1. Understanding Science 3. The Scientific Method © Colin Frayn, 2012 www.frayn.net

2. Introduction © Colin Frayn, 2012 www.frayn.net How do we gain knowledge? –Intuition? –Guesswork? –Dogma? –Word-of-mouth? –Heuristics? What are the drawbacks of these? How can we improve on them?

3. The Steps 1.Define a Question 2.Form Hypotheses 3.Perform Experiments 4.Analyse the Data 5.Interpret the Data 6.Peer Review 7.Repeat the Experiment 8.Continual Review © Colin Frayn, 2012 www.frayn.net

4. Define a Question Question must be… –Specific –Measurable –Objective –Meaningful –Open But not necessarily… –Naturalistic Science is methodologically naturalistic © Colin Frayn, 2012 www.frayn.net

5. Form Hypotheses Possible models to answer our question –Dark matter Elementary particles? Massive objects –Dead stars, wandering planets? Supermassive objects –Black holes? Models to guide our investigation –How far is the moon? Arm’s length? Hundreds of km? Light years? How do we know which measurements to take? © Colin Frayn, 2012 www.frayn.net

6. Perform Experiments Is it plausible? –Back-of-the-envelope calculations –Rough estimates for what we expect to see Set the parameters –We need our result to be Statistically significant Unbiased Unambiguous © Colin Frayn, 2012 www.frayn.net

7. Analyse the Data This is a very tricky area –Scientists often get this wrong Statistical significance –Agreeing with our hypothesis vs. –Confirming our hypothesis Randomised, placebo-controlled trials –The gold-standard for medicine –Avoiding accidental or deliberate bias © Colin Frayn, 2012 www.frayn.net

8. Interpret the Data Bayes’ Theorem –Do our results support our hypothesis? –Remember to include previous results! Publication bias –Half the story is worse than none “Texas sharpshooter” fallacy –Outcomes should be clear before the study is performed © Colin Frayn, 2012 www.frayn.net

9. Peer Review Humans commit errors –Confirmation bias –Measurement errors –Methodological errors –Omissions –Ignorance Peer review reduces errors –Humans think differently –Incentivises accuracy © Colin Frayn, 2012 www.frayn.net

10. Repeat the Experiment Unknown unknowns –E.g. germ theory Anomalies sometimes occur –Repeat to confirm We’re looking for underlying laws –Laws should be consistent –Tests should be repeatable Different assumptions –Different mistakes –Different equipment –Different analysis –Pons & Fleischmann © Colin Frayn, 2012 www.frayn.net

11. Continual Review Science does not dictate truths –…it assigns probabilities Theories are always open to review –By anyone –Though there is a burden of proof Experiments should be repeated –Better equipment –New methods –Different background assumptions © Colin Frayn, 2012 www.frayn.net

12. Criticisms Thomas Kuhn (1922-1996) –“The Structure of Scientific Revolutions” (1962) –Truth by consensus –Progress in stages Paradigm Shifts Paul Feyerabend (1924-1994) –“Against Method” (1975) Didn’t believe in a “scientific method” –Proponent of scientific anarchism –Paradigm shifts “break the rules” © Colin Frayn, 2012 www.frayn.net

13. Summary What is the Scientific Method? –A process –Designed (over centuries) to… Minimise error Maximise discovery of true information Build coherent Theories Understand weaknesses in our models Iteratively improve on our knowledge © Colin Frayn, 2012 www.frayn.net