2 - Reliability

• Reliability

• Validity

• Scales

PRINCIPLES of Psychological Testing: (3)

Reliability

refers to the accuracy, dependability, consistency, or repeatability

Reliability

👉 If you take the test again, you should get similar results.

• Test-Retest

• Alternate/Parallel Forms

• Internal Consistency

• Interrater

(4) Types of RELIABILITY

Test-Retest Reliability

-measures consistency over time

Test-Retest Reliability

correlation coefficient indicates the relationship between scores obtained by individuals within the same group on two administration of the test

Test-Retest Reliability

👉 Same test

👉 Same group of people

👉 Two different times

decrease

Test-retest correlations tend to _____ as the interval between the test administration lengthens

(The longer the gap between the first and second test, the more likely scores will differ.)

Alternate/Parallel-Forms Reliability

computed by comparing consistency of scores of individuals within the same group on two alternate but equivalent forms of the same test

Alternate/Parallel-Forms Reliability

• Two forms (A & B) of the test

• Same content/skills measured

• Similar difficulty

• Administered to the same group

If scores are similar → high reliability

Alternate/Parallel-Forms Reliability

two tests that are equivalent in difficulty level can be administered on the same day

• or very close to each other without concern about a practice effect

Internal Consistency

Provide an estimate of test score reliability that indicates the consistency of responses to the different items or parts of a test during a single test and administration

Internal Consistency

👉 Do the questions agree with each other?

Internal Consistency

A 20-item anxiety test is given once to students. If all items show similar results and measure anxiety consistently, the test has high ___________

• Split-half

• Interim Consistency

2 Common Measure of INTERNAL CONSISTENCY:

Split-half

👉 divided into two equal halves, (even, odd)

👉 scores of both halves are compared,

👉 to check if they give similar results.

If both halves give similar scores, the test is reliable.

It is given only once (single administration).

• Spearman-Brown Formula

Split-half:

Spearman-Brown formula

used to correct the computed reliability based on shorter length

Spearman-Brown formula

yields an estimate of what reliability would be if it were observed on the test’s full length

Spearman–Brown formula

A 40-item test is split into two 20-item halves and correlated. The reliability becomes lower because each half is shorter.

• The ____________ is used to estimate the reliability of the complete 40-item test.

Interim Consistency

assesses the extent to which the items on a test are related to each other and to the total score

Interim Consistency

provides an estimate of the average intercorrelations between all of the items on a test

• Kuder-Richardson (KR) formula 20

• Cronbach’s Alpha

Interim Consistency: (2)

Kuder-Richardson (KR) formula 20

for two response answers such as true or false; yes or no

Cronbach’s Alpha

more than 2 alternatives are computed

• Multiple choice, rating scales

Interrater Reliability

refers to the degree of agreement between two or more independent judges

Interrater Reliability

high level of subjectivity

Alternate/Parallel-Forms Reliability

Set A and Set B test (on the same day)

Test-Retest Reliability

Pre-test and Post-test

Interrater Reliability

Singing competition

Internal Consistency (Split-Half Reliability)

Dividing 100 test items to odd and even numbers