Placement testing

Abstract: This is an invaluable guide to language testing and assessment. For many teachers of English language learners, the field of assessment is foreign territory. Assessment has its own culture, traditions, and terminology. This training guide is intended to help classroom teachers become more comfortable creating and using assessments. "A Practical Guide to Assessing English Language Learners" provides helpful insights into the practice and terminology of assessment. The text focuses on providing the cornerstones of good assessments - usefulness, validity, reliability, practicality, washback, authenticity, transparency, and security - and techniques for testing. It devotes a chapter to the assessment of each of the four main skill areas (reading, writing, listening, and speaking) and also covers placement testing (such as using TOEFL[registered] and MELAB), diagnostic testing, evaluation, and instructional decision-making with regard to testing. Tips to improve students' test-taking strategies are offered, and each chapter ends with a helpful list of Ten Things to Remember, as well as informative case studies featuring two teachers and their assessment decisions.

Abstract: The need for an educated work force and a college education has increased over the last several decades. However, a large number of entering college students are not prepared for college-level course work. There is a widespread need for remedial education at colleges and universities across the country, increasing costs to students and the public for education that students should have successfully completed in high school. Although recent literature often documents this problem, for several years the public has failed to fully recognize and understand this issue, and schools in general have been slow to respond to the need for change. These factors have led to imminent federal legislation requiring greater accountability and increased testing in public schools. "In 1995, nearly all public two-year institutions and 81 percent of public four-year institutions offered remedial courses" (Kirst, 1998, p. 76). In Georgia, "30 percent of.. students who graduated with college preparatory diplomas in 1995 took remedial courses in college" (Sandham, 1998c, p. 25). In New York, "only 13 percent of CUNY community college students pass[ed] three basic skills tests measuring 11th grade proficiency" (Sandham, 1998c, p. 25). College remediation rates for students were 46% in Maryland and 60% in Florida (Malooney, 1996). The California State University System reported that "47 percent of freshmen had to take remedial English, and 54 percent enrolled in remedial math" (Kirst, 1998, p. 76). At some California campuses in the system, 80% to 90% of the freshmen needed remedial education (California Community Colleges, 1995; Manzo, 1996; Ponessa, 1996), These remediation rates exist in a system that is supposed to take the top 30% of high school graduating classes in the state. In 1995, the National Center for Education Statistics (NCES) found that 29% of all freshmen required remedial education at 4-year colleges and universities. At community colleges "the figure was 41 percent" (Ignash, 1997, p. 9). The "chain of blame" is a metaphor for the discourse taking place in the educational community. It describes how "universities blame the high schools, the high schools blame the middle schools, and the middle schools blame the elementary schools for poor [student] preparation" (Ponessa, 1996, p. 31). A lack of adequate funding and overcrowding in classes are often cited as reasons for poor results (Ignash 1997, p. 5). Educators express concerns that student failure to take college preparatory courses, grade inflation, and a lack of academic rigor in high school courses all contribute to the need for remediation in college. College administrators report that "there's been so much grade inflation, the colleges don't know what they've got until they get it" (Sandham, 1998c, p. 25). Recommendations for "more rigorous work requirements in order to reduce grade inflation" (Bandy, 1985, p. 88) have been advanced, and many educators also advocate "an end to social promotion and an emphasis on intervention" (Feldman, 1997, p. 9). College teachers report that students "have gone through high school math classes without gaining a real understanding of the subject matter. `We've found that they've failed in high school, but somehow there's a C on their report card"' (Ponessa 1996, p. 31). English instructors in high schools state: "Our basic is kind of like remedial, and our so-called honors is more like a college-prep borderlining on basic....If we go strictly by the book, half of my honors class would have failed" (Ponessa, 1996, p. 32). A prior study conducted by Lappan and Phillips (1984) found that nearly 70% of students enrolled in intermediate algebra at a university had taken 3 to 4 years of math in high school at the algebra 1 level and above. In the same study, another 42% of students in elementary algebra had taken 2 to 3 years of college preparatory math. Many of these students earned C grades in their high school college preparatory courses. …

Abstract: This chapter describes the concepts of scaling, norming, and equating, as they pertain to assessment in higher education. Issues associated with scaling, norming, and equating in admissions testing, placement testing, and assessments used for outcomes and accountability are then discussed. Detailed examples of how scaling, norming, and equating procedures are used in the ACT (ACT, 2015a) and SAT (College Board, 2016c) admissions testing programs, and the Advanced Placement program are provided. A variety of scores and scales are used with educational tests. Raw scores can be as simple as a sum of the item scores, or so complicated that they depend on the entire pattern of item responses, as is sometimes the case with item response theory (IRT)-based scoring. Theoretical considerations for scoring, scaling, norming, and equating are described. Incorporating normative information begins with the administration of the test to a norm group. The implementation of test equating requires a process for collecting data, referred to as an equating design.