How valuable are you? That’s a question that many large organisations ask about their employees every year in the ritual of the performance review. If they asked the question honestly that might be alright, but many of them impose unjustifiable assumptions about the distribution of normality. For example, they might dictate that each department within the organisation identifies a certain proportion of under performers and over performers.

I’m no statisitician, but this seems to be a woeful misinterpretation of the normal distribution. Wikipedia helpfully explains that: “a normal distribution is often used as the first approximation to describe real-valued random variables that cluster around a single mean value.”

Can performance be characterised as a “random value”? If so, then the organisation has bigger problem than fair performance reviews.

The Wikipedia entry then goes on to describe how “the normal distribution arises from the central limit theorem” which holds when dealing with “a large number” of samples. There are a number of ways of working out a reasonable value of “large” – with varying degrees of statistical validity. There is no reason to think that the few tens of people that make up a team or department is large enough to provide a statistically valid sample, and so there is no indication that the normal distribution will apply.

“All my reports have performed at least as well as expected,” you may hear a manager opine “but I still have to give at least one of them a poor result.”

Then it gets worse. Once they’ve squeezed each department under a poorly fitting curve they then pass these details up the management hierarchy. At each level in the organisation managers attempt to map one department against the others to produce a supposedly fair assessment of an individuals performance within their larger peer group. It’s not unknown for this to happen at several levels, each time comparing larger samples, until the data is normalised at the VP level. Then the data gets communicated back down the tree until the worker at the bottom finally gets her performance score.

The worker is expected to take the feedback on the chin, even when this farcical exercise has negatively impacted their performance related bonuses or career prospects. Sure, you can appeal, but no-one wants to look like a whinger. On what grounds would you appeal against such an opaque, subjective process anyway?

So, instead, I think each review should be inscribed with a paraphrased extract from The Hitchhiker’s Guide To The Galaxy: ”Normality has been enforced. Anything you can’t cope with is therefore your own problem.”