Multicollinearity
Multicollinearity
Introduction to
01
Multicollinearity
Consequences of
02
Multicollinearity
Detection of
03 Multicollinearity
Remedies for
04 multicollinearity
MULTICOLLINEARITY
PERFECT
Perfect multicollinearity is a statistical issue that arises in
regression analysis when there's an exact linear relationship
between two or more independent variables in your model. This
means you can perfectly predict the value of one variable based
on the value of another.
TYPES OF
MULTICOLLINEARITY
IMPERFECT
Imperfect multicollinearity, unlike its perfect counterpart,
occurs when there's a high degree of correlation between two
or more independent variables in your regression model, but
the correlation is not perfect (meaning the correlation
coefficient isn't exactly 1 or -1).
CONSEQUENCES
OF
MULTICOLLINEARITY
1. Estimates will remain unbiased
2. The variances and standard errors of
the estimates will increase
3. The computed t-scores will fail
4. Estimates will become very sensitive to
changes in specification
5. The overall fit of the equation and the
estimation of coefficients of
nonmulticollinear variables will be
largely unaffected Severe Multicollinearity increases the variances
of βs
ANALYSIS
t-score for disposable income (Ydi) increases
drastically ( from 1.04 to 7.13) when the liquid
assets variable is dropped from the equation