Distributions
General Stats
Stats Cont
More Stats
Stats on Stats
100

Normal Distribution

what are normal population parameters and equal central tendency (mean, median, mode) 

100

Experimental Design

What is manipulation

Has control and Random Assignment 

100

Correlation

A relationship between two or more variables 


100

Sample

A subset of data (comes from the population)

100

Pearson R Correlation

Looks at the relationship between two or more variables 


Determines the Strength and Direction 

200

Skewed Distribution

What is a violation of population parameters 

This leads to underestimated correlations 

Leads to Type II Errors 

200

Non- Experimental Design

Only contains observation 


- no control or random assignment 

200
Bivariate Correlation 
Ignores the influence of other variables 
200

Population

The entire population which you are studying 

200

Control

Part of an Experimental Design


All variables are under the same influence 

300

Positively Skewed Distributions

The tail is headed in the positive direction 
300

Correlation

Tells you the strength and direction 


.43 is a moderate and positive relationship 

-.43 is a negative/moderate relationship 

300
Multiple Regression

Are more precise because it looks at the influence of other variables 

300

Regressions

Are a predictor between 1x and 1y 

300

r2 change

determines the amount of variance each predictor adds 
400

Negatively Skewed Distribution

Tail is headed in the negative direction 

400
t-test

a difference between two groups 

400
Type 1 Error

You reject the null when you shouldnt have (the null is true)

400

Multiple Regressions

Are between multiple x's .... 1x, 2x, 3x, etc 
400

p < .05

95% confident 

5% doubt


"I am 95 confident that this is not due to chance." 

500

What are the characteristics of a normal distribution 

1 SD -- 68%

2 SD -- 95%

3 SD-- 99%

500

Random Selection

Ensures Representativeness; everyone has an equal opportunity to be selected 

500

Type 2 Error

You fail to reject the null -- you say there is no difference when there actually is 

500

Benefits of a Factorial ANOVA

Reduce Type I error

Look at the interactions  

500

Ceiling and Floor Effect 

Ceiling -- when the mode is the highest possible value 

Floor - when the mode is the lowest possible value 

M
e
n
u