5713997588_6cafe7b28c

No matter what your role as an information worker – data matters. Data helps us understand our customers, health, friends, family, profitability, employees and the list goes on and on. As the world continues to have more computers, cameras, and sensors we will continue to have more data and have a better opportunity to either improve the quality of life or the alternative.

At one time executives had to be sold into the value of data driven decisioning versus their gut. These days that is less of a problem. Now the challenge is to not get lost in the data and making the wrong interpretation. As we all know the phrase “lies, damn lies, and statistics.”

A common confusion is the difference between causation and correlation. Things are correlated if two things have a mutual relationship between them. This can be a positive or negative correlation. Something can be correlated but not caused by something else and this is often the case. An example could be decreased meat sales correlated to declining home prices. Instead an economic downturn negatively caused both of these.

Things are caused if one thing causes another thing to change in a positive or negative direction. For example, you increase product sales when you increase Facebook ad spends. If you understand one thing causes another then you can act on that one thing and be confident that the other thing will be impacted accordingly.

People are quick to think that because things are correlated that one thing causes the other. Making decisions on things that are correlated but not caused by can lead to bad results. However, it does give us a starting point where we can do testing whether it is A/B testing if software or other testing to determine if one thing is indeed caused by another thing or if they are merely correlated.

The other item to be aware about is that there are varying degrees of causation and correlation. Generally causation or correlation do not have a one-to-one relationship. Sometimes this relationship can be greater than one-to-one but often times it is less.

WLP Tip No. 5: Next time you read the results of a study or setup your own research study make sure you identify if the results represent causation or are merely correlation and additional research is needed to determine causation.

Have an awesome week and remember to do something today to supercharge your success.

As always appreciate your feedback, emails, comments, likes, and re-tweets!

photo credit: side-by-side via photopin (license)