URSI 403--Doing Urban Research -- Lecture Notes 6
V. Analyzing the Data
A. Single variable (descriptive statistics)
- Bar graph & histogram
- Measures of central tendency
- Mean (average)
- Median (midpoint)
- Mode (most frequent)
- Measures of dispersion
- variance
- standard deviation
B. Relationships between variables (measures of association)
- Non-parametric statistics
- Parametric statistics
- Assumptions
- i. bivariate normal (linear relation, x & y normally distributed)
- ii. homoscedasticity (equal variance)
- iii. measurement error in y alone (x independent of error term)
- iv. random sampling
- General linear regression
- i. basis of correlation, multiple regression, analysis of variance,
and factor analysis
- ii. estimate line that best fits observations y=a+bx
- iii. "least squares" estimate--minimize variance (deviation
squared) of x on y.
- iv. slope (b) is covariance (sum of joint variance) divided by
sum of squared deviation in x.
© 1996 A.J.Filipovitch
Revised 1 September 96