Design and scoring Standardized test

some standardized testing uses multiple-choice tests, relatively inexpensive score, form of assessment can used.

standardized testing can composed of multiple-choice questions, true-false questions, essay questions, authentic assessments, or other form of assessment. multiple-choice , true-false items chosen because can given , scored inexpensively, quickly, , reliably through using special answer sheets can read computer or via computer-adaptive testing. standardized tests have short-answer or essay writing components assigned score independent evaluators use rubrics (rules or guidelines) , benchmark papers (examples of papers each possible score) determine grade given response. not standardized tests involve answering questions; authentic assessment athletic skills take form of running set amount of time or dribbling ball distance.

most national , international assessments, however, not evaluated people; people used score items not able scored computer (such essays). example, graduate record exam computer-adaptive assessment requires no scoring people except writing portion.

the term normative assessment refers process of comparing 1 test-taker or peers. norm-referenced test (nrt) type of test, assessment, or evaluation yields estimate of position of tested individual in predefined population. estimate derived analysis of test scores , other relevant data sample drawn population. type of test identifies whether test taker performed better or worse other students taking test. criterion-referenced test (crt) style of test uses test scores show whether or not test takers performed on given task, not how performed compared other test takers. tests , quizzes written school teachers can considered criterion-referenced tests. in case, objective see whether student has learned material.

scoring issues

human scoring relatively expensive , variable, why computer scoring preferred when feasible. example, critics poorly paid employees score tests badly. agreement between scorers can vary between 60 , 85 percent, depending on test , scoring session. states pay have 2 or more scorers read each paper; if scores not agree, paper passed additional scorers.

open-ended components of tests small proportion of test. commonly, major academic test includes both human-scored , computer-scored sections.

score

there 2 types of standardized test score interpretations: norm-referenced score interpretation or criterion-referenced score interpretation.

norm-referenced score interpretations compare test-takers sample of peers. goal rank students being better or worse other students. norm-referenced test score interpretations associated traditional education. students perform better others pass test, , students perform worse others fail test.
criterion-referenced score interpretations compare test-takers criterion (a formal definition of content), regardless of scores of other examinees. these may described standards-based assessments, aligned standards-based education reform movement. criterion-referenced score interpretations concerned solely whether or not particular student s answer correct , complete. under criterion-referenced systems, possible students pass test, or students fail test.

either of these systems can used in standardized testing. important standardized testing whether students asked equivalent questions, under equivalent circumstances, , graded equally. in standardized test, if given answer correct 1 student, correct students. graders not accept answer enough 1 student reject same answer inadequate student.