Teleconsultation in School Settings: Linking Classroom Teachers and Behavior Analysts Through Web-Based Technology

Abstract

This paper describes a technically driven, collaborative approach to assessing the function of problem behavior using web-based technology. A case example is provided to illustrate the process used in this pilot project. A school team conducted a functional analysis with a child who demonstrated challenging behaviors in a preschool setting. Behavior analysts at a university setting provided the school team with initial workshop trainings, on-site visits, e-mail and phone communication, as well as live web-based feedback on functional analysis sessions. The school personnel implemented the functional analysis with high fidelity and scored the data reliably. Outcomes of the project suggest that there is great potential for collaboration via the use of web-based technologies for ongoing assessment and development of effective interventions. However, an empirical evaluation of this model should be conducted before wide-scale adoption is recommended.

The use of functional behavior assessment procedures has become a more common practice in public schools as a result of both legislation and best practice guidelines. Functional behavior assessment can be classified into three types (indirect assessment, descriptive assessment, and functional analysis) that can be organized on a continuum from least to most precise and easiest to most difficult to perform (Neef & Peterson, 2007). The most precise form of functional behavior assessment is a functional analysis, which involves direct observation of problem behavior and systematic manipulation of its antecedents and consequences. The functional analysis is often considered the “gold standard” of functional behavior assessment because it is the only form of assessment that verifies the function of problem behavior (Bijou, Peterson, & Ault, 1968; Lennox & Miltenberger, 1989; Mace, 1994). However, the time, effort, and professional expertise required to conduct and interpret functional analyses are often cited as problems with the approach, limiting the widespread use of functional analysis by school personnel and other practitioners (cf., Spreat & Connelly, 1996).

Because a fair amount of expertise is required to conduct functional analyses, research has examined various ways to train individuals to conduct them (e.g., Baker, Hanley, & Matthews, 2006; Iwata et al., 2000; Moore, Edwards, Sterling-Turner, Riley, DuBard, & McGoerge, 2002; Moore & Fisher, 2007; Najdowski, Wallace, Doney, & Ghezzi, 2003; Wallace, Doney, Mintz-Resudek, & Tarbox, 2004). These studies suggest that a variety of individuals with diverse backgrounds can conduct functional analyses and that a range of training formats can be effective. Less research has been conducted on teaching the analytical skills necessary to make decisions during and after a functional analysis. For example, a fairly specialized skill set is required to summarize and interpret data from functional analyses and to design functional analysis conditions to meet the idiosyncratic needs of individuals with problem behavior. Therefore, it might be wise to consider how to forge collaborative relationships between behavior analysts and school personnel. Such relationships would eliminate the need for large numbers of school personnel to learn the complex set of skills essential for conducting and interpreting functional analyses. However, sustaining ongoing collaborative projects may present a challenge, particularly for schools located in rural settings. Mileage reimbursement costs and the time required to travel to rural sites may prohibit ongoing consultation and collaboration between behavior analysts and school personnel. Alternatives to face-to-face interactions are needed if such collaborations are to be feasible.

Some researchers and practitioners—particularly those within the medical field—have begun to examine alternatives to conventional (i.e., face-to-face) means for service delivery. Barretto, Wacker, Harding, Lee, and Berg (2006) described one model, specific to functional analysis. The experimenters conducted brief functional analyses of problem behavior for 75 individuals with disabilities during a 3-year period. Specifically, the authors coached parents to conduct functional analyses through a fiber-optic telecommunication system referred to as the Iowa Communication Network (ICN) that connected most of the hospitals and high schools located in Iowa. This allowed families to forego making long trips to the university for the evaluation.

Idaho is an example of a mostly rural state that could benefit from a service delivery model like that in Iowa. To illustrate, Idaho consists of 83,557 square miles (as big as Massachusetts, New York, New Jersey, and New Hampshire combined), yet the population is only 1.3 million people (U.S. Census Bureau, 2000). Forty-eight percent of the counties in Idaho are designated “rural,” and 35% of counties are designated “frontier” (meaning there are fewer than six people per square mile, and there is no population center nearby). The largest city in Idaho is Boise, which has a population of 185,787. To further complicate matters, Idaho has 13 national forests and several mountain ranges, making travel across the state somewhat daunting, especially in the winter months. Although Idahoans would likely benefit from the telemedicine model described by Barretto et al. (2006), the necessary infrastructure is unavailable. At a cost of $2.3 million for development of the ICN (Iowa Communication Network Facts), establishing a similar statewide fiber-optic cable network in the state of Idaho is not foreseeable in the near future. It is likely that other rural states face similar financial constraints; therefore, alternative teleconsultation methods need to be explored.

Machalicek et al. (2009) described the use of less-expensive and more easily accessible videoconferencing software and equipment (i.e., laptop computers, Internet connection, and web cameras) to conduct functional analyses for 2 participants who demonstrated challenging behaviors. Although this approach seems promising, the analyses were conducted by graduate students in special education rather than by school personnel or practitioners. In addition, sessions were broadcast from one room in the school building to another room in the same building, limiting the determination of whether this training method could be effective over greater physical distances between two facilities.

Web-based teleconferencing technology is more readily available and probably more affordable for behavior analysts and school districts than fiber optic networks. Thus, it may be useful to explore such technologies as a viable alternative for teleconsultation. The purpose of this article is to describe a technically-driven, collaborative approach to assessing the function of problem behavior using web-based technology. Specifically, behavior analysts collaborated with school personnel in rural Idaho to provide ongoing consultation and collaboration for conducting functional analyses of problem behavior. Information presented in this article aims to provide others with a example of how ongoing collaborative relationships can be established between behavior analysts and school personnel using currently-available technology to disseminate high quality behavior analysis services to children. Our goal is to encourage other researchers and practitioners to further evaluate this approach.

Pilot Project Description

School District, Project Origination, and Project Members

The current pilot project was initiated by a director of special education services (fifth author) in a rural school district in Idaho. The school district served just over 4,000 students and had five elementary schools, two middle schools, one high school, and one middle/high school in four different small towns. The district was also home to a charter school and a residential treatment facility for students with emotional/behavioral disorders and for abused children. Just over 40% of the student population was of Hispanic descent, while the vast majority of remaining students were Caucasian. A very small percentage of the student population was American Indian. Approximately 62% of the students participated in a free lunch program.

The district's special education director approached behavior analysts employed in the special education program at a local university to request staff training on behavior management. The school personnel wanted to learn more about functional analysis and reinforcement-based interventions and requested a training workshop on these procedures. The university personnel agreed to conduct training workshops with the school staff but suggested that ongoing consultation/collaboration might be useful to assist school personnel in implementing functional analyses for the first time. Both parties agreed that such ongoing consultation would be difficult, given that the school district was located in a very rural setting, approximately 100 miles away from the university. When the university personnel explained the current technologies being used for telemedicine, the director expressed interest in conducting a pilot project to test the feasibility of the process. The special education director sought and received district funding to implement a web-based teleconsultation model within her school district.

The behavior analysts were both associated with a local university and were Board Certified Behavior Analysts (BCBAs) with experience in implementing functional analyses and reinforcement-based interventions. The senior behavior analyst was a faculty member who had received her Ph.D. in special education. The junior behavior analyst was a doctoral student in the special education program who had received a M.A. degree in special education/applied behavior analysis.

Preliminary Instruction

The initial training and ongoing consultation/collaboration, conducted by the behavior analysts, combined both face-to-face interactions (for initial trainings and rapport building) and web-based consultation (for ongoing interactions). The initial training consisted of two full-day workshops that were conducted about 1 month apart. Fifteen school personnel participated in the workshops. Workshop attendance was voluntary, and not all participants attended both workshops. The school team who collaborated with the behavior analysts in the case example described later in the paper attended both workshops. The first workshop consisted of instruction on how to conduct a functional analysis. The second workshop consisted of instruction on function-based interventions. The behavior analysts described and modeled the assessment and treatment procedures and provided extensive packets of information for the workshop participants (e.g., a CD-ROM containing the presentation, selected research articles, videotaped examples, sample data sheets). A brief written test was given following each workshop to evaluate the school personnel's mastery of workshop content. All materials from these workshops are available for review from the first author.

Behavior analysts collaborated with school personnel in rural Idaho to provide ongoing consultation and collaboration for conducting functional analyses of problem behavior.

Technology and Equipment

An independent technology corporation provided the initial equipment setup and ongoing technical assistance to the school personnel and behavior analysts. Employees of the district had limited expertise in this type of technology, and the university program did not have the necessary technology expertise. The company had an established record of professionally selling, consulting, designing, installing, and servicing a variety of specialized technologies in the state. The additional consultation services provided by the outside technology company were paid for by the school district. The system used to connect the university with the district classroom was similar to a surveillance system used in convenience stores and consisted of the following features: two cameras in the district classroom settings, a digital video recorder (DVR) located at a central site in the district, a monitor for the DVR, and a remote access software program installed on the university personnel's computers and select district computers. The cameras were Ganz ZC-D6000™ series with high-resolution color (see www.cbcamerica.com). The DVR was a NVDV3-4000™ model with MPEG-4 Digital Video compression, 4-channel video input, 4-channel audio input, 200 Gigabyte hard drive storage capacity, embedded linux operating system, and remote access and monitoring (see www.nuvico.com). The DVR had the capacity to backup recorded data to CD-RW, DVD, external hard drive, or USB flash drive. The classroom teacher could control when the cameras and DVR were on. Anytime the cameras and DVR were turned on, the behavior analysts could log onto the Internet via the password-protected, remote-access software program and view the classroom activities. The remote access and management program had the following features: both static and dynamic IP options for network (LAN, WAN, Internet or PSTN); live monitoring for multiple cameras from multiple DVR's; the possibility of camera control functions (e.g., zoom and pan features, although these were not active in this project because the cameras did not have this capability); remote audio capability; and remote search, playback, and backup options. The program settings could be configured from a remote site.

Although the university had two DS3 Internet connections each with 45 megabit per second bandwidths, the school district had only a DSL Internet connection. Therefore, the upload speed was somewhat slow (compared to a high-speed, cable connection), and the quality of audio was poor when the behavior analysts viewed video via the Internet. (However, when the classroom personnel activated the DVR and recorded experimental sessions, audio quality was high.) Although the behavior analysts could hear the classroom audio, the system did not permit behavior analysts to broadcast audio back to the school personnel. As a result of these two constraints, a telephone connection was used for audio. The behavior analysts viewed sessions via the web while simultaneously talking into and listening via a telephone connection. The school personnel conducted sessions while wearing a wireless telephone headset to receive instructions and ongoing feedback throughout the sessions and to talk to the behavior analysts.

The total technology costs for the pilot project were approximately $5,000, including the cameras, surveillance system, DVR, headsets, long-distance cell phone charges, and all set-up fees/technology support. The school district's technology support staff also provided some additional support within their currently existing job descriptions at no extra cost. The behavior analysts used their own computers, at no extra cost to the project.

Case Example

School Team

Following the workshops, one teacher (third author) and a speech language pathologist (SLP, fourth author) volunteered to participate in ongoing consultation via web-based teleconsultation. Another teacher in the district also volunteered to participate; however, she did not have any students who would benefit from a functional analysis of problem behavior at the time of the project. The classroom teacher taught in an early intervention special education program in the district. She had her A.A. degree in early childhood education, a B.S. in health education, a B.A. in elementary education, and 7 years of teaching experience. The SLP provided speech/language services to children in the early intervention program. She provided services primarily in the classroom, using a collaborative model with the classroom teacher. She had her M.S. in speech and language pathology and 1 year of professional experience prior to the project. The teacher and SLP attended and actively participated in both training workshops. On the end-of-workshop written tests, one member of the school team scored 100% and 94%, respectively, while the second member scored 87% and 59%, respectively.

Child Participant

After the training workshops and the equipment installation, the school team identified a child who engaged in problem behavior and for whom it was appropriate to conduct a functional analysis. Ethan was a 4-year-old boy who received early intervention services in a preschool classroom under the category of developmental delays due to his poor verbal skills and fine motor abilities. Ethan was selected for participation in the project due to difficulty following instructions, making transitions, and staying on task. He also engaged in other problem behavior such as crying, screaming, and aggression. Prior to the project, Ethan spent most of his school day in a room outside the classroom where he could scream without disrupting the other students.

After the child participant was identified, the school personnel contacted the behavior analysts to begin the collaborative assessment process. The participant's parents, as well as the parents of the other children in the classroom, signed a state-approved “exchange of information” form to give consent for the project staff to videotape all students in the classroom and to broadcast the video to the behavior analysts. In addition, the teacher placed a sign on the door to the classroom when sessions were in progress to alert others that videotaping was in progress and that, by entering the room, the individual consented to being videotaped.

Problem Identification, Target Behaviors, and Hypothesis Development

The classroom teacher and the SLP conducted a functional behavior assessment interview with the rest of the preschool intervention team and the participant's parents (O'Neill, Horner, Albin, Storey, & Sprague, 1997). Following this, the school personnel and behavior analysts together defined the specific target behaviors. Target behaviors identified and defined were work refusals, off-task behavior, aggression (hitting and kicking), spitting, whining, crying, and tantrums. The teacher and SLP then collected a variety of observational data in the classroom by using scatter-plots (Touchette, MacDonald, & Langer, 1985; O'Neill et al., 1997) and antecedent-behavior-consequence (ABC) charts (O'Neill et al., 1997). The school team had received these materials at the workshops conducted by the university team. The teacher and SLP compiled all of this information and hypothesized that Ethan's problem behaviors probably served a variety of functions. After discussing the results with the behavior analysts, the school team decided that conducting a functional analysis would help further clarify those function(s).

Hypothesis Testing (Experimental Manipulations)

The teacher and SLP conducted the functional analysis in the early intervention classroom during regularly scheduled classroom routines. The classroom had two full-time instructional aides in addition to the classroom teacher. The aides' primary role during this project was to maintain classroom routines and activities while the classroom teacher and the SLP conducted the functional analysis sessions. The functional analysis was conducted using a multielement design with procedures similar to those described by Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994). All sessions were 5 min. Four conditions were conducted – free play, attention, demand, and tangible. The first condition, free play, consisted of allowing Ethan to play with toys of his choice and providing him with one-on-one adult attention. All occurrences of problem behavior were ignored, and neutral blocking was used if necessary (to prevent injury to self and/or others). This condition served as a control condition against which all other conditions were compared. The second condition, attention, consisted of allowing Ethan to play with toys of his choice again. However, Ethan was required to play alone, and adult attention was diverted from him. Contingent upon the occurrence of problem behavior, the adult provided Ethan with about 10 s of attention (e.g., “Ethan don't do that. You need to play nicely…like this.”). The purpose of this condition was to determine if Ethan's problem behavior was maintained by attention. The third condition, demand, consisted of prompting Ethan to engage in an activity of the adult's choice—an activity that was somewhat challenging for him to complete. This included activities such as coloring, name writing, and completing a maze worksheet. Contingent upon the occurrence of problem behavior, the task and the adult's attention were removed for about 10 s. The purpose of this condition was to determine if Ethan's problem behavior was maintained by escape from demands. The fourth condition, tangible, consisted of allowing Ethan to play with a non-preferred item. The adult had access to the child's preferred items, attention was delivered noncontingently at least once every 30 s, and no task demands were presented. Contingent upon the occurrence of the target behaviors, Ethan received access to the preferred items for approximately 10 s. After this time expired, the adult returned the non-preferred item and again interacted with the preferred items. The purpose of this condition was to determine if Ethan's problem behavior was maintained by access to preferred items.

Coaching from the behavior analysts occurred in person during two initial on-site visits and remotely for the remaining sessions using the web-based technology. When coaching was done remotely, the behavior analysts logged on to the web-based system to observe the functional analysis live and to provide coaching during the sessions.

Data Collection

The classroom teacher turned on the DVR to record the functional analysis sessions while broadcasting them to the university. The teacher and SLP designed the data sheet that they used to collect data on Ethan's problem behavior. They collected data via 10-s partial interval recording while replaying the sessions recorded on the DVR. (Data sheets are available on the Behavior Analysis in Practice website, http://www.abainter-national.org/BAinPractice.asp). The teacher and SLP created graphs of the data using Excel™ software. These activities were covered in the initial workshops that the teacher and SLP attended, but the behavior analysts provided ongoing consultation to the school team throughout the data collection and scoring process. For example, the teacher and SLP designed the data collection sheet and sent it via e-mail to the behavior analysts for feedback and input. Then, the teacher and SLP scored and graphed the data and sent an electronic copy of the graph to the behavior analysts. The behavior analysts assisted in interpreting the resulting data.

Summary of Outcomes

Child outcomes.

The behavior analysts and the school personnel met via telephone to discuss the results of Ethan's functional analysis, shown in Figure 1. The target behaviors remained at zero during all free-play conditions. Levels of the target behaviors were consistently higher in the tangible condition relative to the free-play condition. These results suggested that his problem behavior was maintained by positive reinforcement in the form of access to preferred items. The intervention agreed upon by the team was that Ethan should be provided with access to preferred items (e.g., a train) contingent upon appropriately engaging with less-preferred activities. Praise also was delivered for completing classroom work and for staying on task. These interventions were discussed between the behavior analysts and the school team after the assessment process had concluded. The behavior analysts asked questions of the school team to facilitate the development of a function-based intervention for Ethan. The intervention selected was consistent with those discussed in the initial workshops and during the ongoing consultation process. The school team then implemented these interventions in the classroom throughout the school day. The behavior analysts remained available for consultation and observed the school team implementing the intervention on occasion. Although the school team reported through e-mails that Ethan's behavior had improved, the classroom staff did not collect data consistently enough to formally evaluate the intervention.

Figure 1.

Percentage of intervals with target behaviors across free play, attention, demand, and tangible conditions for Ethan.

Interobserver agreement.

The behavior analysts independently scored the functional analysis data on problem behavior by using the save and playback feature through the remote system or by using a CD-ROM copy of the sessions that had been recorded by the school personnel. Because the remote management system could only save video footage to be used with the playback feature for a short term period (within 7 days), the method of playback to score interobserver agreement was selected based on whether the session could be scored within a few days after the session was conducted (playback feature) or needed to be scored more than a week after the session was conducted (CD-ROM copy). Interobserver agreement (IOA) was calculated by comparing the data collected by the behavior analysts to those collected by the school personnel. IOA for the occurrence of the target behaviors was scored for 100% of sessions and was calculated using an interval-by-interval comparison of the data sheets. The total number of agreements was divided by the sum of the total number of agreements plus disagreements and multiplied by 100%. The overall mean agreement was 99% (range, 93% to 100%).

Fidelity of functional analysis implementation.

Assessment fidelity data were collected by the behavior analysts via the CD-ROM copy of the sessions. Fidelity data were scored for 100% of sessions using a checklist developed by the university team (available for download on the Behavior Analysis in Practice website, http://www.abainternational.org/BAinPractice.asp). Fidelity percentages were calculated by taking the number of steps completed correctly and dividing it by the total number of steps possible and multiplying by 100%. Assessment fidelity was 100% on all occasions. IOA on fidelity was also completed for 33% of sessions and was calculated on an indicator-by-indicator comparison. The total number of agreements was divided by the sum of agreements plus disagreements and multiplied by 100%. Agreement was 100%.

Conclusions

We described the use of relatively inexpensive, web-based technologies for providing ongoing behavioral consultation. Despite the physical distance between the school district and the behavior analysts, school personnel were able to implement a functional analysis with high fidelity, collect observational data of child performance during the functional analysis, and collaborate with the behavior analysts to analyze and interpret the data. High percentages of assessment fidelity were likely due in part to the ongoing web-based consultation provided by the behavior analysts. Verbal prompting and feedback were provided while the school team conducted the sessions for the explicit purpose of maintaining high assessment fidelity.

The outcome of this pilot project, which is consistent with the results of prior research in this area (Barretto et al., 2006; Machalicek et al., 2009), suggest that teleconsultation can be delivered though less-expensive, web-based technology to school personnel who have little experience with data collection or functional analysis. A controlled evaluation of this approach is needed, however, to determine if the outcomes obtained in this pilot can be replicated with other school personnel.

We encountered a number of problems while conducting this pilot project. First, school personnel did not collect data in a systematic manner during the intervention. The staff reported that the interval recording system was fairly labor intensive and not always practical for ongoing monitoring. The behavior analysts had hoped to work with the school team to develop a more efficient data collection system that could be used to monitor the child's progress. However, the school year came to a close before this could be accomplished. Anecdotal comments from Ethan's teacher, classroom aides, and other school personnel (e.g., school psychologist) suggested the intervention was very successful and quickly reduced problem behavior. Such reductions in problem behavior may reduce the motivation of school team members to continue collecting data because the behavior is no longer viewed as a problem. Further research and collaboration between school teams and behavior analysts are needed to determine the most efficient forms of ongoing data collection and the supports that need to be in place for their continued use.

Second, the school district had only a DSL Internet connection. We found that this was simply not fast enough to transmit high-quality video and audio. A faster connection, such as a cable modem, is probably necessary to obtain the level of quality of audio and video needed to provide the best behavioral consultation. The poor quality transmission we experienced sometimes made it difficult for us to see exactly what was occurring in the classroom (i.e., due to the “jerky” quality of the video). We also found that having a camera that could pan and zoom would have been extremely useful. This would have allowed us to follow the child around the room more easily and zoom in to observe what he was doing during the analysis. Although the cost of pan- and zoom-capable cameras is much higher than the cameras the school purchased, it was our opinion by the end of the project that the money would have been well spent on such capabilities. Continued, ongoing research with improved technologies will be necessary to determine how best to make use of available technology.

Third, only 3 of the 15 individuals who attended one or both of the initial workshops volunteered to participate in the pilot project. There are many possible reasons why the others did not volunteer. For example, some teachers may have been skeptical of the technology in their classrooms and the idea of being observed by the behavior analysts on an ongoing basis; others may have been concerned about the added responsibility and time commitment; other workshop attendees did not have classrooms in which they could work with students on an ongoing basis in this capacity (e.g., the school administrator). However, individuals who did not volunteer to participate in the pilot project were not asked about the reasons for this decision.

Given the limited scope and descriptive nature of this pilot project, a systematic evaluation of this model is needed before we can recommend that it be implemented on a broad scale. Further research also is needed to determine the type and level of support required for school personnel to maintain a high degree of assessment and treatment fidelity. For example, it was unclear what role the initial workshops played in skill acquisition by the school personnel, or how well they would have performed without the behavior analysts' ongoing consultation. Similarly, the role of the on-site coaching that occurred prior to the web-based consultation is unclear.

Implications for Practice

To promote further evaluation of this collaborative approach, some guidelines for implementing the model are provided in Table 1. The school team reported feeling intimidated by the presence of the camera and DVR in the classroom, initially. However, they also reported that the rapport that was built via the training workshops helped them feel more comfortable with the process. In the end, the school team reported finding a valuable “friend in education” in the behavior analysts because the positive feedback and coaching from the behavior analysts helped them to provide a sophisticated level of service delivery that students such as Ethan need in order to be successful in their school.

Table 1.

Tips for teleconsultation.

Tip	Implication
Practice data collection techniques	Strengthens data scoring skills Strengthens working relationships Helps school personnel work more effectively with other students
Record all sessions electronically	Allows for continual review and discussion of the integrity of sessions being conducted Provides multiple opportunities for data scoring practice and feedback
Complement teleconsultation with on-site consultation	Establishes initial rapport
Provide intense support initially and gradually fade over time	Increases likelihood of ongoing success
Arrange release time for individuals involved in the consultation process	Helps provide structured time for adequate planning and implementation of the consultation model
Invest in quality technology and equipment	Helps prevent equipment and technology failure Enhances the teleconsultation process

Scoring the data proved to be a very challenging aspect of this project for school personnel. The collaborative model of service delivery seemed especially important for both data collection and for interpreting the data. Digitally recording the functional analysis sessions on the DVR was crucial to the data collection process. This allowed the school team to review and discuss the integrity of the procedures and the data collection process. Thus, behavior analysts should be prepared to invest this level of support early in the collaborative partnership. It was possible to fade these supports as the consultation relationship matured and school personnel gained more experience over time.

High-speed Internet connections also are critical when using the type of technologies described in this project. As noted previously, the school district had only a DSL Internet connection, which was not fast enough to transmit high-quality video and audio. A camera that could pan and zoom also would have been useful. Sometimes, Ethan displayed subtle problem behaviors that were difficult for the university-based data collectors to observe.

Time constraints were one of the most challenging aspects of this project for both the school team and the behavior analysts. It was often difficult for the team members to schedule time to meet virtually or over the telephone to complete this process. It was clear to us that both the school team and behavior analysts needed to have release time dedicated to this process in order to best collaborate. Those wishing to implement a consultation model of this sort may want to consider how such release time can be provided to all team members. Although this may result in additional up-front costs, it would be more cost-efficient in the long run if it reduces the lag between referral and intervention and improves the consultation process. More research is needed to better understand the time commitments necessary for on-going collaboration.

In sum, we have described a pilot project that enabled a school district and behavior analysts to work collaboratively to provide effective training for school personnel and behavior-analytic services for children with disabilities. More research is needed on how these partnerships can be forged and the components necessary for effective collaboration in conducting functional analyses and function-based treatments. The development of new technologies provides promise for a range of consultation services that were once difficult to administer due to distance and cost constraints.