In many empirical projects, strategies of one function (purchasing, manufacturing, etc.) are regressed against firm performance. This supposedly suggests if a particular strategy contributes to firm performance. In doing this we are ignoring the contributions of all other functions. It could be that the marketing strategy is what creates great performance for a firm and the poor purchasing actually brings down the performance by a bit. However empirical results would show that the poor purchasing to cause high performance.
Can such methodologies be held valid? Are there other ways to link functional strategy to firm performance?