The Use of Response Prompting and Frames for Teaching Sentence Writing to Students With Moderate Intellectual Disability

Abstract

In the current study, we examined the effects of response prompting strategies (i.e., constant time delay, system of least prompts) and frames on sentence writing for three participants, ages 7 to 12, with moderate intellectual disability. We used a concurrent multiple probe across behaviors design to evaluate the efficacy of the intervention package and posttest probes to assess generalized responding to untrained stimulation. During intervention, the teacher taught two students to construct sentences using selection-based software and another to generate handwritten responses across three different writing frames (i.e., I want _________, I see _____, The _____ is ______). Our findings suggest that the package was effective and produced variable levels of maintenance and generalized responding for all three participants.

Keywords

moderate to severe disability written expression sentence writing autism spectrum disorder

Written expression is a complex topography of communication that involves the simultaneous execution of a range of skills (e.g., spelling, adherence to grammatical conventions, naming). Researchers have suggested that individuals with moderate to severe intellectual disability (MSD) have difficulty acquiring functional repertoires in written expression (Dockrell, Ricketts, Charman, & Lindsay, 2014; Pennington & Delano, 2014; Purrazzella & Mechling, 2013). This presents a formidable barrier to inclusion for these individuals as writing is considered by a broad community of readers to be essential to daily activities. For example, writing is often required across a range of educational (Mercer & Mercer, 1998), postsecondary (Stevens, 2005), and social contexts (e.g., texting, social media). Furthermore, writing may offer additional benefits for students with and without disabilities. First, written expression may serve as a form of communication for individuals who do not use vocal speech. Lavigna (1977) taught individuals with MSD to exchange written words with a communication partner to access preferred items. Second, the acquisition of early writing skills may put learners’ writing responses in contact with the community of readers who may serve to shape the development of more complex communication skills.

Despite the importance of written expression, few research teams have evaluated strategies for improving the written expression of students with MSD. The majority of investigations have involved the application of response prompting techniques (e.g., time delay, system of least prompts [SLP]) and/or technology (e.g., computer software, power point, Smart Boards) to teach a range of writing skills including spelling (Purrazzella & Mechling, 2013; Stewart, Hayashi, & Saunders, 2010; Stromer, Mackay, Howell, & McVay, 1996), story construction (Pennington, Ault, Schuster, & Sanders, 2011; Pennington, Collins, Stenhoff, Turner, & Gunselman, 2014; Pennington, Stenhoff, Gibson, & Ballou, 2012), text message construction (Pennington, Saadatzi, Welch, & Scott, 2014), writing personal letters (Collins, Branson, Hall, & Rankin, 2001), and writing resume cover letters (Pennington, Delano, & Scott, 2014). Interestingly, the literature reflects a paucity of interest in the evaluation of procedures for the instruction of sentence writing skills to students with MSD. These skills are critical as sentences provide the reader with information as to purpose or function of the writers’ response. For example, the written word “puppy” may provide little meaning to a reader, whereas the sentence “I want a puppy” or “I see a puppy” may have a particular evocative effect on a reader’s response. The ability to generate or construct sentences helps writers wield their vocabulary with increased precision and serves as a foundational skill for the development of more complex narratives (Kame’enui & Simmons, 1990; Saddler & Asaro-Saddler, 2009).

Only two research teams have targeted sentence construction for students with MSD (Basil & Reyes, 2003; Yamamoto & Miya, 1999). In sentence construction tasks, individuals learn to select words to create a sentence that matches a sample (e.g., picture, spoken phrase). Sentence construction tasks are a common component of instruction for users of augmentative and alternative communication (AAC). For example, the Picture Exchange Communication System (Frost & Bondy, 1994) incorporates sentence construction as a part of the training package (i.e., Phases 4, 5, and 6). These construction tasks differ from traditional generative writing in that they provide a means for students to produce increasingly complex responses in the absence of a spelling repertoire. This is important as many students with MSD may present difficulties in acquiring spelling skills. Furthermore, data suggest that some individuals may acquire generative writing responses following instruction on story construction tasks (Pennington et al., 2011; Pennington, Collins, et al., 2014). Basil and Reyes (2003) demonstrated the effectiveness of a multimedia package on the acquisition of sentence writing skills for six students with MSD. Students selected words from a computer array to construct a sentence describing a picture. Following a correct response, the computer software provided auditory feedback (i.e., the spoken sentence, fanfare). Similarly, Yamamoto and Miya (1999) taught students to construct sentences by selecting Japanese characters from a computer array. Participants received an auditory “fanfare” following correct responses and a beep tone followed by the presentation of the correct response following errors. Both of these promising efforts demonstrated that students with MSD could learn sentence construction skills. Unfortunately, both research teams used pre- and posttest measures to evaluate their interventions with a small number of participants, resulting in a failure to demonstrate a functional relation between the intervention and student outcomes. Furthermore, both instructional packages incorporated trial and error learning in lieu of the more efficient response prompting strategies (e.g., constant time delay [CTD], SLP), which generally results in few errors during instruction (Collins, 2012). Finally, neither research team assessed generalization across untrained stimuli.

One potential strategy for teaching students with MSD to write sentences, involves the application of response prompting strategies in the context of sentence frames (Skinner, 1957). A frame can be defined as a scripted portion of a sentence (e.g., I want, I see) that can be completed to produce a targeted written or spoken response. When using frames, an instructor teaches a student to complete a common frame across multiple stimuli. For example, a student may be asked to complete an “I see” frame in response to multiple pictures. After learning to use the frame across several pictures, the student may apply the newly acquired frame in writing about novel stimuli. Researchers have used frames (e.g., Can I have, I want) to teach students to make requests using spoken sentences and have suggested that these frames have facilitated generalized responding (Betz, Higbee, Kelley, Sellers, & Pollard 2011; Hernandez, Hanley, Ingvarsson, & Tiger 2007). In light of the effectiveness of using frames to teach spoken responses, we sought to address the following research question:

Research Question 1: What are the effects of an instructional package including response prompting and frames on the sentence writing responses of students with moderate intellectual disability (MID)?

Method

Participants

Three children, ages 7 to 12 years, with MID participated in the study. The participants attended a self-contained special education classroom within a public elementary school. Holly was a 7-year-old female with autism spectrum disorder (ASD; Gilliam Autism Rating Scale = 103), childhood apraxia, and a reported full-scale IQ score of 47. Holly’s most recent formal language assessment (Preschool Language Scale–Fourth Edition) indicated a significant delay in expressive communication. Niles was a 12-year-old male with ASD (Autism Diagnostic Observation Schedule Comparison Score = 9) and a reported full-scale IQ score of 43. Formal language assessment data were not available. Due to comorbid physical impairments, both Holly and Niles had no prior history in producing manual print, and both used selection-based computer software during academic tasks. Prior to the study, both students could touch a cell within a computer array to make a selection. Burt was an 8-year-old male with apraxia and a reported full-scale IQ of 53. Burt’s scores on the Clinical Evaluation of Language Fundamentals–Fourth Edition, including those on Formulated Sentences and Sentence Structure subtests, indicated that he was in the “very low range” of communication functioning. Prior to the study, he could write single words and two-word combinations (e.g., red ball) with approximated spelling and copy text but did not produce complete sentences. All three participants emitted one- to three-word spoken requests and labels, and received intervention from a speech and language pathologist as a part of their educational program for 60, 80, and 100 min per week, respectively. Although formal reading assessments were not available, the classroom teacher identified all three students as primarily sight word readers. Prior to this investigation, Holly, Niles, and Burt produced one- or two-word utterances (e.g., big car), and did not use capitalization and punctuation conventions during writing tasks. The teacher identified these weaknesses as barriers to progress in developing more complex writing skills.

Settings and Materials

A teacher with a master’s degree in special education conducted all sessions in the special education classroom in a one-to-one instructional arrangement at a small table located in the back of the classroom. During each session, the participant sat next to the teacher with his or her back to the rest of the classroom. Six other students were present in the classroom and received instruction from paraprofessionals.

Instructional materials included various edibles and tangibles identified as preferred via stimulus preference assessments, pictured stimuli already present in the students’ tact repertoires, and model sentences presented on index cards. To promote generalization, we selected three exemplars of each pictured stimulus for writing tasks involving the “I see” frame and three different written models for each picture for use during tasks involving the “The ___ is” frame (e.g., The whale is big, The whale is blue, The whale is wet). During the investigation, Holly and Niles used a digital word bank presented on an iPad™ from which they selected words to construct sentences. The word bank was a 3 × 3 grid of nine familiar words (i.e., assessed prior to probe sessions) constructed using the software program Clicker Sentences (Crick Software Inc.). The word banks included the words for the target frame (e.g., I, want, a), the word corresponding to the selected preferred item or presented picture, and words that served as distracters. At the beginning of each session, the software program randomly organized the placement of the same set of words. We used this feature to reduce the potential influence of word order during tasks. Burt used paper and pencil for all writing tasks.

Dependent Variable

During probe and intervention conditions, we collected data on participants’ construction of complete sentences. We defined a complete sentence as (a) corresponding to the presented stimulus (e.g., “I want the candy” in the presence of candy), (b) including the target frame (i.e., I want, I see, The ___ is), (c) starting with a capitalized word (i.e., only for Burt, the Clicker Sentences™ program automatically capitalized the first word in a sentence), and (d) ending with a period. Data were reported and graphed as the percent of complete sentences.

A second observer, either the first or third author, collected dependent variable reliability data during 28%, 30%, and 37% of sessions for Holly, Niles, and Burt, respectively. During at least 25% of both probe and treatment conditions, the observer sat behind the teacher and participant while recording the student’s handwritten or typed response. We used the point-by-point method to calculate interobserver agreement across raters. The number of agreements were divided by the total number of agreements plus disagreements and multiplied by 100%. We determined the mean agreement across participants to be 99% (i.e., 100%, 98%, 100%).

A second observer collected treatment integrity data on the teacher’s implementation of the instructional package for 31% of probe and 36% of intervention sessions. During probe sessions, data were collected on six teacher responses: (a) conducting a preference assessment, (b) obtaining the student’s attention, (c) presenting a directive plus target stimulus, (d) waiting 10 s for the student to respond, (e) checking for completion, and (f) delivering praise. During intervention trials, data were collected on 13 potential responses for Holly and Niles and six responses for Burt. The discrepancy in the number of steps accounted for the differences in prompting procedures (i.e., CTD, SLP). The responses were (a) conducting a preference assessment, (b) obtaining the student’s attention, (c) delivering a directive plus target stimulus, (d) waiting 10 s for the student to respond, (e) administering prompt procedures (i.e., three to 10 potential steps were scored), and (f) delivering praise or a tangible contingent on an unprompted or prompted correct response. Treatment integrity data were 100% across probe and 99% across intervention sessions.

Experimental Design and Procedures

We used a concurrent multiple probe (Horner & Baer, 1978) across behaviors to evaluate the efficacy of an intervention package on the participants’ acquisition of sentence writing and construction responses. We selected this design because of its utility in demonstrating a functional relation with nonreversible behaviors. Furthermore, the teacher reported that writing was a nonpreferred task for the participants, so we sought to reduce potentially inhibitive testing effects by limiting the number of times students were asked to engage in tasks without instruction during baseline conditions. We also conducted probes after training to assess generalized responding to untrained stimuli.

Teacher training

Prior to conducting full probe and intervention sessions, we trained the teacher to implement instructional procedures using behavioral skills training (BST; Parsons & Reid, 1995). BST is a package that has been used to teach practitioners to implement a range of interventions for students with disabilities (Rosales, Stone, & Rehfeldt, 2009; Sarokoff & Sturmey, 2004), and is comprised of the presentation of instructions, modeling, rehearsal, and feedback. Two days prior to BST training sessions, we provided the teacher with a list of operationalized procedures. We asked the teacher if she had any questions, responded to her questions, and then directed her to perform three trials using the procedure with a nonparticipating student from the classroom. During these trials, she performed SLP procedures with 75% accuracy and CTD procedures with 83% accuracy. Subsequently, we reviewed the procedures, modeled the procedure with a confederate student, and then asked the teacher to perform the procedure for three-trial blocks. We delivered feedback after each block and continued until she performed three trials with 100% fidelity.

Probe conditions

During probe sessions, the teacher presented all three sets of instructional stimuli. The teacher presented Holly and Niles with five opportunities per stimulus set (i.e., total of 15 trials per probe session), whereas Burt received three trials per stimulus set (i.e., nine trials per probe session). Burt received fewer trials because handwriting required greater physical response effort. During probes for the use of the “I want” frame, the teacher first conducted a preference assessment (multiple stimulus without replacement [MSWO]; DeLeon & Iwata, 1996) to identify an item in which the student might be motivated to make a written request. The MSWO has been used widely to identify potential reinforcers for teaching requesting (e.g., O’Reilly et al., 2012; Valentino & Shillingsburg, 2011). Subsequently, she placed the preferred item on the table in front of the student, and stated, “Write a sentence to tell me what you want.” During probes for the use of the “I see” frame, the teacher presented a picture and the directive “Write a sentence telling me what you see.” During probes for the use of the “The ___ is” frame, the teacher presented a picture and the directive, “Write a sentence telling me about the________.” During all probes, the teacher waited 10 s for the participants to respond. If a participant did not start writing within 10 s, paused for 10 s, or vocally indicated completion, the trial was ended, and the teacher provided praise for sitting.

Intervention conditions

During intervention conditions, the teacher taught a single skill until the participant met criterion (i.e., 100% accuracy across three consecutive sessions). Intervention sessions were identical to probe sessions with two exceptions: (a) The teacher administered response prompting procedures if students did not respond within 10 s or made an error and (b) students received a tangible (i.e., I want frame) or praise (i.e., I see and The ___ is frames) contingent on prompted and unprompted correct responses. We used two different response prompting procedures. For Holly and Niles, we used a SLP (Collins, 2012) procedure because the controlling prompt (i.e., physically maneuvering the student’s finger to the correct word on the iPad™) did not require the student to attend to the word selected and a written model of the complete sentence did not consistently evoke a correct response. Therefore, we elected to present a hierarchy of prompts (i.e., model, physical) to ensure that Holly and Niles had an opportunity to respond to a model while ensuring performance of the targeted response. At the onset of each trial, the teacher presented a directive to write plus the targeted instructional stimulus and waited for 10 s for the student to respond. If the student did not respond within 10 s or made an error, the teacher erased the participant’s last selection (i.e., if an error was made), presented a printed model of the correct sentence, pointed to each word on the card, and waited for the participant to make the corresponding selection. If a participant made an error in the presence of the model, the teacher physically guided the participant to complete the sentence. Upon completion of prompted and unprompted sentences, the teacher provided a preferred stimulus or praise. For Burt, we used a CTD (Collins, 2012) procedure because the teacher reported that he had previously copied text with high levels of accuracy, thus only a single prompt was necessary to produce an accurate response. During the first session, the teacher conducted three 0-s delay trials. The teacher presented a directive to write plus the targeted instructional stimulus, immediately followed by the printed model of a sentence. The teacher delivered a preferred stimulus or praise contingent on the completion of prompted responses. In subsequent sessions, the teacher presented the printed model of the sentence contingent on the performance of an error or failure to respond within 10 s. Again, the teacher delivered a preferred stimulus or praise for prompted and unprompted sentences.

Maintenance and generalization probes

Following intervention on each frame, we conducted maintenance probes on previously mastered targets using procedures that were identical to those in full probe sessions. The teacher conducted generalization probes immediately following instruction on the final frame for each participant. For Holly and Niles, we added three new words to each of their arrays. We conducted generalization probes using procedures identical to full probes, with two exceptions. First, during probes for generalized written requests, we conducted a preference assessment using novel stimuli (i.e., chip, hug, high five). Second, during probes for generalized descriptive sentences, we asked participants to construct sentences about novel but familiar stimuli pictured in a book (i.e., shark, bird).

Results

Prior to intervention, participants did not consistently generate or construct targeted sentences. Following intervention, the participants produced three types of sentences and maintained responding following intervention (see Figures 1 to 3). Prior to intervention using the “I want” frame, Holly did not construct a single complete sentence. Following the introduction of intervention, she did not produce a sentence until the seventh session. She then met criterion after the next three sessions. Following the introduction of intervention on the “I see” and “The ___is” frames, she constructed sentences during the first session, and met criterion within Sessions 4 and 5, respectively. She maintained criterion levels of responding across all maintenance probes. During generalization probes (see Table 1), she constructed one complete sentence and two incomplete sentences (i.e., I want the chip, see fish, fish is hungry). Prior to intervention, Niles also did not construct a single complete sentence. During intervention on the “I want” frame, he constructed one sentence following the second day of training and rapidly met criterion within 7 days. During intervention on the “I see” frame, he produced two sentences on the second day of training, and again, his performance rapidly accelerated to criterion (i.e., within five days). Finally, using the “The ___ is ___” frame, he immediately produced a sentence on the first day of training but failed to produce sentences during the next four sessions. On Day 6, he constructed all five sentences but demonstrated variability in performance (i.e., 80%–100%) during the next seven sessions. He finally met criterion on Day 14 of intervention. He maintained variable levels of responding during maintenance probes. During generalization probes, he constructed two complete sentences and one incomplete sentence (i.e., I want the tickle, penguin, The shark is blue). Finally, Burt generated no sentences during baseline sessions. Following the introduction of intervention on the “I want” frame, Burt generated one sentence on the first day of intervention and rapidly met criterion by Day 5. During intervention using the “I see” frame, Burt met criterion within three days. This suggests that Burt might have acquired the response during the presentation of 0-s delay trials. Finally, Burt generated two sentences on the second day of intervention but displayed variability in responding (i.e., 33%–100%) until meeting criterion on Day 15. Burt also maintained variable levels of responding during maintenance probes. During generalization probes, he constructed one complete sentence and two incomplete sentences (i.e., I want the candy, I the shark, The shark black).

Figure 1.

Percent of complete sentences for Holly across probe and intervention conditions.

Figure 2.

Percent of complete sentences for Niles across probe and intervention conditions.

Figure 3.

Percent of complete sentences for Burt across probe and intervention conditions.

Table 1.

Generalized Responses to Untrained Stimuli.

Participant	Frames
Participant	I want_____	I see_____	The ___is ____
Holly
Generalization	I want the chip	see fish	Fish is hungry
Niles
Generalization	I want the tickle	penguin	The shark is blue
Burt
Generalization	I want the candy	I the shark	The shark black

Discussion

The development of sentence writing skills is critical for all learners as they help clarify a writer’s intent, and ultimately help them gain access to desirable and specific outcomes. This functional approach is especially critical for learners in which writing may be more difficult due to physical or intellectual disabilities. Writing must be perceived as meaningful by the writer if they are to persist during writing tasks and ultimately develop more complex writing skills. In the current study, we evaluated the effectiveness of two instructional packages comprised of response prompting and frames in teaching sentence production for three students with MID. For two of the students, we used the SLP and assistive technology to teach sentence construction tasks, whereas for one student, we used CTD to teach the generation of handwritten sentences. Our findings suggest that the procedures were effective in establishing sentence production and some generalized responding for all three children. This is not surprising in light of the previous work using similar procedures to establish spoken sentences in participants with similar characteristics (Hernandez et al., 2007). Although all three participants demonstrated the emergence of generalized responses, none generated criterion levels of responding (i.e., complete sentence) during all three tasks. This may have been due in part to (a) a small number of instructional sessions on each frame and (b) the limited number of exemplars (i.e., three) trained to instructional sessions.

Interestingly, all three participants produced complete sentences when asked to request novel items. Furthermore, they maintained high levels (i.e., 100%) of performance on requesting tasks following intervention. These findings may reflect the contingencies associated with each task. The requesting task required students to produce a written request, resulting in direct access to a tangible reinforcer, whereas the describing tasks (e.g., I see, The ___ is ____) resulted in access to praise. This could potentially inform practice as it might indicate the need for incorporating requesting tasks in early writing instruction. An early focus on requesting is consistent with data on language development as young children generally acquire simple requests before naming objects (Carpenter, Mastergeorge, & Coggins, 1983). Furthermore, these functional requesting tasks may help early writers understand the writer (i.e., speaker) and reader (i.e., listener) relationship as they capitalize on powerful motivational variables. The consideration of these motivational variables is well established as a key component to teaching other functional classes of language (e.g., labeling, responding to the language of others; Michael, 1988). These findings may also suggest that the participants had stronger requesting than labeling repertoires prior to the study. More research is required to evaluate the impact of preintervention vocal repertoires and motivational variables on the effectiveness of the current instructional packages.

In addition, it is unclear as to whether all components of the intervention were necessary to produce the observed effects. It may have been the case that participants might have met criterion without the use of frames. This is plausible as response prompting strategies have long been used in the instruction of a range of skills (Touchette, 1971; Walker, 2008; Wolery et al., 1992). Furthermore, the presentation of varied models (i.e., no frames or mixed frames) during instruction might have produced more generalized responding. Future investigations will be necessary to parse out the most active components of this and similar instructional packages.

Finally, it is important to note that these procedures are consistent with recommendations for all struggling writers and may be used to facilitate the development of sentence writing repertoires in students without MSD (Graham, Harris, & Larsen, 2001). This is critical as data have consistently illuminated a pattern of low performance in the area of written expression by a majority of students (U.S. Department of Education, Institute of Educational Sciences, National Center for Education Statistics, National Assessment of Educational Progress, 2011). The current package may offer a solution to teachers charged with helping their struggling writers overcome the hurdle of developing basic writing skills.

Limitations

Several additional limitations must be noted. First, it is important to note that in the current study, we used two different procedures matched to the skills of participants. This limited the number of demonstrations of effect for a single package. However, we did use an experimental design (i.e., multiple probe across behaviors) that permitted the number of demonstrations of effect required to show a functional relation for each package. Second, we did not test participants’ performance on generalization tasks prior to intervention. Therefore, our results related to generalized responding must be considered with caution. Finally, our operational definition of a complete sentence may not have been sensitive to participants’ progress in developing sentences. That is, a participant’s response “dog” would be scored the same as “see the dog.” Future research should address the challenges in the measurement of smaller functional units within written responses.

Conclusion

Overall, our findings demonstrated the effectiveness of an instructional package in teaching students with MSD to construct complete sentences for different purposes (i.e., requesting, describing). These results may have direct implication for future research and practice as only a few studies addressing basic writing skills for this population exist. Furthermore, we demonstrated the modification of instructional procedures for students with two different ways of responding during tasks. This emulates the requirement of many special educators to develop instructional programs for diverse learners under their care. Finally, we suggest that the importance of written expression for all learners serves as an impetus for increased investigations into this area. Practitioners need more demonstrations of how to move their students with MSD forward in this complex area so that each child may more fully engage in a range of educational and social opportunities and perhaps, one day, write their own story.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Basil

Reyes

(2003). Acquisition of literacy skills by children with severe disability. Child Language Teaching & Therapy, 19, 27–49.

Betz

A. M.

Higbee

T. S.

Kelley

K. N.

Sellers

T. P.

Pollard

J. S.

(2011). Increasing response variability of mand frames with script training and extinction. Journal of Applied Behavior Analysis, 44, 357–362.

Carpenter

R. L.

Mastergeorge

A. M.

Coggins

T. E.

(1983). The acquisition of communicative intentions in infants eight to fifteen months of age. Language and Speech, 26, 101–116.

Collins

B. C.

(2012). Systematic instruction for students with moderate and severe disabilities. Baltimore, MD: Paul H. Brookes.

Collins

B. C.

Branson

T. A.

Hall

Rankin

S. W.

(2001). Teaching secondary students with moderate disabilities in an inclusive academic classroom setting. Journal of Developmental and Physical Disabilities, 13, 41–59.

DeLeon

I. G.

Iwata

B. A.

(1996). Evaluation of a multiple-stimulus presentation format for assessing reinforcer preferences. Journal of Applied Behavior Analysis, 29, 519–533.

Dockrell

J. E.

Ricketts

Charman

Lindsay

(2014). Exploring writing products in students with language impairments and autism spectrum disorders. Learning and Instruction, 32, 81–90.

Frost

L. A.

Bondy

A. S.

(1994). The picture exchange communication system. Focus on Autism and Other Developmental Disabilities, 9, 1–19.

Graham

Harris

K. R.

Larsen

(2001). Prevention and intervention of writing difficulties for students with learning disabilities. Learning Disabilities Research & Practice, 16, 74–84.

10.

Hernandez

Hanley

G. P.

Ingvarsson

E. T.

Tiger

J. H.

(2007). A preliminary evaluation of the emergence of novel mand forms. Journal of Applied Behavior Analysis, 40, 137–156.

11.

Horner

R. D.

Baer

D. M.

(1978). Multiple-probe technique: A variation of the multiple baseline. Journal of Applied Behavior Analysis, 11, 189–196.

12.

Kame’enui

E. J.

Simmons

D. C.

(1990). Designing instructional strategies. Englewood Cliffs, NJ: Merrill.

13.

Lavigna

(1977). Communication training in mute autistic adolescents using the written word. Journal of Autism and Childhood Schizophrenia, 7, 135–149.

14.

Mercer

C. D.

Mercer

A. R.

(1998). Teaching students with learning problems. Upper Saddle River, NJ: Prentice Hall.

15.

Michael

(1988). Establishing operations and the mand. The Analysis of Verbal Behavior, 6, 3–9.

16.

O’Reilly

Aguilar

Fragale

Lang

Edrisinha

Sigafoos

Didden

(2012). Effects of a motivating operation manipulation on the maintenance of mands. Journal of Applied Behavior Analysis, 45, 443–447.

17.

Parsons

M. B.

Reid

D. H.

(1995). Training residential supervisors to provide feedback for maintaining staff teaching skills with people who have severe disabilities. Journal of Applied Behavior Analysis, 28, 317–322.

18.

Pennington

R. C.

Ault

M. J.

Schuster

J. W.

Sanders

(2011). Using response prompting and assistive technology to teach story-writing to students with autism. Assistive Technology Outcomes and Benefits, 7, 24–38.

19.

Pennington

R. C.

Collins

B. C.

Stenhoff

D. M.

Turner

Gunselman

(2014). Using simultaneous prompting to teach generative writing to students with autism. Education and Training in Autism and Developmental Disabilities, 49, 396–414.

20.

Pennington

R. C.

Delano

M. D.

Scott

(2014). An intervention for improving resume cover letter writing skills of students with intellectual disabilities. Journal of Applied Behavior Analysis, 47, 1–5.

21.

Pennington

R. C.

Delano

M. E.

(2014). Teaching written expression to students with intellectual disabilities. In Browder

D. M.

Spooner

(Eds.), More language arts, math, and science for students with severe disabilities (pp. 127–145). Baltimore, MD: Paul H. Brookes.

22.

Pennington

R. C.

Saadatzi

Welch

K. C.

Scott

(2014). An investigation of robot delivered instruction to teach texting to students with intellectual disabilities. Journal of Special Education Technology, 29, 49–58.

23.

Pennington

R. C.

Stenhoff

D. M.

Gibson

Ballou

(2012). Using simultaneous prompting to teach story writing to a student with autism. Education & Treatment of Children, 35, 389–406.

24.

Purrazzella

Mechling

L. C.

(2013). Evaluation of manual spelling, observational and incidental learning using computer-based instruction with a tablet PC, large screen projection, and a forward chaining procedure. Education and Training in Autism and Developmental Disabilities, 48, 218–235.

25.

Rosales

Stone

Rehfeldt

R. A.

(2009). The effects of behavioral skills training on implementation of the picture exchange communication system. Journal of Applied Behavior Analysis, 42, 541–549.

26.

Saddler

Asaro-Saddler

(2009). Writing better sentences: Sentence-combining instruction in the classroom. Preventing School Failure: Alternative Education for Children and Youth, 54, 159–163.

27.

Sarokoff

R. A.

Sturmey

(2004). The effects of behavioral skills training on staff implementation of discrete-trial teaching. Journal of Applied Behavior Analysis, 37, 535–538.

28.

Skinner

B. F.

(1957). Verbal behavior. Acton, MA: Copley.

29.

Stevens

(2005). What communication skills do employers want? Silicon valley recruiters respond. Journal of Employment Counseling, 42, 2–9.

30.

Stewart

Hayashi

Saunders

(2010). Enhancing vowel discrimination using constructed spelling. The Analysis of Verbal Behavior, 26, 57–64.

31.

Stromer

Mackay

H. A.

Howell

S. R.

McVay

A. A.

(1996). Teaching computer-based spelling to individuals with developmental and hearing disabilities: Transfer of stimulus control to writing tasks. Journal of Applied Behavior Analysis, 29, 25–42.

32.

Touchette

P. E.

(1971). Transfer of stimulus control: Measuring the moment of transfer. Journal of the Experimental Analysis of Behavior, 15, 347–354.

33.

U.S. Department of Education, Institute of Educational Sciences, National Center for Education Statistics, National Assessment of Educational Progress. (2011). Writing assessment. Retrieved from http://nationsreportcard.gov/report_archive.aspx

34.

Valentino

A. L.

Shillingsburg

M. A.

(2011). Acquisition of mands, tacts, and intraverbals through sign exposure in an individual with autism. The Analysis of Verbal Behavior, 27, 95–101.

35.

Walker

(2008). Constant and progressive time delay procedures for teaching children with autism: A literature review. Journal of Autism and Developmental Disorders, 38, 261–275.

36.

Wolery

Holcombe

Cybriwsky

Doyle

P. M.

Schuster

J. W.

Ault

M. J.

Gast

D. L.

(1992). Constant time delay with discrete responses: A review of effectiveness and demographic, procedural, and methodological parameters. Research in Developmental Disabilities, 13, 239–266.

37.

Yamamoto

Miya

(1999). Acquisition and transfer of sentence construction in autistic students: Analysis by computer-based teaching. Research in Developmental Disabilities, 20, 355–377.