Karteikarten
General
Karten in diesem Stapel (23)
- 1Test method facets environment that influences student’s scores & performance
- 2- Construct irrelevant(time of the day, temperature, unintended background noise)
- 3- Construct relevantability to write a given number of words, relationship between input and desired response)
- 4Test construct• We cannot measure the test construct but have to make it measurable = operationalize it.
• = what we want to have info about (e.g. fluency)
• Construct definition can be
o Syllabus-based
o Theory-based
o operational
• e.g. “Students can build arguments from graphical information - 5Method vs. construct• Method (=essay) vs. construct (=build argument)
• Same method can be used for testing different constructs (e.g. essay)
• One construct can be tested with different methods
• We need the right method for a certain construct
• Method = vehicle by which we can elicit performance
• Test method effect = the effect that a method has on scores, should be minimized - 6Info Test Types- An achievement test cannot be used as a proficiency test
- A proficiency test can also be used as a placement test
- Diagnostic tests can be short
- An achievement test can have diagnostic value
- Proficiency tests are usually long
- Placement tests can be short
- New matura = Primary purpose is proficiency, so it is a proficiency test - 7Direct testing Advantageseasier to make, assessment easier
- 8Indirect testing Problem and Advantage• Problem: relationship between performance on test tasks and real performance of the skills is unclear
• Advantage: a huge variety of tasks can be used (in contrast to direct testing - 9Discrete point testing• Testing one element at a time. Series of items used.
• Almost always indirect - 10Integrative testing• Student has to combine many language elements
• Almost always direct - 11Norm-referenced testing Advantage• Advantage: not as restricted as criterion referenced testing
- 12• Reliability coefficiento Ideally = 1 test yields precisely the same results twice
o Worst case = 0
o Usually = 0.70 – 0.99
o To calculate it: You need two sets of scores for comparison
Test-retest-method = make a group take the same test twice drawback: students will remember
Alternate forms method = use two different forms of the same test often not available
Split half method = Test is split in two halves. Student gets a score for each half. - 13Test items• Item = a testing point, receives a 0 or 1 score.
- 14• Task= a collection of items
- 15• Methodtasks + scoring (counting our judging)
- 16• Elicitation modewhat we give students to make them produce the target form.
Can be
o oral (e.g. students listen to a monologue) or
o written (students read a question) - 17• Response modehow students should respond to the tasks. Can be
o oral (e.g. students describe sth) or
o written (e.g. students fill in the blank) - 18• Items in an oral exam (example):o Warm-up (e.g. greetings, non-scored item)
o Level-check questions (tests the objectives)
o Probe (tests a bit beyond the objectives)
o Wind-down (e.g. praising, non-scored item) - 19MC questions MC-items are receptive items, also called selective response items = test takes choose from a set of responses = the supply
The one correct option is the key the other options are distractors - 20 Item facility= item difficulty for the students
- 21 Item discriminationextent to which an item differentiates between high- and low-ability students
- 22 Distractor efficiencyextent to which a distractor lures a sufficient number of test-takers
- 23integrating tests vs. integrative tests• Integrating tests (= e.g. tests where students need to read something in order to do a writing task, reading necessary on a low level) vs. integrative tests (= e.g. test where two aspects are tested at once, see above)
Informationen
- Beschreibung
- Testing and Assessment Karteikarten am General, erstellt von Lisza Neumeier am 21/05/2019.
- Erstellt von
- Lisza Neumeier
- Aufrufe
- 2
- Bewertung