Be Aware of Selection Bias
I wrote recently about the mistaken perception of older members not being tech savvy. (You can read that here.)
One of my astute readers, Will Slade at Protech, pointed out the following (I'm paraphrasing): "It's possible that the reason many associations perceive older members to be not tech savvy is that the older members will complain while the younger members will just walk away."
This is an important point! Which leads us to selection bias.
Selection bias occurs when we don't have a truly "random" selection of data points under consideration. In the case above, only the older people are complaining and so the assumption is "Our older members are having more trouble." This may or may not be true, but because they're the only ones we're hearing from, we're drawing conclusions based on incomplete data.
I believe associations (and ALL humans) make this mistake quite frequently. Between selection bias and recency bias, too often we're making decisions about design, programs, pricing, and more, based on faulty or incomplete data.
So be careful about how you're using the data you have on hand to make decisions. If the data is incomplete or poorly collected, it may steer you in the wrong direction!
![]()
Wes's Wednesday Wisdom Archives
Write it down!
Write it down! If you’re in the market for a new AMS (or any new […]
Why “weeding the garden” is so important
Why “weeding the garden” is so important One of the most important activities for maintaining […]
It’s time to stop excusing association professionals
It’s time to stop excusing association professionals I worked in associations for nearly 10 years, […]
More demos is not better
More demos is not better When I work with clients on selecting a new system, […]
The database owner should be “passionate”
The database owner should be “passionate” I’m frequently asked by my clients: “Who should we […]
Discipline is required
Discipline is required Last week I discussed the importance of taking action. To manage data […]
Action is required
Action is required Over my many years of consulting, one thing I’ve noticed about many associations […]
Some data analytics tips from the DAN “Science Fair”
Some data analytics tips from the DAN “Science Fair” Last week I had the opportunity […]
Are all your processes frictionless?
Are all your processes frictionless? I’m not a huge fan of buzzwords, but I love […]
Trends don’t need perfect data
Trends don’t need perfect data When it comes to analyzing data trends (changes in data […]
