Test for multicollinearity
Multicollinearity is the strong correlation between the independent
variables. Some correlation between them is highly expected as all
variables are ‘causing factors’ to tree biomass. However, ‘strong
correlation’ causes ‘strong multicollinearity’ by which the true effect
of estimated regression coefficients would be lost. Thus, we can no
longer depend on standard statistical tests. In such circumstances, a
variable that has less explanatory power with the dependent variable
must be removed from the regression model. Nevertheless, while doing so,
there should not have strong collinearity between the remaining variable
with other potential explanatory variables. A variance inflation factor
(VIF) quantifies how much the variance is inflated. It is a measure of
how much the variance of the estimated regression coefficient bk is
”inflated” by the existence of correlation among the predictor variables
in the model. A VIF of 1 means that there is no correlation between the
kth predictor and the remaining predictor variables,
and therefore the variance of bk is not inflated at all. The general
rule of thumb is that VIFs exceeding 4 warrant further investigations,
while VIFs exceeding 10 are signs of serious multicollinearity requiring
correction (Belsley et al ., 1980).
3. Results