this means that the variance differs in different segments of the data. In time-series data, this would mean that the variance is different in different time periods. In cross-sectional multivariate data, this might mean that the variance of some outcome variable differs for different value ranges of the outcome variable, or predictor variables. For example, in a model to predict housing prices, variance might be greater for high value houses, or for houses in a given neighborhood.
Week #30 – Heteroscedasticity
- July 28, 2014
- , 7:44 pm
Heteroscedasticity generally means unequal variation of data, e.g. unequal variance. More specifically,