I'm being asked to give a lecture on logistic regression to undergraduate psychology students.
I understand what logistic regression is, and how to use it, but I'm not sure I can explain the stats in plain English. I've looked online, but haven't found explanations that I feel are sufficiently 'basic' for undergraduates to grasp.
Does anyone have links to a site/paper/presentation that explains this in VERY BASIC English (no stats knowledge required)?
When I teach Introductory Psychology, I sometimes explain regression models intuitively to explain how researchers found a particular finding. It goes something like this. Imagine somebody is behind screen, we know nothing about them. Our job is to guess how tall they are (i.e., dependent variable). What do we guess? 8 feet / 2.5 m? (Class clearly says no to an absurd guess, students quickly suggest about the average). Yes, knowing nothing, our best guess is the average height (i.e., constant). But now we learn something about the person, say gender. What can we do with our guess? (Students usually catch on very quickly that we modify our guess up or down.) We can keep learning more and more facts about the person (i.e., variables), like country of ancestry and diet. A multiple regression is a statistical technique for writing the equation that nudges our guesses up and down in the best possible way. I’ve personally never gone on to distinguish a linear regression from a logistic regression, because that’s more technical than I’ve needed to be. But if I did, I’d probably go on to imagine somebody else behind a screen and you need to guess if they are male or female. We learn facts, like height, nail length, average voice pitch, etc. A logistic regression is just like a multiple regression, but it’s a statistical way to write an equation to guess category membership. Hope this helps some, Jennifer! ~ Kevin
You have a problem if there is NO background in statistics. The best I can do is recommend my favorite set of notes. They are attached. Best, David Booth
Thanks all for your responses. I've read through a lot of paper, lectures, and internet pages and came up with the attached, which I think is about as simple as I can make it.
Please feel free to use yourselves (please let me know/mention me if you do). Also VERY happy to get comments - hope I haven't written anything incorrect!
- Badges
- Science topic
- Similar topics
- Psychology