Reliability and Validity

Question 1

Reliability means what about the number?

Answer 1

You can TRUST the number.

Question 2

True scores are only ________ and cannot be ________.

Answer 2

Theoretical and Measured

Question 3

X = T + E

Answer 3

X = True score + Error score

Question 4

As reliability goes ______ error goes _____.

Answer 4

up and down

Question 5

Interval Consistency =

Answer 5

Split half and coefficient alpha

Question 6

If the coefficient alpha is 0.9 it shows:

Answer 6

Internal Homogeneoty

Question 7

N•rxx/1+(N-1)rxx
or
2•r/1+r

Answer 7

Spearman Brown

Question 8

With Spearman Brown N=

Answer 8

N= change in length of the test (new divided by old).

Question 9

If you have a reliability of 0.45 and want one of 0.6, how much should you lengthen the test?

Answer 9

N=rd(1-rxx)/rxx(1-rd)
rd= reliability desired
you’ll get 1.83 so multiply your items by 1.83 and you’ll get you answer

Question 10

(N/N-1)•(Sx2-£pq/Sx2)

Answer 10

KR20; used when you have false data

Generally displayed as T/F or 1/0

Question 11

Sy•SQRT(1-rxx)

Answer 11

Standard error of the measurement usually a (+,-)

Question 12

r•2/SQRTr11•r22

Answer 12

r12 or the correction for attenuation. Used o find the correlation of there were perfect reliability with no standard error

Question 13

Nonsystematic error is _____.

Answer 13

Random

Question 14

Face Validity shows:

Answer 14

Do the questions look like they’re testing what they say they’re testing for

Question 15

Content Validity shows

Answer 15

Is the content testing what it said it will be testing.

Question 16

Criterion Validity has 2 kinds _______ and ______. It was based on _____.

Answer 16

Concurrent
Predictive
The real world

Question 17

Concurrent Validity

Answer 17

Comparing your test to another test at the same time.

Question 18

Predictive Validity

Answer 18

Predicting something in the future. Ex: SAT scores predicting college GPA