Featured Articles From State Newsletters
This section of Communiqué Online reprints articles from state newsletters whose editors have nominated them as worthy of national attention. This month’s article was originally published in The Connecticut School Psychologist (2010, Vol. 16, No. 3, pp. 7–22), the newsletter of the Connecticut Association of School Psychologists. The article was nominated by Heidi MacDonald, editor of The Connecticut School Psychologist. For membership and other information about the Connecticut Association of School Psychologists, visit www.caspweb.org.
The author of this article, Terry D. Williams, PhD, is a bilingual school psychologist and licensed psychologist who works in the New Britain (CT) Schools. His professional interests include doing research, consulting, and continuing to work with children. Dr. Williams would like to thank his colleagues for helping to enrich his professional experience, and most of all the children who teach him about courage and resilience on a daily basis.
Articles from state newsletters that are reprinted in Communiqué Online have been nominated by their newsletter editors. They are reprinted here as submitted by their editors, without further review or editing by Communiqué or Communiqué Online staff.
Assessment of CLD Students Within a Multitiered System
By Terry D. Williams
Consider the following scenarios:
An experienced first grade teacher in a working class community has several children struggling in reading in her class. These students speak Spanish as well as English in their homes. The teacher has tried many interventions, including commercially produced curricula designed to meet the needs of native English speakers, to no avail. The teacher suspects the students have LD and makes a referral for testing.
A psychiatrist in a mental health clinic interviews a child who migrated from Puerto Rico two months earlier. The child is withdrawn, uncomfortable with his fellow students in school, and is not sleeping well. It is noteworthy to the psychiatrist that during the interview the child does not maintain direct eye contact. The psychiatrist diagnoses the child as autistic.
A Planning and Placement Team (PPT; Connecticut’s version of an IEP team) determines that a culturally and linguistically diverse student should have speech and language services based upon a comprehensive evaluation and cognitive testing that suggests that verbal areas are lower than non verbal areas.
A school psychologist decides to test a bilingual child in English based on a teacher’s statement that the child "speaks English well."
The above scenarios illustrate decisions made about assessment data whether the data was derived from classroom observation, interview, standardized test results and/or input from a teacher. Unfortunately, the assessments did not consider cultural and/or linguistic factors which may have led to a better understanding of the data and more helpful decisions about helping the child in question.
This paper was written to provide useful information and practical assessment guidelines, in particular to school psychologists and other pupil personnel staff, to those implementing RTI (Response to Intervention) with culturally and linguistically (CLD) diverse students. In addition to guidelines for nondiscriminatory assessment of CLD students, this writer wants to advocate for uniformity in the implementation of assessment guidelines in the form of a district-wide Best Practices manual that would be regularly used by school personnel. District-based guidelines and ongoing staff training are the best ways to avoid the kinds of outcomes described above.
This paper includes the following sections: Definitions; Changing Demography in the US; What is Nondiscriminatory Assessment?; Factors that Lead to Nondiscriminatory Assessment; Why Do We Need to be Concerned with Nondiscriminatory Assessment?; A Broader View of Assessment; Responsibilities of the Assessor; Non-Biased Assessment Through the Tiers; The Importance of a District-Wide Best Practices Manual; and Conclusion.
At this point some definitions are in order. A culturally and linguistically diverse (CLD) student is a speaker of a language other than English who is in the process of acquiring English (Klingner, de Shonewise, de Onis, & Barletta, 2008). In this paper CLD will be used interchangeably with English Language Learner (ELL). Multi-Tiered System refers to the multiple tiers of intervention service delivery in a Response to Intervention (RTI) program. "RTI" is the practice of (1) providing high quality instruction/intervention matched to student needs and (2) using learning rate over time and level of performance to (3) make important educational decisions" (NASDSE, 2006). Assessment is one of the three systems necessary to make RTI function well. Hintze (as cited in Murdica and Anderson, 2009) suggests that these three systems, which include the assessment, instructional, and data management and decision-making systems, share close inter-relationships. Suggestions for nondiscriminatory assessment will be made for Universal Screening, Progress Monitoring and Comprehensive Evaluations.
Changing Demography in the U.S.
The changing demography in the U.S. underscores the growing likelihood that school personnel will need to assess CLD students in the near future (Rhodes, Ochoa, and Ortiz, 2005). In 2004-2005 nearly 10.5% of the U.S. student population was English Language Learners. Spanish-speaking students constitute 77% of the ELL school-age population (Rhodes et al., 2005). The Connecticut State Department of Education (CSDE) provided information on the number of ELLs in each of 148 districts in Connecticut for the 2007-2008 school year. The data showed a range of 1-1374 with the larger cities systems having the largest number of ELLs. It also important to point out that each school district had at least one ELL, and some of the larger towns had a sizeable number of ELLs (e.g., East Hartford 476).
Other pertinent information about the ELL population includes the following: The academic performance of the ELL population is well below that of their peers: ELLs have a high dropout rate (NEA Policy Brief, 2008); According to the National Research Council (2002) a child’s race and ethnicity are significantly related to the probability that the child will be inappropriately identified as disabled; and It has been pointed out that CLD children are both under and overrepresented in special education. A plausible explanation for disproportionality is the notion that since ELLs tend to underachieve in comparison with their mainstream counterparts, this may put pressure on school personnel to give them extra assistance. Consequently, educators may see special education as the only viable alternative or that the student’s difficulties are due to a learning disability. On the other hand, school personnel may be uncomfortable with making a referral to special education as they attribute the academic problem as being related to second language acquisition when in fact the student may have a learning disability (Orosco, de shonewise, de Onis, Klingner, & Hoover, 2008). Cloud (2010), at a presentation entitled Distinguishing Second Language Acquisition from Learning Disabilities, pointed out that ELLs tend to be underrepresented in special education in schools with large ELL population and overrepresented in schools with a small ELL population.
What Is Nondiscriminatory Assessment?
Nondiscriminatory (used interchangeably with unbiased) assessment is not a single procedure or measure, but includes a wide range of attempts to uncover, as equitably as possible, pertinent information upon which decisions about performance and functioning can be fairly made. According to Ortiz the process of nondiscriminatory assessment should begin with the assumption that the student’s difficulties are extrinsic in nature and therefore attributable to circumstances that are external to the child. The goal of this kind of assessment is to make sure that potential causes of a student’s learning or behavioral problems (e.g. low motivation, anxiety, cultural and linguistic differences) have been ruled out as causal factors for presenting learning or behavioral problems (Ortiz, 2002). Chief among the potentially discriminatory factors in assessment is the assumption of comparability of experiences between the norm sample of a measure and the individual student being assessed. Given the heterogeneity of bilingual individuals with regard to linguistic and cultural experience the CLD student is not adequately represented by any existing norm sample (Rhodes, Ochoa and Ortiz, 2005). According to Baca and Clark (1992) 25% of bias is related to an assessment instrument and 75% of bias is related to the interpretation and use of the results of the assessment.
Distinguishing Cultural and Linguistic Difference From Disability
In order to avoid biased assessment and interpretation, multi-tiered teams must learn to distinguish an intrinsic disability from a behavioral difference due to cultural and linguistic diversity. Various behaviors normally exhibited by students in different stages of second language acquisition are similar to those seen in learning and/or behavior disorders. For example, a child may appear inattentive while attempting to acquire a second language or require a longer "wait time" before responding to a question or statement (Hoover, 2009). A child arriving to a new culture may present with behaviors appropriate to her culture of origin, but appear "going against the grain" in the new culture. For example, a child may prefer cooperative group work over independent or competitive tasks. A recently migrated child may exhibit some atypical-appearing behaviors. For example, the fact that the child is not responding when spoken to, failing to talk despite having the skill to do so, and preferring to be alone (as mentioned earlier in the example of the psychiatrist labeling a newly-migrated child as "autistic) may reflect a normal acculturation process (Collier, 2004). Collier defines acculturation as "the process of adaptation to a new cultural environment." Collier indicates that there is considerable evidence that cultural, linguistic and psychological changes occur among populations undergoing acculturation.
Another example of how a linguistic difference could be confused with a learning problem is the situation when a classroom test becomes a test of language ability rather than a test of content knowledge (Bailey and Butler, 2004). An effective way of avoiding this would be to match the assessment measure language with the stage of literacy development of the ELL being assessed. Cloud, Genesee and Hamayan (2009) established a literacy development framework for Reading and Writing which ranges from Beginning to Advanced Supported. This framework describes milestones that students pass through as they learn to read and write in English, from the earliest emergent literacy stage (Beginning) to the advanced level (Advanced Supported). Cloud et al. (2009) point out that their framework reflects the stages or benchmarks for literacy development that have been used by school districts. They suggest that the framework is intended to serve as a guide in developing curriculum, and that school personnel should familiarize themselves with the ESL framework in their particular school district. The ideal situation would be to reduce the linguistic complexity of an assessment, but maintain the rigor of the measure (Abedi & Lord, 2001).
So how does one distinguish between a learning difference and learning disability? Hoover (2009) states that similar academic or social-emotional problem behaviors have to be noted in both languages and evident in both cultures before an intrinsic disorder can be considered. Hamayan & Damico, Mattes & Omark and Ortiz & Maldonado-Colon (as adapted by Ortiz, 2004) provide the following guidelines for distinguishing language differences from disorders:
- The disorder must be present in the child’s native language (L1) and English (L2) keeping in mind that this condition may occur for other reasons.
- Testing must be conducted in the native language (L1) and/or both the native language and in English.
- Assessments must be conducted using both formal and informal measures
- Language must be assessed in a variety of speaking contexts.
- Patterns of language usage must be described.
- Error patterns must be determined.
- The child’s language performance must be compared to that of other bilingual speakers who have had similar cultural and linguistic experiences.
- Factors which may be contributing to the interruption of development in the native language must be identified.
Why Do We Need to Be Concerned With Nondiscriminatory Assessment?
We should be concerned with nondiscriminatory assessment for three reasons: validity of assessment results, disproportionality of CLD students in special education, and it’s the law. Earlier it was mentioned that bias may result when there are cultural and linguistic differences between a measure’s normative sample and the student being assessed. In addition to bias being a problem in this situation, the validity of assessment results may also be suspect. The bias and lack of validity are not due to any defect in the assessment measure itself. Rather, because the CLD student’s performance will be compared to others of the same age or grade, she very likely will perform less well due to her beginning the acculturation process at a different point from that of the individuals in the normative group. Therefore, standardized measures are measuring the degree to which an individual has become acculturated as well as the degree of English-language proficiency obtained by a student (Rhodes et al., 2005).
Disproportionate representation refers to "unequal proportions of culturally diverse students in special education programs" (Artiles & Trent, 2000). Disproportionality is often determined by calculating a group’s representation in a special education category in comparison with their proportion of the total school-aged population or in reference to the representation of a comparison group, frequently White students (Klingner et al., 2008). The most common ethnic groups overrepresented include African American, Chicano/Latino, American Indian, and a few subgroups of Asian American students (Artiles & Trent, 2000). In summarizing the factors contributing to disproportionate representation of CLD students Rhodes et al., (2005) suggest the following school systemic variables: bilingual and general education instructional practices; referral practices; inadequately trained school psychologists; use of untrained interpreters; insufficient or inadequate assessment practices pertaining to language proficiency, academic achievement and intelligence; and non compliance with state and/or federal guidelines (e.g. exclusionary factors).
With regard to the legal requirements for nondiscriminatory assessment, IDEA 2004 requires that: assessment instruments are selected and administered so as not to be discriminatory on a racial or cultural basis; are provided and administered in the child’s native language and in the form most likely to yield accurate information on what the child knows and can do academically, developmentally and functionally and are used for the purpose for which the assessments or measures are valid and reliable (CSDE, 2009).
RTI: A Broader View of Assessment
When one hears the word "assessment" the thought arises of giving a student a test, scoring it and then interpreting the results. The academic or behavioral problem is seen as existing within the child and the test will help the examiner learn the nature of this intrinsic problem. Cloud et al. (2009) note that assessment is much more than whether the student gets the correct answer, understands the story or is paying attention to the teacher. It is also what you do in response to the results of your assessments. Since there is a reciprocal relationship between assessment and instructional planning, assessment is as much about the teacher as it is about her student. In assessing contextual factors that may contribute to student performance three domains of data were identified by Bernhardt (as cited in Heritage and Yeagley, 2005): demographic data (e.g. grade level, ethnicity, language spoken at home); perception data which can shed light on student, teacher and parent attitudes about learning, teaching, and school programs; and school processes which include information about curriculum, teaching strategies, student discipline, parent communication, teacher qualifications, and other aspects of the school that can have an impact on student learning.
In the case of nondiscriminatory assessment of CLDs within an RTI context, the assessor needs to begin with exploring extrinsic factors that may contribute to the learning problem. Practitioners need to develop hypotheses that pertain to the student’s unique experiential background within the context of the learning environment. Often it is the interaction between experiential and learning factors that contributes to a poor match between instruction needed and instruction delivered. Data that can be useful in identifying contextual information include review of educational records, direct observation of instruction, review of the content and relevancy of the curriculum, examination of the match between student needs and instruction, and interviews with parents, teachers, the student in question and finally medical records (Ortiz, 2002).
Garcia & Ortiz (2008) have suggested that the examination of the student-learning environment interaction is not sufficient for RTI to be successful with the CLD student. These writers go beyond the above parameters to include systemic variables such as sociopolitical, cultural, and linguistic contexts that influence educational processes. Specifically, these variables include teacher, classroom, school and community characteristics in addition to student variables such as culture, language status and proficiency.
Responsibilities of the Assessor
Once a school system decides to participate in a multi-tiered program which includes regular assessments, the system takes on the role of test developer. The Individual with Disabilities Educational Improvement Act (IDEIA) mandates that all assessments be technically sound. By "technically sound" the assessment must demonstrate validity, reliability and fairness (Braden, Kubiszyn, and Ortiz, 2007). Assessments must also adhere to standards set forth in the Standards for Educational and Psychological Testing (American Educational Research Association, American Psychological Association and National Council on Measurement in Education, 1999) and the Code of Fair Testing practices in Education (Joint Committee on Testing Practices, 2004). Therefore, there are legal, professional and ethical responsibilities that require that test developers ensure fairness in the assessments of CLD students.
In discussing validity (the extent to which an assessment measures what it is intended to measure) and reliability (the degree to which assessment results are consistent over repeated applications) Heritage and Yeagley (2005) point out that demonstrating reliability doesn’t necessarily ensure validity. For example, a Math assessment that focuses on multiplication may provide highly consistent results, but be a poor indicator of Math problem-solving skills. Assessment validity can be seen from two perspectives: (1) that the assessment is valid for use in a multi-tiered response to intervention model and (2) that the uses and inferences from the assessment scores are applied in a manner consistent with the purposes for which the assessment was validated (Hoover, 2009). In addition to validity, reliability, and fairness, an assessment measure must demonstrate fidelity. Hoover (2009) refers to fidelity in multi-tiered response to intervention as the "proper implementation of assessment practices, procedures, and devices along with accurately drawing proper inferences from obtained assessment scores."
Non-Biased Assessment Through the Tiers
While RTI (or Scientific Research-Based Intervention as it is called in Connecticut) represents a new paradigm designed to avoid the "wait to fail" models of the past, the salient issues for CLD children remain the same as those that confronted prereferral teams of the past. Specifically, the question for the assessor is to what extent are the cultural and linguistic needs of the students being incorporated within each Tier (Hoover, 2008). There is concern in the literature that CLD students in a multi-tiered system may be exposed to a "one size fits all" standardization of instruction and assessment methods (Hoover, 2009; Garcia, 2009; Garcia & Ortiz, 2008, Gerber, 2005). Some of the reasons for this include a paucity of teachers certified to teach ELLs, inadequate training of pupil service personnel in assessing CLD students, erroneous assumptions about the learning needs of CLD students, and district-specific practice and budgetary issues. Waxman, Tellez and Walberg’s research (as cited in Brown & Doolittle, 2008) suggest that less than 20% of the 56% of public school teachers in the US who have at least one ELL in their class are certified to teach ELLs. Consequently, most teachers lack the experience and expertise to teach reading and other subjects to CLD students. In addition most multidisciplinary school teams, which have the responsibility to make special education eligibility decisions for ELLs, lack the training and experience in distinguishing a learning difference from a learning disability (Klingner, Artiles, & Barletta, 2006). In a survey of school psychologists in 8 states 83% of school psychologists self-reported that they were less than adequately trained by their university programs to conduct bilingual assessments. It is worth noting that these same school psychologists indicated that they had actually conducted assessment of CLD students (Ochoa, Rivera & Ford as cited in Rhodes, et al. 2005). With regard to the contribution of erroneous assumptions which may contribute to a "one size fits all" view Klingner, Barletta, & Hoover (2008) cite the following assumptions that can be problematic when applied to CLD students:
- "Evidence-based instruction" is good for everyone. Klingner and Edwards (as cited in Klingner et al., 2008) note that numerous instructional approaches recommended as being "evidence-based" have not been validated with ELLs.
- Learning to read in one’s second language is similar to learning to read in one’s first language; therefore instructional approaches that have been found to be effective with mainstream English speaking students are appropriate for ELLs. For example ELLs benefit from additional oral language instruction.
- Students who fail to respond to research-based instruction have some kind of learning problem or internal deficit. Klingner and Edwards (as cited in Klingner et al., 2008) note that there could be many reasons why a child may not respond to a particular instructional approach. For example, the method may not be effective with this particular child and a different approach could yield better results.
With regard to district-specific practice and budgetary issues as potentially contributing to resistance to diversifying assessment practices and possibly promoting a "one size fits all" mentality is a study which examined the current assessment practices in Northeastern School Districts. This study (Madeus, Renaldi, Bigaj, & Chafouleas, 2009) found that the choice of assessment techniques in a district was based on past use and reputation. This suggests that school personnel accustomed to implementing the same assessment methods with all children will want to continue with those same procedures despite research calling for more culturally and linguistically appropriate methods for diverse learners. In addition, the purchase of additional assessment materials was also driven by a set budget in the district. Purchasing revised editions of currently norm-referenced measures constituted the most frequent use of budgeted funds. Thus, there are other contextual forces that undermine a more flexible, child specific assessment paradigm.
The contemporary model of assessment within a multi-tiered response to intervention system is a process that runs along a continuum, progressing from initial screening of all students, to the monitoring of progress of a select group of struggling learners, to individual comprehensive assessment for those who do not respond to previous evidence-based interventions (Hoover, 2009). Below relevant issues for CLD students within RTI assessment decision points will be discussed.
Universal screenings are assessments that are given routinely to all students in a grade, within a school, or district (C.S.D.E., 2008). Universal screenings, used in essential academic areas, identify each student’s level of proficiency (usually three times per year). The screening data are used to (1) assess the adequacy of the curriculum and instructional process and (2) identify those students who need further intervention at Tier II (NASDSE, 2006). In the case of the CLD student, effective assessment (as well as instruction) must be "linguistically and culturally congruent." Specifically, Delpit, Gay, Macedo & Bartolome (as cited in Brown and Doolittle, 2008) suggest that the teacher must know the student’s level of proficiency in his/her first (L1) and second language (L2) and provide "culturally relevant" curricula that reflect the background experiences of the student. Therefore, "linguistically and culturally congruent" means that the student’s language and culture are not viewed as liabilities, but instead are seen as strengths upon which to build a learning experience. In the event that a CLD student has been identified as struggling in school the assessor/assessment team must examine the instructional program and determine the match between the demands of the curriculum and the student’s current level of proficiency in the language of instruction. Moreover, it is important to examine the student’s achievement relative to his/her "true peers" (children with a similar language proficiencies, culture and experiential background). If several "true peers" are also having difficulty this may be an indication that the instruction may have to be changed to better meet the needs of the students (Brown, & Doolittle, 2008). The technical review committee of the National Center on Response to Intervention (NCRI) (www.rti4success.org) rates universal Reading and Math screening measures on criteria including reliability, validity, generalizability and whether the measure provides disaggregated data (i.e. data reported separately for specific sub-populations) relative to validity and reliability for diverse populations. The review committee rates the measures on a scale ranging from "convincing evidence," "partially convincing evidence," unconvincing evidence," to "no evidence submitted."
Progress monitoring, which can occur at Tiers I, II, and III, is a research-based method involving the assessment of student progress (academic and/or behavioral) on a regular basis. The primary purposes of progress monitoring are to determine both a student’s level of achievement (e.g. number of words read correctly) as well as rate of improvement (e.g. does the student show a consistent increase in number of words read correctly). The goal of progress monitoring is to use this data to provide more effective instruction to learners. Progress monitoring can be used to assess progress of both individual learners and/or entire classrooms of students (Hoover, 2008). Hoover (as adapted by Klingner et al., 2008) lists the following as among the tasks necessary for nondiscriminatory progress monitoring:
- Cross cultural interviews and classroom observations are conducted.
- Cross-validation of information from the home and family settings corroborates progress monitoring data.
- Culturally responsive progress monitoring methods are implemented.
- Effects of environmental and cultural influences on suspected problems are documented.
- Progress monitoring is completed by culturally and/or linguistically competent persons.
- Linguistically appropriate goals and services are included in progress monitoring.
- Previous instructional programs and student progress in those programs are considered.
- Home and family information is documented and accurate.
- Student’s language dominance and English proficiency are determined.
- Translators/interpreters are properly used if necessary in the progress monitoring.
- Alternative and authentic forms of progress monitoring are included.
- Evidence-based adaptive behavior interventions are attempted and documented
The NCRI technical committee reviews (www.rti4success.org) and rates progress monitoring measures on variables such as reliability, validity, being sensitive of student improvement, norms disaggregated for diverse learners and disaggregated reliability and validity data. As was the case with the universal screening ratings, the measures are rated from providing "convincing evidence," to "no evidence submitted." It is important to note that the NCRI did not indicate how validity and reliability data were disaggregated (i.e. how specifically the data were broken down).
In examining NCRI data, the reader should note that only 2 of 9 screening measures (i.e. Predictive Assessment of Reading and STAR Early Literacy) were rated as providing "convincing evidence" with regard to disaggregation of reliability and validity data for diverse learners. Regarding progress monitoring instruments there were no norms disaggregated for diverse learners for the reviewed measures. Additionally, there were only 2 of 9 measures (i.e. AIMSweb and STAR Early Literacy) found to provide "convincing evidence" of disaggregation of validity and reliability data for diverse learners. It is important to note that disaggregation does not necessarily mean that the reference group with whom a bilingual child is compared adequately matches the learner with regard to language proficiency and other acculturation factors.
In addition to assessing student performance, a comprehensive assessment within a RTI context also needs to include an appraisal of school personnel in affording scientifically validated interventions at Tiers I and II (i.e. in a three-tiered RTI program). Evaluation teams need to carefully delineate the components of the interventions and determine through direct observation whether the strategy was implemented according to predetermined guidelines. Finally, assessment only should be implemented in all areas that are potentially related to the disability in question (NASDE, 2006). Unlike universal screening and progress monitoring (which primarily measures the effects of instruction) the comprehensive assessment process examines abilities intrinsic to the learner and may result in a referral to special education. Since there has been a close examination of evidence-based practices in Tiers I and II there can be more confidence that the learning difficulty is related to an intrinsic learning issue as opposed to ineffective teaching or an inappropriate curriculum (Hoover, 2009).
Ortiz (2002) has recommended the following as a framework for a nondiscriminatory assessment of CLD learners:
- Assess and evaluate the learning ecology.
- Assess and evaluate language proficiency.
- Assess and evaluate opportunity for learning.
- Assess and evaluate educationally relevant cultural and linguistic factors.
- Evaluate, revise, and re-test hypotheses.
- Determine the need for language(s) of assessment.
- Reduce bias in traditional testing practices.
- Utilize authentic and alternative assessment procedures.
- Evaluate and interpret all data within the context of the learning ecology.
- Link assessment to intervention.
The reader is referred to Rhodes et al., (2005) and Flanagan, Ortiz and Alfonso, (2007) for specific guidelines on how to use standardized tests in a nondiscriminatory fashion.
The Importance of Authentic Assessments
In contrast to assessments which provide a volume of scores, authentic assessment is seen in student products. Work products often reflect a combination of skills within a student’s learning. These collective tasks generally provide information about a variety of skills such as use of higher order thinking, and social and behavior management skills in different classroom settings. Collections of work samples may provide: patterns of student strengths or weaknesses; information on the experiential background of the learner; a thorough exploration of an area of academic or social-emotional strength/weakness; and cultural and linguistic needs specific to the learner that may be identified and supported in the classroom (Hoover, 2009). For example, observing a small group of 4th graders collectively writing their own "book" based on a lesson in class one is able to assess both literacy and social skills. Baca and Cervantes (2004) list the following as useful with culturally and linguistically diverse populations: analytic teaching, criterion referenced tests, curriculum-based assessments, dynamic assessment, interviews, language samples, narrative analysis, testing to the limits, observations, rating scales, review of records, and work samples. Cloud et al., (2009), who consider observation of students during planned activities as the most powerful method of assessment, list the following as important authentic assessment tools: observation, reading and writing conferences, running records, dialogue journals, and learning logs. The reader is referred to this reference for specific guidelines on implementing these assessments. These writers assert that assessing literacy skills, especially with regard to setting up literacy plans for a student, involves more than reading and writing as we generally think of them. This kind of assessment involves collecting information about skills such as written language directly, such as word-decoding skills or the ability to understand written text. Additionally, many of the skills assessed involve literacy more indirectly. Specifically, these skills are referred to as "underlying language competencies" because they are not specific to written language. For example, these kind of competencies include labeling, sequencing, understanding cause-effect relationships, inferring consequences, and predicting. These competencies are important in learning to read and write because storybooks and other written texts will require these skills (Cloud et al., 2009). In summary, while many of the authentic assessment tasks lead to fewer scores to interpret, they provide educators with insights into a student’s application of interrelated skills. Combining instructionally embedded assessment with more standardized measures (e.g. AIMSweb) provides CLD learners a more comprehensive overview of learner needs, and at the same time helps multi-tiered teams to make the most informed decisions possible to assist struggling learners (Hoover, 2009).
After reviewing the literature one notes that there is wide agreement that Curriculum-Based Measurement (CBM) is a highly reliable and valid assessment practice (Fuchs & Fuchs, 2007; 2008; Saenz, 2008; Hoover, 2008; Braden et al., 2007; Blatchley & Lau, 2010). CBM can be used for screening as well as progress monitoring. With a CLD population CBM data can be used to inform instruction (e.g. Are most ELLs making progress or only some?). Given its ability to yield scores in target domains (e.g. words read correctly) CBM can reduce subjectivity regarding students not making progress (Saenz, 2008). One of the main limitations of CBM with the CLD population is the fact that there are few studies of its use with ELLs (Saenz, 2008; Hoover & Barletta, 2008).
The Value of a District-Wide Best Practices Manual
At the beginning of this paper the reader was asked to consider some scenarios that reflect a failure on the part of professionals to consider linguistic and cultural differences in the CLD population. For example, a teacher referring a CLD child for testing child who doesn’t respond favorably to commercially packaged instruction offered to all students may not understand that the child’s cultural and linguistic differences may require a measure validated for the CLD population or more in the way of authentic assessments before making any decisions about special education referrals. A PPT team referring a child for speech services based on a low verbal score on a cognitive measure may not understand that a lower verbal score may be related to second language acquisition factors rather than a speech disability. Finally, a school psychologist deciding to test a CLD student in English based on a teacher’s statement about his conversational skills in English may not understand that oral language proficiency in English is only one part of overall proficiency in a language. The other component of proficiency is Cognitive Academic Language Proficiency (CALP) (Cummins, 1980). Proficiency in CALP is necessary to prevent the assessment measure from becoming a test of language as opposed to an equitable measure of knowledge in a content area. In this writer’s opinion the way to avoid the above scenarios is to create uniformity for teachers, administrators, pupil personnel, and multidisciplinary teams in the form of a Best Practices Manual for the assessment of CLD students within a multi-tiered system. After the completion of this manual it would be necessary for staff to be trained on an ongoing basis to solidify concepts and make necessary changes in the manual.
This paper is a call to widen our assessment lens when it comes to assessment of a CLD child. To accomplish this the RTI team first needs to go beyond looking for a problem as residing in the child and consider contextual variables (culture, classroom instruction, family circumstances, district policy and support for RTI, previous schooling) in trying to understand a learning difficulty. Since there is a reciprocal relationship between assessment and instruction one cannot look at one and not consider the other.
One of the goals in writing this is to help all pupil service personnel better understand and desire to learn more about necessary cultural and linguistic considerations in their work with ELLs. It was stressed that assessments at any level should include both measures with data points and those that allow authentic application of skills in the form or work products. Cloud et al., (2009) provide specific guidelines for their recommended assessment tools. The guidelines proposed by these authors help to structure the assessments so that there may be some uniformity in their application.
The recommendation was made that districts develop a Best Practices Manual in the assessment and instruction of CLD students within a multi-tiered system. For something like this to "stick" it would be necessary for ongoing staff training and development. It would also seem important that there is significant collaboration among teachers, school psychologists, counselors, speech and language pathologists, ESL teachers, and administrators in sharing concerns, ideas about practice and efforts to refine services to CLD students. As would be the case with any district-wide endeavor administrative support would be necessary.
Finally, concerns with a "one size fits all" approach to assessment were discussed (Hoover, 2009; Garcia, 2009; Garcia & Ortiz, 2008, Gerber, 2005). This point of view entails using identical assessments and interventions with all students regardless of cultural and linguistic status. Perhaps instead of saying "what works" as if every student had the same needs one should say "what works with whom, in what context and under what circumstances" (Saenz, 2008).
Abedi, J., & Lord, C. (2001). The language factor in mathematics tests. Applied Measurement In Education, 14, 219-234.
American Educational Research Association, American Psychological Association and National Council on Measurement in Education. (1999). Standards for Educational and Psychological Tests. Washington, DC: American Psychological Association.
Artiles, A. J., & Trent, S. C. (2000). Representation of culturally/linguistically diverse students. In C.R. Reynolds & E. Fletcher-Jantzen (Eds.), Encyclopedia of Special Education, Vol. 1 (2nd Ed., pp. 513-517). New York: Wiley.
Baca, L., & Clark, C. (1992). EXITO: A dynamic team assessment approach for culturally diverse students. Minneapolis, MN: Council for Exceptional Children.
Bailey, A.L., & Butler, F.A. (2004). Ethical considerations in the assessment of the language and content knowledge of U.S. school-age English Learners. Language Assessment Quarterly, 1 (2&3), 177-193.
Braden, J., Kubiszyn, T., & Ortiz, S. (2007). Response to Intervention: Ensuring Reliability, Validity and Fairness. Presentation at the NASP Continuing Education Workshop. New York City.
Blatchley, L., A., & Lau, M.Y. (2010, May). Culturally Competent Assessment of English Language Learners for Special Education Services. Communique, 38 (7), pp.27-29. National Association of School Psychologists.
Brown, J. E., & Doolittle, J. (2008). A Cultural, Linguisitic, and Ecological Framework for Response to Intervention with English Language Learners. Denver, CO: National Centerfor Culturally Responsive Educational Systems (NCCREST). (Available from http:www.nccrest.org)
Cloud, N., Genesee, F., & Hamayan, E. (2009). Literacy Instruction for English Language Learners: A Teacher’s Guide to Research-Based Practices. Portsmouth, NH: Heinemann.
Cloud, N. (2010, January). Distinguishing Second Language Acquisition from Learning Disabilities. Special Education Resource Center (SERC) sponsored training, Cromwell, CT.
Code of Fair Testing Practices in Education. (2004). Washington DC: Joint Committee on Testing Practices.
Connecticut State Department of Education (CSDE). (2008). Using Scientific Research-Based Interventions: Improving Education for all Students. Connecticut State Department of Education Bureau of School and District Improvement.
Connecticut State Dept. of Education (CSDE). (2009). 2009 Guidelines for Identifying Children with Learning Disabilities: Executive Summary. Hartford, CT: Author
Cummins, J. (1980). The entry and exit fallacy in bilingual education. NABE Journal, 4(3), 25-59.
Fuchs, D., & Fuchs, L.S. (2007). The role of assessment in the three-tier approach to reading instruction. In D. Haager, J. Klingner, & Vaughn (Eds.), Evidence-based reading practices for response to intervention (pp. 29-42). Baltimore, MD: Brookes Publishing.
Garcia, S., B., & Ortiz, A., (2008). A Framework for Culturally and Linguistically Responsive Design of Response-to-Intervention Models. Multiple Voices for Ethinically Diverse Exceptional Learners, pp. 24-41.
Garcia, S. B. (2009). Implications of RTI for Bilingual and ELL Students. Paper presented at the 2009 NYSABE Conference. Tarrytown, N.Y.
Gerber, M. (2005). Teachers are still the test: Limitations of response to intervention strategies for identifying children with learning disabilities. Journal of Learning Disabilities, 38, 516-524.
Heritage, M. & Yeagley, R. (2005). Data Use and School Improvement: Challenges and Prospects. In Herman, J.L., & Haertel, E. (Eds.), 104th Yearbook of the National Society for the Study of Education. Chicago: University of Chicago Press.
Hoover, J.J. Data-Driven Decision Making in a Multi-Tiered Model. (2008). In Klingner, J.K., Hoover, J.J., & Baca, L.M. (Eds.), Why Do English Language Learners Struggle with Reading? (pp. 75-92). Thousand Oaks, California: Corwin Press.
Hoover, J. (2009). RTI Assessment Essentials for Struggling Learners. Thousand Oaks California: Corwin Press.
Klingner, J. K., & Barletta, L. M. (2008).Response to Intervention Models and English Language Learners. . In Klingner, J.K., Hoover, J.J., & Baca, L.M. (Eds.), Why Do English Language Learners Struggle with Reading? (pp. 37-56). Thousand Oaks, California: Corwin Press.
Klingner, J., K., Artiles, A., & Barletta, L. M. (2006). English Language Learners who struggle with reading: Language acquisition or learning disabilities? Journal of Learning Disabilities, 39, 108-128.
Klingner, J. J., deShonewise, E. A., de Onis, C., & Barletta, L.M. (2008). Misconceptions About the Second Language Process. In Klingner, J.K., Hoover, J.J., & Baca, L.M. (Eds.), Why Do English Language Learners Struggle with Reading? (pp. 17-36). Thousand Oaks, California: Corwin Press.
Madeus, J., Rinaldi, C., Bigaj, S., & Chafouleas, M. (2009). An Examination of Current Assessment Practices in Northeastern School Districts. Assessment for Effective Intervention, Vol. 34, #2, pp. 86-93.
Murdica, P. & Anderson, P. (2009, December). An Elementary Case Study: Scientific Research-Based Interventions (SRBI) and Consideration of Learning Disability. Special Education Resource Center (SERC) sponsored training, Cromwell CT.
National Association of State Directors of Special Education. (2006). Response to Intervention: Policy considerations and implementation. Alexandria, VA: Author.
National Education Association (NEA) Policy Brief. (2008). English Language Learners Face Unique Challenges. Washington, DC: NEA Education and Policy Department.
National Research Council. (2002). Minority students in special and gifted education. Washington, DC: National Academy Press.
Orosco, M.J., de Shonewise, A., de Onis, C., Klingner, J.K., & Hoover, J.J. (2008). Distinguishing Between Language Acquisition and Learning Disabilities Among English Language Learners. In Klingner, J.K., Hoover, J.J., & Baca, L.M. (Eds.), Why Do English Language Learners Struggle with Reading (pp. 5-16)? Thousand Oaks, California: Corwin Press.
Ortiz, S. (2002). Best Practices in Nondiscriminatory Assessment. In A. Thomas & J. Grimes (Eds.), Best Practices in School Psychology IV (pp. 1321-1336). Bethesda, MD: National Association of School Psychologists.
Rhodes, R. L., Ochoa, S.H., & Ortiz, S.O. (2005). Assessing Culturally and Linguistically Diverse Students: A Practical Guide. New York, NY: The Guilford Press.
Saenz, L. (2008). Using CBM to Progress Monitor English Language Learners. National Center on Response to Intervention. Retrieved from http://www..rti4success.org