Term
|
Definition
hypothetical factor that cannot be observed directly but is inferred from certain behaviours and is assumed to follow from certain circumstances
-mental rotation task - determined construct (visual imagery) by measuring response time |
|
|
Term
| how to evaluate measures (2) |
|
Definition
1. reliability (repeatable)
-inverse relationship with measurement error (observed score approximates true score)
2. validity (measures what its designed to)
-assumes reliability |
|
|
Term
| types of validity in testing measurement (5) |
|
Definition
| 1. content 2. criterion 3. construct 4. convergent 5. discriminant |
|
|
Term
| def: content validity + eg |
|
Definition
test items "make sense" in terms of construct (NOT face validity - seems valid to test takers)
-eg heart rate to measure performance anxiety |
|
|
Term
| def: criterion validity + eg |
|
Definition
accurately forecasts future behavoir
-eg GRE should predict future success in grad students |
|
|
Term
| def: construct validity + eg |
|
Definition
whether test adequately measures a construct (operational definition = important)
-eg IQ test measures intelligence |
|
|
Term
| def: convergent validity + eg |
|
Definition
scores on a test are designed to measure one construct are correlated with s cores on another test theoretically related to the construct
-eg written langage scores + language comprehension scores |
|
|
Term
| def: discriminant validity + eg |
|
Definition
scores on a test designed to measure one construct are UNrelated with scores on another test theoretically UNrelated to the construct
-eg spatial reasoning scores + problem solving scores |
|
|
Term
|
Definition
used for group classification, categorical data
-eg gender |
|
|
Term
|
Definition
rankings, ordering is meaningful but metric is not
-eg 1st 2nd 3rd in a race |
|
|
Term
|
Definition
equal interval rankings, arbitrary 0
-eg temperature |
|
|
Term
|
Definition
equal interval rankings with a true zero, each increase reflects the same change in the underlying measure
-eg height |
|
|
Term
| define: population/sample |
|
Definition
1. all members of a defined group
2. subset of population |
|
|
Term
| define: descriptive/inferential statistics |
|
Definition
1. summarize data
2. draw conclusions - applied to broader population |
|
|
Term
| descriptive stats: measures of central tendency (3) |
|
Definition
1. mean: Xhat = sumX/N
2. mode: most frequent score
3. median: location = (n+1)/2. use if outliers! |
|
|
Term
| descriptive stats: measures of variability (4) |
|
Definition
1. range: highest - lowest score
2. standard deviation - sq of E(xhat-x)^2/(n-1). estimate average deviation from mean
3. variance. sd^2. important for anova
4. IQR. range 25th-75th percentile - use if outliers |
|
|
Term
|
Definition
1. no diference between puplations --> reject/fail to reject
2. researchers prediction |
|
|
Term
|
Definition
| probability that research outcome is due to chance |
|
|
Term
|
Definition
-reject null hypothesis when it is true
-likelihood = alpha
-think you have a real effect when you dont |
|
|
Term
|
Definition
-fail to reject null hypothesis when it is right
-dont find significant effect in the sample when there is one in the population
-possibly due to unreliable/insensitive measures |
|
|
Term
| structure of analysis of significance |
|
Definition
| (systematic variance + error variance)/error variance |
|
|
