BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:When experimentation meets limits\, but simple correlation is unin
 formative: modelling developmental influences and other complex phenomena 
 - Professor Mark Haggard\, Dept. of Experimental Psychology\, Cambridge
DTSTART:20111111T163000Z
DTEND:20111111T180000Z
UID:TALK32711@talks.cam.ac.uk
CONTACT:Nichola Daily
DESCRIPTION:In recent years there have been blistering critiques from clin
 ical triallists and Bayesian theorists of the way statistics are used in g
 eneral science and these have now entered the mainstream (eg Ben Goldacre 
 in the Guardian). This highlights the need for wider/deeper understanding 
 of the principles of method. The problem is greatest where control obliges
  us to handle multivariate data. Contributions of multiple regression and 
 factor analysis are combined in the technique of structural equation model
 ling (SEM). A rational approach is possible by adhering to the methodologi
 cal principle of reconciling empirical adequacy (goodness of fit) with par
 simony (simplicity and structural elegance of model). The application of b
 asic principles of method will be illustrated by the rational choice and r
 epresentation of cascade model of developmental influences in 500 children
  with ear and hearing problems.
LOCATION:Ground Floor Lecture Theatre\, Department of Experimental Psychol
 ogy
END:VEVENT
END:VCALENDAR