A strong correlation (>0.85) between the independent variables included in a multivariate linear regression model indicates which assumption may have been violated?

A strong correlation greater than 0.85 among independent variables in a multivariate linear regression model indicates that multicollinearity may be present. Multicollinearity occurs when two or more independent variables are highly correlated, which can lead to instability in the estimation of coefficients, increased standard errors, and difficulties in determining the individual effect of each predictor on the dependent variable.

When multicollinearity is present, it complicates the interpretation of the regression model because it becomes challenging to discern which variable is truly affecting the dependent variable. This could potentially render some of the predictors statistically insignificant, even if they are indeed important for the model.

The assumption regarding normal distribution of the dependent variable, residual analysis, and homoscedasticity relate to different aspects of regression analysis. The normal distribution assumption pertains to the distribution of residuals rather than the correlation between independent variables. Residual analysis focuses on examining the residuals for patterns that might indicate assumptions are violated. Homoscedasticity relates to the assumption that the variance of errors is constant across all levels of an independent variable, which is also separate from concerns about multicollinearity. Thus, the identification of a strong correlation among independent variables clearly points to potential issues with multicollinearity specifically