If the Question Is Ad Effects, the Answer Is ‘Not Elasticities’
本文指出用对数回归中的弹性系数比较广告效果是错误的，并推荐使用线性回归中的广告系数进行此类比较。
<h3>ABSTRACT</h3> Advertising sales elasticity is often used to compare advertising effects (across years, campaigns, media, regions and brands). Broadbent describes why using ‘elasticity’ in the form of a coefficient in a logarithmic regression, is wrong; and why seeing ‘elasticity as the proportion of sales due to advertising is misleading. He recommends using the advertising coefficient in a linear regression for such comparisons.