Stats: module 5

statistical hypothesis

two mutually exclusive and exhaustive mathematical statements about the outcome of the analysis
- null hypothesis
- Alternate hypothesis (directional/ non directional)

Non-directional hypothesis

- not sure in treatment with increase or decrease
- error in both directions

directional hypothesis

have to be really sure that it is directional, has to have directional words

type 1 error

say something worked when it really didnt

type 2 error

we say there was no change in our experiment when there should have be one

statistical power

the probability of finding an effect or relationship is one exists
2 ways to change
- reduce noise: variability, the things interfering with being about to see a change
- increase sample size, the bigger the sample the easier to see a change if there is one to be seen

Power

the ability of a test to detect a real effect in a population based on a sample taken from that population

correlation

variables that are related or associated
- the closer to +1/-1 the more related the correlation is
0 = no correlation
- negative correlation does not mean bad: it means that as one variable goes up the other goes down

Pearson product-moment correlation coefficient (r)

the extent to which the direction and size of deviations from the mean in one variable are related to the direction and size of deviations from the mean in another variable
- one independent and one dependent variable

to determine r:

- compute the mean and standard deviation for each pair of variables
- calculate z-scores for pair of variables and multiply them
enter values into formula to calculate

common variance

r squared

multiple correlation (R)

more than 1 independent variable, and one dependent variable

partial correlation

examines the relationship between and independent variable on a dependent variable after removing the influence of another independent variable

regression analysis

- the relationship between dependent variable and one or more independent variables
- involves prediction

bivariate regression

- simple linear regression
- if we know one variable we can predict the other

multiple regression

- examines the ability of multiple independent variables to predict a dependent variable and their relative contributions