SPECIAL CONSIDERATIONS WITH RESPONSE TO INTERVENTION AND INSTRUCTION FOR STUDENTS WITH DIVERSE BACKGROUNDS

Response to Intervention (RtI) has been heralded as having significant promise for improving outcomes, reducing disproportional placement in special education, and making education more culturally responsive for culturally and linguistically diverse (CLD) youth. This article reviews the major recent empirical and qualitative research findings with respect to its impact, particularly with regard to CLD youth. Extant policies are reviewed, in addition to best‐practice recommendations by experts in the field of culturally responsive RtI. The major conclusion of this article is that despite some recent high‐quality studies, insufficient research currently exists to support full implementation of an RtI model with CLD youth or those who have individual diversity. In many cases, researchers need to minimally disaggregate their findings and more clearly explain their samples and methods, moving beyond providing simple ethnicity counts as proxies for culture. Future studies should include fully described contextual variables and consider the science to pedagogical practice gap. In addition, outcomes for youth need to be documented as part of assessing whether RtI is having the desired effect on producing a more equitable educational milieu. Finally, additional research is needed in subject areas beyond reading and should be conducted more frequently with CLD adolescents. © 2012 Wiley Periodicals, Inc.

Response to Intervention (RtI), sometimes called Response to Instruction, has significant promise for improving the educational opportunities of students who are culturally and linguistically diverse (CLD; McKinney, Bartholomew, & Gray, [27]; National Center for Culturally Responsive Education Systems [NCCRESt], 2005; Xu & Drame, [47]). One of these hopes is RtI's potential to reduce disproportionate placement of ethnic minority youth in special education. Historically, the disproportionate placement of these students as well as the disproportionate use of expulsion and suspension is very concerning (Artiles, Kozleski, Trent, Osher, & Ortiz, [1]; McKinney et al., [27]). Over the last 30 years (Klingner et al., [22]), in both case law and scholarly study, the focus has been on successful and equitable special education assessment and placement. Equitable assessment and placement are still complicated by subjective, "soft" special education eligibility categories that involve clinical judgment in areas such as emotional disability, specific learning disability or mild mental retardation/cognitive disability. Although RtI could potentially provide additional evidence regarding the presence of more subjective disabilities (e.g., specific learning disability), some researchers question whether RtI will simply shift children into different "soft" categories rather than reducing overrepresentation more generally (Ciolfi & Ryan, [9]). Others are troubled that students such as English learners (ELs) may currently be misclassified within special education placements (Zamora, [48]). For example, García and Ortiz ([15]) assert in their review that ELs tend to be underrepresented in the early grades for specific learning disabilities and overrepresented in later years (see also Vaughn, n.d.).

Variation in laws across states and with regard to RtI's place in general education further complicates its role in addressing disproportionate placement. Thus, the impact of RtI on disproportionate placement may be quite different from state to state. Specific to disproportional placement, there is concern that each state has its own requirements governing whether RtI is included and/or required in special education evaluations or within individual special education categories (Zirkel, [49]). Therefore, to consider the effects of RtI on equitable assessment and placement, analysis at the state level may be most appropriate. For example, California has one of the largest concentrations of Latino students and significant numbers of other diverse populations (U.S. Census Quick Facts, [43]). A recent study (Bouman, [5]) involving a survey of 142 school districts in the state found that those districts that had implemented RtI did not have significantly lower placement rates than non‐RtI districts. African Americans were overrepresented in the emotional disability, developmental delay, and mental retardation/cognitive disability categories. In the respondent districts, African American, Latinos, and Native Americans were overrepresented in the specific learning disability category, whereas the weighted risk ratio for European Americans was decreased. Asian Americans were significantly underrepresented, even in school districts that used RtI. Interestingly, the use of RtI, traditional testing, the full implementation of an RtI model, the use of processing tests, or the provision of Tier 2 had no significant effects on placement rates. Although, overall, fewer children were placed for specific learning disabilities (the mean decreased from 4.91% to 4.05%, p <.01), disproportionate placement continued for children who were CLD, and it actually increased for African Americans over a 5‐year period (2002‐2003 through 2007‐2008), even after the institution of RtI.

In a detailed qualitative study examining the implementation and use of an RtI model in a school with students who were Latina/Latino ELs, Orosco and Klingner ([32]) found that using an RtI model gave false confidence that assessment (whether progress monitoring or formal assessment for placement) and instructional practice were aligned. Orosco and Klingner ([32]) have demonstrated that teachers too easily blur prereferral processes and RtI, as suggested by Kavale, Kauffman, Bachmeir, and LeFever (2008), and wrongly assume that their implementation removes bias from assessments and special education determinations. In addition, research (Bouman, [5]; Orosco & Klingner, [32]) suggests that the use of RtI with ethnic minority populations may not yet be developed to the point of resolving issues as complex as disproportional special education placement and culturally responsive alignment of RtI to determinations made about students in actual practice. Both studies reflect the difficulties of applying research to practice. Small‐scale, heavily controlled research studies are not the same as large‐scale implementation (Reynolds & Shaywitz, [36]) at the state or even the school level.

Mellard and Johnson ([28]) regard RtI as a method that could be used to organize all of the competing mandates and policies to which schools must adhere and thus make possible the achievement of policy coherence. They warn that policy incoherence is a prevailing problem, with many schools interpreting and implementing each mandate and policy with which they must comply separately or as an add‐on. For students who are CLD, added policies governing additional educational provisions for them, such as Title I and Title III, make the importance of seeking policy coherence all the more salient. Klingner and Edwards ([23]) further explain that strategies and frameworks need to consider sociocultural contexts to be considered culturally responsive as well as to consider "what works with whom?" (p. 112). Because the population of children who are CLD is growing exponentially, with some estimates suggesting that these students will be the majority as soon as 2023 (Frey, [13]; see also Passel, [34], for a discussion more specific to the demographics of immigrant youth), it is imperative that a framework such as RtI be investigated fully both for its promise and for its current limitations for implementation with students who are CLD and/or who have individual diversity, such as lower socioeconomic status. What follows is a tier‐by‐tier discussion of culturally responsive RtI centering on the current challenges and possibilities of moving research into practice. Highlighted for each tier will be current known limitations in the empirical data and methods of promise. Although significant limitations exist, RtI is a phenomenon mentioned in many state laws (Zirkel, [49]) and is already widely being used (Kavale et al., [21]), including with CLD students (García & Ortiz, [15]), despite gaps in the research base.

A solid RtI plan has a strong Tier 1 as its basis, and full implementation takes 5 to 7 years (Batsche, [3]). Tier 1, in theory, is supposed to support approximately 80% of the general school population in reaching basic achievement benchmarks (Batsche et al., [4]). If schools fail to provide quality instruction for 80% of all students as measured by achievement outcomes, making conclusions about a child's individual failure to respond rather than exploring an inadequate Tier 1 as potentially more causal would be very questionable practice (Burns, Jacob, & Wagner, [6]). To provide culturally responsive RtI, schools must ensure that any differences noted are due to the individual students and not systemic problems with providing culturally responsive teaching and instructional differentiation (Orosco & Klingner, [32]). Batsche ([3]) points out that it is important to disaggregate data and see if there are different predictable outcomes based on groupings such as ethnicity, EL, and other categories featured in No Child Left Behind.

A significant concern with using an RtI model is that the first tier is not always fully implemented or analyzed at the classroom level (Klingner & Edwards, [23]). Relatively little research has been done on ensuring that Tier 1, typically involving universal screening and differentiated, high‐quality evidence‐based instruction, is provided to all students (Xu & Drame, [47]). Even less research exists regarding culturally responsive provisions at Tier 1 (García & Ortiz, [15]; Xu & Drame, [47]). The efforts that exist almost exclusively focus on components of reading proficiency in the early grades (Mellard & Johnson, [28]). Studies featuring secondary students and RtI are conducted considerably less frequently (Vaughn et al., [46]). Several researchers (Orosco & Klingner, [32]; VanDerHeyden, Witt, & Gilbertson, [44]) have noted that with the EL population, most studies included interventions and programs implemented by the researchers rather than actual educators in the field attempting the provision of RtI.

An intentionally culturally responsive educational system must be built (Klingner et al. 2005) to include intensive professional development on culturally responsive effective practices that go beyond cultural appreciation activities (García & Ortiz, [16]). Training and professional development within the selected RtI model needs to be ongoing, consciously avoiding the "train and hope" approach (Barnes & Harlacher, [2], p. 427). Within Tier 1, it should not be presumed that students, including CLD learners, have sufficient background knowledge to use a packaged curriculum, so that teachers must strive to provide comprehensible input (Orosco & Klingner, [32]). To provide such comprehensible input, teachers must possess a deep understanding of second language acquisition, the relationship of native language proficiency to the development of English, assessment of proficiency in the native language and English, sociocultural influences on learning, effective first and second language instruction, informal assessment strategies that can be used to monitor progress, particularly in language and literacy development, and effective strategies for working with culturally and linguistically diverse parents. (Ortiz, 2001 , p. 2)

Teachers (Mills, [30]) and psychologists (Meyers, [29]) must examine their own views and practices around issues of social justice, race, and students from disparate backgrounds. Such an examination of social justice in educational practice cannot be piecemeal or an add‐on (Mills, [30]), and basic culturally responsive educator training in this regard is currently lacking (Xu & Drame, [47]).

Measurement Issues. Additional research is also needed to discern which outcomes should be used to measure success at Tier 1. Issues with the predictive validity of common universal screeners used with students across economic disadvantage, limited English proficiency, disability status, and race/ethnicity were noted in a recent, well‐designed study (Hosp, Hosp, & Dole, [20]). In this study, in which floor issues in the earlier grades were also noted, the authors caution psychologists against relying on the results of an individual screening measure. Haager ([18]) suggests that an overreliance on screening measures beyond their predictive utility with ELs noted in some early studies is not warranted. A possible partial solution or stopgap to this dilemma may be to supplement with local norms (Linan‐Thompson, Cirino, & Vaughn, [25]). Other recent research (Marchant, Hardin, & Paulson, [26]) has demonstrated that when disaggregated achievement data are considered across states, gaps in scores have different meanings in different states. Knowledge of an existing gap by itself is not very meaningful. Finally, as some have pointed out, most ELs are native born, many are second and third generation, and Tier 1 needs to further meet their needs (Zamora, [48]).

Other needs at Tier 1 involve students living in poverty. As Marchant et al. ([26]) have pointed out, the achievement gap associated with lower socioeconomic status is perhaps more robust and meaningful than the gap associated with race. They found that the poverty achievement gap was a better gauge of a state's general educational achievement than differences due to ethnicity alone. Growing numbers of students are living in poverty in the United States, and Rothstein ([39]) noted that minority and low‐income children might have more vision and hearing problems than their middle‐class peers might. A higher incidence of untreated or undertreated oral health problems, lead exposure, and asthma (Rothstein, [39]) may also affect school attendance rates and the child's ability to attend and respond to interventions. Rothstein also notes that interruptions in instruction caused by frequent moves can also be disruptive to educational progress. Building strong relationships with parents and caregivers and understanding the reasons children may not attend school regularly or be able to fully pay attention while there would also be important considerations in providing a comprehensive and culturally responsive Tier 1 design (Artiles et al., [1]). Particularly given the aforementioned link between economically disadvantaged students and hearing and vision problems, individuals with Tier 1 universal screening scores below the cut‐value should minimally have vision and hearing checked (i.e., do not wait for screening programmed by grade level) prior to choosing interventions (Burns et al., [6]).

The cultural responsiveness of classroom instruction is paramount, and Klingner and Edwards ([23]) recommend that good design should include findings from qualitative studies on effective diverse classrooms. With regard to culturally responsive classroom instruction, an important element in adequate Tier 1 design is that "research has shown the effectiveness of structured cooperative methods for English learners" (Calderón, Slavin, & Sánchez, [7], p. 113). Practitioners may wish to consider comprehensive and contextual formats that move away from a deficit model of understanding diverse learners. Methods of promise in this regard include frameworks such as the Sheltered Instruction Observation Protocol model (Echevarría & Graves, [12]) and Universal Design for Learning (sometimes termed Universal Design for Instruction; see Sopko, [42] and the CAST Universal Design website provided in Table 1). Comprehensive methods such as these, rather than piecemeal approaches and interventions, may better assist schools to achieve coherence (Mellard & Johnson, [28]) in providing systemic change in school culture and to promote movement away from deficit models toward implementation of culturally responsive RtI. These systemic factors may be most germane in developing culturally responsive RtI (García & Ortiz, [15]). Table 1 provides web resources for programs that are generally systematic and broadly conceptualized. It also provides resources appropriate for culturally responsive educators and researchers.

1 Helpful Websites for Culturally Responsive RtI

Although much of empirical RtI research, especially on CLD students, is focused on reading, it is important to note that researchers are still studying the mechanisms involved in the acquisition of reading and debating whether a cumulative language approach (language skills interact with phonological knowledge and literacy knowledge) or a phonological approach (vocabulary influencing phonological knowledge, which then directly relates to the acquisition of reading) provides the best predictive model (Poe, Burchinal & Roberts, 2004). Questions remain regarding what is best at Tier 1, especially for CLD children. Poe et al. ([35]) found that in an African American sample (N = 77) of young, low‐income students, the cumulative language approach was more predictive of achievement. Language skills, particularly in the second grade, and phonological skills, when children are first learning to read, were each important. Different models and strategies may be useful for different populations and subpopulations, age groups, or sociocultural contexts. The cumulative language approach, which is more contextual, may be more appropriate and predictive of positive outcomes for children from diverse backgrounds. If so, schools need to be equipped and able to provide a variety of programming and differentiation. Interestingly, Poe et al. also found evidence that access to high‐quality childcare or responsive, stimulating care in the home may positively impact reading skills, as it was predictive with her sample. Minimally, when designing Tier 1 and when going from Tier 1 to Tier 2, efforts should be made to incorporate the parents, including those who are CLD and/or have individual diversity. Others have also emphasized the importance of providing adequate CLD responsive interventions with parent involvement (García & Ortiz, [16]).

In Tier 1, it is important to consider factors that may impede instruction. As already discussed in the section on Tier 1 Limitations, sensory and health concerns must also be addressed. Another concern is the lack of instruction precipitated by disproportionate use of suspension and expulsion for ethnic minority youth. A promising method to address this concern would incorporate a culturally responsive positive behavior support program into the RtI process (Ciolfi & Ryan, [9]; Klingner et al., [22]; McKinney et al., [27]). Reynolds and Shaywitz ([36]) discuss creating a true primary prevention program by including both academic and behavioral screenings in their list of best practices for RtI. A culturally responsive behavior support program would include disaggregating data to determine whether interventions were beneficial for all school populations and would include contextual features embedded into the policies and daily practices of the schools. Highlighting this importance is a recent study by Skiba et al. ([41]). In this large‐scale examination of the discipline practices of 364 elementary and middle schools, they found that African Americans were two to three times more likely to be referred to the office than European Americans. In addition, Latinos and African American students were more likely to receive much harsher punishments than their European American counterparts were for the same infractions.

Batsche and colleagues (2006) have explained that Tier 2 serves approximately 15% of the student population not responding to Tier 1. Students in Tier 2 require additional instruction and targeted interventions. The authors specify that typically, the strategies utilized in Tier 2 are conceptualized through a standard treatment protocol, a problem‐solving model, or a combination of both. Students participate in the planned and progressed monitored interventions for a predetermined period before further educational decisions are made.

At Tier 2, a common concern is that RtI is often yoked to special education and eligibility, in particular, for the identification and placement of specific learning disabilities (Zirkel, [49]). More specific to Tier 2, Barnes and Harlacher ([2]), in delineating RtI as a set of principles with associated features, have specified that simply noticing a problem, providing intervention, and checking its progress may not be RtI. They suggest that true RtI must consider a systemic schoolwide proactive approach involving effective practices. In this view, random acts of intervention, devoid of context and lacking systematically applied prevention components, would not constitute RtI. Without a multitiered system based on evidence and research and without the policy coherence discussed by Mellard and Johnson ([28]), schools will need to consider whether they actually are providing a true Tier 2. Critics of RtI (Kavale et al., [21]; Reynolds & Shaywitz, [36]) have called for a distinction between RtI as a prereferral and prevention model versus RtI as an identification model. Critics tend to endorse the former rather than the latter.

Sanetti, Gritter, and Dobey (2011) comprehensively reviewed the extant literature in four major peer‐reviewed school psychology journals appearing between1995 and 2008. They focused their content analysis on reviewing treatment integrity issues. Of the studies suitable for their review, most focused on academics or behavioral disruptions as the dependent measure, and the majority focused on younger students. Most of these studies did not operationally define the intervention (independent variable), making it very difficult for those who wished to use them in their work or to replicate their findings. Some of the studies did provide a reference to the intervention used. One third of the studies inadequately defined the intervention used or failed to provide sufficient sources for the intervention. Only half of the studies assessed by Sanetti et al. reported quantified treatment integrity data.

For those working with CLD students, a further concern is the appropriate selection of intervention practices and methodologies developed and field‐tested with a variety of diverse learners. Klingner and Edwards ([23]) found that most research provides insufficient demographic data about the participants and suggest that studies report "information about language proficiency, ethnicity, life experiences (e.g., socioeconomic, specific family background, immigration status)... [and] data should be disaggregated to show how interventions might differentially affect students from diverse backgrounds'' (p. 111). They also report concern that ELs are often excluded from studies altogether because of limited English proficiency, yet findings are reported to be applicable to all students. Thus, few studies typically provide sufficiently disaggregated field‐testing. Additionally, populations such as ELs vary tremendously, even among students who have the same heritage language. For example, ELs have a host of English language levels, varying acculturation issues, presence or absence of culture shock, and other within‐group differences (Díaz‐Rico, [11], also see Collier's website referenced in Table 1) making it extremely difficult to ensure generalizability, even when proxy demographic information is used in place of culture, such as subject ethnicity (Klingner et al., [22]).

McKinney et al. ([27]) recommend using the standard protocol first in Tier 2 and then moving to the problem‐solving model for students who are CLD. As Barnes and Harlacher ([2]) discussed, the dual discrepant criteria to judge nonresponsiveness as described Fuchs and Fuchs ([14]) and associates may have the most value in determining the success of RtI with individual students. Although some research has been done in this area (see Linan‐Thompson et al., [25], for a discussion), there are lingering questions regarding the validity and reliability of screening and progress monitoring measures with students who are CLD. The extra measure of also monitoring the student's rate of growth (learning rate) during diagnostic teaching and intervention as described by Fuchs and Fuchs ([14]) would provide students who are CLD an additional measure of protection to assure that the correct determination was made. Linan‐Thompson et al. ([25]) have pointed out that ELs often were not always included when benchmarks were normed, so if only reaching the selected standardized benchmark were used to make determinations about responsiveness, they would be at a distinct disadvantage and possibly could be falsely identified as needing Tier 2 intervention. However, alternatively using the dual discrepant criteria, if a student continues to fall below the expected aim line but demonstrated a large amount of continuous growth (measured by slope or comparing the student's slope to those of her or his peers), this student would be considered to have responded to the intervention. Students would need to fall below expectations and additionally not show growth on more than one intervention to be deemed nonresponsive.

Although there are lingering questions of how benchmarks should be set for ELs, Linan‐Thompson et al. ([25]) found some support in their empirical study for a version of a dual discrepant model having predictive utility when compared with other RtI models in an EL sample. Gravois and Rosenfield ([17]) reported that an appropriately trained team receiving follow‐up training and consultation could reduce the disproportionate referral and placement of CLD (mostly African American in this sample) students. Phonological awareness training in the language of instruction paired with a token economy had some preliminary support in a small study for improving reading in first‐grade ELs (Healy, Vanderwood, & Edleston, 2005). Vaughn et al. ([46]) found that combining targeted reading interventions with vocabulary and comprehension professional development for teachers had some benefits in mostly word‐level outcome measures for a large, diverse sample of secondary students.

To improve culturally responsive service provision at Tier 2, it would also be important to consider the pitfalls described in the excellent qualitative study conducted by Orosco and Klingner ([32]). In this study, teachers rather than researchers attempted to implement RtI with a large percentage of Latina/Latino ELs. Major themes that evolved from this research include "Misalignment in Instruction and Assessment, Negative Schooling Culture, Inadequate Teacher Preparation, and Limited Resources" (p. 276). This research demonstrated that simply selecting a research‐validated curriculum and using "generic" progress monitoring are insufficient for students who are CLD. Practitioners should also consider incorporating local norms into progress monitoring benchmarks (Linan‐Thompson et. al., 2007).

Batsche et al. ([4]) designate Tier 3 as intensive, sustained interventions, often serving a small portion of the school population (approximately 5%). At Tier 3, a possible benefit of a fully implemented RtI model is the reduction of teachers being the primary or only source of referrals for students who need extensive intervention (Mellard & Johnson, [28]). Research has demonstrated that systemic approaches, which include classroom‐wide and individual intervention and progress monitoring, may be better referral sources than teachers alone (VanDerHeyden et al., [44]). Tier 3 could include special education in some models (García & Ortiz, [15]). In traditional models of special education service delivery, critiques often mentioned have included that special education is frequently a separate system (Ciolfi &Ryan, 2011; Mellard & Johnson, [28]).

An important question to ask in evaluating the efficacy of such models as RtI is whether outcomes in general and special education (often in RtI Tier 3) for CLD students are improving. Artiles et al. ([1]) and Cartledge and Dukes ([8]), in reviewing special education outcomes for CLD children and African Americans, respectively, paint a glum picture of outcomes. Achievement, completion of school, and participation in post‐high school opportunities continue to be challenges for students who are given a special education label in the school setting. Also, of concern in these reviews is that African Americans are twice as likely as European Americans to be placed in more restrictive, self‐contained settings, thus denying greater numbers of these students extended opportunities to interact with their typical peers, to have consistent exposure to the general curriculum and content, and to gain needed credits for graduation. They also stated that in affluent school districts with fewer African American children, the percentage of African American children labeled as needing special education was reported to be higher. Similar concerns have been reported for Latinos (Zamora, [48]).

Not much has been changed or addressed in Tier 3 to distinguish it from the traditional, separate, deficit‐based systems afforded by special education. Furthermore, children who are CLD generally still have very poor outcomes when marginalized in special education. Some of the most basic issues that need to be addressed when developing a culturally responsive Tier 3 include ensuring that exit criteria exist, disaggregating outcomes for all learners, carefully balancing all of the pullouts, and ensuring that students are getting sufficient exposure to the content required to graduate. In short, is there a benefit that outweighs the risks and stigma of being labeled as "special education" for students who are CLD? Labeling indiscriminately or to "at least get some help" for a struggling student sometimes makes it too easy to blame students and their parents for the perceived intrinsic deficit without forcing schools to look at their piecemeal and incoherent policies, practices, and packaged curriculums that may be institutionalizing disproportional placement (Artiles et al. 2010; Klingner et al., [22]) and lack of instruction (Orosco & Klingner, [32]). Ciolfi and Ryan ([9]) have stated that until quality individualized academic assistance, protection from suspension and expulsion, and behavioral support are more readily available, disproportional placement will likely continue because many see special education as the only route to needed services.

In a researcher‐implemented study of a systemic model of assessment and intervention, VanDerHeyden et al. ([44]) evaluated the implementation of STEEP (System to Enhance Educational Performance) and found more proportionate special education placements by ethnicity and gender and an overall reduction in the number of students referred for evaluation. One of the positive aspects noted in this implementation is that the intervention targeted not only individual students but was also provided to entire classrooms.

After noting that no federal guidelines exist to clarify additional procedures required for an equitable special education assessment for ELs, Haager ([18]) examined the extant literature base on RtI with EL students. She found that little is known empirically as to what constitutes sound instruction for ELs and that most studies focused on word‐level instruction, with little research considering how those learning English derive meaning from what they read in the new language. Haager concluded that fidelity of implementation is both an important methodological concern for research with ELs and an important instructional issue. With regard to research methodology, she asserted that careful attention and control of fidelity ensures that treatment outcomes can be attributed to the studied methods. Inclusion of fidelity measures when assessing students for Tier 3 is thus imperative to achieve a culturally responsive RTI framework.

Although beyond the scope of this article, readers with an interest in actual testing procedures with CLD youth are encouraged to read the chapter by Rhodes ([37]), in which he delineates several of the issues of concern from a neuropsychological perspective, including specifics about individual measures and considerations when using them with a variety of CLD populations. Rhodes, Ochoa, and Ortiz (2005) provide a very comprehensive and detailed manual for assessing CLD children, which specifies the importance of assessment in both languages with ELs, different procedures to use depending on the combinations of proficiency in each language, considerations for cultural loading and linguistic demands for commonly administered norm‐referenced measures and individual subtests, measuring acculturation, and much more for readers who desire additional information on psychoeducational assessments. However, two recent studies (Cormier, McGrew, & Evans, [10]; Kranzler, Flores, & Coady, [24]) raise important questions about the validity and use of the cross battery approach and the Culture‐Language Interpretive Matrices, relegating the cornerstone of Rhodes et al.'s (2005) approach into an area of promise needing further empirical validation.

Burns et al. ([6]) provide an excellent review of legal and ethical concerns involved when using RtI as part of an assessment for specific learning disabilities (SLD). In their comprehensive review, they found that using primarily RtI data to determine SLD is likely "indefensible" (p. 274) and much more research is needed specifically on how RtI translates into practice and fidelity of implementation. Other concerns addressed by these researchers revolved around the lack of research with CLD youth.

If an EL student is found eligible and an Individualized Education Plan (IEP) is devised, Echevarría and Graves ([12]) recommend that IEPs include: 1. Assessment in both native and English language skills to determine language competence 2. Goals related both to the development of English and the native language when possible 3. Instructional practices that are known to be effective for English learners, including those that require the active participation of the learner [and]4. A focus on outcomes, not simply process. (p. 24)

A few additional suggestions are as follows:

• Observe the student in the classroom and across settings (Klingner & Edwards, [23]).
• Carefully review the student's schedule to ensure coherence (Mellard & Johnson, [28]). If the student is also an EL and/or receives multiple special education services, such as speech or occupational therapy, make sure pull‐out is not overwhelming and not conducted during primary content instruction time.
• Carefully document limitations of methods, including whether assessment or intervention procedures are not yet validated for children with the student's background in all written documents, so readers can make their own conclusions.
• Because multidisciplinary teams are not always accurate (VanDerHeyden et. al., 2007), consider data over "wanting to get more help" when discussing making a referral for special education for students who are CLD. If children are responding to an intervention, carefully consider how this progress can be supported in the general education setting.

As always, the clinician, school psychologist, or educator has a responsibility to report the limitations and cautions of the measures she or he is using when interpreting tests, conducting interventions, or analyzing data. If the measure, intervention, or analysis has not gone beyond the inclusion of CLD children in research, then conclusions are reported with caution, and full disclosure of what is known about the use of the method with children who are CLD is provided so that readers of the report or IEP can make their own conclusions. If a nonstandard or exploratory/experimental method is utilized for intervention or assessment, it should be identified as such. Combinations of measures, observations, diagnostic teaching and dynamic assessments, and parent and teacher observations and interviews may be even more valuable when assessing or providing interventions to a student who is CLD (Klingner & Edwards, [23]). The often‐stated refrain in many psychoeducational reports of "eligibility and placement decisions should not be made based on this report alone" takes on even greater significance when working with culturally and linguistically diverse populations. Context, careful observation across settings, and careful review of records and the student's history provide interpretive information and are essential. Klingner et al. ([22]) perhaps stated it best: The combination of historical racism and extremely ambiguous definitions, policies, and practices places the most vulnerable students at increased risk of inappropriate labeling and isolation. These serious outcomes reveal the fallacy of applying special education's categorical mindset to what is essentially a continuum of human behavior. (p. 7)

Culturally responsive RtI, if it is to succeed and be beneficial to all learners, must strive to be more than a new form of a categorical mindset or a deficit model. Additional research that is contextual, detailed, precise, and provides sufficient specificity of the samples is needed before we can assert with certainty that RtI can produce positive outcomes and equitable education for all learners.