Saturday, December 13, 2014

Introduction to Statistics

The main purpose of statistics is description (to explain) and inference (to predict).

There are 2 broad categories of statistics: 1.) descriptive; 2.) inferential.

The goal of descriptive statistics can be looked at as data reduction. Descriptive statistics uses as few descriptions (i.e. numbers) as possible to provide the most information regarding the sample. In order to summarize data, measurements of central tendency and measures of dispersion common.

The goal of intferential statistics is generalizability to the population. Inferential statistics are used to reason from a sample to the population, in other words, from the particular to the general.

According to the law of large numbers, if the sample is large enough, the sample information is likely to be accurate so larger samples are usually better.

univariate - describes 1 variable
bivariate - describes 2 variables simultaneously
multivariate - describes 3 or more variables simultaneously

"directional hypothesis"

measure of central tendency
nominal - mode
ordinal - mode/median
metric - mode/median/mean

measures of dispersion
greater dispersion = greater spread

Inferential statistics - the level of measurement you have will dictate the type of test you will use; test you could use are the t-test, ANOVA, chi-square, regression

Statistical Software
Intro to STATA


No comments:

Post a Comment