Hogg A. The effectiveness of task clarification, positive reinforcement and corrective feedback in changing courtesy among police staff. Journal of Organizational Behavior Management. 1997;17:65–99. doi: 10.1300/j075v17n01_04. [DOI] [Google Scholar]

Articles from Behavior Analysis in Practice are provided here courtesy of Association for Behavior Analysis International

RESOURCES