Deck 5: Reliability and Other Desired Characteristics

Question 1

A reliability coefficient (r) and a standard error of measurement (SEM) were computed on a test with 10 items.What would happen to those statistics if the test were increased to forty items?

A)r would decrease, SEM would increase.
B)r would decrease, SEM would stay the same.
C)r would increase, SEM would decrease.
D)r would increase, SEM would stay the same.

Answer 1

C

Question 2

The standard error of measurement tends to be smallest for which of the following scores?

A)average
B)extremely high
C)extremely low
D)high and low

Answer 2

A

Question 3

The split-half method provides a measure of

A)equivalence.
B)internal consistency.
C)stability.
D)external reliability.

Answer 3

B

Question 4

Which of the following types of reliability provides an inflated reliability coefficient for a speeded test?

A)equivalent-forms (with a time interval)
B)equivalent-forms (without a time interval)
C)split-half method
D)test-retest method

Question 5

The standard error of measurement is especially useful for which of the following calculations?

A)comparing the reliability of different tests
B)converting raw scores to standard scores
C)estimating validity coefficients
D)interpreting individual test scores

Question 6

Reliability can be best determined by

A)analyzing the overall assessment plan.
B)correlating assessment scores.
C)comparing test scores to a criterion.
D)comparing the errors examiners make.

Question 7

A set of test scores is most likely to provide a small reliability coefficient when there is a large

A)number of test items.
B)number of test scores.
C)spread of scores.
D)standard error of measurement.

Question 8

The term reliability is closest in meaning to which of the following terms?

A)consistency
B)objectivity
C)practicality
D)validity

Question 9

Which of the following test characteristics would remain most constant over different groups?

A)reliability coefficient
B)standard error of measurement
C)standard deviation
D)validity coefficient

Question 10

The reliability of a criterion-referenced, performance mastery test focuses on which of the following?

A)decision consistency
B)spread of scores
C)stability of test scores
D)standard error of measurement

Question 11

Which of the following methods of estimating reliability takes into account the greatest number of types of consistency in scores?

A)test-retest (immediate)
B)test-retest (time interval)
C)equivalent-forms (immediate)
D)equivalent-forms (time interval)

Question 12

The standard error of measurement refers to the error involved in which of the following?

A)any assessment
B)computing the standard deviation
C)setting mastery standards
D)standardized testing

Question 13

The term reliability, as used in testing and assessment, refers to which of the following?

A)accuracy of test construction
B)method of test interpretation
C)test or assessment results
D)method of test construction

Question 14

If a test has a standard deviation of 8 and an equivalent-form reliability of .75, the standard error of measurement is

A)2.
B)3.
C)4.
D)6.

Question 15

Whenever the reliability coefficient is low, the standard error of measurement will be:

A)zero.
B)low.
C)high.
D)unchanged.

Question 16

Which of the following methods of estimating reliability is easiest to obtain?

A)equivalent-forms (with a time interval)
B)equivalent-forms (without a time interval)
C)split-half method
D)test-retest method

Question 17

As the reliability coefficient (r) increases the SEM

A)increases.
B)decreases.
C)stays the same.
D)increases at first then decreases to zero.

Question 18

Which of the following types of reliability is the best to assess when students take a pretest and posttest?

A)equivalent-forms method
B)Kuder-Richardson method
C)split-half method
D)D interrater reliability method

Question 19

Which of the following types of reliability provides a measure of internal consistency?

A)test blueprint analysis method
B)equivalent-forms with no time interval
C)Kuder-Richardson method
D)test-retest with a one-month interval

Question 20

A set of test scores may be classified as which of the following?

A)inconsistent and accurate
B)unreliable and valid
C)inaccurate and valid
D)reliable and invalid

Question 21

As SEM increases, what happens to the confidence band?

A)It widens.
B)It narrows.
C)It stays the same.
D)It automatically goes to zero.

Question 22

An achievement test is most useful if it is

A)easy to score.
B)comprised of different question types.
C)reliable and valid.
D)given in essay form.

Question 23

As reliability increases, what happens to the confidence band?

A)It widens.
B)It narrows.
C)It stays the same.
D)It automatically goes to zero.

Question 24

Which of the following would be the best procedure to use for improving the reliability of a classroom test?

A)making items easier
B)making items more difficult
C)increasing the number of items or tasks
D)using more extended-response tasks

Question 25

An SEM for a given test is 3.A student gets a score of 75 on the test.What are the limits of the confidence band on this student's score?

A)69-71
B)72-78
C)79-82
D)80-86