Term
|
Definition
| summarizing a group of scores |
|
|
Term
|
Definition
| drawing conclusions based on scores but going beyond them |
|
|
Term
| ratio scale (level of meas.) |
|
Definition
| equal-interval variable that has an absolute zero point (meaning if the value is 0 the variable is completely absent) |
|
|
Term
|
Definition
| also called ORDINAL variable |
|
|
Term
|
Definition
|
|
Term
|
Definition
| most representative value of a group of scores |
|
|
Term
|
Definition
| random selection (most ideal) |
|
|
Term
| Changing raw to Z and vice versa |
|
Definition
| Z= X-M/SD ... X= (Z)(SD)+M |
|
|
Term
| % of scores above/below a particular score |
|
Definition
| Change raw score to Z score, find on Z table in % Mean to Z or % in tail |
|
|
Term
| calc. Z score based on percentage |
|
Definition
| draw graph, shade, estimate Z where shade stops (in SDs), look up in table |
|
|
Term
|
Definition
| predicts a particular direction |
|
|
Term
|
Definition
| does not predict specific direction |
|
|
Term
|
Definition
| for directional hypo (region of the CD where the null is rejected is all on one side) |
|
|
Term
|
Definition
| for nondirectional hypo (region of the CD where the null is rejected is on both sides) |
|
|
Term
|
Definition
+1.96 and -1.96 for %5 (two-tailed)
+1.64 and -1.64 for %5 (one-tailed)
+2.58 and -2.58 for 1% (two-tailed)
+2.33 and -2.33 for 1% (one-tailed) |
|
|
Term
|
Definition
| COMPARISON DISTRIBUTION when testing hypos. involving ONE sample of MORE than ONE person |
|
|
Term
|
Definition
mean, spread (calc. using variance and SD), shape
1. M of DoM is same as pop. M
2. Variance of DoM = pop. variance divided by # individuals in each sample
3. if 30+ individuals or dist. of population on individ. is normal, then shape of DoM will be normal |
|
|
Term
|
Definition
measure of difference (lack of overlap) between populations. INCREASES AS MEANS GET FURTHER APART (essentially, difference between the means)
standardized effect size: M1-M2/pop. SD |
|
|
Term
| Conventions of effect size |
|
Definition
small .20
medium .50
large .80 |
|
|
Term
|
Definition
| d= effect size (M1-M2/pop. SD) |
|
|
Term
|
Definition
| combining effect sizes from different studies |
|
|
Term
|
Definition
| probability that the research hypo will give stats. signif. result if H1 is TRUE |
|
|
Term
| Factors that determine power in a study |
|
Definition
| Effect size (bigger effect size, greater power) (smaller SD, bigger effect size) |
|
|
Term
|
Definition
| M1=M2 + (d, predicted effect size)(pop. SD) |
|
|
Term
|
Definition
| rejecting null when it's true |
|
|
Term
|
Definition
| BETA not rejecting the null when you should |
|
|
