NOTES
I'm going to start with a quick mention of deduction, just for completeness, then move onto induction for the main body of the talk - first about some of the ways induction can be misused - mistaken causality, data collection biases, and coincidence, and a few of the ways to overcome these weaknesses - using independent data to test your conclusions , and designing tests to distinguish between hypotheses with some theoretical basis.
Throughout the talk, it should be helpful to remember one of the goals of reasoning - to find the handles that help control things, and avoid trying to control things in unproductive ways - like cutting off a balloon's string to make it stop flying.