Question

1. Two versions of a test are administered to a reference group of participants and based upon the correlation acrosss participants between the two occasions a determination is made about the test.

a. construct validity

b. test-retest reliabilty

c. criterion-referenced measurement

d. Spearman Brown prophecy formula

e. Cronbach alpha reliabilty formula

f. Criterion-related predictive validity

g. Criterion-related concurrent validity

h. inter-rater reliabilty

i. Content validity

j. Alternate or equilvalent forms of reliabilty

l. Cronbacj alpha validity coefficient

m. Kuder-Richardson reliabilty formula

2. A test is administered one time to a reference group and each item is scored right or wrong. Based upon how each item contributes to the total score a determination is made concerning the total score on the test.

a. construct validity

b. test-retest reliabilty

c. criterion-referenced measurement

d. Spearman Brown prophecy formula

e. Cronbach alpha reliabilty formula

f. Criterion-related predictive validity

g. Criterion-related concurrent validity

h. inter-rater reliabilty

i. Content validity

j. Alternate or equilvalent forms of reliabilty

l. Cronbacj alpha validity coefficient

m. Kuder-Richardson reliabilty formula

3. Two tests are administered to a reference group. One is a new intelligence test and the second is the Stanford-Binet intelligence test. A correlation is computed between the two tests to make a determination.

a. construct validity

b. test-retest reliabilty

c. criterion-referenced measurement

d. Spearman Brown prophecy formula

e. Cronbach alpha reliabilty formula

f. Criterion-related predictive validity

g. Criterion-related concurrent validity

h. inter-rater reliabilty

i. Content validity

j. Alternate or equilvalent forms of reliabilty

l. Cronbacj alpha validity coefficient

m. Kuder-Richardson reliabilty formula

4. A test is administered ine time to a reference group and the test is divided into two parts. The parts are sored seperately, a correlation is computed, and a determination is made using a special estimation technique. What is the special estimation tehnique called?

a. construct validity

b. test-retest reliabilty

c. criterion-referenced measurement

d. Spearman Brown prophecy formula

e. Cronbach alpha reliabilty formula

f. Criterion-related predictive validity

g. Criterion-related concurrent validity

h. inter-rater reliabilty

i. Content validity

j. Alternate or equilvalent forms of reliabilty

l. Cronbacj alpha validity coefficient

m. Kuder-Richardson reliabilty formula

5. A potentially new test (The Goldwater r test of conservatism) is administered to a reference group along with several existing tests. The conservatism test correlated -.78 with intelligence, -.33 with amount of education, and +.45 with hostility all of which were consistent with a prominent theory. A determination was made.

a. construct validity

b. test-retest reliabilty

c. criterion-referenced measurement

d. Spearman Brown prophecy formula

e. Cronbach alpha reliabilty formula

f. Criterion-related predictive validity

g. Criterion-related concurrent validity

h. inter-rater reliabilty

i. Content validity

j. Alternate or equilvalent forms of reliabilty

l. Cronbacj alpha validity coefficient

m. Kuder-Richardson reliabilty formula

Answer 1

1. Appropriate test based on correlation when two version of test administered to reference group:

INTER RATER RELIABILITY is the best test used in this condition since it is used to observe the same behaviour independently between two groups and compare the results

EXAMPLE: Job performance assessment by office managers. Suppose if an employee receives a score of 9 from one group of managers and a score of 2 from other group of managers. Then inter rater reliability is used to determine that something is wrong with method of scoring and thus exposes the issue so that it can be corrected.

3. Appropriate test based on correlation when two tests are administered to a reference group. One is a new intelligence test and the second is the Stanford-Binet intelligence test.

Alternate and equivalent forms of reliability is used where two tests are given to same sample of reference group and estimate of reliability is calculated from two groups. On a simple note you are trying to find out that test 1 is measures the same thing as test 2.

In other words you want to know if two test scores stay the same.

4. Special estimation technique used when test is administered in onee time to a reference group and the test is divided into two parts.

Spearman brown prophecy formula is used in evaluating test reliability.

It takes split half correlation as input which only estimates reliability of half length test and implements correction to find true estimate of full length test.

rho*= n*rho / [1+(n-1)*rho]

Here formula takes split half reliability (rho) and number of tests (n) as input and produces full length equivalent (rho*)