Communicating statistics: overcoming the trouble with p-values and confidence intervals

Seminar of the UCL Network of Applied Statistician in Health (NASH) on 21 May 2024. This was a double bill with Hilary Watt (Senior Teaching Fellow and Stats Teaching Expert at Imperial College) and Peter Martin (Associate Professor in Biostatistics, UCL). The presentations offer conceptual explanations of p-values and confidence intervals that steer away from common poor standards of applied statistical interpretations. If it's your job to teach students hypothesis tests and/or to explain to your collaborators why p<0.05 does NOT mean "Hurray, we found an effect", this seminar is for you! (Presentation slides are available as attachments.)