I’ve recently been reading a lot of weasel words. Those words which mean one thing, but which imply something else. The other day I was reading an article in a highly respected academic journal that I will not name in this post, for reasons which will become clear shortly.
The article was about using academic research as evidence to develop talent retention strategies. At one point the article states that research has shown that “reducing turnover rates is linked to sales growth and improved employee morale.” So far, so good.
The word “linked” is one of those weasel words. It means that one thing is associated with another. In statistical terms it means that one thing is correlated to another. However, we often assume that the word linked means that one thing caused the other.
In our example, people might infer that reduced turnover rates cause sales growth and improved employee morale. However, that is not the case. All we know is that reduced employee turnover, sales growth and improved employee morale happen together. In fact, it might be that sales growth reduces turnover, because people are making big bonuses and are happy to stay at the organization. When sales are bad, the company is in trouble and people might be tempted to leave for greener pastures.
If you assume a causal relationship, that is that reducing turnover increases sales growth, you might be tempted to put a lot of money, time and attention into reducing turnover rates, making a potentially costly and ineffective move.
How can you avoid this? Look for evidence but watch for weasel words. Don’t assume causality when it isn’t there.