The reverse is true as well. When a stat has been in the low range of the graph and then goes slightly up, it should be validated, even when only a slight rise:
“Also, it’s a bit mean to nag around about a rise. ‘But it isn’t much of a rise, you’re really in too low a range to have a rise count…’
“A rise is a rise. They at least got more. Now, better organizing, they will get more than that. Week by week it goes up.
“Similarly to discount a fall just because stats are high high high is folly. They could do week before last’s as they did it. So what was wrong that they couldn’t do it again? If they got exhausted at it week before last they need more help, obviously. Or better organization.”4
And it was not just the down stat not confronted! The problem with this particular graph occurred not only at the low point, but at the high point! What wasn’t confronted, what wasn’t handled, was not just the low week, but also in the high week!
Investigating Up Statistics
Something very effective was occurring through the upswing but nobody found out what it was! Then something changed and the stat crashed out the bottom. Why is it that when things aren’t going well everyone wants to know why? The lower it goes, the more agitated they can get.
But what happens when a stat goes up? If the statistic is rising, everyone relaxes. And what happens? The statistic drops because nobody found out why it was going up in the first place! And when it drops, what do they say? “Well, of course it dropped. It was up so high.” Which is, naturally, totally “reasonable.” In other words, it was expected to fall, and of course it fell.
An inspection and investigation of an up statistic should be at least as interesting, if not more so, than the down stat. After all, it’s positive! The time to investigate is on the upswing when it’s going well. That’s a lot more fun, and it’s a vital action. It’s hardly ever done to the degree that it should be. When things are going well, you want to find out why. And when they’re not going well, you’re going to find out why also.
Statistics rise or fall because of positive or negative changes. Statistics are a matter of changes. Some kind of positive change in some operating procedure was introduced near the point that started things going on the climb. If a stat drops, something changed. Whatever positive factor was put in was taken out or altered.
“When statistics change radically for better or for worse look for the last major alteration or broad general action just before it and it is usually the reason.
“Example: Letter out statistic falls and falls. In investigating, look for the last major change in that area and if possible cancel it and the statistic will then rise.”
Mr. Hubbard further explains, in the same issue, how he arrived at this management concept:
“I learned this while researching the life force of plants. Every time I saw a research bed of plants worsen, I queried what routine had been varied and found invariably some big change had been made that wasn’t usual.
“It is change that changes things for better or for worse. That’s the simplicity of the natural law.
“If you want to hold a constant condition, don’t change anything.
“If you are trying to improve something make changes cautiously and keep a record of what is changed. Then you watch statistics and if they decline you hastily wipe out the last change. And if they improve you reinforce the change that began it.”5
Of equal importance to changes are comparisons. Comparison is as vital in stats as it is in evaluating anything.
“Statistics must be studied and judged alongside the other related statistics.
“A rising income graph can even be shown sometimes as an actual threat to an organization if the delivery stats are down and stay down. It means the organization is selling and not delivering and may very well crash shortly.”6
4Hubbard, “Reading Statistics,” Policy Letter of 5 May 1971, Organization Executive Course.
5Hubbard, “Statistics, Actions To Take, Statistic Changes,” Policy Letter of 1 February 1966, Organization Executive Course.
6Hubbard, “Statistical Judgment,” Policy Letter of 9 February 1970, Organization Executive Course.