i don't understand the difference between heteroskedasticity in AR model and serial correlation

can someone help? heteroskedasticity: variance of the residuals in one period is depentend on varaince of the residuals in the preecing period (usally heteroskedasticity means residual variance is related to the level of independent variables…) serial correlation: serial correlation , residuals are correlated? in

Serial Correlation: Residual value is dependent on previous value. Can easily be spotted in a graph. Heteroskedasticity: Variance is dependent on value. So if value increases, variance increases. Can easily be spotted in a graph. Conclusion: Look at some graphs showing the limitation. You’ll get it.

heteroskedasticity = variance of residuals is directly related to the level/value of independent variables serial correlation = autocorrelation = variance of the residuals in one period is dependent on varaince of the residuals in the preceding period

Both of your responses are wrong.

bmw, CONDITIONAL heterskedaticity is when the variance is directly related to the level of ind variables, not necessarily all heterskedasticity… you can easily identify these things on scatter plots