Giving the Gift of Deviation

Here is some unsolicited advice to all gift-givers. If a loved one asks for the gift of a scale to check his or her weight, don’t be the one to give it to them. It doesn’t matter how cool the scale is; it is a no-win gift that sends the wrong message, which is, “Hey, chubby, I agree — you really could use this! Great gift idea.

I fell into this trap, lured by a nifty new scale that doesn’t show your actual weight, but only a loss or gain from the start date of your diet. The device promises to keep you motivated and help you lose weight faster, improve your self-image, focus on positive change, and feel good about yourself. Clearly, the design of this scale is far superior to those talking ones with real attitude. You know, the ones that snarl, “Hey! One at a time!”

OK, so I wasn’t the one who gave this gift; rather, I was the knucklehead who asked for it. (Note to self: mother was right. Always ask for jewelry). In my defense — and given the work I do — the concept makes complete sense to me: show only the degree to which a quantitative value (pounds gained or lost, for example) differs from a primary value (starting weight in pounds).

A Deviation Graph works exactly the same way.

The graph below (which could also be a line graph) displays — as does my fancy new scale — only the direct variance of weight lost or gained.

deviation graph 01

The advantage of showing changes this way is that you don’t have to do any mental calculations of how much you lost or gained in a week — the number’s right there in front of you.

Now, let’s consider a different example of how you might use a deviation graph to display data. You want to show the number of full-time employees (FTE’s) actually employed in each department compared to budget figures. Such information is typically displayed like this:

deviation graph 02

If the audience for your report needs to know only how much the actual FTE’s vary from budget, this graph causes them to work a lot harder than is really necessary. They have to not only compare each bar within and between departments, but calculate differences as well. By displaying the differences directly in a deviation graph, as in the example below, you can eliminate the mental calculation and gymnastics required of the viewer. If your viewer is seeing variance only, (s)he can tell at a glance which departments are over or under budget on FTE’s.

deviation graph 03

In some cases, it may make more sense to display the variance as a percentage. Let’s return for a moment to the first example of pounds lost and a different display with an important difference. If, for example you are a counselor or clinician leading a weight-loss group, a graph that displays the information as a percentage of weight loss rather than pounds (as in the first example) is most likely a better way to compare weight loss from individual to individual in a group. Consider the graph below, which displays data in that way.

deviation graph 04

If you have ever seen the television show The Biggest Loser (my guilty pleasure), this concept will be familiar to you. As demonstrated by the show, if one person starts out weighing 500 pounds and another person starts out weighing 250 pounds and they each lose 125 pounds, their accomplishments are very different. One contestant has lost 25% of starting weight, the other, 50%. Viewing the data in a graph that directly shows the variance percentage is the only fair and logical method. (Not to mention that when there is a $250,000 grand prize, you want to be certain that the rules really are logical and just. You wouldn’t want to send anyone off on a Ben and Jerry’s binge after all that hard work.)

In the end, my best piece of advice may be the one I started with: a weight scale is probably a poor gift choice, especially when you are trying to convey your unconditional love. Note to my hubby: in the future, ignore my gift request, and remember that diamonds truly are a girl’s best friend.


Leave a Reply

Your email address will not be published. Required fields are marked *