Exploratory factor analysis - Avoiding multicollinearity in risk factor studies: An example

The present study used factor scores from an exploratory factor analysis of 265 herds describing five underlying structures in typical Danish slaughter pig production. The objective of the present study was to investigate whether chronic pleuritis prevalence could be described in a simpler and more objective manner, using factor scores from an exploratory factor analysis, than by an ordinary risk factor study. The results showed that connections between type of herd, herd size, type of feed, all in all out production and type of floor were described in the risk factor study using factor scores, but could not be estimated in the ordinary risk factor study, due to multicollinearity.