« Best of the Force: For Strategic Rewards, There Must First Be a Strategy... | Main | Best of the Force: Line of Sight - Framing It Right and Taking it Beyond Rewards »


Feed You can follow this conversation by subscribing to the comment feed for this post.

I am no statistician, but have come to understand from the experts that the median is a more 'stable' and preferred measure relative to the mean. No dissenting view here.

In the matter of the mode, it seems to me that too many organizations make reference to what 'most others' are doing, some suggesting that a modal stance of sorts is the 'safe' basis of their choices / proposals. In that regard they appear to be suggesting that the mode is reported. I am yet to see the modal statistics reported in well-known salary surveys (there are just too many surveys out there -- a thousand plus from WorldatWork alone --- for me to be absolute about this), although modal ‘practices’ may be discerned in some surveys.

We know that the herd instinct brings up the question (from bosses, boards, and others): what are most others doing? And the question forces some (or many) to present data and practices as being what 'most others' are doing. The question then is: Where someone reports that some statistic or pay level as what 'most others' are doing / paying, are they choosing the mean or median or other statistic as basis for their suggestion?

If "median-based" stats are available, I would be shocked if someone is using the average (for the reasons that you outlined above).

E.K. -

That has been my personal experience as well - that tracking and setting pay practices against the mean will typically produce a bumpier ride, more zigs and zags than using the (typically) more stable mean.

Excellent broader point about the mode. Although it rarely shows up in surveys of pay rates, it does dominate discussion about (and research on) pay practices - appealing to our instincts to flee to the middle of the herd (which only appears to be safer...)

Thanks for the thoughtful comments.

Robert -

I once had a colleague in the consulting field(now retired) who insisted on using the mean for the very reasons that most of us avoid it. He believed it important that his data set and data-based decisions reflected the outliers and the extremes in the data set.

While I appreciated his position, in the end we did agree to disagree.

I think there are survey users who continue to rely on the mean - some because they don't have a sound understanding of statistics but a few (like my colleague) who understand quite well and knowingly choose a different indicator than most.

Appreciate the comment

The comments to this entry are closed.

About The Author

  • More Info Here
    Compensation consultant Ann Bares is the Managing Partner of Altura Consulting Group. Ann has more than 20 years of experience consulting with organizations in the areas of compensation and performance management.

Compensation Force Spot Survey

Enter your email address:

Delivered by FeedBurner

Search This Site


  • Get this widget from Widgetbox