Learning Styles in the Age of Differentiated Instruction

Abstract

The concept of learning styles has tremendous logical and intuitive appeal, and educators' desire to focus on learning styles is understandable. Recently, a growing emphasis on differentiated instruction may have further increased teachers' tendency to look at learning styles as an instructionally relevant variable when individualizing instruction in increasingly heterogeneous classrooms. We discuss the overlapping concepts of individualized instruction and differentiated instruction, briefly review the evidence base for learning styles, and argue that instruction should indeed be individualized and differentiated. We conclude that there is insufficient evidence, however, to support learning styles as an instructionally useful concept when planning and delivering appropriately individualized and differentiated instruction.

The idea that people learn things differently has tremendous intuitive appeal. It is not difficult to argue, for example, that among the myriad skills people master over their lifespan, some things are learned more quickly than others, skills are mastered with greatly varying amounts of practice, and the acquisition of some skills demands different types and levels of instruction and support. Moreover, different people learn to read, write, solve mathematical computation problems, hit a baseball, and bake a cake to hugely discrepant levels of success or mastery. An understandable outgrowth of this generally accepted logic is that humans must have some discernible way or method of acquiring information or mastering skills that suits them best: a learning style. In education, there has been no shortage of controversy about learning styles, with fundamental questions centering on quite basic issues. Do learning styles exist? Can learning styles be assessed and established reliably? If so, does the assessment of learning styles lead to instruction that serves students better?

A huge volume of literature appeared in the late 1970s through the 1990s regarding learning styles, with much of the literature focused on debate about whether science supports the construct and its utility for educators. More recently, the notion of learning styles has received perhaps unintended attention as the concept of differentiated instruction has become a mantra for schools and classrooms nationwide. Differentiated instruction, broadly defined as “varying instruction to meet the individual needs of all students” (Tomlinson, 1999), typically includes a focus on individual students' learning profiles. In most models, the term learning profile has come to include learning styles (e.g., Anderson, 2007; Tomlinson, 1999). Further, learning styles and modality-based instruction continue to work their way into the parlance of teacher education, particularly in practitioner-oriented journals that might be presumed to have greater impact on practice. This often takes the form of subtle endorsement of learning styles or modality-based instruction through suggestions that being aware of a student's learning style is necessary in order to individualize instruction (e.g., Murawski & Hughes, 2009; Regan, 2009). In this way, learning styles have clearly become a part of teacher lore.

Regarding students with disabilities, matching instruction to individual students' strengths and needs has been a hallmark—indeed a defining characteristic—of modern-day special education (Hallahan, Kauffman, & Pullen, 2009; Kauffman & Landrum, 2006). Students are identified with disabilities and provided with special education services when it is determined that they cannot receive an appropriate education through instruction offered to typically developing students. Special education, then, is “specially designed instruction that meets the unusual needs of an exceptional student” (Hallahan et al., 2009, p. 12). But does the constellation of unique learning needs of students with disabilities comprise a learning profile? And is part of that profile a learning style?

The purpose of this article is to explore the concept of learning styles, and specifically to discuss the extent to which learning styles represent an area of meaningful focus for educators charged with teaching atypical learners. In subsequent sections, we (a) provide an overview of terminology and overlapping concepts that may contribute to confusion regarding the importance of learning styles, (b) summarize literature reflecting debate over the empirical basis for learning styles, and (c) argue that while there are meaningful differences in how students should be taught based on their strengths and needs, the most important and instructionally relevant variables do not include learning styles.

TERMINOLOGY AND DEFINITIONS OF OVERLAPPING CONSTRUCTS

Educators have played a prominent role in the field of learning disabilities for nearly a century; prior to the early 1900s, the study and treatment of disabilities was primarily the domain of medicine (Hallahan, Lloyd, Kauffman, Weiss, & Martinez, 2005). According to Hallahan et al. (2005), multimodal instruction for students with learning disabilities—instruction geared toward a student's preferred or strongest learning modality—has been discussed by scholars for nearly as long, since at least the 1930s. Throughout the evolution of the field, ongoing debates about appropriate definition, identification, and service provision for students with learning difficulties have focused on broad and overlapping concepts such as individualizing instruction and matching instruction to individual strengths and needs. In the context of these discussions, a number of terms have been used, and new terminology has evolved recently that has potentially resulted in greater confusion about appropriate interventions. We consider two important and overlapping terms here: (a) individualized instruction and (b) differentiated instruction. Individualized instruction, in our view, represents perhaps the most fundamental and defining characteristic of special education, and has a long tradition in special education policy and practice. Differentiated instruction, in contrast, represents a relatively recent response to the growing trend of including students with disabilities in general education, which demands individualizing within increasingly heterogeneous classrooms. Whether learning styles should play role in individualizing or differentiating instruction remains controversial. We consider each concept in turn.

Individualized Instruction

Kauffman, Mock, Tankersley, and Landrum (2008) discussed the misperception that individualized instruction might imply little more than a one-on-one instructional arrangement. Instead, they note that individualization refers simply to the matching of instruction to individually identified needs. Indeed, individualized instruction may be delivered one-on-one, to a small group of students, or even in the context of whole-class instruction. Hallahan et al. (2009) credited Itard and Seguin, regarded as among the first pioneers of modern special education, with promoting the fundamental ideas that formed the basis of individualized instruction. Hallahan et al. suggested that individualized instruction refers to instruction in which “the child's characteristics, rather than prescribed academic content, provide the basis for teaching techniques” (2009, p. 25). For students with identified disabilities, a full educational evaluation should lead to the design of instructional programs that target individual students' strengths and needs. This is typically accomplished through the Individualized Education Program (IEP) development process. We see at least two key ideas at work in the IEP development process. One involves determining what to teach, and the other involves determining how to teach. Determining what to teach involves assessing children's skill sets across academic and preacademic (and vocational and pre-vocational), social-behavioral, and functional skill domains. Matching the skills and strengths children bring to bear with their life, vocational, and independent living goals provides a framework for planning an instructional program.

Once students' present levels of achievement and skill strengths and needs are established relative to their goals, a second purpose of the IEP process is to document accommodations that are necessary to make learning appropriate and accessible to students with disabilities (e.g., Haager & Klingner, 2005). Increasingly, this is accomplished in the general education classroom, with appropriate supports, modifications, and accommodations provided in a way that maximizes the extent to which students with disabilities are educated in general education environments with their nondisabled peers. When an appropriate education cannot be achieved in the general education classroom, even with these supplementary supports and services, placement in a different educational environment for part of the day is considered. Regardless of setting, accommodations to typical assessment and instruction are generally necessary to meet the individual needs of students with disabilities. Hallahan et al. (2009) defined accommodations as “changes in the delivery of instruction, type of student performance, or method of assessment which do not significantly change the content or conceptual difficulty of the curriculum” (p. 64). Accommodations generally include such things as allowing the use of a calculator, reading a test aloud to students, creating assessments with fewer tests items, and providing extended time on assignments.

Providing individualized instruction through the use of accommodations for students with disabilities is required by law, but we would argue that instruction needs to be individualized for many students, including any who struggle in a given domain or academic content area. Instructional materials, grouping arrangements, and instructional interactions themselves (e.g., questioning, corrective feedback, repetition, nature and intensity of prompts) are often adapted based on needs identified through prior assessment and observations of classroom performance.

To summarize, instruction is individualized when (a) it is planned in a way that builds on what individual students currently know and can do and targets meaningful goals regarding what they need to learn next; and (b) accommodations and modifications to teaching and testing routines are made in order to provide students with full and meaningful access to the content they need to learn. While we argue that these two components form the basis of individualized instruction, they can also be seen as critical building blocks for the more broadly applied concept of differentiated instruction, through which teachers seek to maximize learning in classrooms that are increasingly heterogeneous.

Differentiated Instruction

Differentiated instruction is a pedagogical approach to teaching and learning for students of differing readiness levels, interests, and modes of learning within the same classroom (Stradling & Saunders, 1993; Tomlinson et al., 2003). As Stradling and Saunders (1993) stated, differentiated instruction is “the process of matching learning targets, tasks, activities, resources, and learning support to individual learners' needs, styles, and rates of learning” (p. 129). Tomlinson (1999) suggested that differentiated instruction is designed to provide various learning opportunities for students who differ in their readiness levels (what they know, understand, and can do in relation to the content), their interests (affinity, curiosity, or passion for a topic), and their learning profiles (which may be shaped by their intelligence preferences, gender, culture, or learning style). Tomlinson further suggested that by differentiating instruction, teachers can (a) challenge all learners by providing varied levels of difficulty, (b) vary the degree of scaffolding, and (c) vary the way in which students work. The intent of differentiated instruction is to maximize each student's growth and individual success by meeting each student where he or she is at the time and assisting them in the learning process. Differentiation is based on a set of beliefs that (a) students who are the same age differ in their readiness to learn, their experiences, and their life circumstances; (b) differences are significant enough to impact what students learn, the pace at which they learn, and the support they need from teachers; (c) students learn best when connections can be made between the curriculum and interests or life experiences; and (d) teachers should attempt to maximize each student's learning. Differentiated instruction is proactive, student centered, dynamic, and rooted in assessment. It also emphasizes multiple approaches to teaching content and the use of flexible grouping (Tomlinson, 1999).

Rock, Gregg, Ellis, and Gable (2008) explained the theoretical framework of differentiated instruction, based on Tomlinson's work, through four guiding principles and seven essential beliefs. The four guiding principals include

(a) A focus on essential ideas and skills in each content area, (b) responsiveness to individual student differences, (c) integration of assessment and instruction, and (d) ongoing adjustment of content, process, and products to meet the individual students' levels of prior knowledge, critical thinking, and expression styles (p. 33).

Furthermore, Rock et al. (2008) described seven essential beliefs about differentiated instruction, again based on Tomlinson's work, which include (a) experiences in life and readiness to learn differ significantly among same-aged students; (b) these difference have a considerable effect on their learning; (c) students' learning is heightened when teachers challenge them beyond their independent level; (d) learning is more effective when related to real-life scenarios; (e) student learning is enhanced by authentic learning opportunities; (f) student learning is enhanced when they are respected and valued by their teachers, school, and community; and (g) the ultimate goal of education is to recognize and promote the abilities of each student. Based on these assumptions, teachers can differentiate instruction by adjusting (a) content, (b) process, and/or (c) products based on students' readiness, interest, and learning profiles (Tomlinson, 1999; Rock et al., 2008).

We find striking similarities between recent conceptualizations of differentiation and traditional special education for students with disabilities. Consider the framework typically associated with differentiation: the modification of content, process, and product. In the context of traditional special education models, content is modified for some students with disabilities based on strengths, needs, and appropriate goals for school, employment, and independent living. Instead of a traditional academic curriculum, the content that forms the basis of curriculum for some students with disabilities is modified to include functional skills, or vocational and prevocational skills, that other students may not need. Processes may be modified when it is determined that typical instructional methods and materials are not appropriate for some students with disabilities. Smaller instructional groups may be formed. Instruction may be made more explicit and more direct and may be delivered in smaller, more frequent doses. Finally, products may be modified in that students with disabilities might be required to complete different assignments or respond to alternate assessments to demonstrate their mastery of content.

Consider further the recommended bases on which teachers are encouraged to differentiate: students' readiness, interests, and learning profiles. For the first two of these, readiness and interests, the similarities between differentiation and special education are again striking. We take readiness to mean simply meeting students where they are, planning instruction based on careful and thorough assessment of what students know and need to learn next. Indeed this is a fundamental part of the IEP process. Matching instruction to student interests has long been a part of working with struggling learners, the most classic example being the need for reading material that appeals to older readers whose reading skills are far below their chronological age (e.g., “high interest, low vocabulary” books). Despite these similarities, we are equally struck by the final basis on which teachers are encouraged to differentiate: students' learning profile. As noted earlier, Tomlinson (1999) suggested that students' learning profiles may be shaped by intelligence preferences, gender, culture, and learning style. Given a history of significant controversy and debate, particularly regarding students with disabilities, the inclusion of learning styles in this list is of perhaps greatest concern when we contrast differentiation with special education. ¹

Learning Styles

The term “learning styles” has appeared in the education literature for at least 40 years (see Dunn & Dunn, 1979), although the concept itself has been controversial almost from the start in special education in particular (e.g., Dunn, 1983, 1990; Kavale & Forness, 1987, 1990; Kavale & LeFever, 2007; Lovelace, 2005; Stahl, 1999). According to Dunn (1983), learning style is based on the concept that individuals differ significantly in the way (or style) that they concentrate, absorb, and retain new information. This style, as Dunn (1983) described, “comprises a combination of environmental, emotional, sociological, physical, and psychological elements that permit individuals to receive, store, and use knowledge” (p. 496). Dunn and Dunn (1979) described 18 different elements involved in diagnosing an individual's learning style and suggested that the majority of individuals have between six and fourteen elements that strongly affect their learning style (Dunn, 1983). For example, under environmental elements, Dunn and Dunn (1979) and Dunn (1983) argued that individuals respond differently to instruction based on the temperature and lighting of the room in which instruction occurs and the formality of the physical environment of a classroom. In addition, Dunn (1983) argued that emotional elements such as motivation, persistence, responsibility, and structure affect the way that individuals respond to instruction, suggesting, for example, that “impersistent students often need ‘breaks’ while they are learning” (p. 498). The sociological elements are based on the idea that some students work better individually while others work better collaboratively. The physical elements focus around perceptual strengths, intake, time of day, and need for mobility. Regarding perceptual strengths specifically, Dunn and Dunn (1979) claimed that 20% to 30% of students appear to be auditory, 40% are visual, and 30% to 40% are either tactual/kinesthetic, visual tactual, or some combination of the four major senses. Based on this, they argued that when instruction is predominantly of one form (e.g., lecture, or lecture/discussion) teachers should not be surprised that “so few students achieve as well as we believe they should” (p. 240). Therefore, Dunn (1983) stressed the importance of matching instruction to students' perceptual strengths.

The psychological elements of learning style include global versus analytic learners, left-versus right-brain learners, and impulsive versus reflective learners. Therefore, the way teachers introduce lessons (sequentially for the analytic learner or describing the big picture for global learner, for example) should be based on the learning styles of individuals or groups of students. Further, the way teachers solicit feedback from students would differ; impulsive students, who will often call out answers, and reflective students, who will rarely volunteer information, require different instructional strategies according to Dunn (1983).

LEARNING STYLES: THE EVIDENCE BASE

While there are a number of models of learning styles and instruments for assessing them (e.g., Canfield & Lafferty, 1970; Gregorc, 1979; Kolb, 1981), the Dunn and Dunn model (1993, 1999) has received the greatest attention, especially in relation to students with disabilities. Perhaps most important in this regard was Dunn's (1983) description of her model of learning styles and methods of assessing learning styles, which appeared in Exceptional Children, arguably the most prominent research journal in special education. In that article, Dunn posited that learning styles could be established for at least two distinct groups of exceptional students: those who are gifted and talented and those who are “underachieving.” She noted that students “with low reading achievement preferred an informal environment when studying or learning; were adult-motivated rather than self-motivated; functioned best in the late morning; and preferred learning through their tactile and kinesthetic senses” (p. 501). She also noted in this paper “by 1978 we had revealed that many poorly reading children seemed to prefer low light” (p. 497). Finally, Dunn claimed that when students were taught with instructional strategies or materials that complemented their learning styles increased academic achievement, improved attitudes toward school, and a reduction in discipline problems occurred.

A second important paper in the learning styles debate was Kavale and Forness' (1987) report of the results of a meta-analysis that summarized 39 research studies examining the efficacy of modality-based instruction in special education. Kavale and Forness reported two major findings. First, the establishment of a modality preference yielded an effect size of .512, suggesting that “on average, 70% of subjects demonstrating a modality preference could be differentiated clearly on the basis of their test scores while 30% could not” (pp. 231–232). Second, teaching to a preferred modality resulted in an average effect size of .144, suggesting that, in general, modality-based instruction resulted in a gain of 6 percentile ranks. Kavale and Forness concluded that “no appreciable gain was found by differentiating instruction according to modality preference” (p. 238).

Dunn (1990) responded with a critical analysis of the Kavale and Forness (1987) meta-analysis, citing what she claimed were numerous examples of their “bias,” but it is worth noting that the original Kavale and Forness meta-analysis was not a review or evaluation of Dunn's work per se, but merely a comprehensive review of studies of modality-based instruction. In her critique, however, Dunn termed their search and circumscription of the literature base “capricious at best” (p. 352), noting that only two of her own studies had been included. Kavale and Forness (1990) responded to this specific criticism by noting that “when even a cursory examination revealed a study to be so inadequate that its data were essentially meaningless, it was eliminated from consideration. This is the reason that only two of Dunn's studies were included in our analysis” (p. 358).

Subsequent to these exchanges, the debate continued with a meta-analysis by Dunn, Griggs, Olson, Bailey, and Gorman (1995); a critique of this meta-analysis by Kavale, Hirshoren, and Forness (1998); a meta-analysis by Lovelace (2005); and a critique of Lovelace by Kavale and LeFever (2007). These exchanges reiterated many of the earlier concerns noted by Kavale and Forness (1990) regarding statistical issues or matters related to the interpretation of effect sizes. For example, although Dunn et al. (1995) reported a mean effect size of .755 based on a review of 36 studies, Kavale et al. (1998) noted that the conclusion by Dunn et al. (1995) that interventions implemented over the course of a year or more had greater effect, with a mean effect size of 1.345 across two studies, was flawed because the effect size for a typical child receiving typical instruction for one year would be 1.00, so those studies whose mean effect size was 1.345 a year or more later probably had much smaller true effects. Beyond these technical concerns, we find more troubling the methods used by Dunn et al. (1995) and others to locate studies for their meta-analyses: “The search process began with the identification of descriptors for a computer-based search of the Dissertation Abstracts International, Research in Education from 1980 to 1990” (p. 355). We wonder specifically why a search for studies would begin with a search for dissertations. Indeed the literature ultimately retained for review included 35 dissertations and 1 published study. Additional sources searched were the Annotated Bibliography of Research (1992, 1995) and Research on the Dunn and Dunn Model (1992, 1995), both of which are documents produced by the Center for the Study of Learning and Teaching Styles at St. John's University. Not surprisingly, given the search methods used, 20 of the dissertations retained for review were from St. John's University, where R. Dunn was a faculty member. The one published study examined the impact of learning style preferences (specifically perceptual preferences) on employee training effectiveness, and was published in Human Resource Development Quarterly. One can conclude logically from this search process and the studies located that either Dunn et al. did not conduct a full and comprehensive literature search, or that very, very few school-based studies of learning styles have been published in educational journals (indeed none in the body of literature used for this meta-analysis).

While Kavale and colleagues have been perhaps most prominent in their critical appraisal of learning styles' empirical basis, others have concurred with their general conclusions. Stahl (1999), for example, was particularly skeptical about the value of learning styles in the context of reading instruction, and has taken particular exception to arguments that have been advanced by Çarbo (1983) with regard to students with disabilities. Çarbo argued essentially that reading achievement is dramatically improved when reading programs match students' learning (or reading) styles. Çarbo (1988) argued that phonics instruction was neither necessary nor effective for teaching students to read. While Çarbo later conceded that phonics instruction is a necessary component of good reading instruction, she continued to caution that phonics “is most appropriate for students whose reading styles match the phonics method” (2005, p. 48).

Partly in response to Çarbo's writings, Stahl (1999) provided an overview of research reviews on learning styles from which he concludes “one cannot reliably measure children's reading styles and even if one could, matching children to reading programs by learning styles does not improve their learning” (p. 2). Furthermore, Stahl (1999) highlighted the problem of citing a preponderance of unpublished studies when promoting learning-styles–based instruction. In summary, despite a wealth of published papers espousing learning styles, there remains a dearth of published research in support of matching instruction to learning styles. The only reviews of which we are aware that provide support for a learning styles model (Çarbo, 1983; Dunn et al., 1995; Lovelace, 2005) rely heavily on unpublished reports (which lack the check-point of peer review), and too often include a preponderance of unpublished dissertations from a single university.

We note further that quality indicators for research in special education were published in 2005 (see Odom et al., 2005), and increasingly scholars are attending to these indicators both in publishing their own studies and in producing quantitative syntheses of literature (see Cook, Tankersley, & Landrum, 2009). We encourage future reviewers to attend carefully and more explicitly to the methodological quality of studies when reviewing learning styles literature, and we encourage readers to evaluate existing studies of the impact of teaching based on learning styles with an equally critical eye toward methodological soundness.

APPROPRIATELY MATCHING INSTRUCTION TO INDIVIDUAL NEEDS

If instruction is to be effective, it must be matched to individual needs. Where we differ with those who would espouse learning style as a relevant instructional variable is in delineating individual needs that are instructionally relevant. Consider the example of reading. We know that to become proficient readers, students must learn the letters of the alphabet and the sounds those letters make (e.g., Snow, Burns, & Griffin, 1998; National Reading Panel, 2000). This is followed by an essentially developmental sequence in which students must master increasingly sophisticated phonological skills and acquire orthographic knowledge as they learn to segment spoken words into parts and blend parts of words together. Thus, faced with the task of teaching reading to a group of youngsters, a teacher must ascertain through careful assessment which skills a student has, which are emerging, and which are lacking entirely. Based on this, instruction may be individualized so that students are taught content at their instructional level. Note that large or small groups may be formed based on similar content needs, and one-on-one instruction may even be provided in small doses for students with particularly acute skill deficits.

Determining instructional levels, students' strengths and needs, and the level of content to be taught is only the first step in individualizing. Suppose a student has an identified disability—a learning disability or attention deficit disorder, for example. Teachers may need to further individualize in any number of ways. They might use direct instruction and provide repeated practice in learning to say the sounds in a word slowly (e.g., “m-m-m-a-a-a-n-n-n”) or to blend sounds together to form words (e.g., blending together the /a/ sound and the /m/ sound to form the word “am”) (e.g., Reading Mastery Plus, 2002). They might provide extended opportunities for distributed practice or offer more frequent or overt positive reinforcement for correct responses for students who do not respond readily or cannot attend successfully for meaningful periods of instruction, while allowing students who master skills more quickly to continue to move through the curricular sequence. Again, note that individualization carries the hallmark of deciding what children know and need to know, and then modifying instruction for those who struggle.

Differentiated instruction adds to the notion of individualization primarily in scope and breadth of application. Differentiated instruction goes beyond the basic concepts of individualization and provides additional guidelines for teachers dealing with a diverse classroom. As we have noted, differentiated instruction is a teacher's response to students' individual needs. Teachers can differentiate the (a) content that is being taught, (b) process that is being used to teach the content, and/or (c) learning product that is expected according to students' readiness, interests, and learning profiles through a variety of instructional strategies (Tomlinson, 1999). For example, content can be differentiated by providing text at varied reading levels or by providing examples based on individual students' interests. The process, or the way in which the content is taught, can be differentiated by using cooperative grouping activities and assigning different roles to students based on their learning characteristics or developing activities that seek multiple perspectives on the content being taught. Finally, products can be differentiated by tiered assignments or providing a range of formats for a final project (i.e. poster, presentation, or performance). All of these suggestions build on the underlying concepts of individualized instruction. Differentiated instruction merely promotes more overtly that educators must think about individualizing instruction in the context of content, process, and product.

Stahl (1999) supported this notion when he argued that although children are in fact quite different in their personalities and preferences, research demonstrates that these differences have little to do with how successful a reader or writer they will become. Stahl suggested instead that differences in exposure to oral language or written text are far more important and likely to impact the development of literacy skills. Therefore, Stahl suggested, “we ought to think about different methods being appropriate for children at different stages in their development” (p. 4). Tomlinson (1999; 2003) supported this notion by stressing the importance of differentiating based on students' readiness, but also stressed the importance of differentiating based on students' interests. Stahl cautioned against doing this all the time and stressed the importance of using different methods for different goals. He noted that

approaches that involve the children reading books of their own choice are important to develop motivated readers. But whole language approaches, which rely largely on children to choose materials they read, tend not to be as effective as more teacher-directed approaches for developing children's word recognition or comprehension (p. 5).

As noted previously, Tomlinson (1999, 2003) promoted differentiating instruction based on students' learning profiles (which includes the student's learning style). However, Tomlinson (2009) cautioned education leaders that students' learning profiles should not be used as a replacement for their readiness needs, and suggested that having the option to do an assignment via different learning modes (i.e. visual, auditory, or kinesthetic) will have little impact if the student is unable to read the textbook. Indeed, Stahl warned that educators trying to make this type of differentiation work without focused attention on readiness and needs may waste valuable instructional time and energy that could be better spent on implementing research-based practices. Based on this premise, Kavale and LeFever (2007) posed the question: “How does the Dunn and Dunn Model fare in the context of instructional effectiveness?” (p. 95) and reported that even if the findings from heavily criticized meta-analyses and reviews are accepted their model reveals modest efficacy when compared with other instructional methods (see Kavale, 2007 for a detailed description). Faced with increasing heterogeneity in classrooms, arguments to use proven or promising evidence-based practices over those with little to no empirical support or even modest efficacy seem more compelling than ever.

CONCLUSION

The history of special education has included a focus on individualizing instruction since its earliest days. While special education has failed to live up to its full promise by most accounts, there have been periods of significant progress and success in our history, and we know that special education done right can produce dramatic, positive results for students with disabilities (e.g., Kauffman, Bantz, & McCullough, 2002). We argue that special education that is true to its foundational concepts is most likely to meet with success.

Differentiated instruction represents the newest incarnation of schools' attempts to meet the needs of a diverse student population. Among the more prominent models of differentiation is Tomlinson's (2003) model, which proposes that teachers modify content, process, or products to meet the varied needs of students. We agree that differentiation is clearly necessary, especially in the context of increasing diversity in classrooms. The need to modify content, process, and product is predicated on variations in students' readiness, interests, and “learning profiles.” Acknowledging the need to differentiate, we nonetheless concur with Rock et al. (2008), who offered advice on determining students' readiness, interests, preferences, strengths, and needs. They suggested that while teachers might consider students' styles of thinking, they should “not confuse this with learning styles. In fact, we urge resisting the temptation to try to match instructional methods with students' preferred modalities because research does not support such a practice” (p. 35).

The intuitive appeal of learning styles probably means that debate will not go away, and changing the mindset of teachers and teacher educators with regard to learning styles is a herculean task. Perhaps science will ultimately carry the day. If so, it is our view that (a) the dearth of published research studies on learning styles-based instruction, (b) the preponderance of unpublished studies that form the “empirical basis” for learning styles, and (c) the prevalence of dissertations from a single university in this unpublished literature base will lead researchers, teacher educators, and ultimately teachers to the following conclusions. It is wise to individualize instruction. Differentiation provides one framework for individualizing in the context of a heterogeneous classroom. Focusing on students' learning styles adds little, if anything, of educational benefit to this process.

Notes

1. For the purposes of this paper, we ignore for now the concept of “intelligence preferences,” which we assume refers to the theory of multiple intelligences, although we are aware of no empirical evidence that this theory carries any instructional relevance for teachers (see Lloyd & Hallahan, 2007; Stahl, 1999; Willingham, 2004).