영어교육론 - Assessment(2)

* types of assessment

 

1. formative / summative assessment (+ high stake standardized tests)

① formative assessment - in the course, assess developmental process with feedback - on-going evaluation

② summative assessment - summarize and measure what students have grasped(learned) by product with summative feedback at the end of the course(결과에 대한 피드백), total score, percentile rank

 

2. informal / formal assessment(=test - that is carefully designed, time-constrained, and has scoring rubrics)

① informal assessment - unplanned, spontaneous, and incidental comments to give diagnostic or formative feedback without recording results, judgement of their achievements, or specific scoring or grading formats

② formal assessment - planned, systematic techniques constructed to give teacher and student an appraisal(평가) of student achievement (+ all tests are formal assessments, but not all formal assessments are tests)

 

3. norm-referenced / criterion-referenced test

① norm-referenced test - 상대 평가(place Ss in rank order)

② criterion-referenced test - 절대 평가(designed to give Ss feedback according to the developmental progress)

 

4. proficiency / achievement test (+ placement and diagnostic test(+한 번에 하나의 skill만))

① proficiency test - global competence in a language - low content validity - 한 번에 여러 개 skills를 평가

② achievement test - determine acquisition of objectives at the end of a period of instruction

 

5. discrete-point / integrative assessment(testing)

① discrete-point assessment - assess only a single independent linguistic point at a time, out of context

② integrative assessment - assess combination of many linguistic knowledges(skills) at a time, contextualized

 

* integrative test - in terms of the number of skills{ex) reading, listening} assessed

1. cloze test - a reading passage that is mutilated by the deletion of word, integrative(requires many abilities)

① deletion - ⑴ random deletion(fixed ratio deletion), ⑵ rational deletion(selective deletion)

 

+ scoring difficulty(due to contextual inferencing) and washback is larger in rational deletion

② scoring method - ⑴ exact word method(high scoring reliability), ⑵ acceptable word method [using content schema(pragmatic expectancy), formal schema(linguistic expectancy), creativity, and strategic competence]

 

2. c-test - the second half of every word is deleted - high scoring reliability

반응형

* alternatives in assessment(=alternative assessment)

1. traits(performance-based assessment) 

⑴ assess both process and product

⑵ directly observable performance(direct testing - 단, 이는 selected response도 가능)

⑶ constructed response items

⑷ integration of language skills

⑸ losses in practicality and reliability are made up for in higher validity, washback, authenticity

 

2. types(procedure) (+ interviews and observation)

① portfolios - a purposeful collection of students' performances that demonstrates their efforts, progress, and achievements - student-centered, responsibility and autonomy, interaction between T and S, ample feedback

② journals(dialogue journals) - written conversation between T and Ss - pay little attention to structure, form, or correctness, but focus on Ss' learning progress, affective states(create rapport) / + journal은 일기장 같은 거

③ conferences - one-on-one interaction between T and Ss, direct and formative feedback and assessment

④ self or peer assessment - direct involvement of students, encourage high autonomy(Ss take control of their learning) and motivation, T provides feedback(washback), use checklist as a guideline

 

* Items

1. selected-response / constructed response test formats(items)

① selected-response items - 이미 정해진 답 중에 고르는 것

② constructed-response items - Ss respond to a series of open-ended questions with the target language and Ts need expert judgement in the scoring

 

2. item analysis (적절한 문항인지) - internal consistency(reliability)와 연관

① item facility(난이도) - 0.0(모두 틀림), 1.0(모두 맞춤) - the number of correct responses divided bt that of total test takers

② distractors analysis(efficiency) - 오답지가 제 기능을 하고 있는지 - 한 명도 안 고른 오답은 제외

③ item discrimination(변별도) - 상위권과 하위권을 얼마나 구분해 내느냐 - 0.25~0.35가 변별력이 있고 0이하 일 때는 internal consistency에서 오류

④ item response distribution - 오답지 분석과, 변별도를 동시에 - 하위권 학생이 상위권 학생보다 정답률이 더 높은 문제(internal consistency에서 오류)를 제외

+ distractors - incorrect alternatives(대안) or options(보기) in multiple choice test items, stems(질문)

 

3. multiple-choice test(indirect test - infer competence) tasks - principles of making alternatives

① inadvertent clues, which allow students to answer correctly without the proper knowledge는 금지

⑴ option의 grammatical/semantic category를 맞추기 / ⑵ world knowledge 활용금지

⑶ passage dependency - 주어진 문맥 안에서만 해결 / ⑷ item dependency - 다른 문제에서 힌트를 얻으면 X

② context를 포함하여 authentic해야한다.