The effect of error correction and goal setting with reinforcement on the acquisition of tacts of form and function of unknown nouns for individuals with autism.
A tact is a verbal response that is under the control of the physical or social environment and maintained by generalized conditioned reinforcement (Skinner, 1957); colloquially (and very loosely) speaking, labeling something in the environment may be considered a tact. For example a child who sees a cat enter the room and says, "kitty" or a person who hears a kettle boiling and says "It's hot" may very well be emitting tact responses, so long as those responses are maintained by generalized conditioned reinforcement.
A single object can have multiple tacts. For example, a speaker may emit tacts for a ball that include tacts of its form (e.g., "It's a ball!"), its function (e.g., "You play with it."), its feature (e.g., "It's round."), and its category (e.g., "It's a toy.") (Partington & Sundberg, 1988). For example, a screwdriver (form) can be used to turn screws (function) and be yellow (feature) and a tool (class). The ability to communicate effectively requires an understanding of form, content (including knowledge of object function), and use of language (Lahey, 1988). Individuals with autism may require direct instruction to learn the form, function and use associated with a single word.
Incorporating fluency building into the complement of instruction is emerging as a practice to promote skill development for individuals with autism (Ross & Greer, 2003; Kubina & Wolfe, 2005). Fluency instruction has been demonstrated to be effective in teaching a wider range of skills for a variety of learners, including learners with significant disabilities (White, 1986; Lindsley, 1992). Fluency has also been shown to be an effective teaching strategy in the development of language skills for individuals with autism. King, Moors & Fabrizio (2003) used fluency instruction with a young learner with autism to develop preposition skills. One of the unique aspects of this research was that through fluency instruction the child learned how to respond to prepositions in multiple ways, demonstrating flexibility in the use of language. Fluency is achieved when a participant can exhibit a behavior with accuracy and speed (Howell & Lorson-Howell, 1990). This study evaluated the effects of error correction and goal setting with reinforcement on the acquisition of novel form and function labeling.
PARTICIPANTS AND SETTING
Three participants (Jeff, Linda, and Leo) with autism and co-occurring mental retardation served as participants for this study. Each had an extensive mand (1 )repertoire and used speech to request desired items and activities. All three participants could answer simple personal questions (e.g., What is your name?, Who is your teacher?). Jeff, Linda, and Leo all initiated verbal conversational exchanges with others, although these conversations were rather simple and limited. The age, ethnicity, and intellectual ability for the participants are listed in Table 1. All three participants attended a self-contained day school for children with autism in an urban setting. The school served approximately 100 children with Pervasive Developmental Disorders. None of the participants had participated in fluency instruction prior to this study.
Table 1. Participant Characteristics Participant Age Ethnicity Intellectual Ability Jeff 7 Hispanic Mild to Moderate Mental Retardation Linda 18 African Mild to American Moderate Mental Retardation Leo 20 Hispanic Borderline to Mild Mental Retardation
The intervention occurred in either a therapy room typically used for one-to-one instruction by educators and therapists, or in the participants' normal classroom. During instruction, the participants sat facing an adult instructor. Data collectors sat behind the participants and out of the participants' visual field.
The generalization probes occurred in the participants' classrooms, in other locations on the school campus, or in the community where the items occurred naturally. For example, the probe for "gas can" occurred in the parking lot next to a car, and the generalization probes for "cookie sheet" and "peeler" occurred in the school's kitchen.
The materials used for this study included photographs of unknown items (see Table 2), reinforcement items, and electronic timers. To assess whether or not the participants could emit tacts for item names, photographs of the items were printed on 4" by 4" flashcards. Prior to baseline each participant was shown the picture of the items and asked the question, "What is it?" Of all the items that the participants tacted incorrectly, 10 incorrect items were randomly selected for each participant for use in the study. Teachers familiar with each participant were asked to name a reinforcing activity to be used with each participant. The activity that was chosen for use with Jeff was an edible or table-top game, while Linda and Leo's chosen activity was social interaction. A digital timer ensured accurate timings for each session.
Table 2. Unknown objects for each participant Participant Unknown Nouns Jeff Kettle Hanger Telescope Gas can Can opener Flash Blender Peeler Binoculars Stethoscope drive Linda Peeler Cookie Flash drive Gas can Nail sheet clippers Binoculars Q-tip Stethoscope Elevator Recycle bin Leo ATM Shoe Flash drive Push pin Car jack horn Escalator Bobby Grill Propane Peeler pin cleaner tank
RESEARCH DESIGN, INTEROBSERVER AGREEMENT, AND PROCEDURAL INTEGRITY
This study employed a multiple probe design with systematic replication of the intervention across participants (Cuvo, 1979; Horner & Baer, 1978; Murphey & Bryan, 1980). The multiple probe design is a variation of the multiple baseline design (Baer, Wolf, & Risley 1968). Probe trials were conducted intermittently across the interventions rather than on a daily basis (Tawney & Gast, 1984). An independent observer collected inter-observer agreement data on the dependent variable during 64% of sessions. An agreement was defined as both observers recording the correct occurrence or nonoccurrence of the dependent variable. A disagreement was defined as one observer scoring a response as having correctly occurred and the other observer scoring the response as having not correctly occurred. Interobserver agreement was calculated by dividing the number of agreements by the number of agreements plus disagreements and converting this ratio to a percentage. The total mean agreement for all participants across form and function was 98%, with a range of 75% to 100%.
Procedural integrity data were collected for both independent variables as well as the provision of assent. Data for participant assent, goal setting, and delivery of reinforcement were collected for all sessions by the researchers via a simple checklist contained on the data collection tool. Procedural integrity was 100% across all of these variables
The form tact was the first behavior addressed followed by function tact. Each subject participated in both form and function tact programs. A correct response required an intelligible vocal tact of either the form or the function within two seconds of presentation of the picture. In the function phase one function tact was trained but any correct function was accepted as correct (e.g., peeler was trained as "to peel food", but "to peel potatoes" was accepted; elevator was trained as "to go up and down floors", but "to go up and down in a building" was accepted). An incorrect response was a vocal response of anything other than the tact of the form or function or no response within two seconds. Each timing, the number of correct tacts, and the number of incorrect tacts were recorded. Data were graphed on a Standard Celeration Chart and an equal interval graph the same day they were recorded.
ERROR CORRECTION AND GOAL SETTING PROCEDURES
The independent variable used for this study consisted of two parts a) echoic error correction and b) goal setting with reinforcement. Upon completion of each timing post baseline, participants engaged in an error correction procedure. The error correction procedure consisted of the instructor showing the picture of each noun that had been missed during the previous session and verbally naming the form or function, depending on the experimental phase. The participant was then asked to verbally repeat the form or function of the item.
Prior to each instructional session, a goal was set for the participant. The goal for each session was to achieve the same or better rate per minute of the highest score from the previous session. The instructor said the goal aloud and also explained how the participant could earn their preferred item or activity. For example: "You got two correct last time. To get your (reinforcer) you need to get two or more correct." If a participant met or exceeded their goal, they immediately had access to their preferred item or activity. If a participant met their goal during the first or second timing they were asked if they would like to try again. If the participant did not achieve their goal by the third timing, they were thanked for trying and returned to class or went on to the next activity listed on their schedule.
Prior to baseline, each participant assented to participate by verbally responding an affirmation to the question "do you want to do some work?" The participant was shown approximately 30 photographs of unique objects. They were asked, "What is this?" Any verbal response, correct or incorrect, resulted in a move to the next photograph. If the participant did not answer, the researcher moved on to the next photograph after two seconds. Ten items that the participants had previously labeled incorrectly were randomly selected for intervention, giving each participant a set of 10 unknown items.
Baseline sessions consisted of one-minute timings. Participants were asked if they would like to do some work, and once they assented, they were accompanied to a room or went to a work area within their classroom. The 10 photographs previously missed were randomly ordered, and the participant was asked while looking at the each photograph, "What is it?' (form tact) or "What is this used for?" (function tact). If the participant provided any verbal response at the presentation of the photograph the next photograph was presented. If a participant did not respond, the photograph changed after two seconds of presentation. Participants never completed more than three timings during any session. Participants received minimal verbal praise (for example, a simple "thank you") for engaging in timings during baseline. Data were collected on a frequency recording sheet indicating the dependent variable of either the form or function of the 10 items.
Intervention consisted of goal setting with reinforcement and an echoic error correction procedure. Two phases were included in the research. The first phase addressed the tact for the form (object label); the second phase addressed the tact for the function of the object. Prior to the timing a goal was set for each participant of emitting either one correct response (if no corrects occurred during prior timings) or emitting greater than or equal to the number of correct responses from the previous session (if any corrects had occurred in the prior timing). The goal was stated prior to each timing, and the available reinforcer for achieving the goal was labeled for the participant. The photographs were randomly ordered for each timing and a one-minute timing was completed. Prior to each timing for the form (object label), participants were told they would be asked to say what the item was. Prior to each timing for the function of the object, they were told to say what the item was used for. At the presentation of the first card of each timing the researcher stated either, "What is it?" or "What is it used for?" If the pre-set goal was achieved, reinforcement was immediately delivered after the one-minute timing ended, and timed practice stopped for that session. If the participant did not achieve his or her goal during the first timing, then before the next timing echoic error correction was delivered.
The error correction procedure consisted of the researcher pairing the photograph of any missed items from the prior timing with a vocal label of either the item's form or function and requesting that the participant repeat the form or function. Error correction continued until participants could tact either the form or the function of each previously missed item with 100% accuracy. Fluency of the behavior was assessed individually and defined as when the behavior was being exhibited both with accuracy and speed (Howell & Lorson-Howell, 1990).
To assess the ability of the participants to tact targeted noun forms or functions outside of the training environment, a generalization probe occurred after each intervention phase. The intent of the probes was to ensure that the participants could tact in much more natural settings. For the generalization probes, participants were accompanied to the natural environment where the item would be used, and asked about either the form or function of the actual item in that environment. In the natural environment the instructor would ask, "What is this?" for the form probe, or, "What is this" and "What is it used for?" for the function probe. The generalization probe after the function phase was cumulative in that at the end of the form phase both form and function were assessed.
Social comparison data for the actual items that were used in the research were collected to allow for a comparison of the rates achieved by participants. Normative rates were used to assess the outcomes of the participants against a standard. Normative comparison rates of form and function tact emission were obtained by timing the performance of three typically developing adults (Kazdin, 1977). The mean rates were 28 correct and 9 incorrect for form and 20 correct and 3 incorrect for function. The normative rates were acquired after the intervention stopped.
Figure 1 shows the rate of performance for tacting noun object labels (forms). Linda and Leo demonstrated an immediate response to intervention as shown through their error rate decreasing and their correct rate increasing. During the first three sessions after the intervention started, Jeff was still not providing a verbal response, and instead made hand motions that approximated the functions of the actual items (e.g., putting hands up to eyes and looking through them for binoculars). Because of this, Jeff experienced a brief modeling phase that lasted three sessions to ensure that he understood the response topography (vocalization) expected. During this modeling phase, an adult confederate sat next to Jeff and completed a one-minute timing while Jeff watched. Jeff then completed a one-minute timing. Following the modeling phase, Jeff returned to the intervention conditions as previously implemented.
[FIGURE 1 OMITTED]
Jeff, Linda, and Leo all increased their rates of tact emission for noun forms. Linda achieved a rate comparable to the members of the social comparison group in five sessions for tacting noun forms, while Leo achieved a rate comparable to the members of the social comparison group in eight sessions. Jeff did not achieve a rate comparable to the members of the social comparison group for form by the time the experiment ended.
Figure 2 shows the participants' rates of correctly and incorrectly tacting the functions of nouns. For two sessions after intervention started, Linda continued tacting the object labels (forms) of the items rather than the items' functions. This resulted in her emitting no correct responses during those two timings. During the third timing she began to emit the correct function tacts, often paired with their forms. For example she would say "clippers to cut your nails" or "cookie sheet to bake with." Leo's first data points in baseline also represent his tacting the form of the item rather than its function. Upon the second presentation of the photographs and the mand, "What do you use these for?" he began to tact the items' functions. While Leo's baseline performance improved slightly across the course of the phase, the error correction procedure and goal setting resulted in both an increase in his rate of correct responding and a concomitant decrease in his rate of incorrect responding. Jeff also emitted form tacts rather than function tacts during baseline in the function phase. Upon introduction of the correction procedure and goal setting with reinforcement, he immediately started emitting the function tact for some of the objects. Linda achieved a rate comparable to the members of the social comparison group in three sessions for tacting the function of objects, and Leo achieved a rate comparable to the members of the social comparison group in five sessions. Jeff did not achieve a rate comparable to the members of the social comparison group for tacting function.
[FIGURE 2 OMITTED]
In the first generalization probe, Jeff tacted 8 of the 10 items and Linda and Leo tacted all 10 forms of the items in the natural setting. Linda and Leo were able to tact form and function of all 10 during the second generalization probe. Jeff was able to tact 8 of the 10 form items and 4 of the 10 function items during the second generalization probe.
The ability to tact the form, content/function, and use of language is necessary for effective communication (Lahey, 1988). This study demonstrated one method of instruction that proved effective in promoting these skills. All three participants in the current study increased their ability to tact both the form and function of previously unknown items and two participants' rates of responding mirrored those of typically developing adults. This skill gain occurred with minimal instruction and the performance improvement generalized to more natural contexts. Simple correction procedures and timings with improvement goals produced meaningful results.
Timings, although novel to the participants, became a regularly scheduled component within their instructional day. Each participant assented to completing the timed practices, and within a brief period of time Linda began requesting to engage in timed practice with a smile and positive tone in her voice. Leo also provided assent for each session. Initially Jeff would only tolerate one or two timings per session, often declining to participate for additional timings. As the research progressed, Jeff too tolerated multiple timings. Late in the research, on two occasions when only two occurred in one day, he asked, "Do one more?" This may be an indication of his desire to continue participating and an affinity for the timed practices.
Performance standards were difficult to define for the three participants, and indeed each achieved a different maximum rate of correct responding. Social comparison rates were obtained after the intervention finished and both Leos' and Linda's performance exceeded that of members of the comparison group for both form and function tacts. Jeff did not achieve rates similar to those of members of the comparison group.
Jeff, Leo, and Linda all provided the form tact during the baseline for the function phase, and Linda provided the form tact during the two initial sessions for intervention of function phase. Although the question, "What do you use this for?" was provided as an antecedent for each session, the question exerted no discriminative control initially for emission of a function tact. The inability to flexibly use language is core to the disability of autism (National Research Council, 2001), and this may have affected the participants' ability to acquire these particular skills. Given only limited instructional opportunities, participants with autism may learn one aspect of object referencing but not gain the additional linguistic aspects. For this reason, providing instruction across multiple aspects of object tacting seems critical.
Although Leo's performance during the baseline phase for function showed an upward trend, the level change and separation between correct and incorrect response rates demonstrated the effectiveness of the intervention. Prior to intervention, Leo correctly tacted the function of only a few items. The error and correct ratio was stable as his speed increased indicating that his accuracy was not improving, only his speed. It was only after the implementation of the intervention that he began to tact the correct function for all of the items
Linda's performance during the initial intervention sessions of the function phase showed a low rate of correct responding. During these sessions, she continued to tact the form of the noun as opposed to the noun's function. Two sessions after implementation of the error correction and goal setting procedure she began to emit the tact for both the form (object label) and the function of the object. (e.g., "A Q-tip is to clean your ears") This low level and subsequent paired responding implies that she was generalizing the tact relations reinforced during the form phase to the function phase. This resulted initially in incorrect responding, followed by correct but paired responding of both form and function. For this reason it is hypothesized that the primary stimulus controlling her responding may have been the presentation of the picture, rather than the discriminative stimulus specified during the function phase
There are several limitations to this study. Only two of the three subjects were able to achieve normative rates of performance. Further evaluation is needed to discern why the effect was not as strong with Jeff as it was for Linda and Leo. It should be noted that Jeff had more difficulty attending during the timings and often looked away from the photographs and attended to other things in the environment. Thus Jeff's attaining lower rates of correct responding than either Leo or Linda may have related to his limited repertoire of learning behaviors such as attending to tasks. Further research may help discover what pre-requisite skills may enhance participants' progress when using this instructional strategy.
The novelty of timings may also have had an effect. Prior to this study none of the subjects had participated in fluency timings. Jeff had difficulty understanding the expected topography of the response. In addition, given that intervention for tacting nouns' functions followed intervention for tacting nouns' forms for each subject, prior training may have affected acquisition of the function tacts. While certainly a confound in terms of experimental control, from a clinical perspective this may be seen as a positive considering that additional learning tasks can be presented successively, and this successive presentation may facilitate more rapid acquisition,.
This study used a two-part intervention to produce positive effects in the tacting repertoires of three children with significant developmental disability. This two-pronged intervention resulted in an effect for all participants. However, there is no way to discern if the positive gains made by the participants can be ascribed to both parts of the intervention or to one part only. Future research can help answer that question. Fortunately, these two strategies used together can work quite well in applied settings.
Baer, D., Wolf, M., & Risley, R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1, 91 - 97.
Cuvo, A.J. (1979). Multiple-baseline design in instructional research: Pitfalls of measurement and procedural advantages. American Journal of Mental Deficiency, 84, 219-228.
Horner, R.D., & Baer, D.M. (1978). Multiple probe technique: A variation of the multiple baseline. Journal of Applied Behavior Analysis, 11(1), 189-196.
Howell, K.W., & Lorson-Howell, K. A. (1990). What's the hurry: Fluency in the classroom. Teaching Exceptional Children, 22(3), 20-27.
Kazdin, A. E. (1977). Assessing the clinical or applied importance of behavior change through social validation. Behavior Modification, 1, 427-452.
King, A., Moors, A., & Fabrizio (2003). Concurrently teaching multiple verbal operants related to preposition use to a child with autism. Journal of Precision Teaching and Celeration, 19(1), 38 40.
Kubina, R.M. & Wolfe, P. (2005). Potential applications of behavioral fluency for learners with autism. Exceptionality, 13(1), 35-44.
Lahey, M. (1988). Language disorders and language development. New York: MacMillan.
Lindsley, O. R. (1992). Precision teaching: Discoveries and effects. Journal of Applied Behavior Analysis, 25, 51-57
Murphey, R.J., & Bryan, A.J. (1980). Multiple-baseline and multiple-probe designs:
Practical alternatives for special education assessment and evaluation. The Journal of Applied Behavior Analysis,4(3), 325-335.
Lord, C. & McGee, J. (Eds.). (2001) Educating children with autism. National Academy Press; Washington DC.
Partington, J.W. & Sundberg, M.L. (1988) The assessment of basic language and learning skills: Anassessment, curriculum guide, and tracking system for children with autism or other developmental disabilities. Danville, CA: Behavior Analysts, Inc.
Ross, D.E. & Greer, D. R. (2003). Generalized imitation and the mand: inducing first instances of speech in young children with autism. Research in Developmental Disabilities, 24(1), 58-74.
Skinner, B. F. (1957). Verbal behavior. New York: Appleton Century-Crofts.
Tawney, J.W., & Gast, D.L. (1984). Single subject research in special education. Columbus, Ohio: Charles E. Merrill Publishing Company.
White, O. R. (1986). Precision teaching -- precision learning. Exceptional Children, 52, 522-534
* AUTHOR CONTACT INFORMATION
PATRICIA WRIGHT, PHD, MPH
Easter Seals, Inc.
233 S Wacker, Suite 2400
Chicago IL 60606
ROBERT ALEXANDER, MA, BCBA
Chicago Behavior Development
4677 N. Virginia Ave, 1NChicago, IL 60625
NICHOLAS MILES, MS, BCBA
Butterfly Effects, LLC
750 East Sample Road, Building 2, Suite 102
Pompano Beach, FL 33065
Patricia Wright - Easter Seals, Inc Nicholas Miles - Chicago Behavior Development Robert Alexander - Butterfly Effects, LLC
|Printer friendly Cite/link Email Feedback|
|Author:||Wright, Patricia; Miles, Nicholas; Alexander, Robert|
|Publication:||The Journal of Speech-Language Pathology and Applied Behavior Analysis|
|Date:||Aug 1, 2012|
|Previous Article:||Effects of a speaker immersion procedure on the production of verbal operants.|
|Next Article:||Characteristics of naturalistic language intervention strategies.|