IntroductiontoBiostatistics
Stat4625/5625
Introduction
OferH
arel
Departm
entofStatistics
University
ofCo
nnecticut
Introduction–p.1/15
IntroductiontoBiostatistics
TimeandPlace:M-W-F11:00-12:00,CL
AS344
Onoccasio
nswe
willm
eetintheteaching
(com
puter)
labCL
AS339
Instructor:OferH
arel
Office:C
LAS320
Phone:486-6989
Office
hours:M-W
9:00-10:00
orby
appointment
Email:[email protected]
Web:
www.stat.uconn.edu/˜oharel
Introduction–p.2/15
IntroductiontoBiostatistics
Teaching
Assistant:JenniferB
oyko
Office:C
LAS317
Office
hours:by
appointment
Email:[email protected]
Introduction–p.3/15
IntroductiontoBiostatistics
Grading:As
follows
Hom
ework:
!"#
Class
participation:
$#
Project%
$#
Exam
1:%$#(WeekofFeb7-11)
Exam
2:%$#(WeekofMar21-25)
FinalExam:&
"#(According
totheschedule)
StudentsofStat4625
andStat5625
willbe
evaluated
usingdifferentcurves.
Hom
ework:Assignmentswillbe
announcedinclassand
placed
ontheclasswe
b-site.A
ssignm
entsareduein
classoftheassign
date.
Introduction–p.4/15
WhatisStatistics?
Statisticaltechniques
arebeingused
inmanyaspectsofourlife.
Surveysforelections,consumerreports,productsatisfaction
etc
Theeffectsofdrugs
Productquality
Econom
etrics
Statistics(Dictionary.com
)–Themathematicsofthecollection,
organizationandinterpretationofnumericaldata,especiallythe
analysisofpopulationcharacteristicsby
inferencefromsampling.
Introduction–p.5/15
WhatisStatistics?
Therearemanydifferentdefinitionsforstatistics,certain
conceptsappearinmostdefinitions:variation,uncertainty,
inference,science.
Inourdailylife
therearemanyexam
ples
forthe
useofStatistics:
Parentsofachild
with
genetic
defectconsidertohave
anotherchild.Theywillbase
theirdecisionon
thechance
ofthenextchild
willhave
thesamedefect.
Tochoose
thebesttherapy,adoctormustchoosebetween
severalpossibilities
Introduction–p.6/15
WhatisStatistics?
Inan
experim
enttoinvestigatewhetherafood
additiveis
carcinogenic(enhancesthechance
forcancer),theUSD
Ahasanimalstreated
with
andwithoutthe
additive.
Doessm
okingcausecancer?
Indesigningandplanning
medicalcarefacilities,thereisa
need
totake
intoaccountchangingneedsofmedicalcare.
Introduction–p.7/15
Biostatistics
Biostatisticsisthestudyofstatisticsappliedtobiologicalareas.
Biologicallaboratoryexperim
ents,m
edicalresearch,and
health
services
research
aresomeofthedifferentareas.
Wefocuson
Biostatisticsforseveralreasons:
Somestatisticalmethodology
isused
moreheavilyinthe
medicalresearch
Exam
ples
aredrawnfrombiological,m
edicalandhealth
services
areas
Theaudience
isdifferent,therearestatisticians
and
non-statisticians
inthisclass
Introduction–p.8/15
OurGoals
Introduce
thewo
rldofbiostatistics
Talkaboutdifferentstudies
types
Use
biostatisticsmethodology
Applyourknowledgetorealdataproblems
Makeinference
Introduction–p.9/15
Studytypes–definitions
Anobservationalstudy
collectsdatafromexistingsituation.
Thedatacollectiondoes
notintentionallyinterferewith
the
runningofthesystem
.
Noticethattheactofobservationmightaffectthesystem
.
Anexperim
entisastudyinwhich
aninvestigatordeliberately
setsoneormorefactorstoaspecificlevel.
Ingeneral,experim
entsleadstostrongerscientificinferences
than
doobservationalstudies.
Introduction–p.10/15
Studytypes–definitions
Alaboratoryexperim
entisan
experim
entthattakes
placein
anenvironm
entw
hereexperim
entalm
anipulationis
facilitated.
Inthiscase,m
ostofthe
variables
ofinterestcanbe
controlledveryclosely(temp,airqualityetc.)
Acomparativeexperim
entisan
experim
entthatcom
pares
twoormorelevels,treatmentsortechniques.
Experim
ent/study
unitisthesm
allestuniton
which
anexperim
entorstudy
isperformed.
Introduction–p.11/15
Studytypes–definitions
Anexperim
entisacrossoverexperimentifthe
sameunit
receive
morethan
onetreatmentorisinvestigated
under
morethan
onecondition.The
differenttreatmentsaregiven
duringnon-overlappingtim
eperiods.
Therisks
with
thistype
ofexperim
entarecrossovereffects,a
change
oftheexperim
entalunitovertime,permanent
physiologicalchangeinhumansandanimals,longertim
e,higherriskfordropout
Aclinicalstudyisonethattakesplaceinasetting
ofclinical
medicine.
Introduction–p.12/15
Definitions
Acohortofpeopleisagroupofpeoplewhose
mem
bershipis
clearly
defined.Forexam
ple,ThestudentsenrolledtoStat
4625/5625forS
pring2011
Anendpointisaclearly
defined
outcom
eoreventassociated
with
enexperim
entalorstudy
unit.Thefinalgradeofthe
above
Aprospective
studyisoneinwhich
acohortofpeopleis
followe
dforthe
occurrenceornon-occurrenceofspecified
endpoints,eventsormeasurements.
Intheanalysisofprospective
studytheoccurrenceofthe
endpointisoftenrelatedtocohortmeasurementsinthe
beginningofthestudy
Introduction–p.13/15
Definitions
Baselinecharacteristicsarevalues
collected
atthetim
eof
entry
tothestudy.
Retrospectivestudyisoneinwhich
peoplehaving
aparticularoutcomeorendpointareidentified
andstudied.For
exam
ple,cancerregistry.
ACase-controlstudy
selectsallcases
thatmeetsaspecific
criteria.A
groupcalledcontrol,thatserves
asacomparison
groupisalso
selected.The
twogroups
arethan
compared.
AMatched
case-controlstudy
matched
thecasesand
controlsaccordingtosomecharacteristic.
Introduction–p.14/15
Definitions
Longitudinalstudy
collectsinformationon
studyunits
overa
specificperiodoftim
e,whilecross-sectionalstudy
collects
informationon
studyunits
atafixed
time.
Aplacebotreatmentisdesigned
toappearexactly
likethe
activetreatmentbuttobe
devoidoftheactivepartofthe
treatment
ThePlaceboeffectresults
fromthebeliefthatone
hasbeen
treated
ratherthan
having
experienced
actualchanges.
Astudyisasingleblindifthesubjectsareunawareofwhich
treatmenttheyarereceiving.Itisdoubleblindifinaddition
thosewho
evaluatethestudy,do
notknowthegroup
assignment.
Introduction–p.15/15