So, are consumers buying celebrity brands or are they buying brands that offer lower price points? Past data says consumers care about price points very much, branding not so much.  So, is this a cause of celebrity causation being confused with correlation?  It is a common misconception in statistics gathering, causation and correlation, did something actually cause an effect or is it just correlated with the results you are looking at. In this case, it sure appears that celebrity branding is a correlation to consumer choosing the lower price point product, not a causation of why someone bought a product.