Stat 105 Fall, 2015 Notes for unit #12
1
Stat105,Fall2015Unit#12
WerethebillionsofdollarsthatgovernmentsspentontheantiviraldrugTamiflu
toprotecttheirpopulationsfromafluepandemicawasteofmoney?.an
internationalgroupofresearchershassaidthattheanswermaywellbe
Stat 105, 2015, Unit #3
1
Stat 105, Fall, 2015 Unit #3 Review of Hypothesis testing, confidence bounds
I.
An historical look at the development of hypothesis testing.
A. Karl Pearson (1857-1936)
1) First person to articulate the modern statistical view of
Stat 105 fall semester, 2015 Unit #6
Stat 105 Unit #6: Dose Response
A. TheophrastusBombastusvonHohenheim,Paracelsus
16th Century physician of questionable background confronted with
the most pressing medical problem of the day, a mysterious sexually tran
Stat 105, 2015, Unit #2
Stat 105, Fall semester, 2015 Unit #2
I. Medical Science before 1960
A. Typical published study would deal with case reports
or a series of cases:
Patient characteristics, gender, age, demographics.
Presenting symptoms
Differential
Stat 105 Fall, 2015 Unit #4
Stat 105 Fall, 2015 Unit #4
Multilinear Regression
I.
Linear Regression with one variable: Y = A0 + BX + error
A. Least squares algorithm minimizes the sum of the vertical
distances from the line to each point.
1) Passes throug
Stat 105 class notes for Fall semester, 2015
1
Stat 105 Fall semester, 2015 Class Notes Unit #5
I What is the definition of Cause and Effect?
A. Aristotle: Everything has a cause, and we cannot understand something
until we know its cause.
B. David Hume (
Stat 105 Fall, 2015 Class Notes for Unit #7
1
Stat 105 Unit #7
No (mathematical) model is correct, but some models are useful.
George Box (1919-2013)
I. Generalized Linear Models (GLM)
A. Sometimes, theoretical or biological considerations lead to models
Stat 105 Fall semester, 2015 Unit #8
Stat 105 Unit #8 Multiple Testing
I.
A.
B.
C.
D.
E.
The problem of multiple testing
Neyman-Pearson Hypothesis Testing assumes there is only one question
being tested.
1) a well defined null hypothesis, H0
2) a well def
Stat 105 Fall, 2015 Notes for Unit #11
1
Stat105classnotesforUnit#11
Most decision-making statistical work has nothing to do with probability or
mathematics: it is mostly a matter of getting ones hands dirty in the search for good
basic data and cleaning
Stat 105 Fall Semester, 2015, Unit #10
Stat 105, Fall, 2015 Unit #10Sensitivity and Selectivity
I. Measuring symptoms to diagnose disease.
A, Dr. X has a new method for diagnosing early stage cancer.
The method has correctly identified cancer in over 300
Stat 105 Fall Semester, 2015, Unit #1
1
STAT 105 LECTURE NOTES FOR UNIT #1
David Salsburg:
I can be reached at
david.salsburg@yale.edu
My home telephone number is
860.443.2577
The nature of pure mathematics is such that the conclusions follow inevitably f