Using Functional Analysis to Improve Reading Instruction for Students with Learning Disabilities and Emotional/Behavioral Disorders

Abstract

Addressed to teachers of students with learning disabilities and emotional/behavioral disorders, this article describes an approach to reading instruction in which teachers use principles of functional analysis to generate hypotheses for the causes of reading failure. Teachers then use those hypotheses to guide the selection and implementation of interventions to improve reading success. Five possible causes of reading difficulties are described, and each is accompanied by suggested interventions for reading.

Reading may be the most discussed topic in American education today. Reports that alarming numbers of children are not learning to read are the impetus for this discussion, and the never-ending argument about how to teach reading is at the center of the debate. The Nation’s Report Card on the reading skills of fourth graders for the year 2000 states that 63% of the students read at the basic skills level, and only 32% read at the proficient or expected level (Donahue, Finnegan, Lutkus, Allen, & Campbell, 2001), These students are unlikely to catch up in their skills in the normal flow of elementary school curriculum. The literature reports that first-grade students with reading difficulties will most likely remain poor readers (Torgesen, 1998), and 74% of third graders who are behind in reading still have difficulties in the ninth grade (Lyon, 1995). Of concern to special educators are those children who have disabilities that make learning to read especially difficult. It is estimated that 80% of students with learning disabilities struggle with reading (Lyon, 1995), and students with emotional/behavioral disorders often have disabilities in reading as well (Coleman, 1996).

Although reading instruction has always been a cornerstone of the elementary curriculum, the nation’s teachers have never reached consensus about how to teach this essential skill so that all children can learn. This ambivalence may have its biggest impact on early learners. Many students who are eventually identified as having reading disabilities may have deficits due to inadequate or inappropriate instruction during the beginning reading years, rather than actual dysfunctions in the psychological processes used for reading (Carnine, Silbert, & Kameenui, 1997). These students may seem to have learning disabilities when they really fail to read because of ineffective instruction. For students with emotional or behavior disorders, the difficulties are compounded by inappropriate or unproductive learning behaviors. Teachers need to consider the reasons for student failure to plan and implement instruction for students who are not learning to read. We propose that the principles of functional analysis commonly used in behavior therapy can be used for this purpose.

Functional analysis can be described as procedures to determine the function of behavior; the reason a behavior occurs can often be discerned by discovering what triggers it and what reinforces it (Schill, Kratochwill, & Gardner, 1996). In work with students with emotional and behavior disorders, functional analysis is used to study inappropriate or unproductive student behavior, to design and implement behavioral support plans that include interventions to alter the behavior, and to assess the outcomes of those interventions. The process involves defining the behavior; observing its frequency, intensity, and duration; and then choosing and implementing one or more interventions, based on a hypothesis about the function of that behavior that can be used before or after the behavior to alter it. Data are collected and analyzed to ascertain the effectiveness of the intervention and to make decisions about continuing or altering the hypothesis and the intervention (Artesani, 2001). In short, functional analysis is used to determine what the student achieves by engaging in the behavior and what external factors might be manipulated to change the behavior.

Researchers in school psychology have proposed applying these same principles of functional analysis to academic difficulties. Daly, Witt, Martens, and Dool (1997) describe a model in which five possible hypotheses for academic deficits can be tested as functional reasons why students fail. These authors then propose instructional and behavioral interventions that can be used before and after academic performance to address the difficulties defined by each hypothesis. The five hypotheses for student academic failure are the following:

• They do not want to do the task.
• They have not spent enough time doing the task.
• They have not had enough help to do the task.
• They have not had to do the task that way before.
• The task is too difficult (Daly, Witt, Martens, & Dool, 1997 p. 556).

Each of these hypotheses can be tested, and each can provide valuable information to teachers about how and why students are not learning. The following sections describe application of the model to reading instruction.

1. They do not want to do the task.

   Teachers have long known that students may fail because they lack the skills, or they may fail because they lack motivation to try. The hypothesis that students are unsuccessful because they are not motivated can be tested by providing incentives to work. If a student is successful when a salient incentive is available, then appropriate skills are obviously in place. If the student is not successful, then either the student does not have the necessary skills or the incentive was not motivating enough for that student.

   Reading is motivational when the material is interesting to the students and when they are successful. The conventional wisdom is that reading instruction should take place at each student’s instructional level, that being the level at which a student is 91-95% successful at word recognition and 75% successful at answering literal comprehension questions about the reading. If students are routinely more successful than this, then they are at an independent reading level and are not likely to gain new reading skills. If students are less successful than the instructional level, then they are struggling with too many words and are likely to become frustrated. If students are working at the right level with interesting materials, then Hypothesis 1 can be tested by offering appropriate and appealing incentives for effort and completion of work.

2. They have not spent enough time doing the task.

   This hypothesis refers to the amount of time students spend practicing a new skill. Research has shown the importance of high rates of active responding when students are learning or practicing new skills (Greenwood, Hart, Walker, & Risley, 1994). Active responding can be described as any response-whether spoken, written, or demonstrated in some other way-that is initiated by a teacher prompt and is observable to the teacher or other students. Active responding requires that students attend to the teacher and become actively engaged in practicing new learning. Active responding goes beyond the concepts of on-task behavior and engaged time to refer to the actual frequency of student responses that can be monitored by the teacher. To test Hypothesis 2, students should be observed to determine their rates of active responding. If rates are low, chances are that increasing opportunities to respond to teacher prompts will have a positive impact on learning.

   Learning to read requires significant amounts of teacher modeling and student responding, especially if students have learning difficulties. Beginning with pre-readers who are learning letter names and sounds and continuing with students who are learning higher level word recognition and comprehension skills, significant effects can occur when teachers model new skills and then prompt high rates of student practice responses (Carnine, Silbert, & Kameenui, 1997). One tactic for accomplishing this that also benefits students with attention deficits or other behavior difficulties is to teach students to use choral responses to teacher prompts. Choral responding, also called group or unison oral responding, serves several purposes. First, students are not singled out and can therefore risk responding without fear of failure. Second, students can be motivated as a group to respond enthusiastically and rapidly; and third, students who are unsure of the response receive immediate feedback from other responders and quickly learn the correct response (Reid, 1995). Teachers make efficient use of choral responding when they give high rates of praise to the group for correct response, or when they re-teach following incorrect responses. The overall effect is that students are so involved in successful learning that they have little time for unproductive behavior. These high rates of practice responses prepare students to be successful when they move to independent practice.

3. They have not had enough help to do the task.

   Sometimes students are unsuccessful because they have not had enough help to master the required skills. The teacher can promote mastery by manipulating two aspects of instruction: providing feedback and matching instructional procedures to student skill levels. Hypothesis 3 can be tested by increasing or improving each of these aspects and measuring resulting changes in student learning.

   Teachers provide feedback when they either affirm or correct student responses, so that accuracy is achieved quickly and efficiently. When students respond correctly, the teacher should affirm the response by saying “Correct,” “That’s right,” “Good answer,” or something similar. When students offer an incorrect response or no response, the teacher should re-teach the skill, model the response, prompt the students to respond, and then provide the appropriate feedback for the response. High rates of feedback let teachers incorporate another essential aspect of effective instruction: high rates of praise. Research has verified that students who receive high rates of positive feedback learn faster, retain more, and have fewer behavior problems in class (Eisenberger & Cameron, 1996).

   The second aspect that can be manipulated is the match between instructional procedures and student skill levels. Increasing feedback by itself will not be effective if students are taught inappropriately or do not grasp the skills or concepts. Matching instruction to student skills requires that teachers understand students’ present levels of education performance. At the acquisition stage, teacher modeling with high rates of response and feedback is essential. As students achieve initial mastery of a skill, the instruction should focus on fluency, or accuracy at an appropriate rate. Fluency increases when students confidently practice a skill in authentic situations with high rates of success. Mismatches between instruction and student learning happen when students are asked to improve fluency in a skill that they have not mastered or when they are not practicing skills to achieve fluency.

   Each aspect of reading requires an appropriate match between instruction and student performance. Whether learning phonemic awareness, letter names and sounds, fluency, or comprehension, students need appropriate teacher modeling and feedback for responses and enough successful practice to become good at the skill.

4. They have not had to do the task that way before.

   Effective teaching and learning require a match between instructional methods, learning materials, and practice activities (Heward, 2000). If methods and materials do not compel students to practice skills in the way they will actually use them, then little mastery or generalization will occur.

   For example, imagine a teacher assigning students an illustrated one-page story to read and then asking the students to answer several questions, as a means to practice oral or silent reading comprehension. In this scenario, the students get all the right answers to the comprehension questions, but the teacher does not realize that they found them by looking at the illustration. In this case the students did not actually practice reading comprehension because it was easier to get the right answers in another way.

   This hypothesis can be tested by carefully reviewing instructional materials to determine if they help students master skills and then making changes in the materials if they do not. Researchers have determined that materials should provide sufficient examples and non-examples to help students differentiate between correct and incorrect responses, and should also require students to use the target skill to respond correctly (Carnine, Silbert, & Kameenui, 1997). Combined with Hypothesis 2, this means that students need to spend enough time practicing new learning to achieve mastery and must be practicing with the right kind of materials.

5. The task is too difficult.

   The final hypothesis is that materials are simply too difficult. As mentioned in Hypothesis 1, students are most likely to learn when they are challenged, but they should make errors on no more than 5 to 9% of responses. At this level, students continue to advance in new learning, without becoming frustrated with frequent failure. Matching materials to student skill levels can be difficult to accomplish in a general education classroom, given the normal range of abilities, but teachers must consider this hypothesis when planning for individual students. Teachers must continue to adapt and modify curriculum as indicated by individual students’ IEPs and make sure that students are working at their instructional levels and at a good pace.

Using the five hypotheses to improve reading instruction

The instructional considerations indicated in the five hypotheses can be used to troubleshoot a reading program to improve student outcomes. We introduce this section by reviewing the latest government research on what is required to learn to read, and we then suggest ways to use the five hypotheses to improve reading instruction.

The requirements for learning to read

In 2000 the National Reading Panel (NRP) published the results of a 2-year analysis of more than 100,000 studies of reading instruction. These results indicated that children must master skills in four areas to become good readers: phonemic awareness, phonics, fluency, and comprehension. Phonemic awareness is described by the NRP as the ability to hear and manipulate the sounds that make up spoken language; phonics is the understanding of the correlation between sounds and letters. Fluency is the ability to read with accuracy, speed, and expression, and comprehension is the ability to understand and enjoy what is read (National Reading Panel, 2000). When students are not successful readers, the five hypotheses can be tested for each of these reading skill areas to determine what changes should be made in reading instruction.

Improving phonemic awareness instruction

Phonemic awareness instruction consists of teaching students to hear and identify the individual sounds in words (Spector, 1995). This is a requisite for most students if they are to learn and understand the letter-sound correspondences inherent in phonics and thus become good decoders of the written word (Torgeson, Wagner, & Rashotte, 1994). Instruction usually involves segmenting words into sounds, combining sounds into words, and identifying and isolating sounds in like and unlike words. In the accompanying table we list ways the five hypotheses may be addressed for phonemic awareness.

Improving phonics instruction

Phonics in the NRP definition refers to understanding the relationship between sounds and letters that allows readers to decode and understand the written word. This skill is the primary method for recognizing words in print, and proficient word recognition is the foundation of fluency, comprehension, and reading enjoyment (Chard & Osborn, 1999). Phonics instruction teaches students to analyze words according to their sound-spelling structure and thus to learn to read more and more new words without assistance. Research has shown that, rather than becoming fluent because they skim over words, good readers actually pay close attention to each word as they read (Rayner & Pollatsek, 1989). Instruction in this essential skill can be evaluated and refined according to the five hypotheses, as shown in the table.

Improving reading fluency

Fluency is a combination of speed and accuracy and is a direct result of accurate practice. Students improve reading fluency as they practice reading stories at appropriate instructional or independent levels. The teacher’s role in improving fluency is to provide a wide variety of reading materials and then monitor students to make sure that they are not making and practicing errors. That can only be done by regularly listening to individual students read and keeping track of progress. Fluency is usually measured by the number of words read correctly per minute, but can also be more subjectively measured by listening to students read with expression. Both aspects of fluency can be assessed and improved through careful instruction, as described in the table.

Improving reading comprehension

Understanding and enjoying what is read are the purposes of reading. The skills from the NRP report mentioned above are the tools good readers use to make sense of what they read, and how they learn from and enjoy the printed word. Comprehension informs, and information fuels thought and self-expression through speaking, writing, and doing.

Literal, inferential, critical, and creative comprehension can all be assessed and taught. Literal comprehension refers to understanding what the author said; inferential comprehension is understanding what the author meant. Critical comprehension involves evaluating what the author said, and creative comprehension means adding to what the author said to extend or enhance the story (Reid, 1997). Instruction for comprehension can be evaluated and improved as described in the table.

Case examples of using the five hypotheses to improve reading

The following student case examples illustrate how the five hypotheses can be used to improve instruction for reading. Notice that the cases involve teachers collaborating to find the best ways to help their students.

Lydia. Lydia is a third grader with learning disabilities who has difficulty increasing her reading fluency. She can decode words at an acceptable level of accuracy but just doesn’t seem to increase the rate and accuracy of oral reading. She is healthy, attends school regularly, and lives in a stable family, so the difficulty seems to be with the methods being used to help her improve her reading skills. Her teacher began working with the special education staff to discuss ideas that might help Lydia. The group decided to consider each of the hypotheses to discover the most likely difficulty.

Lydia is a willing worker and seldom complains, so the classroom teacher ruled out Hypothesis 1. After a careful review of her test results and the difficulty of the materials, Hypothesis 5 did not seem to define the problem either. The instructional materials appeared to complement her assessed abilities.

The team then considered Hypothesis 2 to determine if Lydia was getting enough practice to increase fluency. Observation in the classroom revealed that Lydia had no opportunities to read aloud to someone who could time her and evaluate her accuracy. The teacher usually had students take turns reading aloud in groups but had no routine in place for regularly assessing and improving fluency. After some discussion with the team, the classroom teacher decided to include regular opportunities for students to time and record their reading rates, to encourage them to improve in this area. The teacher also began listening to each student read at least once each week. The team decided to evaluate the effects of these changes by monitoring improvements in Lydia’s fluency.

Jerry. When Jerry was 14, he experienced some serious emotional difficulties and was institutionalized in a state hospital for several months. When, he returned to the eighth grade in his home community, Jerry’s general education English teacher approached his resource room special education teacher to discuss Jerry’s difficulty reading eighth-grade-level literature. They agreed that a functional analysis of his reading difficulties would be useful in determining precisely why he was not successful in the general education literature curriculum. The teachers systematically considered each of the five hypotheses and narrowed their professional guesses to (a) he does not want to do the task, or (b) he has not had enough help to do the task, since inattention and lack of motivation had been related to academic failure for Jerry in the past. They decided that the task was too difficult because standardized tests suggested that Jerry had word attack and comprehension skills at, or just below, grade level. Most likely, his emotional difficulties were interfering with his learning. The first step for the teachers was to choose one of these hypotheses and test it. With help from his mother, they chose to let him earn time playing computer games after school in the resource room when he completed his work satisfactorily. They wanted to test whether lack of motivation was contributing to his reading difficulties. But this strategy did not increase the number of reading assignments Jerry satisfactorily completed. They then arranged for him to spend his daily homeroom time working individually with a volunteer. The personal relationship that developed from their interactions and the assistance Jerry received did result in his turning in acceptable assignments on time.

Conclusion

Functional analysis is a common process for assessing and addressing the behavioral needs of individuals. The same process used to identify and define a problem, choose a hypothesis, design an intervention, implement the intervention, and evaluate the results can be applied to reading instruction for individuals with learning disabilities and emotional/behavioral disorders. Inherent in using this process to improve reading instruction is the requirement for teachers to know their students’ present levels of reading performance and then to use methods and materials that match students’ instructional levels. Teachers should use functional analysis to determine if reading instruction is optimally efficient and effective and then employ research-based methods. These processes can help teachers avoid inadequate or inappropriate instruction during the beginning reading years.