Chapter 4

Question 1

What’s sufficient for a ‘good test’?

Answer 1

Reliability AND Validity

Question 2

Reliability

Answer 2

Consistency. Does not have to be valid. (can be consistently wrong!)

Question 3

Validity

Answer 3

How well a test measures the definition of its concept

Question 4

norm-referenced testing and assessment

Answer 4

Derive meaning from test scores by comparing an individual’s test score with a group of testtakers.

Question 5

Normative sample

Answer 5

group of people’s performance analyzed for reference to evaluate individual testtakers

Question 6

Norming

Answer 6

process of deriving norms

Question 7

Stratified Sampling

Answer 7

Samples that include different subgroups of a population

Question 8

Purposive Sampling

Answer 8

Selecting a sample that you think is representative (eg. Cleveland to infer nation)

Question 9

Incidental sampling

Answer 9

Also convenience sampling. Using sample that’s convenient

Question 10

Standardization sample vs. Normative sample

Answer 10

After standardization, there might be changes in normative sample to make in representative (eg. more Hispanic immigration, greater proportion needed). They could be different.

Question 11

Percentile

Answer 11

Expresses the percentage of people whose score falls below a raw score.

Question 12

Why might percentiles be deceiving for something normally distributed?

Answer 12

In the middle of distribution, raw scores might differ slightly but have large percentile differences.

Opposite for extremes of scores.

Question 13

Age-equivalent scores or Age norms

Answer 13

Average performance of samples who were at various ages.

(eg. height, mental age (highly debated))

Question 14

Grade norms

Answer 14

Average test performance in a given school grade.

Question 15

What are age norms and grade norms referred to more generally as?

Answer 15

Developmental norms

Question 16

National norms

Answer 16

Nationally representative of a population.

Question 17

National anchor norms

Equipercentile Method

Answer 17

Equivalency table for scores on two different tests, providing stability to test scores.

Equipercentile is using percentile scores to calculate equivalency.

Need same sample to take both the tests.

Question 18

Local norms

Answer 18

Modifying a test to fit the purpose of a local population (?)

Question 19

Fixed reference group scoring system

Answer 19

eg. SAT.

There are scaled scores to calibrate. Purpose is so different years can compare scores (2021 score of 10 means the same in 1995 score of 10 even if raw scores differ)

Question 20

Criterion-referenced testing and assessment

Answer 20

A set standard. Different to norm-referenced. Not relative to others.

eg. get higher than a certain raw score to get liscense.