Come on feel the noize

Capture

“Hey there little fella, we don’t want any trouble, hand over your noise quietly and you might not fall down some stairs, understand?”

If you’ve know anything about systems thinking and that, you know it’s chock full of counter intuitive truths.

Here are some truths about one aspect of it all:

  • when you “eyeball data” by looking at it in a table, graph or chart, you’ll see patterns where there are none
  • just because a number is different this months from last, doesn’t mean anything has changed
  • the more data you have, the more noise there is to hide any signal
  • look for a “trend” and you’ll find one, whether one exists or not

Here’s my new year’s challenge for you…

1: Get your measures out. The measures you use to understand your system. Get them, in the usual form they are captured in, in graph, or pie chart, table or words. Go on, now, then come back with them in your hand.

2: Point at the noise with your finger. The noise you have identified and are now consciously and explicitly separating from the signal, so you know what the limits of that noise is, so you know where it is and can identify and ignore it.

3: Point at the signal with your finger. The signal you have identified as everything that isn’t noise, and are now consciously and explicitly using as the information that tells you about what is happening with that measure.

4: If you can’t do this Screw up your chart graph or table and throw it in the bin. It’s dangerous to you.

If you cannot describe the actual method you use to separate noise from signal, an operational description of method, you don’t have one and aren’t doing it. And don’t try and kid yourself that somehow you’re doing it anyway, or that it isn’t a problem in your special part of the world. There’s more noise than signal everywhere.

I swore I would never do a post on control charts as I don’t have the patience. I don’t, not typing all that out. There are much better blog posts and books out there on it all. This is the best description of the whys and wherefores I’ve come across.

But it is really important, so the purpose of this post is to urge you to recognise if you don’t know what is noise or signal, variation, common cause or special cause, then go elsewhere for the detail. Here’s a good place to start. Because you’re here reading this you’ll probably agree already that numbers are important, understanding them is equally important.

If you can’t point at your noise over there, and your signal over here, then your new year’s resolution is to learn how to. If you don’t randomness will fool you for the rest of the year.

dilbert2001182781025

Advertisements
This entry was posted in data, knowledge, statistics and tagged , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s