Event overview
Departmental Seminar Series: Dr Rory Allen
Fisher’s F-ratio is basic to statistics, but as an abstract concept students find it hard to grasp. The “broken stick” is a way to present the F-ratio in a single diagram, in which the null and experimental hypotheses are compared. Occam’s razor is represented by a line in this diagram. The F-ratio is formed from two other lines by taking the ratio of their slopes, and it is immediately clear why a high F-ratio corresponds to a desirable experimental hypothesis: it means that the hypothesis escapes decapitation by Occam’s razor. This approach has other benefits than giving a memorable picture of what the F-ratio is, and how to calculate it. It also allows a clearer explanation of how degrees of freedom are derived by subtraction from the more basic and obvious concept of model parameter numbers. It also yields an intuitively appealing explanation of why the R-squared measure of effect size is misleadingly high, and suggests an elementary demonstration of how and why the formula needs to be changed to derive its unbiassed counterpart, adjusted R-squared.
Dr Allen got his PhD with a study on autism and music at Goldsmiths in 2010, and taught statistics on various MSc psychology programmes at Goldsmiths, University of London, from 2008 to 2016. At that point he retired from teaching and worked full time at writing a statistics book entitled ‘Statistics and Experimental Design for Psychologists’, which was published in October 2017. The talk will be on the methods developed in this book.
Dates & times
Date | Time | Add to calendar |
---|---|---|
25 Oct 2018 | 4:00pm - 5:00pm |
Accessibility
If you are attending an event and need the College to help with any mobility requirements you may have, please contact the event organiser in advance to ensure we can accommodate your needs.