BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
PRODID://PSA//205018
BEGIN:VEVENT
DTSTAMP:20260526T035512
VTIMEZONE:America/New_York
DTSTART:20210218T150000Z
DTEND:20210218T164500Z
UID:205018
SUMMARY:Phil Stat Wars Forum: "Testing with Models That Are Not True" (Christian Hennig)
LOCATION:on-line
DESCRIPTION:Phil Stat Wars Forum: "Testing with Models That Are Not True" (Christian Hennig)\n\n02/18/21 10:00 AM EST\n - 02/18/21 11:45 AM EST\Description:\nChristian Hennig on Testing with Models That Are Not True, at the 6th meeting of the Phil Stat Forum: The Statistics Wars and Their Casualties.\Details:\nIf you would like to constructively engage with us, see this pdf for instructions: https://philstatwars.files.wordpress.com/2020/09/info-phil-stat-forum-september.pdf\n\nTESTING WITH MODELS THAT ARE NOT TRUE\nChristian Hennig (Dept. of Statistical Sciences, University of Bologna)\nABSTRACT: The starting point of my presentation is the apparently popular idea that in order to do hypothesis testing (and more generally frequentist model-based inference) we need to believe that the model is true, and the model assumptions need to be fulfilled. I will argue that this is a misconception. Models are, by their very nature, not "true" in reality. Mathematical results secure favourable characteristics of inference in an artificial model world in which the model assumptions are fulfilled. For using a model in reality we need to ask what happens if the model is violated in a "realistic" way. One key approach is to model a situation in which certain model assumptions of, e.g., the model-based test that we want to apply, are violated, in order to find out what happens then. This, somewhat inconveniently, depends strongly on what we assume, how the model assumptions are violated, whether we make an effort to check them, how we do that, and what alternative actions we take if we find them wanting. I will discuss what we know and what we can't know regarding the appropriateness of the models that we "assume", and how to interpret them appropriately, including new results on conditions for model assumption checking to work well, and on untestable assumptions.\Location:\non-line\n\n,
X-ALT-DESC;FMTTYPE=text/html:Phil Stat Wars Forum: "Testing with Models That Are Not True" (Christian Hennig)<br /><br />02/18/21 10:00 AM EST - 02/18/21 11:45 AM EST<br />Description:<br />Christian Hennig on Testing with Models That Are Not True,&nbsp;at the 6th meeting of the Phil Stat Forum: The Statistics Wars and Their Casualties.<br />Details:<br />If you would like to constructively engage with us, see this pdf for instructions:&nbsp;https://philstatwars.files.wordpress.com/2020/09/info-phil-stat-forum-september.pdf<br />
<br />
TESTING WITH MODELS THAT ARE NOT TRUE<br />
Christian Hennig&nbsp;(Dept. of Statistical Sciences, University of Bologna)<br />
ABSTRACT: The starting point of my presentation is the apparently popular idea that in order to do hypothesis testing (and more generally frequentist model-based inference) we need to believe that the model is true, and the model assumptions need to be fulfilled. I will argue that this is a misconception. Models are, by their very nature, not &quot;true&quot; in reality. Mathematical results secure favourable characteristics of inference in an artificial model world in which the model assumptions are fulfilled. For using a model in reality we need to ask what happens if the model is violated in a &quot;realistic&quot; way. One key approach is to model a situation in which certain model assumptions of, e.g., the model-based test that we want to apply, are violated, in order to find out what happens then. This, somewhat inconveniently, depends strongly on what we assume, how the model assumptions are violated, whether we make an effort to check them, how we do that, and what alternative actions we take if we find them wanting. I will discuss what we know and what we can't know regarding the appropriateness of the models that we &quot;assume&quot;, and how to interpret them appropriately, including new results on conditions for model assumption checking to work well, and on untestable assumptions.<br />Location:<br />on-line<br /><br />,  
PRIORITY:3
TRANSP:TRANSPARENT
BEGIN:VALARM
TRIGGER:-PT5M
ACTION:DISPLAY
DESCRIPTION:Reminder
END:VALARM
END:VEVENT
END:VCALENDAR