Printer Friendly

Teaching preschool children with disabilities to respond to the lures of strangers.

Ensuring children's safety is a concern and priority for both parents and educators. To ensure safety, children must engage in safe practices, recognize unsafe situations, and respond quickly to dangerous situations. Few studies teaching safety skills to preschool children have been reported in the literature. Exceptions include teaching children to make emergency telephone calls (Jones & Kazdin, 1980; Rosenbaum, Creeden, & Drabman, 1981) and react to strangers' invitations (Miltenberger & Thiesse-Duffy, 1988; Poche, Brouwer, & Swearingen, 1981; Poche, Yoder, & Miltenberger, 1988; Stevens & Long, 1982). Some attempt has been made to identify priority safety skills for the preschool curriculum (Collins, Wolery, & Gast, 1991). No studies were found that focused on teaching safety skills to preschoolers with disabilities.

Although adults attempt to ensure that children are safe, many children are abducted each year in the United States. According to the National Center for Missing and Exploited Children (1990), 24,655 children in the United States were reported as missing from June 13, 1984, through March 31, 1990. Preparing children, including those with disabilities, to respond to people who may abduct them is clearly an area of considerable importance. The increased emphasis on community-based instruction in special education (Sailor et al. 1986) may increase the probability that children will meet strangers. Teaching preschool children with disabilities to recognize and respond to such situations has been identified by teachers and parents as a socially valid safety skill (Collins et al., 1991 ).

Four studies have reported teaching preschool children without disabilities to respond to the lures from strangers. Poche et al. (1981 ) successfully used role playing to teach three preschool children without disabilities to respond to multiple exemplars of lures delivered by a variety of strangers. Daily training and probe trials occurred at various sites on the school grounds, and generalization probe trials were conducted in community sites. In a later study, Poche et al. (1988) found a combination of role playing and exposure to an interactive videotape to be more effective in teaching lure responses to children without disabilities than the single use of an interactive videotape or film. Miltenberger and Thiesse-Duffy (1988) found role playing to be more effective than a commercially available safety program (that relied on pictures, dolls, and verbal discussion in a home setting) in teaching 24 preschool children without disabilities to respond appropriately to strangers. Stevens and Long (1982) included a generalized response to the lures of strangers in a social skills training program for four preschool children without disabilities; this program was conducted in a classroom setting and used role playing. Though each of these studies described a role-playing simulation during training, none reported a systematic delivery of prompts.

A safe response to strangers' lures should meet several criteria:

1. Allow children to escape the dangerous situation.

2. Be easy to perform.

3. Be highly generalized across settings.

4. Be maintained over time.

In a telephone interview, The Crimes Against Children Unit of the Police Department in Lexington, Kentucky, stated that they receive daily reports regarding strangers who attempt to lure children or adolescents and that those who successfully escape leave the situation immediately. A training procedure for teaching an effective response should result in rapid acquisition, ensure safety during training, be capable of use with simulations, and employ multiple exemplars (strangers, lures, locations).

The constant-time-delay procedure has been an effective, reliable, and near errorless procedure for teaching many skills to preschool students with disabilities (Alig-Cybriwsky, Wolery, & Gast, 1990; Doyle, Wolery, Ault, & Gast, 1989; Doyle, Wolery, Gast, Ault, & Wiley, 1990). In addition, this procedure has been shown to be more efficient than the system to least prompts (Doyle et al., 1990). Tlte constant-time-delay procedure was chosen for this study because it results in rapid acquisition, minimizes errors during acquisition, and can be used in role play during simulations.

The use of multiple exemplars during training has been shown to promote generalized responding of new skills (Horner, Eberhard, & Sheehan, 1986; Stokes & Baer, 1977), especially if the exemplars sample the range of possible stimuli in the generalization context (Sprague & Homer, 1984). For safety skills, the use of multiple exemplars of intersections or streets has been effective in teaching people with mental retardation to generalize pedestrian skills from training to novel sites (Horner, Jones, & Williams, 1985; Marchetti, McCartney, Drain, Hooper, & Dix, 1983). Because generalization is critical, the use of multiple exemplars of lure types, strangers, and settings during training appears appropriate. Realistic simulations in the classroom (Stevens & Long, 1982) or in viva (Poche et al., 1981) can provide a controlled training setting to ensure the safety of children until the correct response is acquired.

The research questions addressed in this study are:

1. Will simulated training, which combines the constant-time-delay procedure with multiple exemplars of lure types and strangers, result in an increase in the number of correct responses by preschool children with disabilities during in vivo probes?

2. Will this procedure also result in a response that generalizes to novel strangers, lures, and sites?



Four children (age range of 3 years, 5 months to 5 years, 0 months) enrolled in an integrated pre-school served as subjects. Each child was described as having "developmental delays" following administration of the Battelle Developmental Inventory (Newborg, Stock, Wnek, Guidubaldi, & Svinicki, 1984) and Developmental Profile H (Alpern, Boll, & Shearer, 1980). In addition, they had been identified as having speech and language disabilities and were receiving speech therapy.

James was a 5-year, 0-month-old male with a hearing impairment, for which he wore hearing aids. During school assessment, James performed below age level on social-emotional skills (50 months) and communication skills (50 months). Both of these areas had been targeted for intervention. Alex was a 3-year, 5-month-old male who performed below age level on school assessments in the areas of cognition (28-29 months), language and communication skills (32 months), and motor skills (23-28 months). Coping skills were among those skills targeted for intervention. The third student, Mark, was a 3-year, 8-month-old male who performed below age level on school assessment in the areas of cognition (31-34 months) and language/communication skills (30 months). The ability to form complete sentences and to comply with directions were among those skills targeted for intervention. The fourth child, Craig, was a 3-year, 9-month-old male who performed below age level on school assessment in the areas of cognition (38-39 months), language/communication skills (34 months), self-help skills (41-44 months), and motor skills (40-44 months). Speech and language skills were among those skills targeted for intervention.

Responding to the lures of strangers was considered an important intervention target by the preschool staff. Prescreening was conducted with a group of four males and one female chosen by the staff as most apt to respond to the lure of a stranger. The four males selected for the study following prescreening exhibited the following traits:

* They willingly left with a stranger (the second author) for screening, although no introduction had been made or permission had been given by preschool personnel.

* The verbal responses of all children when interviewed were difficult to understand, and sentences were limited in length to five words or less.

* Three of the children responded to the question, "What would you do if a stranger talked to you?" with "I don't know"; one child said "Talk to them."

* Each child sat alone with no sign of panic when left alone for 1 min during the screening process.

Prerequisite skills observed during screening included the ability to imitate verbal and motoric models, remain on task for 5 min, walk independently, wait up to 5 s for a prompt before answering, and follow two-step directions.


All probe and training sessions were conducted on an individual basis. Daily and weekly probe sessions were conducted in multiple community settings. These included various sites along the street on which the preschool was located (a small business district), various areas in an adjoining parking lot and alley, a street designated as a shopping mall, a local YMCA, the entrance to the preschool, and the park grounds of a nearby historical landmark. No setting was used twice during the same week, and novel settings were used for generalization probes. Probe sessions were conducted in the community regardless of weather conditions, which varied in temperature, and ranged from mild and sunny to windy with rain or snow. Safety precautions employed during probes included an inconspicuous person posted between the child and the street to stop a child if he were to approach the street and to intervene if a real stranger approached the child. Neither of these incidents occurred during the investigation.

Simulated classroom training sessions were conducted in various areas of the preschool, varying the site on a daily basis. Areas in which training took place included different points in the classroom (e.g., housekeeping area, kitchen area), the therapy room, and a hall used for hanging coats. Other students and school personnel were present in adjacent areas during these sessions, but were busy with instruction and not distracted by the training sessions. Each in vivo training session was conducted immediately following the daily probe sessions in the same site in which the probe session was conducted.


Multiple "strangers" (N= 26) were used in both probe and training sessions. Strangers for probe sessions were unknown to the children in the study and were recruited from graduate students, research associates, and faculty in the Department of Special Education at the University of Kentucky; from personnel from the Office of Education for Exceptional Children in the Kentucky State Department of Education; and from the families and friends of the investigators. Although all strangers were white, they varied in age, gender, and physical characteristics (height, weight, and hair color). In addition, strangers varied in their manner of dress. Although strangers in daily probes were used repeatedly during training, novel strangers were used in generalization probes. No stranger was used for probes more than once during the same week.

During classroom simulation training, five volunteer strangers were recruited from the local high school and university special education teacher preparation program. Although these "strangers" were known to the students (i.e., they had seen these people before), they were still inappropriate people with whom the target children should leave. These strangers varied according to availability. The stranger from the probe session was used again during in vivo training sessions.

Multiple lures were presented by the strangers during probe and training sessions. As in the study by Poche et al. (1981), three basic types of lures were identified: general ("Would you like to go for a walk?"); authority ("Your mother said for you to go with me."); and incentive ("Would you like to go get some candy?"). Predesignated lures for each type were used during training and daily probes; novel variations were used during generalization probes.

Trainers The investigators used role playing and verbal examples to train the teacher and classroom aide to implement the instructional procedures. They were taught both simulated classroom procedures and in vivo procedures. They demonstrated 100% correct implementation for the procedures before the study began.

Data Collection

During probe sessions, data were collected by the strangers, who received instruction in data-collection procedures from the trainer or investigator before their participation in the study. A pocket cassette tape recorder was carried by the stranger to validate the delivery of the lure and the verbal response of the child. In addition, an observer collected interobserver and procedural reliability data a minimum of once per week per condition for the child in training and intermittently for those children receiving weekly probe sessions (i.e., probes conducted before training began).

Using an adaptation of the system used by Poche et al. (1981), "strangers" scored the quality of responses to a lure during probe sessions according to the number of components present, with a perfect response receiving 5 points. One point was given for each of the following components following the delivery of the lure: (a) correct verbal response ("No"); (b) correct verbal response emitted within 3 s of the lure (latency measure); (c) correct motor response (turning away from the stranger); (d) correct motor response within 3 s of the lure (latency measure); and (e) walking or running in the direction of the teacher and a minimum of 5 ft from the stranger. A response interval of 3 s was allotted for the verbal response and 3 additional s for the motor response. That is, the child had 6 s from the time the stranger delivered the lure to say "No" and move 5 ft (out of reach) away from the stranger. The child was given credit for correctly responding if the motor response occurred simultaneously with or preceded the verbal response. The "strangers" also recorded incidental information (e.g., gestures, utterances) regarding responses.

During constant-time-delay training trials, the trainer recorded data on whether the child responded correctly before or after delivery of the prompt, incorrectly before or after the prompt, or if the child failed to respond after the prompt for both the verbal and motor responses.


Probes. Probe trials were conducted throughout the investigation. These included (a) weekly probe trials before intervention (with 3 consecutive days of probe trials immediately before intervention); (b) daily probe trials during intervention; (c) daily generalization probe trials following criterion performance on daily probe trials; and (d) weekly maintenance probe trials following criterion performance on generalization probe trials. In addition, there were two final follow-up maintenance probe trials spaced at 2-week intervals after criterion was met by the final child.

All probe trials, except generalization probe trials, were conducted in the following manner. A probe session (consisting of a single trial) was conducted daily (Monday through Friday) between 12:30 and 1:00 p.m. following the arrival of children for the afternoon session at the pre-school. During a probe trial, the trainer took the child on a walk with or without other students. When the trainer came to the predesignated probe site, the student was shown something of interest (e.g., toy in store window) and was told to "Wait fight here." At this time, the trainer turned her back to the student and began to walk away. When the trainer was a minimum of 5 ft away, the stranger approached and delivered a predesignated lure. If the child responded with the correct motor response (i.e., he turned and moved toward the trainer) or if the child responded incorrectly or failed to respond, the trainer returned to get the child. This was done to eliminate the possibility of the child's learning to go with strangers. During generalization probe sessions, the trainer was sometimes called away by another individual so that the trainer did not give the command to wait. Because no reinforcement was delivered during probe sessions, the trainer was instructed to respond in a neutral manner to any comments by the child regarding the stranger or lure (e.g., if the child said, "That man asked me to go with him," the trainer would respond with "Oh? He did?"). Probe sessions continued until the child reached criterion, responding correctly with the motor response for 4 out of 4 days and with the verbal response for 3 out of 4 days. The difference in criterion was because the motor response (walking away within 3 s of the lure and being at least 5 ft from stranger) was viewed as the critical safety behavior.

When criterion was met in the daily probe condition, novel strangers, lures, and sites were presented in generalization sessions until criterion was again met. Table 1 shows the combination of novel exemplars used for generalization probe sessions.

Probe sessions to assess maintenance were conducted intermittently for children who had met criterion in the generalization condition until all children met criterion. At that point, follow-up probe trials were conducted across children at intervals of 2 and 4 weeks.

Training. Training sessions were conducted during the afternoon between 1:00 and 4:00 p.m. following the daily probe trial. The time interval between the probe trial and classroom simulation training sessions varied according to the length of the community activity in which the probe trial was delivered. For example, if the probe trial was conducted on the way to swimming class, classroom training was not conducted until the children returned from swimming. Each training session consisted of three massed trials in which each of the three lure types was presented in an intermix sequence, varying the order of lure presentation daily. A constant-time-delay procedure was used to teach the target responses. During the first session, the trainer explained the importance of saying "No" to strangers (without warning against danger or scaring children). This was followed by a demonstration of the correct response to lures that was role played by the teacher, instructional aide, and classroom volunteer or practicum student. The child was then asked to role play the correct response with the trainer within the context of the constant-time-delay procedure.

Following the delivery of each lure, the controlling prompts (verbal model for the verbal response and verbal instruction plus physical guidance for the motor response) were delivered at a O-s-delay interval. On all subsequent days, the trainer announced that it was time to practice the response to strangers and delivered the controlling prompts at a 3-s-delay interval after presentation of the lure. The controlling prompt delivered by the trainer for the verbal component was the verbal model, "No, I have to go find my teacher"; however, the verbal response was scored correct if children only said, "No." Reinforcement (descriptive praise) was given on a continuous schedule (following the verbal response and following the motor response) until the child met criterion (100% correct responding on both verbal and motor components) for 1 day. Reinforcement was then given following the correct combination of the verbal and motor response (FR2) for 2 consecutive days. At this point, reinforcement was withheld to the end of the session (FR6), that is, correct responses to all three lures presented in the session. this criterion remained in effect until criterion was met during daily probe sessions.

If criterion was not met during daily probe sessions 6 days following correct responding in classroom simulation training under the FR6 reinforcement schedule, training sessions were transferred to the probe site immediately following the daily probe session. The same setting and stranger used during the probe session was again used during the invivo training session. Otherwise, in vivo training sessions were conducted identically to the in-classroom simulation sessions and continued until the probe criterion was met.

Experimental Design

A multiple-probe design across subjects was used to evaluate experimental control (Murphy & Bryan, 1980; Tawney & Gast, 1984). Immediately before training, each child received a minimum of 3 consecutive days of probe trials. Because many strangers, lures, and sites were used, these probe trials were equivalent to posttraining generalization probe trials in which novel strangers were used. Before training and following training, children received weekly probe trials. Training in the classroom setting was initiated only after stable performance in the daily probe trials occurred. The multiple-probe design was especially well suited to the purpose of this study because weekly probe trials before training prevented children from being desensitized to the approach of strangers. Yet adequate preinstruction data were collected to identify and control for threats to internal validity (e.g., maturation and history). Further, the weekly probe trials following training provided a measure of response maintenance.


Both interobserver and procedural reliability data were collected across conditions a minimum of once per week. Interobserver reliability scores were obtained by dividing the total number of agreements by the total number of agreements plus disagreements and multiplying by 100. Procedural reliability was calculated by dividing the number of observed teacher behaviors by the number of planned behaviors and multiplying by 100. Both interobserver and procedural reliability data were collected on 52% of all probe sessions and on 39% of all training sessions. During probe sessions, both interobserver and procedural reliability data were 100%. During training, interobserver reliability was 100% and procedural reliability was 99.9% (range of 96%100%). The only procedural error during training was a single failure to reinforce a correct response following the prompt.


Graphed data are shown in Figure 1. The ordinate on the left of the graph shows the number of response components demonstrated by each child during probe trials and corresponds to the closed circles. The ordinate on the right of the graph shows the percentage of correct responses made by each child during training sessions and corresponds to the closed triangles. The abscissa shows the number of probe and training sessions conducted (one single trial probe session and one training session per day).

The study progressed through five phases: (a) baseline data collected during the daily and weekly community probe trials; (b) daily constant time delay training in a classroom simulation with concurrent measurement in the daily, community probe trials; (c) daily in vivo constant-time-delay training conducted after daily, community probe trials; (d) daily generalization probe trials; and (e) weekly maintenance probe trials. Figure 1 shows all children's data for these five phases. As shown in the figure (left ordinate, closed circles), none of the children performed the five components of the response during the daily and weekly, community probe trials before intervention. Two noncritical components (i.e., saying "no" within 3 s with no motor movement) were displayed by James on one probe trial and by Mark on six probe trials.

Implementation of the constant-time-delay training in the classroom resulted in two findings. First, all children rapidly acquired the five-component target response during classroom simulation instruction and maintained high levels of correct responding throughout the condition. James required three sessions to reach 100% unprompted correct performance; Alex required five sessions; Mark required three sessions; and Craig required one session. Second, partial increases in the number of correct responses in the daily, community probe sessions occurred for three of the four children. James performed the verbal components of the response but did not display the motor components of turning away and moving toward the teacher, the two critical safety responses. Alex showed initial increases of the five-component response, but performance was variable and did not maintain. Mark's pattern of performance was similar; an initial increase was noted but did not maintain. Craig's performance in daily probe sessions was not changed as a result of classroom training.

Because of the lack of complete performance and lack of maintenance during the daily, community probe trials, the constant-time-delay training was implemented in community probe sites (in vivo). These training sessions followed each daily probe trial. Children's performance on training trials during in vivo sessions was characterized by 100% unprompted correct performance; however, some decrement in responding occurred for Alex, Mark, and Craig during initial sessions. The in vivo training also resulted in increases during subsequent daily, community probe trials.

For James, two in vivo training sessions resulted in the complete five-component response during daily probe trials. However, when generalization trials were implemented, only partial generalization occurred. One additional in vivo session was implemented, which resulted in complete generalization that maintained across 3.5 months. For Alex, in vivo training resulted in performance of the five-component response on the daily probe trials, but that performance was variable and negatively affected by absences. After six in vivo sessions, Alex consistently performed the motor responses, but not the verbal components of the response. In the 15th in vivo session, an intensity prompt that required him to shout the verbal components during training sessions was implemented. This resulted in only two instances of the verbal behavior in daily probe trials. During the 24th in vivo training session, the number of trials was doubled and resulted in two more instances of the complete five-component response on the daily probe trials. During generalization trials, Alex performed the motor responses, but not the verbal responses. The motor responses maintained at 2- and 4-week follow-up probes. Though the verbal response is desirable, the motor responses are critical for escaping from strangers. For Mark, the in vivo training resulted in immediate increases in the complete response in the daily probe trials; however, his responses dropped after two separate absences. Continued use of in vivo training resulted in complete correct performance of the response in daily probe trials. Mark was withdrawn from the preschool program by his parents (because of school transportation problems) before generalization probe trials could be implemented. For Craig, implementation of the in vivo training resulted in complete performance of the five-component response during daily probe trials. This performance was maintained for 2 months in follow-up probes.

To measure the efficiency of the instruction, the number of minutes of training and children's errors were analyzed. All training sessions for all students totaled 2 hr and 54 min with a mean range of 1 to 4 min per session and a range of 17 to 81 min for children to reach criterion. The training procedures were errorless for two students, James and Mark, and resulted in three errors for Alex (2%) and two errors for Craig (5%).

Following the study, an attempt was made to acquire a measure of social validity. Questions to assess social validity were asked of the preschool staff and parents. The teacher, the two instructional aides, and two parents (the mothers of James and Craig) responded. The specific questions asked were:

* Has the child had an opportunity to respond to a stranger since he has been a participant in the study? If so, what happened?

* Is the child more wary of strangers who come around since he has participated in the study?

* Do you consider the response to strangers a valuable skill?

* Do you think the procedure used was effective? Worth the time?

* Do you think the child has learned this skill? If so, do you think the child could use it in a real situation? Remember it over time?

Without exception, the study was regarded as socially valid, particularly in regard to James and Craig, who were reported to ask permission before going with unknown individuals. In addition, as previously noted, the local police department was contacted to verify the importance of a response to lures; this resulted in validation of the probability of an encounter with a stranger (daily calls received by the department reporting lure attempts by strangers to children across age groups) and identification of the motor component as critical to the target response.


The purpose of this study was to evaluate the effectiveness of simulated training, which combined a constant-time-delay procedure with multiple exemplars of lure types and strangers, to teach a generalized response to the lures of strangers by preschool children with disabilities. Data reveal that this training was effective in teaching the target responses in classroom simulation, in that all children quickly met criterion in responding to multiple lures with few errors. However, this responding did not generalize to community probe sites in which similar lures were presented by multiple strangers. When training was implemented in vivo, three of the four children met criterion; and the remaining child began to consistently respond with the critical motor response of moving away from the stranger. Generalization of responding was then ascertained through probe trials conducted in novel sites with novel strangers and novel lures.

Whereas the data for James, Mark, and Craig show a clear transfer to criterion responding in community probe sites following the implementation of in vivo training, Mark's data appear to have been affected by absences in which responding showed a temporary decrease following each absence. A review trial was required before James reached criterion in generalization probe sessions, whereas Craig displayed immediate responding at criterion level on generalization probe trials. Mark was removed from school by his parents before generalized responding could be assessed.

Although the data for Alex eventually stabilized during in vivo training, Alex did not transfer a consistent verbal response from training to probe sessions. Two supplements to the procedure were implemented in succession for Alex in an attempt to facilitate his making the verbal response ("No"). First, Alex was prompted to shout "No !" during training, which was in contrast to the whisper in which he most often said it. Performance during this time improved, but he did not meet criterion. Second, the number of training trials per session for Alex was increased from 3 to 6. This resulted in two occurrences of criterion responding ("No" plus walking away). Because the motor component was always present during Alex's responding on probe trials, it was decided that he had acquired a functional response (as validated by the police department) and the decision was made to discontinue training and implement generalization probes.

The data support the following statements:

1. The constant-time-delay procedure, when used with multiple exemplars in classroom simulations, is effective in teaching an escape response to a stranger's lures in that setting.

2. Generalization of responding to community sites is facilitated by training in multiple community sites.

3. The use of multiple strangers and lures will result in generalized responding to novel exemplars of strangers and lures. This implies that, when planning the training of skills that will be used in community settings, the trainer should not assume that generalization will take place because responding is at criterion level within the training site; testing of the response in the sites in which the skill has a high probability of occurrence is critical. In addition, an attempt should be made to build in sufficient exemplars of stimuli to result in generalized responding, again testing with novel stimuli to ensure that generalization has occurred.

Several implications exist for further research. First, the training procedure used in this study should be replicated in vivo, omitting the classroom training to determine whether training in simulation before training in vivo is necessary. Second, the procedure should be implemented across other age groups and populations who might be susceptible to the lures of strangers (e.g., adolescents or adults with mental disabilities). And third, the constant-time-delay procedure should be assessed in teaching related safety skills (e.g., response to inappropriate physical approaches, response to offers of drugs, and response to lures delivered on the telephone or when answering a knock on the door).

The acquisition of safety skills increases the probability that children can function safely in less restrictive environments. This, in turn, increases the likelihood of educational agencies adapting community-based instruction and identifying work and recreation sites that are integrated. Safety-skill instruction is critical to responsible community-based instruction. Therefore, if safety skills are to be a priority for instruction, it is imperative that educators use effective, efficient procedures that will result in generalized responding.


Alig-Cybriwsky, C., Wolery, M., & Gast, D.L. (1990). Use of a constant time delay procedure in teaching preschoolers in a group format. Journal of Early Intervention, 4, 99-116.

Alpcto, G.D., Boll, T. 3., & Shearer, M.S. (1980). Manual Developmental Profile H. Aspen, CO: Psychological Development Publications.

Collins, B.C. Wolery, M., & Gast, D.L. (1991). A survey of safety concerns for students with special needs. Education and Training in Mental Retardation, 26, 305-318.

Doyle, P.M., Wolery, M., Ault, M.J., & Gast, D.L. (1989). Establishing conditional discriminations: Concurrent versus isolation-intermix instruction. Research in Developmental Disabilities, 10, 349-362.

Doyle, P.M., Wolery, M., Gast, D.L., Ault, M.J., & Wiley, K. (1990). Comparison of constant time delay and the system of least prompts in teaching preschoolers with developmental delays. Research in Developmental Disabilities, 11, 1-22.

Homer, R.H. Eberhard, J.M., & Shechart, M.R. (1986). Teaching generalized table bussing: The importance of negative teaching examples. Behavior Modification, 10, 457-471.

Homer, R.H., Jones, D.N., & Williams, J.A. (1985). A functional approach to teaching generalized street crossing. Journal of the Association for Persons with Severe Handicaps, 10, 71-78.

Jones, T.T., & Kazdin, A.E. (1980). Teaching children how and when to make emergency telephone calls. Behavior Therapy, 11, 509-521.

Marchetti, A.G., McCartney, J.R., Drain, S., Hooper, M., & Dix, J. (1983). Pedestrian skills training for mentally retarded adults: Comparison of training in two settings. Mental Retardation, 21, 107-110.

Miltenberger, R.G., & Thiesse-Duffy, E. (1988). Evaluation of home based programs for teaching personal satety skills to children. Journal of Applied Behavior Analysis, 21, 81-87.

Murphy, R.J., & Bryan, A.J. (1980). Multiple-baseline and multiple-probe designs: Practical alternatives for special education assessment and evaluation. The Journal of Special Education, 4, 325-335.

National Center for Missing and Exploited Children. (1990, April). Arlington, VA.

Newborg, J., Stock, J.P., Wnek, L., Guidubaldi, J., & Svinicki, J. (1984). Battelle Developmental Inventory. Allen, TX: DLM Teaching Resources.

Poche, C., Brouwer, R. & Swearingen, M. (1981). Teaching self-protection to young children. Journal of Applied Behavior Analysis, 14, 169-171.

Poche, C., Yoder, P., & Miltenberger, R. (1988). Teaching self-protection to children using television techniques. Journal of Applied Behavior Analysis, 21, 253-261.

Rosenbaum, M.S., Creedon, D.L., & Drabman, R.S. (1981). Training preschool children to identify emergency situations and make emergency phone calls. Behavior Therapy, 12, 425-435.

Sailor, W., Halvorsen, A. Anderson, J., Goetz, L., Gee, K., Doering, K., & Hunt, P. (1986). Community intensive instruction. In R. Homer, L. Meyer, & H.D. Fredericks (Eds.), Educational learners with severe handicaps: Exemplary learning strategies (pp. 251-288). Baltimore: Paul H. Brookes.

Sprague, J.R., & Homer, R.H. (1984). The effects of single instance, multiple instance, and general case training on generalized vending machine use by moderately and severely handicapped students. Journal of Applied Behavior Analysis, 17, 273-278.

Stevens, M.L., & Long, L. (1982). Individual and group training of social and health related behaviors in young children. Education and Treatment of Children, 5, 233-248.

Stokes, T.F., & Baer, D.M. (1977). An implicit technology of instruction. Journal of Applied Behavior Analysis, 10. 349-367.

Tawney, J.W., & Gast, D.L. (1984). Single subject research in special education. Columbus, OH: Charles E. Merrill.


DAVID L. GAST (CEC #1143) is Principal Investigator of the SAFE Project and Professor in the Department of Special Education at the University of Georgia, Athens. BELVA C. COLLINS (CEC KY Federation) is an Assistant Professor in the Department of Special Education at the University of Kentucky, Lexington. MARK WOLERY (CEC #104) is a Senior Research Scientist at the Allegheny Singer Research Institute, Pittsburgh, Pennsylvania. REBECCA JONES (CEC #5) is a preschool teacher in the Anderson County Public School District, Anderson County, Kentucky.

This study was supported by the U.S. Department of Education, Grant #H023C90128. The opinions expressed do not necessarily reflect the policy of the U.S. Department of Education, and no official endorsement of the U.S. Department of Education should be inferred. We wish to thank the following people who (along with the authors) agreed to act as strangers during this study: Melinda Ault, Debra Bauder, Tom Branscum, Rita Byrd, Maggie Chiara, Marilyn Coffey, Charles Cole, Courtney Collins, Ted Collins, Toby Collins, Patricia Doyle, Lucy Fleming, Peggy Harrell, Ariane Holcombe, Millie Hughes, Carol Raglan, John Schuster, Sara Sherburne, Renae Sizemore, Danielle Smith, Jeanne Smith, Beverly Stephens, Morgan Stephens, Cheryl Weinzierl, Vince Winterling, student classroom volunteers from Frankfort High School, and practicum students and staff at PUSH Infant and Preschool.

Address all correspondence to David L. Gast, Principal Investigator, Project SAFE, Professor, Department of Special Education, 521 Aderhold Hall, University of Georgia, Athens, GA 30602.

Manuscript received January 1991; revision accepted July 1991.
COPYRIGHT 1993 Council for Exceptional Children
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Gast, David L.; Collins, Belva C.; Wolery, Mark; Jones, Rebecca
Publication:Exceptional Children
Date:Feb 1, 1993
Previous Article:20/20 Analysis: taking a close look at the margins.
Next Article:Two approaches to reading instruction with children with disabilities: does program design make a difference?

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters