14. Three Categories of Language Assessment Selected Response True/False, Matching, Multiple Choice Conference, Portfolios, Self or Peer Assessment Personal Response Constructed Response Fill-In, Short Answer, Performance
15. Selected-Response Assessment (True/False, Matching, and Multiple Choice) Students choose correct answer from a limited set of options. Best used to measure listening an reading skills Quick to administer Scoring is quick, easy and relatively objective Difficult to construct Students do not use any productive language Advantages Disadvantages
16. True False Assessment Advantages Focus on students’ abilities to select from two alternatives Simple and direct indication if a particular point is understood Disadvantages Writers may write tricky answer choices High guessing factor Emphasis on details and unimportant facts Difficult for listening and reading passages Matching Assessment Advantages Low guessing factor Takes up little space Disadvantage Restrictive in measuring ability Multiple Choice Assessment Advantages Low guessing factor Measures wide variety of learning points Can provide useful information about student’s ability Disadvantage Real life language is not multiple choice
17. Constructed-Response Assessment Fill-In, Short- Answer, and Performance Measures productive skills of speaking and writing Can observe interactions of receptive and productive skills Relatively objective Low guessing (bluffing) factor Create problems for subjectivity Scoring is time consuming Advantages Disadvantages
18. Fill-In Assessment Advantages Easy to make and administer Measures ability to produce small amount of language Possibility of assessing interaction between receptive and productive skills Disadvantages Narrow focus Blank may have multiple possibilities Short Answer Assessment Advantage Easy to make and administer Disadvantages Assesses only a few phrases or sentences Multiple answers are possible Performance Advantages Comes close to getting authentic communication Measures students’ abilities to respond to real-life language tasks Counteracts negative washback effects in standardized testing , like bias Disadvantages Difficult to make and time-consuming to administer Various costs are involved Reliability, validity, and test security may be problematic
19. Personal-Response Assessment Conferences, Portfolios, Self and Peer A. Produces language B. Provides: a. individualized assessment, b. ongoing assessment c. rich forms of feedback Difficult to produce and organize B. Scoring is subjective Advantages Disadvantages
20. Portfolios Advantages Strengthen students’ learning Enhance teacher role Improve testing processes Disadvantages Variety in design can be difficult to assess Issues with logistics Subject to interpretation d. Limited reliability e. Questionable validity Conferences Advantages Foster students reflection of their own learning process Develop better self-image Elicit answers Gather information Disadvantages Time consuming Subjective to grade Typically not scored
21. Personal-Response Assessment (continued) Selfand Peer Assessments Advantages Quickly administered Students are directly involved in assessment Help students understanding of the language process Increase motivation to learn Disadvantages Students’ self-estimates are not always accurate May be affected by subjective errors
22. Washback Positive Consequence Assessment corresponds to course goals and objectives. Negative Consequence Assessment does not correspond to course goals and objectives. Feedback Important in diagnostic and achievement testing Essential part of learning process “The assessment procedures used within a particular language program must be directly related to the curriculum if that feedback is to be maximally useful.” (Brown and Hudson, 1998, p.669 )
23. Brown and Hudson (1998) , make two statements about assessments that most teachers should already know: 1. Using multiple sources of information in designing and selecting assessments is a key factor in interpreting assessment. (p. 671) 2. Assessments based on multiple observations are generally more reliable than assessments based on few observations. (p. 671)