Q1. A test report gives both a percentile rank and a list of mastered objectives. The report combines

• (A) Two norm groups only
• (B) Norm-referenced and criterion-referenced information
• (C) Two unrelated tests
• (D) Only formative data

Q2. Which error is common when reading norm-referenced results?

• (A) Reading the rank as a peer comparison
• (B) Using norms from a similar group
• (C) Treating a small percentile difference as a large learning difference
• (D) Checking the test date

Q3. A criterion-referenced test with a very low pass standard may

• (A) Pass students who still lack important skills
• (B) Rank students more accurately
• (C) Become a norm test automatically
• (D) Measure more objectives

Q4. A norm-referenced test with a small, unrepresentative norm group may

• (A) Become more accurate for all
• (B) Prove mastery for every student
• (C) Give misleading comparisons for other students
• (D) Remove the need for scoring

Q5. A school that reports only ranks may hide the fact that

• (A) One student is first
• (B) The whole class performed poorly against the syllabus
• (C) The test was given on time
• (D) The class size is large

Q6. A teacher who sets different pass standards for different classes without reason violates

• (A) Norm-referenced ranking
• (B) Fairness in criterion judgement
• (C) Formative feedback
• (D) Test printing rules

Q7. For a job that requires safe performance, the defensible test type is

• (A) Norm-referenced ranking of applicants only
• (B) A popularity vote
• (C) Criterion-referenced, tied to required performance
• (D) An attendance record

Q8. A gifted programme that admits the top 2 percent uses mainly

• (A) Norm-referenced selection
• (B) Criterion mastery for all
• (C) Diagnostic teaching
• (D) Daily formative quizzes

Q9. A test that reports ‘mastered’ or ‘not yet mastered’ for each skill is most useful for

• (A) Ranking schools nationally
• (B) Awarding a single prize
• (C) Counting enrolment
• (D) Planning the next teaching step

Q10. A student with a high GPA from only easy courses and a student with a slightly lower GPA from hard courses cannot be fairly compared on GPA alone because

• (A) GPA always shows course titles
• (B) GPA counts attendance only
• (C) GPA is never numeric
• (D) GPA does not show course difficulty

Q11. A school adds effort marks to achievement grades. The resulting grade becomes

• (A) More valid for achievement alone
• (B) Easier to compare across schools
• (C) Free from teacher judgement
• (D) A mixed measure, not a pure achievement grade

Q12. A fair grading policy should be

• (A) Shared with students before assessment begins
• (B) Changed secretly after marking
• (C) Based on a single surprise test
• (D) Kept hidden from parents

Q13. A teacher averages scores from very different tasks without weighting. A long project and a short quiz then count

• (A) By their scope automatically
• (B) Only if they are oral
• (C) By student preference
• (D) Equally, even though their scope differs

Q14. A borderline student just below a grade boundary deserves

• (A) An automatic fail without review
• (B) A grade based on age
• (C) A grade based on class size
• (D) A review of the full evidence, not an automatic decision by one mark

Q15. A grading system that allows reassessment after feedback best reflects

• (A) Ranking as a fixed contest
• (B) A single final performance only
• (C) Learning as a process that can improve
• (D) Attendance as achievement

Q16. A school reports grades with written comments on strengths and gaps. This improves on letters alone because it

• (A) Makes ranking faster
• (B) Removes the need for tests
• (C) Guarantees higher grades
• (D) Explains what the grade means for next learning

Q17. When a course uses both a written test and a practical task, the final grade is more valid if

• (A) Only the written test counts
• (B) Only the practical task counts
• (C) Marks are assigned by chance
• (D) Both parts reflect the course objectives in planned proportions

Q18. A GPA requirement for a scholarship may disadvantage students from schools with

• (A) Larger libraries
• (B) Stricter grading standards
• (C) More teachers
• (D) Longer school days

Q19. A standardized test is best used to answer which question?

• (A) What did one student say in class today?
• (B) How does this group compare with a wider reference group?
• (C) What is the best lesson plan for tomorrow?
• (D) Which textbook colour is preferred?

Q20. A test manual should state the test’s purpose, norms and evidence so that users can judge

• (A) Whether the test fits their intended use
• (B) The exact score a student will obtain
• (C) The pass mark for every school
• (D) The best textbook for the course

Q21. A standardized test given to students who have not been taught the content will mainly measure

• (A) Teaching quality in that course
• (B) Student effort in that course alone
• (C) Future potential perfectly
• (D) Prior exposure, not school learning from the course

Q22. A school that uses standardized results to label a whole class as weak without checking teaching may commit

• (A) Perfect use of data
• (B) Over-interpretation of one score source
• (C) A valid diagnosis for every student
• (D) A fair ranking of teachers

Q23. Test security matters for standardized tests because leaked items would

• (A) Destroy comparability and fairness
• (B) Improve the norms
• (C) Make scoring faster
• (D) Reduce test anxiety for all

Q24. A standardized test that is too long may reduce validity because

• (A) Long tests are always more reliable
• (B) Students enjoy long tests more
• (C) Fatigue affects later performance
• (D) Scoring becomes impossible

Q25. A norm group from ten years ago may still be usable if

• (A) The test is printed in a new colour
• (B) More students take it now
• (C) The population and curriculum have not changed much
• (D) Teachers like the test

Q26. A good standardized test report for teachers should include

• (A) What the score means and what to do next
• (B) Only the school average for the year
• (C) Only the raw score with no interpretation
• (D) Only the date of the next test

Q27. A teacher who only gives summative tests will struggle to

• (A) Catch misunderstandings early enough to fix them
• (B) Award final grades
• (C) Rank students at term end
• (D) Report to parents

Q28. A diagnostic test given to a whole class at the start of a unit can help the teacher to

• (A) Award final grades at once
• (B) Group students by specific needs for the unit
• (C) Rank schools in the district
• (D) Replace all teaching

Q29. Formative assessment that only praises without pointing to improvements is weak because it

• (A) Is always summative
• (B) Does not guide the next learning step
• (C) Cannot be given orally
• (D) Takes too much time

Q30. A summative test that samples the full syllabus fairly is stronger than one that samples

• (A) Only the easiest chapter
• (B) All chapters equally
• (C) All objectives in the blueprint
• (D) Both theory and practice

Q31. A student who fails a diagnostic check in fractions but passes whole-number work needs

• (A) A repeat of the entire maths course
• (B) No help until the final test
• (C) Fraction-focused support, not a repeat of all maths
• (D) A lower grade in every subject

Q32. A school that uses formative data in staff meetings to adjust teaching is using assessment for

• (A) Admission only
• (B) Building maintenance
• (C) Public ranking only
• (D) School improvement, not only student grading

Q33. A portfolio that shows drafts, feedback and revisions over a term best supports

• (A) Formative judgement of growth during learning
• (B) A single final rank only
• (C) An admission decision only
• (D) A count of attendance

Q34. A diagnostic test should be followed by a recheck to confirm that

• (A) The student attended school
• (B) The weakness has actually improved after help
• (C) The test was printed correctly
• (D) The class rank changed

Q35. Which practice best balances formative and summative assessment?

• (A) Only one final test with no checks during the term
• (B) Only daily quizzes with no final summary
• (C) Regular low-stakes checks for learning plus a fair final judgement
• (D) No assessment until graduation

Q36. A constructor who writes items that match the blueprint but uses biased language has harmed

• (A) Fairness, even if coverage is good
• (B) Coverage only
• (C) Test length only
• (D) Printing quality only

Q37. A test with a clear blueprint but no tryout may still contain

• (A) Perfect items in all cases
• (B) Faulty items that tryout would have revealed
• (C) No scoring key
• (D) No instructions

Q38. A constructor who keeps an item because it is interesting, even though it does not match any objective, harms

• (A) Content validity
• (B) Test length alone
• (C) Scoring speed alone
• (D) The number of items alone

Q39. A test that allows too little time for slow writers measures partly

• (A) Writing speed, not only subject knowledge
• (B) Only subject knowledge
• (C) Only reading speed in all cases
• (D) Nothing at all

Q40. A scoring rubric for an essay test improves grading mainly by

• (A) Making essays shorter
• (B) Removing the need for a blueprint
• (C) Making criteria explicit and scoring more consistent
• (D) Guaranteeing high marks

Q41. A constructor who reviews items for clarity, bias and key accuracy before printing is doing

• (A) Final quality review
• (B) Syllabus design
• (C) Student counselling
• (D) School inspection

Q42. A test that mixes very easy and very hard items without a plan may fail to

• (A) Measure the middle range of ability well
• (B) Print correctly
• (C) Include a title
• (D) Use four options

Q43. A test constructor should document the test’s purpose, blueprint and tryout results so that future users can

• (A) Print more copies quickly
• (B) Judge the test’s quality and limits
• (C) Change the syllabus
• (D) Rank students faster

Q44. The most defensible reason to drop a tryout item is

• (A) That the constructor dislikes the topic
• (B) That students found it interesting
• (C) Evidence that it is faulty, biased or does not measure the objective
• (D) That it appears late in the paper

Q45. A class mean is 55, the median is 58, and one score is 100. Removing the 100 will most likely

• (A) Raise the mean
• (B) Leave both unchanged
• (C) Lower the median more than the mean
• (D) Lower the mean more than the median

Q46. Two distributions have the same range but one has a larger standard deviation. This means the larger-deviation group has scores

• (A) Identical in every way
• (B) Higher on average
• (C) Lower on average
• (D) More spread around the mean, not just wider extremes

Q47. A score at the 84th percentile in a normal curve is about

• (A) One standard deviation above the mean
• (B) At the mean
• (C) Two deviations below the mean
• (D) At the lowest score

Q48. A teacher reports a class average of 70 with a standard deviation of 2. The class performance is best described as

• (A) Low average with wide spread
• (B) Average with no pattern
• (C) High average with very similar scores
• (D) Impossible to describe

Q49. If a test is made easier, the mean will usually rise and the distribution may become

• (A) Normal in all cases
• (B) Positively skewed in all cases
• (C) Negatively skewed, with scores piling at the high end
• (D) Unchanged in shape

Q50. A quartile deviation is based on

• (A) Only the highest score
• (B) Only the lowest score
• (C) The middle 50 percent of scores
• (D) The full range only

Q51. A teacher who wants to know the most common score in a class should use the

• (A) Mean
• (B) Mode
• (C) Range
• (D) Z-score

Q52. A frequency polygon is useful for

• (A) Listing student names
• (B) Recording attendance
• (C) Setting a pass mark alone
• (D) Showing the shape of a score distribution

Q53. A correlation between study time and test scores does not by itself prove that

• (A) Study time was recorded
• (B) Scores varied
• (C) More study caused the higher scores
• (D) Students studied at all

Q54. A teacher compares a student’s z-score of +1 on Test A and -0.5 on Test B. The student did relatively

• (A) Better on Test B than on Test A
• (B) Better on Test A than on Test B
• (C) Equally well on both
• (D) Worse on both in absolute terms

Q55. A school adopts CIPP because it wants evaluation to inform choices throughout a programme, not only at the end. This reason fits CIPP because CIPP is

• (A) Focused only on final grades
• (B) Decision-oriented across stages
• (C) Limited to one lesson
• (D) Against using objectives

Q56. A Tyler evaluation that finds objectives unmet should lead to

• (A) Blaming students without review
• (B) Hiding the results
• (C) Reviewing teaching, objectives or assessment for mismatch
• (D) Repeating the same teaching unchanged

Q57. In CIPP, context evaluation answers which question?

• (A) How well was it taught day by day?
• (B) What resources are available?
• (C) What were the final outcomes?
• (D) What needs should the programme address?

Q58. In CIPP, input evaluation answers which question?

• (A) What needs exist in the community?
• (B) What outcomes were achieved?
• (C) How was teaching delivered?
• (D) What plan and resources should be used?

Q59. In CIPP, process evaluation answers which question?

• (A) What needs exist before planning?
• (B) Is the programme being implemented as planned?
• (C) What were the final results?
• (D) Which students passed?

Q60. In CIPP, product evaluation answers which question?

• (A) What resources were planned?
• (B) What needs were found?
• (C) How were lessons delivered?
• (D) Were the intended outcomes achieved, and should the programme continue?