I’ve been reading a lot about how election polling works. Because I’m a nerd.
And I’m interested in how organizations try to predict the outcome of elections.
It seems like it would be a real shot in the dark.
What I found is that the accuracy of a prediction (from a poll) depends entirely on your data sources, your methodology, your procedure, your experience, and your analysis.
Why were many 2016 general election polls off by wide margins? Their experience and their methodology had not caught up with the times and the behaviors of the electorate. In particular, it has been studied and determined that many of the polls were not weighted properly for education level of the respondents.
I was “the day before yesterday old” when I learned that, and many other things.
I have learned is that, for the most part, poll results (I’m talking about surveys about predicted voter behavior, not actual vote counts from polling locations) are not raw data. They are not just a summation of answers to questions. They are not “1,000 people said they’d vote for this guy and 970 said they’d vote for that guy, so this guy is xx percentage likely to win.”
I really didn’t know that!
Poll results are carefully processed and analyzed to generate a prediction. Responses are weighted to factor in the ways people surveyed compare with the rest of the electorate and how that compares to the ways other parts of the electorate behave for factors such as gender, education, ethnicity, age, past voting behavior, etc.
The better the data used, the more refined the analysis, the better the accuracy and the result.
#themoreyouknow is, in fact, a valid hashtag!
Rarely are polls completely accurate. There are too many unknowns, and there is always a margin of error when you’re working with anything related to statistics. You can reduce the margin of error with good data and by applying your experience analyzing the type of data you’re working with.
There’s always the possibility of an 11th hour surprise, but there is a reason why campaigns conduct their own polls to decide where and how to allocate resources. Properly run polls overseen by trained professionals yield more useful results than an informal survey of the candidate’s friends and relatives.
Do you want me to guess or do want me to know?
As people come to me and ask me for proposals to make their website or migrate them to a new host or start an email marketing program for them, or grow their existing program, or hire and train a new staff member or make recommendations for marketing, I have to ask them, “Do you want me to guess or do you want me to know?”
If you want me to guess what you need using very little data other than my own experience, well, that’s fast and free.
And it’s probably not very useful.
If you want me to know, within the smallest margin of error, what the best course of action for you will be, I will need data.
Gathering data requires resources.
I’ll need:
- A phone call, during which I ask you lots of questions about your goals, your budgets, your procedures, the limiting factors affecting you business.
- Logins to the software involved in the project I’m going to help with so I (and my team) can see what’s going on before we give recommendations.
- Time to analyze and write out a plan.
Only then can I feel confident in the course of action I’m recommending, and only then should you feel confident. “Why do you offer a free discovery call, then, Katie?” you might ask.
Well, to see if I can tell if it is worth your time and resources to engage me to help you. Maybe there’s a better person to help you. But I won’t know if I’m your person until I do a little preliminary digging.
NOBODY should ever listen to who I think will win an election. I’m not qualified to make that prediction and I have no resources to do so.
It’s the same reason why your uncle might give an enthusiastic thumbs up to a business plan that then flops or your cousin says she hates your logo that your customers adore and that brings you new business.
But if you want to write process documentation, get a new website, start email marketing, or hire a new staff member, I’m your gal and my team’s your team!
Fun & Useful Links
Find your polling place
If you have not yet voted in the U.S. that link has all the info you need.
My favorite cookies have been dethroned
I still love Levain Bakery, but Red’s Scratch Made has pushed them to second place in “KATIE’S GREAT COOKIE ELECTION.” (You really can’t go wrong with either.)
Need help with general IT problems?
Such as: your email isn’t working, somebody hacked your website, you can’t get your IOS to update? Luis Coronado is your guy! My clients love him and have started recommending him to others.
A book that completely swallowed me up last weekend. (I haven’t finished it yet – no spoilers.)
Need social media help? Here are some great resources:
Helen
Kathy
Julie
A book that completely swallowed me up last weekend. (I haven’t finished it yet – no spoilers.)
A book that my friend just loved.
Happy Monday! Let’s do this week!
