reliability/validity

Question 1

test retest reliabilty

Answer 1

• measures stability of a test over time
• same group is given same test on 2 seperate occasions
• if similar/same answers are obtained then has high ER

Question 2

split half reliability

Answer 2

• degree to which all parts of a test equally contribute towards what’s being measured
• test is split into 2 parts (& given 2 ps on same occasion)
• answers 4 each half is compared & if they’re similar has high IR

Question 3

inter rater reliability

Answer 3

• degree 2 which diff raters give constitent estimates of same bev
• rates assess same bev (of same group of ps) independently & then compare their results
• if similar, then is reliable

Question 4

content validity

Answer 4

• degree to which the items on a test are representative of the domain they seek to measure
• ask an expert to look at your results & evaluate whether you measured what you aimed to

Question 5

concurrent validity

Answer 5

• degree 2 which a test relates 2 a similar existing measure
• correlate findings of your exp to a recognised exp that measured same thing, if there’s strong ve+ correlation then is valid

Question 6

predicitve validity

Answer 6

• degree to which results of a test accurately predict a bev/criteaon which will occur in the future
• calculate correlation coefficient between results of test & subsequent target bev, stronger correlation means higher P.V

Question 7

face validity

Answer 7

• degree 2 which test appears 2 measure what it aims 2
• ask people to rate validity of test as it appears to them
• could do this using a likert scale (eg: from extremely suitable for its purpose to completely irrelevant)

Question 8

construct validity

Answer 8

• degree to which a test successfully measures the specific construct or trait that it is supposed to
• need to show that thing your measuring actually exists, can be used to explain a network of (research) findings & predict further rels