STATISTICAL ANALYSIS
75% of statistics are made up on the spot.
-And if you don't get the irony of this quote think really hard!
Section Links: |
See Main DP IA HELP page for Linear Regression & R^2 help |
Choosing the Correct Statistical Test
Possible Suggestions based on number of Independent Variables
- Linear Regression models used for CONTINUOUS NUMERICAL DATA
- Chi-Squared Tests are only used for CATEGORICAL DATA (tallies, frequencies, etc)- can be used in discrete numerical as well at times.
- T-test are used for DISCRETE NUMERICAL DATA generally, must have means to work
- Used for two populations to determine if they are statistically different from each other
- ANOVA test are used for DISCRETE NUMBERICAL DATA
- Used for three or more populations to determine if they are statistically different from each other.
Chi-Squared Tests
Chi-Squared (X2) is a statistical tested used to determine if variables are independent (random) or dependent on each other.
X2 Critical Value Table below:
|
|
T-tests
A T-test is a statistical test used to determine if two means are statistically significant from each other.
|
|