Norms Normative-Based Testing Options and Procedures ELL Testing CONCEPTS TO CONSIDER 2004 IDEA Regs. • Eligibility team should… – “Draw upon information from a variety of sources, including aptitude and achievement tests, parent input, and teacher recommendations, as well as information about the child's physical condition, social or cultural background, and adaptive behavior” Assessment Balance Observation Testing Background Review Factors To Consider In Assessment • Current grade of the student? • Current and previous types of educational programs? • Students language proficiency in both L1 and L2 (and L3 if applicable)? • Was child appropriately placed (or not) as in an ELL program? Review Data in School Records • What is family history/ background? – Need to get information from the family • What is child’s educational background? • Has the child progressed while in the U.S. academically? – How did he/she do before arriving in U.S. • What might have been overlooked by the referral team? Evaluation Procedures • Measure the need for SPED services not the need for bilingual education services. • Psychologists conduct informal and formal language assessments and differentiate between a language disorder from second language acquisition. • Must evaluate if the primary cause meets one of the exclusionary clauses (not just check the boxes) • Choose the appropriate modality to test (formal/ informal; L1/L2/ Nonverbal/ Bilingual) Bilingual Assessment • Is not in L1 or L2 but in both languages simultaneously • Requires knowledge of – Examinee’s Culture – Non discriminatory assessment – Fluency in L1 and L2 MAMBI • MAMBI = Multidimensional assessment model for bilingual individuals • This model Guides choices of assessments looking at all factors of consideration. • Developed by Rhodes, Ochoa, & Ortiz (2005). Language Profiles Profile L1 Proficiency L2 Proficiency Description 1 Minimal Minimal CALP L1 = 1-2 and CALP L2 = 1-2 2 Emergent Minimal CALP L1 = 3 and CALP L2 = 1-2 3 Fluent Minimal CALP L1 = 4-5 and CALP L2 = 1-2 4 Minimal Emergent CALP L1 = 1-2 and CALP L2 = 3 5 Emergent Emergent CALP L1 = 3 and CALP L2 = 3 6 Fluent Emergent CALP L1 = 4-5 and CALP L2 = 3 7 Minimal Fluent CALP L1 = 1-2 and CALP L2 = 4-5 8 Emergent Fluent CALP L1 = 3 and CALP L2 = 4-5 9 Fluent Fluent CALP L1 = 4-5 and CALP L2 = 4-5 MAMBI ESOL/ ESL Services Only Grades: K-4 Language Profile NV L1 1: L1 = 1/ L2 = 1 2: L1 =3 / L2 = 1 4: L1 = 1/ L2 = 3 5: L1 = 3/ L2 = 3 9: L1 = 5/ L2 = 5 BL NV L1 3: L1 = 5 /L2 = 1 6: L1 = 1/ L2 = 5 L2 Grades: 5-7 L2 = Priority Testing Modality Providing the Most Sound Data = Secondary Mode Providing Some Additional Information of Use BL Psychoeducational Assessment Issues Examine individual referrals w/in the context of the systemic patterns of how diverse students are treated at that school. Acknowledge the impact of second language/culture acquisition on the cognitive and socio-emotional development Select and administer tests so as to not discriminate on racial or cultural bias. Understand the limitations and pitfalls associated with the prescribed use of standardized instruments not normed or validated with the population being served. Able to adapt existing assessment tools when necessary and report any deviations from standardization. ELLL PSYCHOMETRIC ISSUES 2004 IDEA regs. . . • Assessments should be … – “selected and administered so as not to be discriminatory on a racial or cultural basis – provided and administered in the child's native language or other mode of communication and in the form most likely to yield – used for the purposes for which the assessments or measures are valid and reliable – administered by trained and knowledgeable personnel” Validity and Tests “When a child’s general background experiences differ from those of the children on whom a test was standardized, then the use of the norms of that test as an index for evaluation that child’s current performance or for predicting future performances may be inappropriate” (Salvia and Ysseldke, 1991). Test bias . . . • Test bias: when one group systematically performs differently from another group on an instrument. • Early test developers felt that the reason that certain culturally groups did worse on certain tests was due to deficits in the cultural group and not problems with the tests. • Test bias relates more to the validity of a test’s scores and not the reliability of the scores. • Test is “culturally loaded” test contains culturally specific elements and expects a certain level of acculturation to do well. Bias in testing stems from… • • • • Cultural content embedded in the test Linguistic demands inherent in any give test Lack of representation within norm samples Changing normative standards to accommodate perceived language needs (translating the test) • Changing normative standards resulting in a change within the developmental nature of the tests (items go from easier to harder) Developmental Nature of Test Items • Test are Developmental/ incremental (easy to Hard) – Learning English may have begun at a different developmental point for ELL. – Students may juggle knowledge b/w L1 and L2 instead of being able to know just one language well. – Students may have received fewer overall hours of exposure to English language than native English speaker. • Example: English Car Their ------------------------------------- Spanish Coche Su Verbal/Nonverbal Bias • Bias in nonverbal tests may occur if the instructions are in English only • Nonverbal tests not normed using stratified acculturation and Language are still biased. • Bias in many tests are revealed more in the verbal tasks than in the nonverbal ones. – May be due to learning styles of demonstration, sample, doing instead of Socratic method – May be due to L1 and L2 differences Normative Stratification • Stratification in the norm using race does not equate to stratification based on culture. • There are currently no tests that stratify based on acculturation. • There are currently no tests that stratify based on amount of time exposed to English. Test Validity • If a test is biased, whatever the reason, you may be testing something other than the construct that you want to examine. • IQ and Achievement tests should measure IQ and Achievement and NOT English Language proficiency!!! Culture-Language Interpretive Matrix (C-LIM) Low Low Mod. High Performance Least Affected Increasing effect of Language Diff. Increasing Effect of Cultural Diff. Performance Most Affected (lang. and Culture) Mod High Kranzler, Flores, & Kody, 2010: Although it makes sense to do it this way current research does not exactly support it. More research is needed. Normative Assessments COGNITIVE IQ and Culture • Scarr (1978): “intelligence tests are not tests of intelligence in some abstract, culture-free way. They are measures of the ability to function intellectually by virtue of knowledge and skills in the culture of which they sample.” • Sample here may be the normative sample; however it may also be the items chosen to be administered as well. 1996: Most common Tests for Bilinguals • • • • WISC (mostly done in all English) Bender Visual-Motor Gestalt Draw-a-Person Leiter Nonverbal Assessment • Conducted PRIMARILY nonverbally – – – – – – UNIT Leiter-R C-TONI Nonverbal DAS-2 Nonverbal SB-5 Nonverbal KABC-2 • May still require some minimal amount of knowledge using receptive language. Native Language Assessment • Most non-English Normative measures are in Spanish only. • Normative Examples Include: – Batería-III – Spanish Version – WISC-IV: Spanish Version – DIAL-3 (has a small cognitive component) • CBM: DIBELS has a Spanish Version Bilingual Assessment • Child can Receive Points for answers in EACH language • Examples Include: – BVAT – Some subtests on the KABC-II Culture-Language Interpretive Matrix (C-LIM) Low Low Mod. High Performance Least Affected Increasing effect of Language Diff. Increasing Effect of Cultural Diff. Performance Most Affected (lang. and Culture) Mod High Kranzler, Flores, & Kody, 2010: Although it makes sense to do it this way current research does not exactly support it. More research is needed. Normative Assessments ACHIEVEMENT Normative-based Achievement Assumptions • Student is like the normative sample. • Student has had opportunity to learn – Content measured – Language used by the test – Skills demanded of the test • Test is measuring ability and not experience Native Language Achievement • Most non-English Normative measures are in Spanish only. • Normative Examples Include: – – – – – Batería-III Bracken Basic Concept Scale: Spanish Version PPVT/TVIP YCAT/PHAI DIAL-3 (more of a screener) • CBM: DIBELS has a Spanish Version Bilingual Assessment Achievement • Child can Receive Points for answers in EACH language • Examples Include: – ROWPVT-SBE – EOWPVT-SBE Nonverbal Achievement • There are no strictly non-verbal achievement tests. • Many times the math computation subtests are so similar from one language to another as to be almost nonverbal. Non-English and Non-Spanish Achievement Assessment • Normative assessment in achievement is pretty much not possible. • Informal methods or CBM methods should be utilized in combination with parent/ teacher interviews, class observations, and child interviews. Native Language Social/ Emotional/Adaptive • Most non-English Normative measures are Spanish translations only. • Normative Examples Include: – – – – BASC- Spanish Translation Vineland- Spanish Translation Conners- Spanish Translation Beck Depression Inventory – Spanish Translation • These tests are only translated into Spanish. They do not include normative data for this population. ELL Normative Testing FACTS TO REVIEW Standards for Testing Individuals of Diverse Language Backgrounds • Testing practice designed to reduce threats to reliability and validity. • Language proficiency should be determined prior to test administration. • Language proficiency should be evaluated across different language tasks. • Any linguistic modifications recommended by test publishers should be in the test manual Standards for Testing Individuals of Diverse Language Backgrounds • When a test is recommended for this population, then developers should provide the info needed for test use and interpretation. • When translating a test, describe methods used in establishing the adequacy of the translation and evidence for reliability and validity data for the translated test’s scores. • When an interpreter is used in testing, the interpreter should be fluent in both languages, should have expertise in translating, and should have a basic understanding of the assessment process.