The purpose of this study was to compare the effects of teacher-directed simultaneous prompting to computer-assisted simultaneous prompting for teaching sight words to 3 elementary school students with intellectual disability. Activities in the computer-assisted condition were designed with Intellitools Classroom Suite software whereas traditional materials (i.e., flashcards) were used in the teacher-directed condition. Treatment conditions were compared using an adapted alternating treatments design. Acquisition of sight words occurred in both conditions for all 3 participants; however, each participant either clearly responded better in the teacher-directed condition or reported a preference for the teacher-directed condition when performance was similar with computer-assisted instruction being more efficient. Practical implications and directions for future research are discussed.
Reading proficiency is a challenge for many individuals with intellectual disability (ID; Onyekuru & Njoku, 2012). Channell, Loveall, and Conners (2013) reported students with ID scored similarly to children 7 to 10 years younger on word recognition tasks. Several factors make reading a difficult task for students with ID, including (a) deficits in memory skills, (b) decreased attention span, (c) difficulty attending to relevant stimuli, (d) distractibility, and (e) delays in social behaviors which affect participation in reading instruction (Onyekuru & Njoku, 2012; Westling & Fox, 2009). Thus, instruction on sight words is the most common form of reading instruction for students with ID (Browder, Wakeman, Spooner, Ahlgrim-Delzell, & Algozzine, 2006).
Sight word recognition is a fundamental element to reading and academic success. Sight words are high-frequency words that a strong reader automatically recognizes without having to decode; this allows the reader to focus more on the meaning of the text rather than each word part (Ehri, 2005). Students’ ability to recognize sight words is correlated with later skills such as reading fluency and comprehension (Ouellette, 2006). The ability to recognize sight words enables students with disabilities to become more active and fluent readers which promotes their access to the general education curriculum (Akcin, 2013).
One of the most researched methods for teaching sight words to students with ID is systematic prompting strategies (Browder et. al., 2006). Simultaneous prompting (SP) is a type of prompting strategy where an instructional cue is presented (e.g., a discriminative stimulus of showing a student the written word her) simultaneously with a controlling prompt (e.g., teacher saying the word her) and then having the student respond immediately (e.g., student saying the word her; Waugh, Alberto, & Frederick, 2011). SP is considered a near errorless learning procedure because the controlling prompt is delivered simultaneously with the discriminative stimulus in every trial (Gibson & Schuster, 1992). In their reviews of the literature, Morse and Schuster (2004) and Waugh et al. (2011) identified a total of 35 studies from 1999 to 2010 that examined the effects of SP. They reported that, in general, SP was an effective intervention. Specifically, across identified studies, SP was used to teach various skills (e.g., literacy, communication, daily living, leisure, vocational skills) to participants of various ages, with a majority of participants identified as having ID and some who had autism (Waugh et al., 2011). Waugh et al. (2011) reported that the majority of selected studies focusing on teaching literacy skills used SP to teach sight word recognition.
Computer-assisted instruction (CAI), a relatively newer method of instruction for students with ID, is growing in popularity due to its effective academic outcomes and positive effect on students’ self-esteem and language skills (Collet-Klingenberg, 2009). One reason that CAI is beneficial for students with disabilities is because it can provide multiple modes of presentation visually and auditorily with images, colors, sound, and elements of interactivity (Mechling, Gast, & Krupa, 2007). Results from recent studies support the efficacy of combining CAI with prompting strategies to teach sight word reading to students with ID or autism (Coleman, Hurley, & Cihak, 2012; Lee & Vail, 2005; Yaw et al., 2011, 2012). Specifically, researchers investigated the effects of using technology to provide another type of systematic prompting, constant time delay. In contrast to simultaneous prompting, with constant time delay, the instructional cue is presented (again, a discriminative stimulus of showing a student the written word, her). Then, after a predetermined delay (e.g., a 2-s delay), the controlling prompt is presented (e.g., computer saying the word her).
Lee and Vail (2005) found combining constant time delay (specifically, a constant time delay of 5 s between the presentation of the instructional cue and the controlling prompt) with CAI improved literacy skills (sight word reading and retell/recall of incedental information) for elementary students with ID or developmental delays. In this study, the specialized software, Word Wizard, was programmed to deliver the controlling prompt after a 5-second delay. Mechling et al. (2007) found PowerPoint-presented 3-s constant time delay effective in teaching grocery sight words to high school students with ID. In addition, Yaw and colleagues found PowerPoint, set up with a modified time delay procedure, to be effective in increasing reading of sight words for a student with autism (Yaw et al., 2011) and another student with ID (Yaw et al., 2012),
Coleman, Hurley, and Cihak (2012) suggested using CAI programmed to provide systematic prompts was beneficial because it enabeled teachers to provide students with structured independent practice without one-on-one supervision from teachers. To directly examine the usefulness of CAI to provide systematic prompting in classrooms, Coleman et al. (2012) compared teacher-directed constant time delay (4-s delay) with computer-assisted constant time delay (also a 4-s delay) to teach recipe sight words to three elementary students with ID. Results indicated that both types of constant time delay instruction led to increases in recipe sight word reading; however, teacher-directed constant time delay was slightly more efficient in terms of trials to criterion for two students whereas computer-assisted constant time delay was more efficient for the remining student. Although teacher-directed intervention was slightly more efficient in terms of trials to criterion, the authors stated benefits of students being able to work independently on tasks that usually required one-on-one adult interaction. However, the authors also argued that these results bolster the importance of teacher-led instruction as a first line of instruction; the use of computer-assisted instruction could then be used for additonal, independent practice of skills. During an extensive literature search, no published studies were located in which the effects of simultaneous prompting combined with CAI for teaching sight word reading to students with ID were located. Furthermore, no studies were located in which Classroom Suite software was used to teach sight words.
Research has demonstrated the effectiveness of using CAI to provide constant time delay as a prompting strategy to teach sight words for students with ID, potentially allowing for independent student practice without extensive teacher supervision. CAI has the potential to expand teacher resources in that it frees the teacher from one-to-one teacher-based instruction by having students practice skills via technology. Additionally, using CAI helps students develop greater fluency in using technology, which may provide them access to other opportunities (e.g., digital media skills, vocational possibilities). SP has been demonstrated an effective prompting strategy for teaching sight word reading to students with ID (Morse & Schuster, 2004; Waugh et al., 2011) and the use of CAI has increased in classroom use; thus, it is timely to examine whether CAI and SP can better teach sight words to students with ID.
Therefore, the purpose of this study was to examine the differential effects of teacher-directed simultaneous prompting compared with computer-assisted simultaneous prompting in terms of effectiveness or efficiency in increasing the number of sight words recognized by elementary school students with ID. Researchers hypothesized that, similar to Coleman et al.’s (2012) findings, both teacher-directed and computer-assisted SP would lead to increases in sight word reading but that teacher-directed SP would be more efficient in terms of number of sessions required to reach criterion.
Three elementary-aged students with ID took part in this study. All three students receive most of their instruction in a self-contained classroom designed for students with intellectual and developmental disabilities who are working on traditional academic instruction. The second author was completing an internship in the students’ classroom working toward initial teacher licensure in special education as well as completion of a Master of Science degree. The classroom is located in an elementary school in a lower income neighborhood in the southeastern United States. Inclusionary criteria for participation in the study included (a) being a student with ID (IQ score ranging from approximately 40 to 55) who received academic instruction in the self-contained classroom, (b) having individualized education program (IEP) goals dedicated to sight word instruction, (c) expressing an interest in working on the project and giving assent, and (d) returning parental consent forms. See Table 1 for a summary of participant information.
At the time of the study, Alice was 9-years old and in the fourth grade. She received special education under the eligibility categories of ID and other health impairment due to having a seizure disorder. Alice’s IQ score was assessed to be 53 on the Wechsler Intelligence Scale for Children III (WISC-III; Wechsler, 1991). Her adaptive behavior standard score was 57 on the Vineland Adaptive Behavior Scales, Second Edition (VABS; Sparrow, Cicchetti, & Balla, 1989). On the Brigance Diagnostic Inventory of Early Development-Second Edition (Brigance, 2004), Alice was able to read 70% of pre-primer words. Alice attended general education for special areas classes (e.g., art, music, physical education) and received occupational therapy 30 min per week. Otherwise, she received instruction in the self-contained special education classroom. According to her teacher, Alice has strong pragmatic language skill but lacks strong formal language skills; she also has trouble with maintaining attention for long periods of time, following multistep directions, and completing seatwork independently. Prior assessment revealed she was able to read 19 words on Rebecca Sitton’s List of 1200 High Frequency Words (Sitton, 1995).
At the onset of the study, Kayla was a 10-year-old, fourth-grade student. Her areas of eligibility are autism and ID. On the WISC-III, Kayla’s IQ score was assessed to be 46. Her adaptive behavior standard score (VABS) was 53. On the Clinical Evaluation of Language Fundamentals-4th edition (Sparrow, Cicchetti, & Balla, 1989), Kayla’s core language score was 42. Vocabulary was a relative strength for her as her standard score on the Peabody Picture Vocabulary Test-III (Dunn & Dunn, 1997), was 80. Kayla received all of her instruction in the self-contained special education classroom other than 50 min per day for special areas classes. She received 30 min of occupational therapy per week along with 4 hr of speech and language therapy per month. According to her teacher, Kayla is frustrated easily and often cries if she perceives tasks as too difficult. Kayla’s teacher also reports that her writing is laborious and slow, she does not demonstrate mastery of phonics or phonemic awareness, and reading is difficult because of her inability to decode different letter sounds. Kayla was able to read 12 words on Rebecca Sitton’s List of 1200 High Frequency Words (Sitton, 1995) prior to the study.
When the study began, Dustin was an 11-year-old, fifth-grade student with eligibilities in autism and ID. Dustin’s IQ score was determined to be 55 using the Cognitive Assessment System (Naglieri & Das, 1997) and his adaptive behavior battery was 55 on the VABS. On the Brigance, his reading and written expression were at the first grade level. Dustin participated in general education special areas classes 50 min per day and received all other instruction in the self-contained special education classroom with consultative occupational therapy services. Prior to the study, Dustin could read 630 words on the Rebecca Sitton’s List of 1200 High Frequency Words (Sitton, 1995). According to his teacher, however, Dustin reads well below grade level and is hesitant to read aloud. His teacher also reports that Dustin can write complex sentences but struggles with spelling and grammar; he frequently answers questions incorrectly or writes inappropriate things (e.g., insulting comments). His teacher further reported that Dustin perseverates on objects in the room or people in the room, and he demonstrates deficits in self-organization, work quality, and ability to listen and respond during classroom instruction.
Classroom Setting and Teachers
All instructional sessions took place in self-contained special education classroom. There was a total of 15 students that spent at least a portion of their school day in the classroom. The students in the classroom ranged from third to fifth grade. On a typical day during the study, a licensed special education teacher and a paraprofessional were present in the classroom. In addition, there was teaching intern present who provided and facilitated study-related instruction to participating students. The teaching intern was a graduate student enrolled in a special education teacher preparation program at a local university. She had already received an undergraduate degree in special education and was participating in a year-long, full-time teaching internship. Her coursework included two instructional methods courses in which prompting procedures were taught explicitly, as well as an assistive technology course. As part of her internship coursework, the teaching intern completed a literature review related to and procedures for implementing systematic prompting strategies and related to conducting research in applied settings.
Rebecca Sitton’s List of 1200 High Frequency Words (Sitton, 1995) was used to identify unknown, target sight words during baseline and for developing a set of 10 target sight words for each participant. This list is used for all students participating in reading instruction in the school system. For the purpose of assessing each participant, a set of flashcards was created with one target sight word written on each flashcard. Data collection sheets were used to record incorrect and correct target sight words read across all phases.
Teacher-directed simultaneous prompting (TDSP)
For each participant, a set of flashcards was created to teach the five target sight words during the TDSP condition. Here, one target sight word was written on each flashcard.
Computer-assisted simultaneous prompting (CASP)
The materials for the CASP condition included two laptop computers with Classroom Suite (Intellitools, Inc., 2007) software installed and headphone sets for each of the two laptops. Each student had his or her own activities designed with target sight words on Classroom Suite. The software was programmed to display each target words individually, three times each. Every time the word was shown on the screen, the software speech synthesizer spoke the word simultaneously. The headphones were used to minimize distractions as well as to help the students hear the words.
Data Collection: Word Selection and Baseline Procedures
The dependent variable was percentage of target sight words read correctly. To select potential target sight words, each participant’s present level of sight word recognition was determined by using information from assessments noted in the student’s IEP along with classroom sight word instruction assessments. Students were working on words from the Sitton list during classroom instruction; therefore, words from the next instructional level were used. Based on this information, 35 words were chosen for each participant that came from the level of the Sitton list that most closely matched assessment information. The interning teacher asked each participant to read the words from his or her selected word list aloud; missed or unknown words were marked as incorrect on data sheets by the interning teacher. In this form, each participant was assessed for three separate baseline sessions (across 3 days) on all 35 words from his or her word list. At the end of the third baseline session, for each participant, the interning teacher identified 10 words that were marked incorrect on all three baseline sessions (thus the percentage of baseline accuracy was 0% for each word set for each participant). It was predicted that accuracy would not increase without intervention, so no further baseline data were collected.
Each participant’s 10 incorrect words were then separated into two sets: one set of five words for instruction with TDSP and one set of five words for instruction with CASP. Because words were selected from the same section of the Sitton list, their difficulty level was similar. To further balance the difficulty level of the word sets, each participant’s 10 words were entered into an online randomization tool (www.randomizer.org) and the first five that were randomly generated were used for TDSP with the remaining words for CASP. See Table 2 for a list of each participant’s word sets.
Two instructional conditions, TDSP and CASP, were compared using an adapted alternating treatments design (Gast, 2010). This design was selected because the researchers wanted to determine if one type of SP would be more effective or efficient. The phases in this study included baseline data collection, technology training, alternating treatments, and a “preferred treatment phase” during which the most effective or preferred condition was implemented (Gast, 2010).
During baseline, data were collected for each student’s percentage of sight words read correctly as described previously. No instruction was provided on the target sight words nor was any feedback given during baseline.
Once baseline data were collected, the participants were trained how to navigate through the Classroom Suite activities. A sample activity was created with five words known by all participants. Known words were selected for training because the purpose was for the students to learn how to operate the software, not learn new words. The instructor used modeling to teach each participant how to begin a session, click buttons to navigate through the activity, select correct responses while orally saying the response, and how to end the session. The importance of paying attention was stressed, and participants were given time to explore the activity and ask questions.
Prior to each instructional session of the alternating treatments and preferred treatment phases, participants were assessed individually on all 10 words which were written on flashcards and shuffled each day for randomization; words read correct and incorrect were recorded on the data sheet. The participants were assessed prior to instruction to determine learning that had been retained from the previous day. Assessment was not repeated following intervention because the objective was for students to retain the information and not to assess information potentially remaining in short-term memory immediately after the intervention session.
Immediately after assessment on all 10 target sight words, participants participated in one of two instructional conditions: TDSP or CASP; students participated in only one instructional condition each day. For each participant, the order in which conditions were presented was randomized using an online randomization tool. The online randomization tool was set to generate 20 sets of two numbers (1, 2). After randomization, the sets were copied and pasted into a word processing document and all ones were replaced with A while twos were replaced with B. For example, random sets of 1, 2; 2, 1; 2, 1 would be replaced with A, B; B, A, B, A. This form of randomization ensured that each condition was presented no more than two consecutive times.
All TDSP instruction occurred in a one-to-one format. During each TDSP session, each of the five target words was presented individually to the participating student. The teaching intern showed the student the word written on a flashcard. The teaching intern immediately said, “This word is [sight word]. What word is this?” The student then repeated the word orally after hearing the teacher say it. Correct responses were followed by verbal praise. Incorrect responses were followed by correction (i.e., teacher said the correct word and had the student repeat it). The five words were presented three times each, in random order, during each TDSP session.
For the CASP condition, Classroom Suite presented words in the same way. A single target sight word appeared at the top of the screen while the program said aloud the word simultaneously with synthesized speech. Although not as clear as human speech, the speech synthesizer within the software program was high quality and resulted in clearly spoken individual words. Three word choices were presented at the bottom of the screen. After receiving the prompt, the participant had to select a word from the choices while saying the word aloud. It is important to note that the topography of participant responses differed between TDSP and CASP. Classroom Suite is a commercially available and relatively inexpensive software program commonly used in special education classrooms. From a strict research perspective, differing responses are not ideal. However, students inherently respond differently in technology-implemented instruction than they do in teacher-directed instruction. Given the recent push to infuse technology into educational programming for all students, the goal was to make a comparison between two types of instruction that are typically used in classrooms.
In Classroom Suite, if the student selected the correct choice, a reinforcer screen (e.g., computer said, “Yay!” while presenting a graphic) was provided followed by the next word. If an incorrect word was chosen, the incorrect word was spoken and the following word was presented. For each participant, Classroom Suite activities presented the five CASP words three times each. Several versions of the activity for each set of words were created to reduce order effects. Each version contained the same words, but words were presented in different orders to minimize participants learning the word order rather than the target words. See Figure 1 for an example of Classroom Suite activities.
Preferred treatment phase
The alternating treatments phase of the study continued for each participant until he or she reached criterion of 100% accuracy for two consecutive sessions out of three consecutive sessions in which 80% or higher responding was obtained in one condition. As is common to alternating treatments studies, once criterion was reached for one condition (the most effective or preferred condition), instructional sessions continued using the preferred treatment condition only (Gast, 2010). The preferred treatment phase was implemented to demonstrate replication of effects. In the present study, when there was a clear difference in performance, based on visual inspection of data, words for which the participants had not yet responded at 100% accuracy were taught using the preferred condition to demonstrate further the effectiveness of the preferred condition. If superiority of one condition was not clearly evident based on participants reaching 100% responding during some sessions for both, participants were asked which condition was preferred and that condition was used to instruct a new set of words (because student had reached 100% responding during some sessions for both sets of previously instructed words).
Interobserver agreement and procedural reliability
The interning teacher used data collection sheets to record correct and incorrect sight words that were read correctly during each assessment. Approximately half of all assessment and instructional sessions across all phases were recorded using a digital recorder for the purposes of evaluating interobserver agreement (IOA). The special education teacher, who had extensive experience with data collection, was familiar with the students, was trained on research procedures and served as the second observer for the purposes of evaluating IOA. The second observer listened to audio recordings for each participant: one during baseline (33.33% for all participants) and at least one third of sessions during intervention sessions for each participant (36.84% for Alice, 40.0% for Kayla, 46.67% for Dustin). The second observer recorded data on student responses using an identical data sheet used by the interning teacher. Data were then compared for each response noting the number of agreements and disagreements. The number of agreements was divided by the number of agreements plus disagreements to yield the percentage of IOA. While conducting IOA, the mentoring teacher also conducted procedural reliability for each intervention session. This was done by a checklist of intervention procedure steps on which the mentoring teacher indicated correct performance. The checklists were different based on the CASP condition having fewer steps (e.g., set up computer, remind student to participate with the activity) than the TDSP condition (e.g., remind the student how to participate, provide flashcard and verbally state the word, reinforce or correct student responses, etc.). The number of steps performed correctly was then divided by the total number of steps to yield a percentage of procedural reliability.
The special education teacher, the paraprofessional, and the three participating students all completed Likert-type questions regarding the social validity of the two instructional conditions: TDSP and CASP. The teacher and paraprofessional were asked to respond to questions about the ease of implementation, effectiveness of the interventions, time efficiency, and likelihood of the interventions being used again. The adults ranked questions by circling one answer for each statement: strongly agree, agree, undecided, disagree, or strongly disagree. The student version had four questions that were read aloud to them by the researcher: (a) two addressed whether or not they liked each instructional condition, and (b) two addressed whether or not they felt they learned more words with each instructional condition. For each question, students were provided three graphics of faces to circle their responses: (a) a face that appeared to be smiling for agree, (b) a neutral face for undecided, and (c) a face that was frowning for disagree. Participants were given a sample item and trained on the meanings of the faces.
Interobserver Agreement and Procedural Reliability
Interobserver agreement and procedural reliability were assessed as described previously. Interobserver agreement was 100%. Procedural reliability was 97.63%.
Effectiveness of the Instructional Conditions
Table 3 displays the participants’ mean percentages of correct responses during each condition. A report of the individual student results follow. According to What Works Clearinghouse (Kratochwill et al., 2010), quality indicators for single subject include visual inspection (e.g., trend, level, immediacy of change) as well as measures indicating effect sizes (e.g., percentage of non-overlapping data, percent exceeding median). Graphic representation, along with explanation of data, is included along with a report of percentage of non-overlapping data (PND). When applied in an alternating treatments design, PND determines relative effectiveness of one condition over the other (Gast, 2010). To determine PND in an alternating treatments study, data point values for each condition are compared. For example, data from the first session of each condition are compared to each other, data from the second session of each condition are compared, and so forth. Gast stated, “Calculation of percentage of non-overlapping data (PND) between conditions being compared is how replication of differences is quantified” (p. 352). For each comparison, the superior condition is noted. The number of comparisons in which one condition is superior is then divided by the total number of comparisons (superior + inferior + equal) to yield a percentage of non-overlapping data (Gast, 2010). Although there are limitations to the utility of using overlap methods for conducting syntheses of single-subject data (Wolery, Busick, Reichow, & Barton, 2010), these measures can be useful, with caution, to examine the immediacy of effect when comparing intervention conditions as in the present study. A report of the PND for each participant follows in Table 3.
During baseline, Alice did not read any of the target sight words correctly. During TDSP, the mean percentage of words read correctly increased to 53.85%, and Alice reached criterion after 12 sessions. During CASP, her mean percentage of sight words read correctly increased to 15.38%, and she did not reach criterion with this condition. Because Alice was unable to reach criterion after 13 sessions with the CASP, the TDSP procedures were implemented to teach those target sight words. Alice’s mean percentage of sight words read correctly in the preferred treatment phase to teach the words she did not meet criterion on was 53.33%, and she required six sessions to reach criterion. Percentage of non-overlapping data was calculated to determine the superiority of each condition. Alice participated in 13 sessions of each intervention. Session-by-session comparisons (as described previously) demonstrated that TDSP was superior 12 out of 13 times, thus PND for TDSP was 92.31% indicating a high level of superiority compared to CASP for Alice. See Figure 2 for Alice’s results.
During baseline, Kayla did not read any of the target sight words correctly. During TDSP, the mean percentage of words read correctly increased to 53.33%, and Kayla reached criterion after 12 sessions. During CASP, her mean percentage of sight words read correctly increased to 31.67%, and she did not reach criterion with this condition. Because Kayla was unable to reach criterion after 12 sessions with the CASP, the TDSP procedures were implemented to teach those target sight words during the preferred treatment phase. Kayla’s mean percentage of sight words read correctly in the preferred treatment phase was 84.0% and she required five sessions to reach criterion during the preferred treatment phase. Kayla participated in 12 sessions of each intervention. To determine PND, session-by-session comparisons were made. Comparisons demonstrated that TDSP was superior nine out of 12 times, thus PND for TDSP was 75% indicating a fairly high level of superiority compared to CASP. See Figure 3 for Kayla’s results.
During baseline, Dustin did not read any of the target sight words correctly. During TDSP, the mean percentage of words read correctly increased to 68.89% and Dustin reached criterion after nine sessions. During CASP, his mean percentage of sight words read correctly increased to 77.78% and he reached criterion after nine sessions. Dustin reached criterion under both conditions after the same number of sessions and his mean percentages were close; therefore, he was able to choose which condition to use for his preferred method. Dustin chose TDSP. He was assigned five new sight words with this condition. Dustin’s mean percentage of sight words read correctly in the preference phase was 70.0% and he required six sessions to reach criterion. PND for Dustin revealed different findings than for the other two participants. Dustin participated in nine sessions for each intervention. During four sessions, CASP was superior, TDSP was superior in one comparison, and performance was equal in four comparison sessions. PND was calculated for CASP with 44.44% superiority, thus indicating moderately better performance in this condition. See Figure 4 for Dustin’s results.
On the social validity instrument, the mentoring teacher and paraprofessional rated each of the questions about ease, effectiveness, and timeliness for CASP with agree and strongly agree. They also answered the reverse questions about CASP being too much work or ineffective with strongly disagree. All three participating students circled smiling faces for every question, indicating that they enjoyed learning the new words with both the teacher and with the computer. Overall, these results indicate that both TDSP and CASP were enjoyable and viewed as effective for the participants.
The purpose of this study was to determine if there was a difference between the acquisition of sight words for elementary school students with ID when taught with teacher-directed simultaneous prompting or with computer-assisted simultaneous prompting. In terms of students learning sight words, both interventions were effective; students made gains in both conditions. However, for two of the students, TDSP was more efficient than CASP because the participants reached criterion in fewer sessions, thus intervention was stopped for the other condition prior to criterion being reached. For the other student, both interventions were equally efficient, but he preferred the TDSP. During the preferred phase (i.e., best alone), all of the participants made gains quickly under TDSP. This further indicates that TDSP an effective instructional procedure for all three participants.
Although specific reasons for differences in performance cannot be gleaned from findings, it is worth mentioning that the type of responding differed for each condition. With TDSP, students read the word aloud to the instructor and received immediate verbal feedback. During CASP, participants selected the target word from three choices on the computer screen. Although both types of responses required active participation, reading the word aloud was similar to the response topography during the assessment component for the study. Additionally, with CASP, there is a possibility that the participants were visually matching the word rather than reading the word. Nonetheless, all participants made some gains in the CASP condition showing that it has merit for further study.
Consistent with other existing research on prompting methods, this intervention showed SP helped the participants to acquire a skill in a timely manner. This study matches the findings of Singleton, Schuster, Morse, & Collins (1999) in that both found SP to be an effective way to teach sight words to student with ID. In addition, this study supports Hudson, Browder, and Wakeman (2013) in that the participants were able to increase their sight word vocabularies when instruction was presented using systematic prompting.
Results of this study support the findings of Coleman et al. (2012); some participants with ID performed better with teacher-directed response prompting procedure whereas some participants performed equally well with computer-assisted response prompting procedure. This study enhanced the research that is currently available in special education about sight word instruction for students with ID. No previous studies were found that directly compared the use of computer-assisted SP to teacher-directed SP. The results of this study should serve as a caution to the field: Despite the push for more computer-assisted instruction in today’s classrooms (Collet-Klingenberg, 2009), computer-assisted instruction should not been viewed as a superior instructional approach. Even when one of the participants was able to reach full mastery on the computer-assisted SP instruction, he still told the instructor that he preferred the instruction and feedback from her.
Although the results of this study provide support for the effectiveness of computer-assisted and teacher-directed sight words instruction for students with ID, this study has several limitations to be addressed. One limitation, inherent to comparing teacher-delivered instruction with computer-delivered instruction, is the type of student responses used in this study. Because a teacher can hear and respond to a student’s oral reading performance whereas Classroom Suite cannot, the type of student responses were different. Although this poses a threat to internal validity in terms of comparison of behaviors, both behaviors are discrete forms and this does represent the way software works in an actual classroom situation. There is a growing trend in education, including special education, to provide technology-assisted instruction using computers or tablet devices. Sometimes, technology is used to replace teacher instruction for delivering content the initial time it is presented. Although technology has many benefits, the authors of the current study do not believe it should be used to supplant quality instruction. Given more advanced (i.e., more expensive and complicated) software that could recognize oral responses and respond exactly as a teacher would respond, a more direct comparison could have been made in this study. However, the piece of software used is more representative of a software package actually available for classroom usage. Responses between technology-implemented instruction and teacher-directed instruction inherently differ and may be part of the reason that teacher-directed instruction remains a superior solution for many students.
Another limitation is in the use of percentage of non-overlapping data to synthesize data. Wolery et al. (2010) stated, “the overlap methods fail to detect all of the characteristics of time series data (i.e., trend and variability), are compromised by the number of data points in the second condition, are not an estimate of the magnitude of effects, and do not use the replication logic inherent to single-subject research” (p. 25). An additional area to note is that there may have been carry-over effects from one treatment condition into the preferred treatment condition. Only one participant completed the preference phase with a new set of five sight words, so the replication of effect was only present for the original words for the other participants. With regard to social validity, the results must be taken with caution. Because of the learning characteristics of the participants, the fact that they responded favorably could indicated that the participants either did not clearly understand the task or wanted to please the researcher since she was the one who read the social validity items to them. Last, due to time constraints, no maintenance or generalization phases could be completed in this study. To draw conclusions about the larger implications and lasting effects of each type of instruction, these phases would need to be completed.
Implications for Future Research
Future research is needed in order to extend and support the findings of this study. Specifically, future research is needed in order to assess the effectiveness of teacher-directed sight word instruction and computer-assisted sight word instruction as well as the difference in sight word acquisition time for both conditions when compared to one another. In addition, how individuals’ cognitive and developmental profiles impact on the differences in the degree of effectiveness in TDSP and/or CASP sight word instructions would provide better understanding of who can benefit from which instruction. Additional research could be done to determine the effectiveness of a combination of teacher-directed instruction with computer-assisted instruction for sight words. Having the double exposure to a set of sight words may have more positive effects than having the conditions used separately. Research on using the computer as a maintenance or generalization tool could also be conducted to determine the effectiveness in that manner. Also, future research is needed to determine which of the conditions leads to maintenance of the acquired sight words over time.
Implications for Practice
The findings in this study have several implications for teachers. Although computer programs, iPads, and iPods are becoming more and more popular and common in the classroom, these results showed that students still benefit and need teacher-directed instruction and prompting in some cases. Even if the students are benefitting from computer-assisted instruction, technology-based instruction should supplement, not supplant quality instruction for students with ID. For example, teachers should introduce new content and provide supervised practice with the content prior to having students practice skills using technology. This would have the benefit of (a) superior instruction provided by a teacher, (b) additional practice without requiring one-to-one teacher time, and (c) a variety of response formats that might help students increase learning and/or generalizing the skill. The results of this study also showed that it is important to solicit student input on instructional procedures. When teachers take into account not only new technology, but also student preferences, they are opening the door to many more opportunities for student independence and self-determination.
Another implication for teachers was the use of time in the classroom when it comes to sight word instruction. Creating the individualized computer activities for the different participants was time consuming initially. However, once one activity was created from scratch, it was less time consuming to edit for other students or for different word arrangements. Also, once the students were trained on the technology, the computer-assisted instruction required no time from the teacher; the students were able to work independently and were exposed to their sight words while the teacher worked with other students. For the teacher-directed instruction, the time put in up front was minimal–only requiring the preparation of data sheets and flashcards. That being said, implementing the teacher-directed instruction took much more time since the teacher played an active role in the exposure of the sight words for the students. Students with ID often require one-on-one instruction. Computer-assisted instruction is one way to give these students a way to practice skills more independently. The results of the present study highlight the importance of critically examining the utility and effectiveness of technology-based instruction for individual students. Although technology can help increase motivation and expand learning opportunities, technology-based instruction should never be used as a replacement for high quality teacher instruction.
Mari Beth Coleman, University of Tennessee; Rebecca A. Cherry, Knox County Schools; Tara C. Moore, Yujeong Park, and David F. Cihak, University of Tennessee.