Term
Define: Sampling Distribution of Means
CENTRAL LIMITS THEOREM |
|
Definition
-Mean is called "means of means" (m sub m)
-Std. dev = std error of the mean
(sigma sub m)/(sq.rt of n)
- is normal if parent pop is normal
- as 'n' decreases, std err decreases |
|
|
Term
|
Definition
|
|
Term
| Define: Frequency Distribution |
|
Definition
- x= individual score
- mean = mu
- std dev = sigma
- determines = prob of ind. score
- based on = actual data |
|
|
Term
| Define: Sampling Distribution |
|
Definition
m = group/sample score
mean = mu sub m
std dev = sigma sub m
determines = prob of groups/samples
based on = theoretical data and CLT |
|
|
Term
|
Definition
1. type of hypothesis
2. levels of measurement of data
3. #/types of variables
4. levels of the independent variable |
|
|
Term
|
Definition
When we REJECT a TRUE null hypothesis
- REJECT when TRUE is NOT type II-
probability = alpha |
|
|
Term
|
Definition
FAILURE to REJECT a FALSE null hypothesis
probability = 1-beta |
|
|
Term
| Define: Level of Significance (alpha) |
|
Definition
- probability of type I error
- prob of rejecting a true null
- generally .05 or .01 |
|
|
Term
| Define: Type II Error (BETA) |
|
Definition
- failure to reject null when it is false
- prob of failing to reject false
- beta related to power
- power = 1-beta |
|
|
Term
|
Definition
1. State null and alternative hypothesis
2. Set criteria to reject the null
- alpha level
- zobs
- direction (one or two-tailed)
3. collect data and compute test stats
4. make conclusion about null and draw conclusion (five o'clock news summary) |
|
|
Term
|
Definition
| Cut off point beyond which we reject the null hypothesis |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
| Fail to reject null when... |
|
Definition
|
|
Term
|
Definition
- random sampling
- independent observation
- value of sigma is unchanged by treatment
- normal sample |
|
|
Term
| Effect size interpretations |
|
Definition
d=.02 = small effect
d=.05 = medium effect
d=.08 = large effect |
|
|
Term
|
Definition
| Probability of correctly rejecting null |
|
|
Term
|
Definition
1. increase "n"
2. increase effect size
3. increase alpha
4. change from two- to one-tailed test |
|
|
Term
|
Definition
1. Draw the null true SDoM graph using sigma-sub-m, and zcv to find mcv. Shade
2. Draw the null false SDoM grpah on top of first graph. Shade power.
3. Find new z.
4. Look up new z to find power. |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
- random sampling
- ind. observation
- normal sample |
|
|
Term
|
Definition
rsquared = (tobs-squared)/
(tobssquared) + df |
|
|
Term
| Define: % variance explained |
|
Definition
| Variance in DV explained by IV |
|
|
