Week 4 Slocum and Tiger 2011 PDF

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2011, 44, 793–805 NUMBER 4 (WINTER 2011) AN ASSESSMENT OF THE EFFICIENCY OF AND CHILD PREFERENCE FOR FORWARD AND BACKWARD CHAINING...

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2011, 44, 793–805 NUMBER 4 (WINTER 2011) AN ASSESSMENT OF THE EFFICIENCY OF AND CHILD PREFERENCE FOR FORWARD AND BACKWARD CHAINING SARAH K. SLOCUM AND JEFFREY H. TIGER LOUISIANA STATE UNIVERSITY Comparative studies of forward and backward chaining have led some to suggest that sensitivity to each teaching procedure may be idiosyncratic across learners and tasks. The purposes of the current study were threefold. First, we assessed differential sensitivity to each chaining procedure within children when presented with multiple learning tasks of similar content but different complexity. Second, we evaluated whether differential sensitivity to a chaining procedure during a brief task predicted differential sensitivity during the teaching of longer tasks. Third, we directly assessed children’s preferences for each teaching procedure via a concurrent-chains preference assessment. Learners acquired all target skills introduced under both chaining conditions, but individual children did not consistently learn more efficiently with either procedure. Short-duration tasks were not predictive of performance in tasks of longer duration. Both chaining procedures were preferred over a baseline condition without prompting, but participants did not demonstrate a preference for either procedure. Key words: backward chaining, concurrent-chains preference assessment, forward chaining, preferences _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _2 Behavior analysts frequently employ response- the literature (Cooper, Heron, & Heward, chaining procedures to teach multistep tasks 2007); the current study focuses on forward that range from food preparation (Schuster, and backward chaining. Gast, Wolery, & Guiltinan, 1988), family-style Forward chaining involves teaching the initial dining (Wilson, Reid, Phillips, & Burgio, step in a task analysis to mastery and then 1984), and self-feeding (Hagopian, Farrell, sequentially teaching additional steps. After & Amari, 1996) to Internet usage (Jerome, a step is mastered and subsequent steps are Frantino, & Sturmey, 2007), playing a game of targeted for teaching, all previous steps along darts (Schleien, Wehman, & Kiernan, 1981), with the current step are required to be making a corsage (Hur & Osborne, 1993), and accurately completed to be considered correct assembling bicycle brakes, meat grinders, and and result in reinforcement delivery. For carburetors (Walls, Zane, & Ellis, 1981). instance, in a hypothetical task that requires Response chaining involves breaking a task Steps A, B, C, and D to be demonstrated in into its component parts via a task analysis and order, an instructor would teach Step A; then then sequentially teaching each individual Steps A and B; then Steps A, B, and C; and component to mastery levels via prompting finally, Steps A, B, C, and D. Typically, an and differential reinforcement. Total task, instructor would deliver reinforcement at the forward chaining, and backward chaining are completion of each successful response (i.e., the three variants of response chaining described in temporal location of reinforcement delivery would vary depending on the required terminal This study was conducted as an undergraduate honors step). Backward chaining involves teaching the thesis by the first author. Sarah K. Slocum is now at the final step of the task analysis initially and Psychology Department of the University of Florida. Correspondence concerning this article should be progressively teaching early components. As addressed to Jeffrey H. Tiger, who is now at the earlier steps are added, all previously taught Department of Psychology, University of Wisconsin– steps and the current step are required to be Milwaukee, 2441 E. Hartford Ave., Garland 219, Milwaukee, Wisconsin 53201 (e-mail: [email protected]). accurately completed in order to be considered doi: 10.1901/jaba.2011.44-793 correct and result in reinforcement delivery. 793 794 SARAH K. SLOCUM and JEFFREY H. TIGER Again, using our hypothetical task requiring retardation. Participants were assigned rand- Steps A, B, C, and D, Step D would be taught omly to either a forward-chaining group or first; then Steps C and D; followed by Steps B, backward-chaining group, and both groups C, and D; and finally, Steps A, B, C, and D. acquired this 18-step task in a similar number The instructor delivers reinforcement at the of trials. completion of the last step. Thus, regardless of Thus, neither forward nor backward chaining the stage of training, reinforcement is delivered has been consistently more efficacious in at the ‘‘natural’’ location (i.e., at the end of the promoting response acquisition. These out- task). comes led Spooner and Spooner (1984) in Given the success of both forward and their review of chaining procedures to surmise, backward chaining in teaching multistep tasks ‘‘it may be that different learners do better with across a variety of populations, including different procedures, and when different tasks persons with intellectual disabilities (Hur & are used, different results are obtainable’’ Osborne, 1993; Walls et al., 1981; Zane, Walls, (p. 123). This summary makes two fundamen- & Thvedt, 1981) as well as persons of typical tal assumptions: (a) All other things being development (Ash & Holding, 1990; Smith, equal, a given individual will consistently 1999; Weiss, 1978), some researchers have demonstrate differential sensitivity to one sought to compare the relative efficiency of teaching procedure with a given task, and (b) these teaching methods (Ash & Holding, 1990; although the histories that result in this Hur & Osborne, 1993; Smith, 1999; Walls differential sensitivity are highly idiosyncratic et al., 1981; Weiss, 1978). Weiss (1978) com- and potentially complex, we should be able to pared forward and backward chaining in the predict an individual’s sensitivity to a teaching acquisition of response chains with undergrad- procedure given their demonstrated sensitivity uate college students given a contrived task. to that procedure in the past. Researchers have Specifically, these authors developed an appa- not empirically validated these assumptions. If ratus that consisted of six buttons and required they are accurate, the identification of differen- participants to press different sequences of tial sensitivity to one teaching procedure should buttons to earn points. They taught each be valuable to teachers who are responsible for participant four six-step sequences using either teaching complex skills, specifically in helping forward or backward chaining and found them to identify the most efficacious teaching forward chaining to result in fewer incorrect procedure possible. If these assumptions are responses and more rapid acquisition. inaccurate, teachers and researchers may be In contrast, Walls et al. (1981) compared expending their energy and resources unneces- forward and backward chaining in assembling a sarily in attempting to compare and predict the bicycle brake, a meat grinder, and a carburetor effectiveness and efficiency of these procedures. with 22 people between the ages of 18 and 46 The initial purpose of this study was to who had been diagnosed with mild to moderate determine if, by holding all other factors intellectual delays and who were from a constant, children would indeed demonstrate a vocational rehabilitation center. The frequency consistent sensitivity to one teaching procedure, of incorrect responses and total training time and if so, if this sensitivity could be identified were similar across backward- and forward- by presenting a brief assessment task that then chaining conditions. Hur and Osborne (1993) could guide teachers’ selections of efficient compared backward and forward chaining in teaching procedures for longer response chains. teaching corsage making to children who had In addition to the relative efficiency of each been diagnosed with moderate to severe mental procedure, teachers also may select teaching  FORWARD AND BACKWARD CHAINING 795 procedures by considering individual children’s with the outcome of the 18-step task). The preferences for the available alternatives. Children, second part involved a preference assessment particularly those with disabilities, rarely are using the concurrent-chains methodology of afforded the opportunity to voice their preferences Hanley (2010), in which we introduced a novel for therapeutic programming. Providing such task and provided participants the opportunity opportunities, in addition to respecting the to select whether the skill was taught using individual’s autonomy, may result in increased forward-, backward-, or no-chaining methods. time on task and may limit the occurrence of problem behavior during instruction (Hanley, METHOD 2010; Mason, McGee, Farmer-Dougan, & Participants and Setting Risley, 1989; Powell & Nelson, 1997; Ringdahl, Vollmer, Marcus, & Roane, 1997). We recruited four participants from a local Assessment of children’s preferences for school for children with developmental and teaching strategies may be complicated, partic- learning disabilities. Each class at this school ularly when individuals have limited vocal offered small student-to-teacher ratios and some one-to-one instruction on a daily basis. All repertoires. Hanley and colleagues described participants, except Daniel, were behind their the use of a concurrent-chains procedure that peers in academic functioning and, according to offered a direct, nonverbal assessment of teacher report, rarely followed two-step instruc- individuals’ preferences and has been effective tions. Further, teachers reported that none of in determining preferences for behavioral the participants had experience with either interventions, classroom behavior-management chaining procedure prior to the study. Bella strategies, and teaching strategies with individ- was a 10-year-old girl who had been diagnosed uals of both typical and atypical development with attention deficit hyperactivity disorder (Hanley, 2010; Hanley, Piazza, Fisher, Con- (ADHD); her teacher reported that she was trucci, & Maglieri, 1997; Heal & Hanley, having trouble focusing during school instruc- 2007; Heal, Hanley, & Layer, 2009; Tiger, tion. Paul was an 11-year-old boy who had been Hanley, & Heal, 2006). This procedure simply diagnosed with ADHD; his teacher reported involves correlating salient stimuli with the that he had trouble concentrating long enough interventions or teaching strategies and then to complete several steps of a task. Daniel was allowing participants to select the correlated an 11-year-old boy with speech delays; his stimuli to gain brief access to the different teacher reported that although he was within interventions or teaching strategies. grade level on academic tasks, he had a stutter We conducted our current study in two and displayed occasional slowed speech. Katie parts. The first involved an efficiency assessment was a 12-year-old girl with learning deficits; her in which participants were taught 3-, 6-, 9-, and teacher reported she was behind grade level in 18-step motor sequences using both forward almost every area of academics. We obtained and backward chaining. We examined the parental consent and daily assent for each child; outcomes of this assessment to determine (a) all study procedures were preapproved by the if children exhibited a consistent differential Institutional Review Board of Louisiana State sensitivity to backward or forward chaining University. The first author conducted all with similar tasks and (b) if we could predict sessions in a vacant room in the school. differential acquisition of the 18-step tasks via learner performance with the tasks of shorter Preexperimental Preference Assessments length (i.e., correspondence between the out- Prior to initiating the formal experiment, we comes of the 3-, 6-, and 9-step comparisons conducted two preference assessments. First, we 796 SARAH K. SLOCUM and JEFFREY H. TIGER Table 1 which each motor sequence would be derived; Operational Definitions of Motor Movements in these were touching one’s nose, eye, and ear; Motor Chains patting one’s head and lap; and clapping one’s Touch head Place either hand on top of the head hands (we present operational definitions of Touch lap Place both hands on top of the thighs these behaviors in Table 1). We then randomly Clap Contact two palms of hands together selected from these movements to design pairs Touch eye Use either pointer finger to contact either eye Touch ear Use either pointer finger to contact either ear of motor sequences that included 3, 6, 9, and Touch nose Use either pointer finger to contact one’s nose 18 steps (we made selections without replace- ment until all six movements were selected and conducted a color preference assessment based then restarted for the 9- and 18-step motor on the procedures described by Heal et al. sequences). We randomly assigned members of (2009) to eliminate bias in the colors associated a motor sequence pair to either forward or with initial-link stimuli during our concurrent- backward chaining in a counterbalanced order chains preference assessment. We used 10 across participants such that any unintended different-colored sheets of construction paper differences in motor sequence difficulty would and presented each sheet to participants in be balanced across participants. For example, pairs. We instructed participants to select the we assigned three-step Motor Sequence A to color they ‘‘liked more’’ on each trial and the forward-chaining condition and three-step provided a brief statement of praise (e.g., Motor Sequence B to the backward-chaining ‘‘thanks’’) after each selection. After we paired condition for Daniel and Paul, and we assigned each colored sheet with each other colored the three-step Motor Sequence B to the sheet, we then ranked color preference based on forward-chaining condition and the three-step selection percentages (i.e., number of times Motor Sequence B to the backward-chaining selected divided by the number of times condition for Bella and Katie. We then presented). We selected three colors that were compared the efficiency of forward and back- ranked similarly from the lower end of the ward chaining sequentially across each motor- preference hierarchy and randomly assigned sequence pair. That is, we taught one three-step each color to be correlated with each condition sequence with forward chaining and one three- in our concurrent-chains preference assessment. step sequence with backward chaining in Next we conducted a leisure-item preference alternating sessions. After both sequences assessment based on the procedures described by reached mastery, we taught both six-step Fisher et al. (1992) to identify activities to which sequences, followed by both nine- and 18-step access was delivered as a reinforcer during sequences. We compared backward and forward teaching conditions. We identified items to chaining across each sequence length in an include via an interview with each participant’s adapted alternating treatments design. classroom teacher and presented items in paired Efficiency assessment. Prior to each pair of arrays to identify highly preferred activities. We teaching sessions, the participant selected a included access to the top three items to be used leisure item from an array of preferred items; we as reinforcers in the study. There were also delivered the selected item as a reinforcer during instances in which participants requested an item the next pair of sessions to ensure that the not in their top three, and we allowed them to quality of reinforcement was identical across obtain that item as reinforcement for that day. chaining conditions. We conducted one to six 10-trial sessions per day within a daily 45-min Procedure session block allocated to each participant, To develop our experimental motor sequenc- typically 4 days per week. Individual session es, we first selected six motor movements from duration depended on the number of steps in  FORWARD AND BACKWARD CHAINING 797 the motor sequence and participant responding in the targeted step and then all remaining during sessions (i.e., sessions that required untargeted steps. The teacher then waited 5 s prompting were longer than sessions with after completion of the trial to deliver an independent responding). Anecdotally, session instruction to initiate the next trial. This durations ranged from about 5 min up to training continued for the first step until the 25 min. In accordance with our experimental participant correctly and independently engaged design, we alternated forward- and backward- in that step on three consecutive trials (mastery chaining sessions in a random and counterbal- criterion). After meeting this mastery criterion, anced order by flipping a coin to determine the teacher then targeted Steps 1 and 2 together, which would be conducted first. then Steps 1, 2, and 3, and so on. Subsequent At the onset of each comparison, we trials were similar except that both the targeted conducted a minimum of three baseline trials step and all previously mastered steps were to ensure that participants could not engage in then required to produce reinforcement. If the the motor sequence prior to instruction. During participant did not initiate the motor sequence, baseline trials, the teacher instructed the engaged in any incorrect responses, or delayed participant to complete a motor sequence by more than 1 s between any previously (e.g., ‘‘Do the — dance’’; we assigned each mastered steps in the sequence, the teacher dance an arbitrary name to facilitate discrimi- provided a model of all currently targeted nation of the experimental conditions). The components (i.e., the current step and any teacher did not provide any consequences for previously mastered steps) and provided the correct or incorrect responding, and after 5 s participant the opportunity to respond again. If she instructed the participant to complete the the correct motor sequence was not emitted motor sequence again, initiating a new trial. following the model, the teacher then physically Following these baseline trials, we initiated guided the participant to complete the targeted instruction of the motor sequence. and any previously mastered steps in the During forward-chaining conditions, we ini- sequence. tially targeted only the first step of the motor Teaching trials were similar during back- sequence. That is, after the instruction, ‘‘Do the ward-chaining conditions, except that we ini- — dance,’’ if the child completed the targeted tially targeted only the terminal step of the steps of the motor sequence independently, the motor sequence. That is, immediately after the teacher delivered praise (e.g., ‘‘Great job!’’), instruction, the teacher physically guided the physically guided the participant to complete all participant to complete all steps except for the untargeted steps, and delivered access to the last step in the sequence. Correct responding preselected leisure item for 30 s. We provided resulted in praise and access to a preferred physical guidance for untrained steps to equate leisure item, and incorrect responding or a exposure to these steps during this condition failure to respond within 5 s resulted in the with the exposure experienced during back- delivery of a model prompt of all steps. If the ward-chaining sessions described below. If the child did not engage in the correct response child did not complete the required step within 5 s of the model prompt, the teacher independently within 5 s of the instruction or physically guided the child to complete the the engaged in an incorrect response, the response. The mastery criterion for that step teacher provided a model prompt that demon- was identical to the criterion used in the strated the required step. If the child did not forward-chaining condition (i.e., three consec- then complete the required step within 5 s, the utive, independently correct trials). After mas- teacher physically guided the child to engage tery on the terminal step had been achieved, 798 SARAH K. SLOCUM and JEFFREY H. TIGER the teacher initiated the next trial with the blocks; instead, terminal-link experiences contin- instruction and immediately physically guided ued until participants mastered the motor the participant to complete all but the last two sequences). In this regard, the speed of each steps and provided a 5-s delay, and so on. participant’s acquisition of the motor sequence Again, all previously mastered steps were determined the number of trials and duration required to be demonstrated with less than 1-s of exposure to each terminal link; participants delay between responses to avoid more intrusive required no more than 20 trials to master a task. prompting. The teacher delivered praise and The teacher conducted control terminal links 30-s access to the identified preferred leisure similarly to baseline sessions, except that she item only after independent responses that terminated sessions based on time. Specifically, we included both the current and all previously measured the duration from the onset of the first mastered steps in the correct order. instruction to the moment the mastery criterion In both forward- and backward-chaining was met during the forward- and backward- conditions, we set the terminal mastery criterion chaining terminal links and yoked the duration of as three consecutive trials of independently control sessions to be equal to the mean of the correct responding of the complete motor previous two chaining terminal links. We initiated sequence following the initial instruction. After control sessions by providing the same initial a participant met the terminal mastery criterion instruction, ‘‘Do the — dance,’’ but did not for a motor sequence, we continued to conduct provide any additional prompting. The teacher training trials for the other motor sequence did not otherwise interact with the participant until it also met mastery. Each participant during these terminal links. Anecdotally, partici- learned the 3-, 6-, 9-, and 18-step sequences in pants generally sat quietly during these periods consecutive order. We determined the relative and awaited the next initial-link opportunity. sensitivity to a chaining procedure based on the We interspersed forced-choice sessions, in differential number of trials required to meet which the teacher presented only one initial-link the terminal mastery criterion in each compar- card on each trial and physically guided the ison. Following mastery of these motor se- participant to complete the initial-link selec- quences, we initiated the preference assessment. tion, in order to promote continued exposure to Preference assessment. We presented three each of the terminal-link conditions. We began colored cards on a table in front of each the preference assessment with three forced- participant to start the initial link of a pre- choice exposure trials (one each for forward ference assessment trial. Selections of (i.e., chaining, backward chaining, and the control touching) one card completed the initial link condition) and included three additional and resulted in the onset of the terminal link of forced-choice exposure trials following every the chain in which the teacher taught a novel five free-choice trials; we considered only three-step motor sequence to completion using selections during free-choice trials with all three one of the two chaining procedures or a control colored cards present in determining preference. procedure, depending on which card the par- Between two and eight preference assessment ticipant selected. The teacher conducted back- trials were conducted per day, with no more ward- and forward-chaining terminal links simi- than five free-choice trials in one day. The larly to those described during the efficiency preference assessment duration ranged from assessments, except that teaching continued 3 days to 9 days. Individual terminal-link until participants met mastery criterion of in- experiences ranged from 2 min to 4 min. We dependently completing the motor sequence terminated the preference assessment following following three consecutive instructions (i.e., free-choice selections of one initial link that sessions were no longer defined by 10-trial totaled six selections greater than the next  FORWARD AND BACKWARD CHAINING 799 closest initial link or following 15 total free- trials scored in agreement, divided this sum by choice trials. the total number of trials, and converted this to a percentage, resulting in mean interobserver Measures and Interobserver Agreement agreement coefficients of 99% (range, 80% to Observers collected data on participant 100%) for Paul, 97% (range, 80% to 100%) for responding on a trial-by-trial basis using Daniel, 99% (range, 80% to 100%) for Bella, pencil-and-paper data sheets. During the effi- and 99% (range, 90% to 100%) for Katie. cacy assessment, observers coded a trial with an In addition, we collected data on partici- independently correct response if the partici- pants’ selections during the initial links of the pant initiated the motor sequence within 5 s of preference assessment. On each trial, an the instruction and completed the correct target observer noted the color of the selected card; step and all previously mastered steps in the selection was defined as contact of the partic- correct sequence with no more than 1 s between ipant’s hand to a colored card. We compared each step. Observers coded a trial with a correct observers’ records of colored cards selected response following a model if the participant during 43%, 25%, 67%, and 25% of initial- initiated the motor sequence within 5 s of the link trials for Paul, Daniel, Bella, and Katie, model prompt and completed the correct target respectively. Observers’ records were scored in step and all previously mastered steps in the agreement if they both coded the same colored correct sequence with no more than 1 s between card selected and in disagreement if their each step. Observers coded a trial with a records did not match; observers agreed on physically guided response if the teacher 100% of the selections across all participants. physically guided the participant to complete the target response sequence. During baseline RESULTS trials only, observers coded a trial with no response or incorrect response if the participant Efficiency Assessment failed to initiate the correct motor sequence We present a summary of each participant’s within 5 s of the instruction or engaged in an acquisition of motor sequences in Figure 1, incorrect response; these data were coded which depicts the number of trials to mastery only during baseline because the programmed given forward- and backward-chaining condi- contingencies during teaching conditions re- tions across each motor sequence comparison. sulted in additional prompting and the code None of the participants engaged in an inde- captured this behavior in the other response pendently correct response during baseline. Paul categories. acquired the three-step motor sequence taught A second observer simultaneously but inde- via backward chaining in 12 fewer trials than the pendently collected data to provide an indicator three-step motor sequence taught with forward of the reliability of measurement during 40%, chaining (20 and 32 trials, respectively). We 55%, 80%, and 64% of sessions during Paul’s, observed a similar pattern during the six-step motor Daniel’s, Bella’s, and Katie’s teaching evalua- sequence comparison in which Paul met the tions, respectively. For the teaching evaluations, mastery criterion in 13 fewer trials given backward we compared observers’ records on a trial-by-trial chaining (39 and 52 trials for the backward- basis; a trial was scored in agreement if both chaining and forward-chaining motor sequences, observers coded the same response category (i.e., respectively). However, we did not observe these independently correct, correct following a mod- differences during the nine-step motor sequence el, physically guided, or no response or incorrect) comparison; he acquired both motor sequences and in disagreement if the observers’ records did following exactly 78 trials. Paul acquired the 18- not match. We then summed the number of step backward-chaining motor sequence in seven 800 SARAH K. SLOCUM and JEFFREY H. TIGER Figure 1. Overall performance of all participants on the 3-, 6-, 9-, and 18-step motor tasks taught with forward and backward chaining. fewer trials than the forward-chaining motor (60 and 57 trials for backward and forward sequence (147 and 155 trials, respectively). chaining, respectively). During the 18-step Daniel demonstrated more variable perfor- comparison, Daniel reached mastery in nine mances across the different motor sequence fewer trials during the forward-chaining condi- lengths. He acquired the three-step backward- tion than during the backward-chaining condi- chaining motor sequence in nine fewer trials tion (128 and 119 trials for backward and than the three-step forward-chaining motor forward chaining, respectively). sequence (11 and 20 trials for backward and Katie acquired both three-step motor se- forward chaining, respectively). Similarly, he quences in the same number of trials with met mastery criteria for the six-step backward- forward and backward chaining (14 trials). She chaining motor sequence in three fewer trials mastered the six-step backward-chaining motor than the six-step forward-chaining motor se- sequence in one fewer trial than the six-step quence (34 and 37 trials for backward and forward-chaining motor sequence (29 and 30 forward chaining, respectively). The reverse trials for backward and forward chaining, pattern was true for the longer motor sequences. respectively). We observed larger differences in During the nine-step motor-sequence compar- her acquisition of the nine- and 18-step motor ison, Daniel met the mastery criterion during sequences. She mastered the nine-step forward- the forward-chaining condition in three fewer chaining motor sequence in 18 fewer trials than trials than the backward-chaining condition the backward-chaining motor sequence (88 and  FORWARD AND BACKWARD CHAINING 801 Table 2 correspondence between the relatively shorter Correspondence Between Outcomes of the Short- and motor sequences and the longer motor se- Long-Chain (18-Step) Motor Sequences quences (i.e., would the determination of Participant 3-step 6-step 9-step 18-step a ‘‘winner’’ from a brief comparison predict children’s sensitivity in one of the longer Paul backward backward tie backward Daniel backward backward forward forward duration comparisons?). These data are shown Katie tie backward forward forward in Table 2. The outcomes of the three-step and Bella forward forward forward forward Percentage 50 50 75 six-step comparisons corresponded with the correspondence outcomes of the 18-step comparison in only two of four cases (50%), which is equivalent to 70 trials for backward and forward chaining, chance. The nine-step comparison may have respectively) and the 18-step motor sequence in been slightly more predictive, with correspon- 35 fewer trials (155 and 120 trials for backward dence in three of four cases (75%); however, the and forward chaining, respectively). small number of participants prohibits any The outcomes for Bella were the most definitive conclusions. consistent across comparisons; she always met Preference Assessment mastery criterion in fewer trials given forward We show each participant’s cumulative chaining, requiring three fewer trials in the initial-link selections from the preference por- three-step motor-sequence comparison (21 and tion of our study in Figure 2; these data 18 trials for backward and forward chaining, represent responding only during free-choice respectively), five fewer trials in the six-step trials. Daniel and Katie alternated between motor sequence comparison (45 and 40 trials forward- and backward-chaining selections for backward and forward chaining, respective- consistently. They both made seven selections ly), 10 fewer trials in the nine-step comparison of forward chaining and eight selections of (94 and 84 trials for backward and forward backward chaining, and neither participant chaining, respectively), and five fewer trials in chose the control condition during the assess- the 18-step comparison (175 and 170 trials for ment. Bella similarly alternated between selec- backward and forward chaining, respectively). tions of forward chaining (six selections) and In summary of the aforementioned results, backward chaining (seven selections), but she backward chaining was associated with fewer selected the control condition on two trials. We trials to mastery in three of four comparisons terminated Daniel’s, Katie’s, and Bella’s assess- for Paul with an equal number of trials in the ments after the 15-trial stop criterion with nine-step comparison. Backward chaining was the determination that they each preferred associated with more rapid acquisition in two backward and forward chaining similarly, and comparisons and slower acquisition in two preferred both chaining conditions over the comparisons for Daniel. Forward chaining was control condition. A preference for one chain- associated with more rapid acquisition in two ing procedure emerged only for Paul. After comparisons, slower acquisition in one com- choosing the control condition on the first trial, parison, and equal acquisition in one compar- he then selected forward chaining in the next ison for Katie. Finally, forward chaining was seven consecutive opportunities. associated with more rapid acquisition in all four comparisons for Bella. DISCUSSION We then examined the outcomes of the 3-step, 6-step, and 9-step comparisons to the We compared the acquisition of motor 18-step comparison to determine the level of sequences across a sample of children with 802 SARAH K. SLOCUM and JEFFREY H. TIGER Figure 2. Cumulative selections during free-choice trials of the concurrent-chains preference assessment. special needs using both forward and backward backward-chaining conditions or in an equal chaining. Across our four participants, this number of trials for the shorter motor sequenc- provided a total of 16 comparisons of forward es, both met mastery criteria more rapidly and backward chaining. Forward chaining was during the forward-chaining conditions of the associated with fewer trials to mastery in eight nine- and 18-step motor sequences. Thus, it did comparisons, backward chaining was associated not appear that there were consistent differences with fewer trials to mastery in six comparisons, in an individual’s sensitivity to either teaching and no difference in trials to mastery was procedure, despite the use of similar tasks. obtained in two comparisons. With the excep- We also sought to determine if differentially tion of the nine- and 18-step motor sequences for rapid acquisition of a shorter motor sequence Katie, we observed marginal differences between could serve as a behavioral assessment to predict backward and forward chaining (M 5 5.79 differential sensitivity in acquisition of a longer difference in trials to mastery; range, 0 to 13). motor sequence. We found correspondence Only one of the four participants demon- between the three- and six-step motor sequences strated a consistent differential sensitivity to a and the 18-step motor sequence for two of the particular teaching condition in each of her four participants (Paul and Bella), and the nine-step comparisons; Bella consistently met the mastery motor sequence was predictive for three of the criterion more quickly under the forward- participants (Daniel, Katie, and Bella). Howev- chaining conditions (M 5 5.75 trials differ- er, given the overall variability within each ence). Paul met the mastery criterion more participant, it seems most appropriate to rapidly in the backward-chaining conditions for conclude that these correspondences were three of the four comparisons. Although Daniel chance occurrences. That is, there were no and Katie met mastery more quickly in the consistencies in individuals’ acquisition of  FORWARD AND BACKWARD CHAINING 803 motor sequences given forward and backward initial links from the backward- and forward- chaining, and thus any differences in the 18- chaining conditions in the present assessment, we step motor sequences could be predicted equally would have been unable to determine if the well by a coin flip or by previous performance. resultant data were indicative of indifference or In total, the results of our efficiency assessment of a failure to discriminate the contingencies ran contrary to Spooner and Spooner’s (1984) associated with the initial links. By including a assertion that there are idiosyncratic differences control condition that was unlikely to occasion in sensitivity to chaining procedures that would selections, we can be fairly confident that our be consistent in individual learners. In the participants were indeed discriminating between current study, we assessed children’s sensitivity the outcomes of their initial-link selections based to each teaching procedure repeatedly using on their minimal selections of the control highly similar tasks. Despite holding the task condition; they were indifferent with regards to and the learner constant, we did not identify which chaining procedure they experienced. any consistencies in sensitivity to one teaching The overall results of this evaluation suggest procedure beyond what would be expected by that (a) there is no consistent difference in task chance. Based on the variability in outcomes of acquisition given instruction consisting of forward comparisons in the literature and the results of or backward chaining between or within partic- the current study, it is likely safe to conclude that ipants, (b) it probably is not possible to use forward- and backward-chaining procedures are differential sensitivity during a brief task to predict similarly effective in establishing behavior chains. differential sensitivity during longer tasks, and (c) In addition to assessing the differential efficiency these procedures are neither differentially efficient of these procedures, we also assessed participants’ nor differentially preferred. preferences for these procedures relative to a control The lack of difference between these proce- condition. All participants preferred either chaining dures, both in terms of efficiency and preference, procedure over the no-chaining control, and three is not an unimportant finding. Rather, these of the four participants displayed no preference for findings indicate that both procedures are effective one chaining procedure over another. That is, at engendering complex chains of new behavior, similar to there being a lack of difference in the and both are preferred by the consumers that efficiency of these procedures, the children who experience them. From a practitioner’s perspec- experienced these teaching procedures were also tive, these results suggest that teachers and indifferent regarding the procedure they received. interventionists should be comfortable imple- The fourth participant, Paul, did meet our menting either procedure with their clientele. preference criteria for forward chaining. However, Our data did not completely rule out the we believe that this may not be an accurate possibility that some tasks may be taught more reflection of his preference; instead, his initial effectively with either forward or backward selection resulted in reinforcement and was chaining. We chose to compare acquisition strengthened immediately and differentially to the across similar motor tasks to rule out differences exclusion of the other options. For instance, at one in task difficulty as a potential confounding point he asked the therapist, ‘‘Why would I switch effect. It may be that greater sensitivity to a to another color if I get my toy with purple [the particular chaining procedure would be identi- color of the initial-link card associated with the fied for a different type of task. For instance, in forward-chaining condition]?’’ making a sandwich, spreading peanut butter These results draw attention to an important across one side of the bread creates a continuous methodological feature of the concurrent-chains visual discriminative stimulus to occasion the procedure, specifically the inclusion of the next response in the task (e.g., putting the knife control condition. Had we included only the down). It may be that such continuous stimuli 804 SARAH K. SLOCUM and JEFFREY H. TIGER exert stronger stimulus control than the presum- individual to remove his or her sweater indepen- ably brief stimuli of a motor movement (e.g., one dently, it is common to guide pulling each arm out can no longer experience touching one’s nose of its sleeve and placing hands under the collar to after that response has ended). These continuous set the occasion for him or her to push the sweater discriminative stimuli may better set the occasion over the head. To minimize the likelihood that this for a subsequent response, although it is unclear level of exposure to future targeted steps did not how this presentation would differentially favor differentially favor backward chaining, we chose to one chaining procedure over the other. include physical guidance on the untrained steps in It also may be the case that some tasks allow both forward- and backward-chaining conditions. more natural or direct sources of reinforcement It is possible that this additional prompting and to result from the completion of the chain using exposure to future targeted steps were in part backward chaining in lieu of socially contrived or responsible for the marginal differences between indirect reinforcement when using forward conditions by resulting in rapid acquisition in both chaining. For instance, in making a peanut conditions. If we had not prompted untrained steps butter sandwich, the immediate natural conse- prior to targeting them for instruction, differences quence of completing a targeted step with between forward and backward chaining may have backward chaining would be access to the been more apparent. Future research may evaluate completed sandwich. By contrast, completing these procedures without additional prompting in an early step via forward chaining (e.g., laying place. pieces of bread side by side) would not result in Finally, although the children did not dem- the same automatic consequence (i.e., complete onstrate a preference for one of these teaching sandwich), and thus teachers would need to rely procedures, it is likely that teachers may have a on delivering another reinforcer (e.g., praise, distinct preference for one procedure over the edible item). Thompson and Iwata (2000) found other. If students are indifferent, then teachers’ that direct contingencies may result in more preferences certainly may be honored when rapid acquisition. By arranging a reinforcement instructional methods are selected. Future re- contingency that was not a direct product of the search may consider the systematic evaluation of behavior (i.e., completion of a motor sequence teachers’ preferences for these procedures. resulted in the teacher delivering a toy), we may have obscured this benefit of backward chaining. The methods of the current study would be REFERENCES applicable to conducting additional comparisons Ash, D. W., & Holding, D. H. (1990). Backward versus of forward and backward chaining with disparate forward chaining in the acquisition of a keyboard skill. Human Factors, 32, 139–146. tasks and more natural reinforcement contin- Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). gencies. It also would be interesting to compare Applied behavior analysis (2nd ed.). Upper Saddle the efficiency of and children’s preferences for River, NJ: Pearson Education. Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, chaining methods relative to total-task presenta- L. P., Owens, J. C., & Slevin, I. (1992). A comparison tion. During total-task presentation, all steps in of two approaches for identifying reinforcers for a multistep task are targeted from the onset of persons with severe and profound disabilities. Journal instruction, and prompts are provided as needed of Applied Behavior Analysis, 25, 491–498. Hagopian, L. P., Farrell, D. A., & Amari, A. (1996). to occasion each component response (e.g., Treating total liquid refusal with backward chaining Kayser, Billingsley, & Neel, 1986; McDonnell and fading. Journal of Applied Behavior Analysis, 29, & McFarland, 1988). 573–575. In practice, it is common to physically guide Hanley, G. P. (2010). Toward effective and preferred programming: A case for the objective measurement learners through the early steps of a task with of social validity with recipients of behavior-change backward chaining. For instance, in teaching an programs. Behavior Analysis in Practice, 3, 13–21.  FORWARD AND BACKWARD CHAINING 805 Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci, Schleien, S. J., Wehman, P., & Kiernan, J. (1981). S. A., & Maglieri, K. A. (1997). Evaluation of client Teaching leisure skills to severely handicapped adults: preference for function-based treatment packages. An age-appropriate darts game. Journal of Applied Journal of Applied Behavior Analysis, 30, 459–473. Behavior Analysis, 14, 513–519. Heal, N. A., & Hanley, G. P. (2007). Evaluating Schuster, J. W., Gast, D. L., Wolery, M., & Guiltinan, S. preschool children’s preferences for motivational (1988). The effectiveness of a constant time-delay systems during instruction. Journal of Applied Behav- procedure to teach chained responses to adolescents ior Analysis, 40, 249–261. with mental retardation. Journal of Applied Behavior Heal, N. A., Hanley, G. P., & Layer, S. A. (2009). An Analysis, 21, 169–178. evaluation of the relative efficacy of and children’s Smith, G. J. (1999). Teaching a long sequence of behavior preferences for teaching strategies that differ in using whole task training, forward chaining, and backward amount of teacher directedness. Journal of Applied chaining. Perceptual and Motor Skills, 89, 951–965. Behavior Analysis, 42, 123–143. Spooner, F., & Spooner, D. (1984). A review of chaining Hur, J., & Osborne, S. (1993). A comparison of forward techniques: Implications for future research and and backward chaining methods used in teaching practice. Education and Training of the Mentally corsage making skills to mentally retarded adults. British Retarded, 19, 114–124. Journal of Developmental Disabilities, 39, 108–117. Thompson, R. H., & Iwata, B. A. (2000). Response acquisition Jerome, J., Frantino, E. P., & Sturmey, P. (2007). The under direct and indirect contingencies of reinforcement. effects of errorless learning and backward chaining on Journal of Applied Behavior Analysis, 33, 1–11. the acquisition of Internet skills in adults with Tiger, J. H., Hanley, G. P., & Heal, N. (2006). The developmental disabilities. Journal of Applied Behavior effectiveness of and preschoolers’ preferences for Analysis, 40, 185–189. variations of multiple-schedule arrangements. Journal Kayser, J. E., Billingsley, F. F., & Neel, R. S. (1986). A of Applied Behavior Analysis, 39, 475–488. comparison of in-context and traditional instructional Walls, R. T., Zane, T., & Ellis, W. D. (1981). Forward approaches: Total task, single trial versus backward and backward chaining and whole task methods. chaining multiple trials. Journal of the Association for Behavior Modification, 5, 61–74. Persons with Severe Handicaps, 11, 28–38. Weiss, K. M. (1978). A comparison of forward and Mason, S. A., McGee, G. G., Farmer-Dougan, V., & backward procedures for the acquisition of response Risley, T. R. (1989). A practical strategy for ongoing chains in humans. Journal of the Experimental Analysis reinforcer assessment. Journal of Applied Behavior of Behavior, 29, 255–259. Analysis, 22, 171–179. Wilson, P. G., Reid, D. H., Phillips, J. F., & Burgio, L. D. McDonnell, J., & McFarland, S. (1988). A comparison of (1984). Normalization of institutional mealtimes for forward and concurrent chaining strategies in teaching profoundly retarded persons: Effects and noneffects of laundromat skills to students with severe handicaps. teaching family-style dining. Journal of Applied Research in Developmental Disabilities, 9, 177–194. Behavior Analysis, 17, 189–201. Powell, S., & Nelson, B. (1997). Effects of choosing Zane, T., Walls, R. T., & Thvedt, J. E. (1981). Prompting academic assignments on a student with attention and fading guidance procedures: Their effect on deficit hyperactivity disorder. Journal of Applied chaining and whole task teaching strategies. Education Behavior Analysis, 30, 181–183. and Training of the Mentally Retarded, 16, 125–134. Ringdahl, J. E., Vollmer, T. R., Marcus, B. A., & Roane, H. S. (1997). An analogue evaluation of environ- Received December 20, 2010 mental enrichment: The role of stimulus preference. Final acceptance April 12, 2011 Journal of Applied Behavior Analysis, 30, 203–216. Action Editor, David Wilder