Friday, 10 January, y
Political Science 3N06 Semester II 2014 Lecture 1: Introduction
- We will be exploring two different types of statistics this semester
- 1) Descriptive Statistics
- They describe the features of a collection of data
o Patterns in particular data
o Different kinds of descriptive statistics
- They provide a description of the distribution of a variable
A) Univariate statistics summarize and describe the characteristics and features of one variable
o They take the (possibly) thousands of data points and reduce them to one or two or three easily
o Mean, median, standard deviation, etc.
Reduce to smaller numbers
Measures of central tendency; (Value that is most typical)
Measures of dispersion; (What number gives me a best sense of the variety of the
responses that exist within this data set?)
B) Bivariate statistics are used to describe the relationship between two variables
o We are not just interested alone, but explaining for example the relationship between income &
health, democracy & development
o Give us a number that can tell us the strength of the association
o C) Multivariate descriptive statistics describe the relationship between three or more variables
- 2) Inferential Statistics
- They allow us to make inferences about populations
- They allow us to explore the probability that what we found to be the case in our sample
o Is true of the larger population from which the sample was drawn
- Allow us to link the smaller population that we can see with the larger population we can’t see?
- Allows us to come up with a probability sample, 90% confident etc. Is also true of the larger population.
- Just as there are different types of statistics, there are also different types of variables
- Certain statistics should only be used with certain types of variables
- A) Independent (cause) and Dependent (effect)
o Important for when we look at measures of association
o At what level has this variable been measured?
- B) Levels of Measurement
o Variables measured at this level have categories that have no numerical relationship to one another
These variables have attributes or possible scores that have no mathematical relationship
to each other.
o E.g. Gender, Religion, Province, Political Orientation
o The categories of the variable can’t be ordered, averaged, etc.
1 Friday, 10 January, y
o Male + female = ?