U8

Question 1

reliability

Answer 1

refers to the extent a measure is consistent and free from error

Question 2

a reliable scale isn’t always _____ but a valid scale is always _____

Answer 2

valid
reliable

Question 3

variance

Answer 3

measure of variability in scores in a sample

Question 4

the ____ the variance the more homogenous the scores

Answer 4

smaller

Question 5

trimmed mean

Answer 5

computed by “trimming away” a certain potion of scores
- used in certain contexts to produce more stable/accurate outcomes

Question 6

measurement error

Answer 6

random events can increase/decrease a score

Question 7

bias error

Answer 7

systematic/constant
present every time consistently

Question 8

classical test theory

Answer 8

X = T (+/-) ME (+/-) BE

X= observed score
T = true score

Question 9

systematic error
(+problem)

Answer 9

• predictable errors of measurement (refection of bias error)
• one direction
• threatens validity of the measurement
  (solved by re-calibrating instrument)

Question 10

random error

Answer 10

• due to change
• not following procedures properly
  (variability in participants)

Question 11

sources of measurement error

Answer 11

• instrument
• participant
• researcher variability
• environment variability

Question 12

correlation might be a measure of _____

Answer 12

reliability

Question 13

less error in reliability score closer true score is to ______

Answer 13

observed score

Question 14

test-retest reliability

Answer 14

measures consistency of measuring instrument (equipment/person) over time/testing sessions

Question 15

interclass correlation coefficient

Answer 15

• similar to PPMC
• bad test-retest reliability
• evaluates association and difference (from variance estimate)
• 2+ measures

Question 16

why is the interclass correlation coefficient limited

Answer 16

• only bivariate
• cannot separate how much error can be expected across tests (variance components)
• score may be poorly correlated but not diff
• scores may be highly correlated but significantly diff

Question 17

standard error of measurement (SEM)

Answer 17

calculates range of expected score deviations from test to test

Question 18

factors affecting test-retest reliability

Answer 18

• learning effect
• rater bias
• test retest intervals

Question 19

learning effect

Answer 19

influences of on test on later test-practice/boredom

Question 20

rater bias
(+prevention)

Answer 20

tester desires a certain outcome
- prevention (masking tester/strict set of method)

Question 21

test-retest intervals

Answer 21

period between assessments
- too short (fatigue/learning effect)
- too long (changes in participant)

Question 22

intra rater reliability

Answer 22

stability of data within 1 individual btwn trials

Question 23

inter rater reliability

Answer 23

stability of observer or instrument measuring same group (aiming to produce generalizable results)

Question 24

alternate forms reliability

Answer 24

• establishing reliability of standardized tests by using 2 forms of a test then observing reliability

Question 25

internal consistnacy

Answer 25

degree to which individual items agree with each other

Question 26

to improve reliability you can _______ # of test items

Answer 26

increase (can have issues w participant focus w too many items)

Question 27

what formula is used to test improved reliability

Answer 27

brown spearman formula

Question 28

split-half reliability

Answer 28

used to establish reliability of questionnaire (taking all items and splitting them into two tests and comparing results)

Question 29

issues w split half reliability

Answer 29

Recall (if interval is too short) or true change (if interval is too long)

Question 30

Cronbach’s alpha

Answer 30

a measure of internal consistency (how closely related a set of items are as a group) form of inter rater