0% found this document useful (0 votes)
142 views46 pages

SAS 201 - Introduction To Research Methods - LECTURE NOTES

Uploaded by

allanopasi1
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
142 views46 pages

SAS 201 - Introduction To Research Methods - LECTURE NOTES

Uploaded by

allanopasi1
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 46

SAS201:I

NTRODUCTI
ONTORESEARCHMETHODS

Tr
aini
ngManual
for2ndYearBA(
Soci
ology
)St
udent
s

Mr
.MosesMut
ua&Dr
.Samuel
Chessa
I
nst
ruct
or

Depar
tmentofSoci
ology&Psy
chol
ogy
Moi
Uni
ver
sit
y

SEPTEMBER2014
Page1of1
Exami
net
hreef
act
orst
oconsi
derwhenr
evi
ewi
ngl
i
ter
atur
e(10mks)

1.I
ntr
oduct
iont
otheResear
chPr
ocess
Whati
sresear
ch?

Def
inet
het
ermr
esear
chhi
ghl
i
ght
ingt
wokeyf
eat
uresofsoci
alsci
ence
r
esear
ch(
5mks
Thewor
d"r
esear
ch"i
sder
ivedf
rom t
heLat
inwor
dmeani
ng"
toknow.
"Resear
chi
s
aboutanswer
ingquest
ions,
suchas:
 WhatdoIwantt
oknow?
 HowdoIwantt
ogai
nknowl
edge?
 WhydoIwantt
oknowi
t?
Usi
ngr
elev
antexampl
es,
discusswhaty
ouunder
standbysoci
alr
esear
ch

Thi
sist
hesy
stemat
icappl
i
cat
ionofsci
ent
if
icappr
oacht
othei
nvest
igat
ionoft
he
occur
renceofcer
tai
nev
ent
sint
ermsofhowt
heyoccur
,whyt
heyoccurandt
her
esul
ts
oft
hei
roccur
rencef
orexampl
epov
ert
y,FGM,
cor
rupt
ion,
cri
me,
etc.I
tist
her
efor
enota
haphazar
d under
taki
ng buta st
ep-
by-
step pr
e-pl
anned act
ivi
tyt
hatendeav
our
sto
under
stand an ev
ent
.Ithas t
he ai
m ofi
nter
pret
ing and est
abl
i
shi
ng f
act
s and
di
scov
eri
ngnew i
nfor
mat
ionaboutt
heev
entandi
nvol
vescol
l
ect
ingi
nfor
mat
ionf
rom
v
ari
oussour
cesusi
ngav
ari
etyofmet
hodsofdat
acol
l
ect
ion.Resear
chcanbeas
si
mpl
easconsul
ti
ngaf
ri
endorascompl
exasdesi
gni
ngandi
mpl
ement
ingaf
ormal
st keSur
udyl
i viv
alSt
rat
egi
esamongWomenLi
vi
ngwi
thHI
V/AI
DSi
nNai
robiSl
ums.

Asanest
abl
i
shedbodyofknowl
edge,r
esear
chi
stheaccumul
ati
onofev
idenceby
r
esear
cher
sov
ert
ime.I
tpr
ovi
desabackgr
oundagai
nstwhi
chcl
aimsf
orknowl
edgeca

Page2of2
nbeev
aluat
ed,andpr
ovi
desawayf
orr
esear
cher
stocommuni
cat
eandcompar
ethei
r
f
indi
ngs.Resear
chasacol
l
ect
ionofmet
hods,
tool
s,andt
echni
quesf
ormst
hebasi
sfor
mostr
esear
cht
ext
sandcour
sesi
nresear
ch;i
tsmet
hodsar
ewhatseti
tapar
tfr
om
ot
herway
sofacqui
ri
ngknowl
edge.Asapur
posef
ulorf
unct
ionalact
ivi
ty,r
esear
ch
mustj
ust
if
yther
esour
cesi
tconsumest
oaccompl
i
shi
tspur
poses.

I
mpor
tance/
rol
eofr
esear
ch
1.I
tof
fer
sawayofexami
ningandunder
standi
ngoper
ati
onsofhumansoci
alaf
fai
rs
e.
g.t
o expl
ain copi
ng mechani
sms underr
etr
enchment
,env
ironment
alef
fect
s of
def
orest
ati
on,et
c.Thi
s ki
nd ofi
nfor
mat
ion i
simpor
tantf
ort
aki
ng appr
opr
iat
e
i
nter
vent
ionact
ion.
2.I
tof
fer
sway
sofsol
vi
ngsoci
alpr
obl
emsi
.e.get
stof
indoutanswer
stosoci
al
pr
obl
emsaf
fect
ingi
ndi
vi
dual
sinsoci
etye.
g.st
reetchi
l
drenpr
obl
ems,
pov
ert
y,et
c.The
wor
kofar
esear
cheri
sther
efor
etoof
fersuggest
ionsbutnotsol
uti
ons.Sol
uti
onsl
i
e
squar
elywi
tht
hegov
ernmentandwel
far
eagenci
esort
hei
ndi
vi
dual
him/
her
sel
f.
3.I
tpr
ovi
despoi
ntsofv
iewandt
echni
calpr
ocedur
es,whi
chuncov
ert
hingswhi
chwe
ar
e notawar
e ofe.
g.wi
tht
he i
mpl
ement
ati
on ofa dev
elopmentpr
oject
,some
unf
oreseen negat
ive ef
fectmayemer
ge t
hatcan be uncov
ered byr
esear
ch.For
exampl
e,wi
tht
hebui
l
dingofadam,mosqui
toesmayst
artbr
eedi
ngandhencespr
ead
ofmal
ari
aandt
her
efor
eincr
easei
nhumanmor
bidi
tyandmor
tal
i
ty.
4.Tot
esti
deasandt
hei
rvi
abi
l
ity
,accept
abi
l
ityandconf
ormi
tywi
tht
hepr
evai
l
ing
ci
rcumst
ances.Forexampl
e,t
otestwhet
heracer
tai
norgi
veni
nter
vent
ionst
rat
egyi
s
accept
abl
etot
hosewhoi
tissupposedt
obenef
it
,whet
heri
tiscompat
ibl
ewi
tht
hei
r
cul
tur
albel
i
efs and t
esti
tsv
iabi
l
ityi
nter
ms ofr
etur
nfori
nvest
mentand t
he
i
ntended/
expect
edout
come.
5.Togauget
heoppor
tuni
tycostofanyf
orgoneal
ter
nat
iveact
ivi
tysot
hatt
hebest
al
ter
nat
ivei
nter
msofcost
sandbenef
it
sisunder
taken.Forexampl
e,i
fiti
sdi
scov
ered
t
hatt
heoppor
tuni
tycostofbui
l
dingadam att
heexpenseofacat
tl
edi
pisv
eryhi
gh,
t
heni
twi
l
lbeadv
isabl
etobui
l
dthedi
p.Thi
ski
ndofi
nfor
mat
ioncanonl
ybegot
ten

Page3of3
t
hroughr
esear
ch.
6.Tot
estt
het
heor
eti
calf
ramewor
kunder
lyi
ngasoci
alphenomenont
obeunder
taken.
Thi
s ev
ent
ual
l
ycont
ri
but
es t
o enhancementofknowl
edge and t
hus ser
ves as a
consi
der
abl
econt
ri
but
iont
oschol
arshi
p.Wecoul
ddoast
udyt
oseewhet
herat
heor
y
l
i
keDi
ff
erent
ial
Associ
ati
onTheor
yofcr
imei
sst
il
lappl
i
cabl
e.
7.Soci
alsci
encer
esear
chf
indi
ngscanbeusedt
ofor
mul
atei
nter
vent
ionpol
i
ciesand
st
rat
egi
est
odealwi
thanypr
obl
em i
dent
if
ied.Assuch,r
esear
chenabl
esmaki
ngof
r
ati
onal
,fact
-baseddeci
sions.Thi
scanbeunear
thedt
hroughev
aluat
iver
esear
ch.
8.Togaugegoalat
tai
nmentandbywhatst
andar
ds/
lev
elst
heyhav
ebeenachi
evede.
g.
t
o det
ermi
ne whet
herf
ree pr
imar
yschooleducat
ion has achi
eved i
tsi
nit
ial
l
yset
obj
ect
ives.
9.Togaugeef
fect
ivenessandpr
ogr
essofpr
oject
/pr
ogr
ammeandt
hef
inali
mpact
.
Thi
scanbeachi
evedt
hrought
ermi
nal
/summat
iveev
aluat
ionoft
hepr
ogr
ammeor
pr
oject
.
10.Tor
aisepeopl
e’
sconsci
ousnessandabi
l
ityt
obesel
f-
cri
ti
cal
,especi
all
ypr
oject
i
mpl
ement
ers.Thi
shel
pst
hem t
obebet
termanager
sandi
mpl
ement
ersi
nsubsequent
under
taki
ngs.
11.Togaugecapaci
tybui
l
dingi
mpactofanygi
vensoci
ali
nter
vent
ionpr
ogr
ammes.For
exampl
e,onemaydoast
udyt
odet
ermi
net
hei
mpactofcer
tai
ntr
aini
ngt
otr
ainees/
to
det
ermi
nei
tscont
ri
but
iont
omanpowerdev
elopment
.
12.I
tpr
ovi
desknowl
edget
hatcanbeusedt
odet
ermi
net
hegener
alappl
i
cabi
l
ityof
obj
ect
ivesandt
heef
fi
ci
entut
il
izat
ionofr
esour
ces.Wi
thoutt
echnol
ogi
calknowl
edge
andpower
,manyr
esour
ceswoul
dgot
owast
esi
nceef
fi
ciencywoul
dbenon-
exi
stent
.
I
magi
neasi
tuat
ionwi
thoutt
het
elephonesy
stem,y
etonewant
stopasssomeur
gent
message?Justi
magi
nehow manyday
sonewoul
dneedt
owal
kfr
om onepl
acet
o
anot
herwi
thoutmot
orcar
s.Assuch,
resear
chensur
esopt
imum andef
fi
cientut
il
izat
ion
ofav
ail
abl
eresour
ces.
13.Resear
chcananddoespr
ovi
dei
nfor
mat
ionont
heaspect
sorar
east
hatneed
ur
genti
nter
vent
ionandhencepr
ior
it
izet
hem f
oract
ion.Forexampl
e,i
tcanbeusedt
o
det
ermi
newhi
chpr
obl
emsaf
fectpeopl
emost
.

Page4of4
Sci
ent
if
icandnon-
sci
ent
if
icr
esear
ch
I
.Sci
ent
if
icr
esear
ch
Sci
ent
if
icr
esear
chi
sthesy
stemat
ic,cont
rol
l
ed,empi
ri
calandcr
it
icali
nvest
igat
ionof
hy
pot
het
ical
proposi
ti
onsaboutt
hepr
esumedr
elat
ionsamongnat
ural
phenomenon.

Br
ief
lydi
scusschar
act
eri
sti
csofsci
ent
if
icmet
hodandhowi
tdi
ff
ersf
rom
ot
herway
sofseeki
ngknowl
edge
Bysy
stemat
icandcont
rol
l
edi
smeantt
hatsci
ent
if
ici
nvest
igat
ioni
ssoor
der
edt
hat
r
esear
cher
scanhav
econf
idencei
ntheout
comeoft
her
esear
ch.I
thast
hef
oll
owi
ng
char
act
eri
sti
cs:
1.Sci
ent
if
icr
esear
chi
sreduct
ive.Thi
smeanst
hatt
ogett
ounder
standt
her
elat
ionshi
p
bet
weenv
ari
abl
esorphenomenon,t
hecompl
exi
tyofr
eal
i
tyi
sreducedi
.e.t
hedet
ail
s
t
hatar
enotr
elev
antar
eignor
ed.Forexampl
e,i
fwear
elooki
ngatt
hequal
i
tyofa
bl
anket
,wemayf
ir
staskour
sel
vest
hepur
poseoft
hebl
anket(
conser
vingwar
mth)
.As
sucht
hemat
eri
al(
wool
,cot
ton,et
c)t
hatmakesi
t,i
srel
evantandhencemustbe
consi
der
ed.Howev
er,i
tscol
ormaynotber
elev
antandhencei
tshoul
dbei
gnor
edso
t
hatwer
educeouranal
ysi
stot
hemat
eri
almaki
ngt
hebl
anket
.
2.Dat
acol
l
ect
ion pr
ocessi
swel
lpl
anned and cont
rol
l
ed t
o enhancev
ali
dit
yand
r
eli
abi
l
ity
.Hence,
sci
ent
if
icr
esear
chi
sor
der
ed,
logi
cal
andsy
stemat
icsucht
hatnost
ep
i
sby
passed.Forexampl
e,onemusthav
eaconci
seunder
standi
ngoft
hepr
obl
em of
st
udybef
orecomi
ngupwi
thaquest
ionnai
ret
hatwoul
dbeputt
other
espondent
sinan
at
temptt
ogett
her
equi
redi
nfor
mat
ion.
3.Ther
esear
chi
sobj
ect
iver
athert
hansubj
ect
ive.Hencer
esear
cher
sar
ecal
l
edupon
t
omi
nimi
zet
hei
rownbi
asesandknowl
edgeaboutt
hephenomenonunder
studyandl
et
t
her
esear
chdat
aspeakf
ori
tsel
f.
4.Ther
esear
chi
sbased/
bui
l
donempi
ri
caldat
ai.
e.eachst
epi
nsci
ent
if
icr
esear
chi
s
basedonobser
vat
ion.Her
e,i
tisassumedt
hatr
eal
i
tyexi
stsout
sidet
heobser
verand
hencer
esear
cher
sar
esupposedt
obeneut
rali
ndoi
ngt
hei
rresear
ch.Ot
her
wise,t
hey

Page5of5
wi
l
lloseobj
ect
ivi
tyandcomeupwi
thdat
athatr
epr
esent
sthei
rownopi
nionsr
ather
t
hant
her
eal
i
ty.
5.Thef
indi
ngsar
ever
if
iabl
eandr
epr
oduci
ble.Assuch,anot
herr
esear
cherdoi
ngt
he
samest
udywi
tht
hesamepar
ti
cipant
sandundert
hesamecondi
ti
onswi
l
lgett
hesame
r
esul
ts/
findi
ngs.

I
I.Non-
sci
ent
if
icr
esear
ch
1.Aut
hor
it
ymet
hod-
thi
sist
hemet
hodt
hatmanypeopl
edependonwhet
heradul
tsor
chi
l
dren.Forexampl
e,chi
l
drenl
ear
nmanyt
hingsf
ir
stf
rom t
hei
rpar
ent
sandsi
gni
fi
cant
ot
her
sar
oundt
hem.Whenadul
tsr
elyont
heknowl
edgeandwi
sdom ofpr
omi
nent
peopl
ethoughtorr
egar
dedashav
ingabet
terknowl
edgeoft
hei
rar
eassuchasdoct
ors,
pol
i
tici
ans,engi
neer
s,l
awy
ers,et
c,t
hismet
hodofacqui
ri
ngknowl
edgei
scal
l
edt
he
aut
hor
it
ymet
hod.Assuch,t
he st
atement
soft
hese peopl
e orexper
ts ar
erar
ely
quest
ioned,butaccept
edasabsol
ute.Hencecer
tai
namountoff
ait
hispl
acedi
nthese
peopl
eassour
cesofknowl
edge.Suchot
herpeopl
eincl
udet
heel
der
lyi
nsoci
eti
es
wher
efor
maleducat
ioni
smi
nimal
,ki
ngsi
nanci
entt
imesandheadsofr
eli
gious
or
gani
zat
ions.
2.My
sti
calmet
hod-
her
e,t
hecor
rect
nessorv
ali
dit
yofknowl
edgei
sassumedt
oresi
de
wi
thasuper
nat
ural
sour
cesuchast
hecasewi
thseer
s,pr
ophet
s,et
c.Hencesour
cesof
knowl
edgear
eregar
dedasaut
hor
it
iesduet
othei
rabi
l
ityt
otr
ansmi
tknowl
edgeort
rut
h
i
mpar
tedt
othem bysuper
nat
uralpower
s/bei
ngs.Themy
sti
calmet
hodofacqui
ri
ng
knowl
edgel
osescr
edi
bil
i
tywhenwecomeupwi
thbet
terway
sofget
ti
ngknowl
edge.
3.I
ntui
ti
ve met
hod-
thi
s means t
hatsomet
imes peopl
e make sense oft
he wor
ld
accor
dingt
owhatt
heyt
hinki
sri
ghtorsui
tst
hem.Forexampl
e,adoct
ormayi
nter
view
apat
ientwi
thoutmaki
ngal
abor
ator
yexami
nat
iony
etar
ri
veataconcl
usi
onaboutt
he
ai
l
mentoft
hepat
ientbasedont
heexpl
anat
iongi
venbyt
hepat
ient
.Thepr
obl
em wi
th
t
hei
ntui
ti
vemet
hodi
sthati
tcannotbet
hesamef
oral
li
ndi
vi
dual
sbecausef
orexampl
e
di
ff
erentdoct
orscancomet
odi
ff
erentconcl
usi
onsaboutt
hesameai
l
mentbasedon
t
hei
rdi
ff
erenti
ntui
ti
ons.Forexampl
e,t
wodoct
orsl
i
steni
ngt
othesamepat
ientcan
makedi
ff
erentconcl
usi
onssuchast
hepat
ienti
ssuf
fer
ingf
rom mal
ari
aort
yphoi
d.

Page6of6
4.Conv
ent
ionalwi
sdom-
thi
sref
lect
swhati
scal
l
edcommonsenseunder
standi
ngof
t
hewor
ldt
hatar
ecommonl
yaccept
edast
rue.Howev
er,commonsensesomet
imes
cont
radi
ctsi
tsel
f.Forex
ampl hatunl
e,weheart i
kepol
esat
tr
actandt
hatbi
rdsoft
he
samef
eat
uref
lockt
oget
her
.Soher
ewhi
chonei
str
ue?
5.Rat
ional
i
sti
cmet
hod-
thi
sisbasedont
hehumanr
easonsi
ncehumanbei
ngar
e
knownt
othi
nkl
ogi
cal
l
yorr
eason.Thebasi
sofknowl
edgei
sthuscor
rectr
easoni
ng
whi
ch enabl
es one t
o know whatmustbe t
rue as a mat
terofpr
inci
ple.Her
e,
obser
vat
ionsofr
eal
i
tyorempi
ri
calst
udi
esar
enotnecessar
ye.
g.i
nmat
hemat
ics.For
exampl
e,onepl
usonei
stwoandnot
hingmor
e.
6.Empi
ri
calmet
hod-
her
e,f
act
sobser
vedi
nnat
urear
ethef
oundat
ionofknowl
edge.
Obj
ect
ivi
tyofobser
vat
ioni
semphasi
zedandonl
ywhati
sobser
vabl
ei.
e.whi
chcanbe
per
cei
vedbyt
hesensesconst
it
utesknowl
edge.Forexampl
e,api
eceofwoodf
loat
son
wat
erwhi
l
ethatofi
ronsi
nks.Forapur
eempi
ri
cist
s,knowl
edgest
opsher
ewi
thout
aski
ngt
her
easonswhyt
heset
womat
eri
alsbehav
eli
ket
hatwhent
heyar
eputi
nto
wat
er.Howev
er,t
hesci
ent
if
icmet
hodwoul
dcombi
nebot
hrat
ional
i
sti
candempi
ri
cal
met
hodandgobey
ondt
oaskandat
temptt
oanswert
hequest
ionwhy
.

2.TheResear
chPr
ocess
I
.Pr
obl
em i
dent
if
icat
ion
Whati
sar
esear
chpr
obl
em?
Ar
esear
chpr
obl
em i
sthesi
tuat
iont
hatcausest
her
esear
chert
ofeelappr
ehensi
ve,
conf
usedandi
l
latease.I
tist
hedemar
cat
ionofapr
obl
em ar
eawi
thi
nacer
tai
n
cont
exti
nvol
vi
ngt
heWHO orWHAT,t
heWHERE,t
heWHEN andt
heWHY oft
he
pr
obl
em si
tuat
ion.I
naconv
ent
ionalsense:apr
obl
em i
sasetofcondi
ti
onsneedi
ng
di
scussi
on,asol
uti
on,andi
nfor
mat
ion,whi
l
eini
tst
echni
calmeani
ngi
timpl
i
est
he
possi
bil
i
tyofempi
ri
cal
inv
est
igat
ion-
thepossi
bil
i
tyofdat
acol
l
ect
ionandanal
ysi
s

Resear
ch can t
husbe ai
med atcl
ari
fyi
ng orsubst
ant
iat
ing an exi
sti
ng t
heor
y,at
cl
ari
fyi
ngcont
radi
ctor
yfi
ndi
ngs,atcor
rect
ingaf
aul
tymet
hodol
ogy
,atcor
rect
ingt
he
i
nadequat
e orunsui
tabl
e use ofst
ati
sti
calt
echni
ques,atr
econci
l
ing conf
li
cti
ng

Page7of7
opi
nions,
oratsol
vi
ngexi
sti
ngpr
act
icalpr
obl
ems.Def
ini
ngar
esear
chpr
obl
em i
nvol
ves
sev
eral
int
err
elat
edst
eps.St
epnumber
 Ascer
tai
nthedeci
sionmaker
’sobj
ect
ives.
 Under
standt
hebackgr
oundoft
hepr
obl
em.
 I
sol
ateandi
dent
if
ythepr
obl
em,
rat
hert
hani
tssy
mpt
oms.
 Det
ermi
net
heuni
tofanal
ysi
s.
 Det
ermi
net
her
elev
antv
ari
abl
es.
 St
atet
her
esear
chquest
ionsorhy
pot
hesesandr
esear
chobj
ect
ives.

a.Sour
cesofr
esear
chpr
obl
ems
1.Obser
vat
ionofr
eal
i
ty-
cont
actwi
tht
hewor
ldandhencedi
rectobser
vat
ionoft
he
wor
ldmayr
aiseani
ndi
vi
dual

scur
iosi
ty.Asar
esul
t,onemaygeti
nter
est
edt
odoa
st
udyt
oar
ri
veatconcl
usi
onaboutev
ent
s,phenomenon,happeni
ngs,et
c.Forexampl
e,
onemayobser
vet
hati
nacer
tai
nar
ea,mostchi
l
drendo notat
tendschool
.The
obser
vermaybei
nter
est
edi
ninv
est
igat
ingt
her
easonsbehi
ndsuchandcomeupwi
th
r
ecommendat
ionst
ohel
pthechi
l
drengot
oschool
.
2.Theor
y-i
fa t
heor
yis cor
rect
;one mayexpectt
hatt
hings happen as pert
he
t
enet
s/pr
oposi
ti
onsoft
het
heor
y.I
fther
eisanyev
entt
hatgoesagai
nstt
het
heor
y,t
hen
onemaybei
nter
est
edi
ninv
est
igat
ingt
her
easonwhyt
hispar
ti
cul
arev
entgoesagai
nst
t
het
heor
y.Forexampl
e,wat
eri
sexpect
edt
ofl
owdownsl
opeduet
ogr
avi
tat
ional
for
ce,
buti
fitf
lowsuphi
l
lthent
hisf
ormsar
easonf
orst
udy
,wi
tht
heai
m ofdet
ermi
ningwhy
t
hewat
eri
sfl
owi
nguphi
l
lagai
nstexpect
ati
ons.
3.Pr
evi
ousr
esear
ch/
studi
es-
dur
ingl
i
ter
atur
erev
iew ornor
malr
eadi
ngofr
esear
ch
f
indi
ngsaboutaphenomenon,onemayf
ind t
hatpr
evi
ousst
udi
escont
radi
ctone
anot
her
.Inaddi
ti
on,pr
evi
ousst
udi
esmaypr
oduceambi
guousr
esul
tst
hatr
equi
re
f
urt
heri
nvest
igat
ion.Mor
eov
er,
ear
li
erst
udi
esmayber
epeat
edwi
thdi
ff
erentgr
oupsof
par
ti
cipant
sorusi
ngdi
ff
erentdat
acol
l
ect
ionmet
hods.Thef
oregoi
ngmayal
sobedone
t
over
if
ywhet
herdat
athatwascol
l
ect
ed10y
ear
sagost
il
lhol
dsatt
hemomentor
somet
hing mayhav
e changed.An exampl
e her
e maybe HI
V/AI
DS pr
eval
ence or
awar
enessandpov
ert
ypr
eval
encer
ates.

Page8of8
4.Pr
act
icalconcer
ns-
exper
iences i
n a cer
tai
n count
ryorcommuni
tymayt
ri
gger
r
esear
ch.Forexampl
e,i
fther
ehasbeenet
hni
cconf
li
cti
nacer
tai
ncount
ryorr
egi
on,
t
hiscanbeasour
ceofr
esear
cht
opi
ctounear
tht
her
easonsbehi
ndi
t.Suchot
her
ev
ent
sthatmayt
ri
ggerr
esear
char
ecr
ime,
unempl
oyment
,pov
ert
y,et
c.
5.Per
sonali
nter
est
s-r
esear
chmaybeunder
takenconsequentofper
sonali
nter
estof
t
her
esear
cher
.Forexampl
e,ani
ndi
vi
dualmaybei
nter
est
edi
ngenderst
udi
esand
hencei
nit
iat
east
udyi
nthesamear
ea.
6.Di
scussi
onwi
thexper
ts-
discussi
onongener
alt
opi
csi
nwor
kshopsorsemi
nar
scan
f
orm abasi
sforr
esear
cht
ounder
standphenomenonbet
ter
.Inaddi
ti
on,di
scussi
ons
wi
thexer
tsofat
opi
cmaycr
eat
eint
erestt
owar
dsr
esear
ch.
7.Themedi
a-i
ssuest
hatar
efr
equent
lyf
eat
uredi
nthemedi
acanbeasour
ceofnew
needt
ounder
takeast
udyf
orexampl
etr
ibalcl
ashesi
nKeny
a,domest
icv
iol
ence,bank
r
obber
ies,
etc.
8.Of
fi
cialdocument
s-of
fi
cialdocument
ssuchasdepar
tment
alf
il
es,
popul
ati
oncensus
r
epor
ts,HI
V/AI
DSpr
eval
encer
epor
ts,cr
imer
epor
ts,r
epor
tsofcommi
ssi
onsofi
nqui
ry,
et
ccanf
orm t
hebasi
sonwhi
cht
oini
ti
atear
esear
ch.Thi
smaybedr
awnf
rom
r
ecommendat
ionsmadei
ntheof
fi
cial
document
s.

b.Char
act
eri
sti
csofagoodr
esear
chpr
obl
em
1.I
tshoul
dbet
imel
yandshoul
dnott
akepl
aceaf
tert
hepr
obl
em hasescal
atedt
o
appoi
ntwher
eiti
shi
ghl
yunl
i
kel
ythatt
hef
indi
ngswi
l
lcont
ri
but
einsol
vi
ngt
hepr
obl
em
orpr
event
ingf
urt
herescal
ati
onoft
hepr
obl
em.
2.I
tshoul
drel
atet
oapr
act
icalpr
obl
em i
nreall
i
feandnotj
usti
magi
nat
ione.
g.pov
ert
y,
cr
ime,
etc.
3.Ther
esear
chpr
obl
em shoul
dbeexper
iencedby
,andr
elat
etoaspeci
fi
cdef
inabl
e
popul
ati
ongr
oupe.
g.Keny
ans,
RedI
ndi
ans,
Keny
anAsi
ans,
Musl
i
ms,
etc.
4.Resear
chpr
obl
em shoul
dpr
ovi
der
esul
tst
ofi
l
lagapi
nexi
sti
ngknowl
edgeby
pr
ovi
dinggr
eat
eri
nsi
ght
s/deeperunder
standi
ngi
ntoaspeci
fi
car
eaofknowl
edge.
5.Resear
choccasi
onedbyt
her
esear
chpr
obl
em shoul
dhav
eapr
act
icalv
aluei
nthat
t
he appl
i
cat
ion oft
he r
ecommendat
ions based on t
he f
indi
ngs shoul
d cont
ri
but
e

Page9of9
t
owar
dsr
ect
if
yinganexi
sti
ngunaccept
abl
esi
tuat
ion.
6.Ar
esear
chpr
obl
em shoul
dbemanageabl
eandcl
ear
lydemar
cat
edi
nor
dert
ofocus
onaspeci
fi
cpr
obl
em,t
aki
ngi
ntoaccountav
ail
abl
eti
me,f
inances,sampl
esi
zeand
abi
l
iti
esoft
her
esear
cher
.
7.The r
esear
ch pr
obl
em shoul
d be based on empi
ri
cal
l
yver
if
iabl
e dat
a orf
act
s
obt
ainedbyst
udy
ingr
eal
i
ty.Forexampl
e,i
sitpossi
blet
ost
udywher
eGodi
sorhowHe
wor
ks?Si
ncet
hisi
snotpr
act
ical
l
yver
if
iabl
e,i
tisbet
terl
eftt
ofai
thr
athert
hanr
esear
ch.
8.A r
esear
chpr
obl
em shoul
dbet
opi
cali
.e.mustber
elev
antt
othecur
rentsoci
al
si
tuat
ionandnotout
dat
ed.
I
I.Resear
chobj
ect
ives
Whati
sar
esear
chobj
ect
ive?
Anobj
ect
ivei
sanyki
ndofdesi
redendorcondi
ti
oni
.e.t
hespeci
fi
caspect
soft
he
phenomenonunder
studyt
hatt
her
esear
cherdesi
rest
obr
ingoutatt
heendoft
hest
udy
.
Obj
ect
ivesar
enor
mal
l
yoft
wot
ypesi
.e.t
hegener
alandspeci
fi
cones.Whyshoul
d
r
esear
chobj
ect
ivesbedev
eloped?Thef
ormul
ati
onofobj
ect
iveswi
l
lhel
pyout
o:
 Focust
hest
udy(
nar
rowi
ngi
tdownt
oessent
ial
s);
 Av
oidt
hecol
l
ect
ionofdat
awhi
char
enotnecessar
yforunder
standi
ngandsol
vi
ng
t
hepr
obl
em y
ouhav
eident
if
ied;
and
 Or
gani
zet
hest
udyi
ncl
ear
lydef
inedpar
tsorphases.
 Faci
l
itat
est
hedev
elopmentofr
esear
chmet
hodol
ogy
 Or
ient
sthecol
l
ect
ion,
anal
ysi
s,i
nter
pret
ati
onandut
il
izat
ionofdat
a.

I
I(a)
.Thegener
alobj
ect
ive/
pur
pose/
aim
Thegener
alobj
ect
ivei
sal
socal
l
edt
hepur
poseoft
hest
udyanddef
inest
heai
msi
na
br
oadway
.Inst
ati
ngt
hegener
al/
pur
poseoft
hest
udy
,ther
esear
cherconv
eyst
he
f
ocusoft
hest
udyi
noneort
wosent
ences.Theuseoft
hegener
alobj
ect
iveort
he
pur
posei
stof
ocust
her
esear
cher
’si
nqui
ryi
ntoapar
ti
cul
arar
eaofknowl
edgei
na
gener
alf
iel
d.Thegener
alobj
ect
ivesmayusesev
eralact
ivev
erbsatt
hesamet
ime.
Thegener
alobj
ect
ive/
pur
poseshoul
dnotbest
atedv
erygener
all
yast
obeuncl
earsuch
ast
hef
oll
owi
ngst
atement
s:

Page10of10
1.Pr
obl
emsofor
phanedchi
l
dren
2.Foodi
nsecur
it
yinI
siol
oDi
str
ict
3.I
nadequat
eresour
cesamongpeasantf
armer
s

Thef
oregoi
ngst
atement
sar
enotcl
earandt
hequest
iont
hatmaybeaskedwoul
dbe:
whataboutt
hepr
obl
ems,f
oodsecur
it
yandt
hei
nadequat
eresour
ces?Assuch,si
nce
t
hef
oregoi
ngst
atement
sar
enotcl
ear
,thegener
alobj
ect
ive/
pur
poseshoul
dbest
ated
i
nsuchamannerast
ofi
tthef
oll
owi
ngcr
it
eri
a:
1.Thepur
poseoft
hest
udymustbecl
earandnotambi
guous
2.I
tshoul
dindi
cat
ethev
ari
abl
esorconcept
sint
hest
udy
3.Wher
epossi
ble,
ther
elat
ionshi
psbet
weent
hev
ari
abl
esshoul
dbest
ated
4.I
tshoul
dindi
cat
ethet
argetpopul
ati
on
5.Thev
ari
abl
esandt
argetpopul
ati
oni
nthegener
alobj
ect
iveorpur
poseoft
he
st
udyshoul
dbet
hesameast
hosei
nthemet
hodol
ogy
6.Cov
ert
hedi
ff
erentaspect
soft
hepr
obl
em andi
tscont
ri
but
ingf
act
orsi
na
coher
entwayandi
nal
ogi
cal
sequence
7.Ar
ecl
ear
lyphr
asedi
noper
ati
onalt
erms,
speci
fyi
ngexact
lywhaty
ouar
egoi
ngt
o
do,
wher
e,andf
orwhatpur
pose;
8.Ar
ereal
i
sti
cconsi
der
ingl
ocal
condi
ti
onsandr
esour
cesav
ail
abl
e
9.Useact
ionv
erbst
hatar
especi
fi
cenought
obeev
aluat
ed.

Thef
oll
owi
ngar
ethel
i
kel
yst
udypur
posesorgener
alobj
ect
ives
1.The gener
alobj
ect
ive/
pur
pose oft
his st
udy i
sto det
ermi
ne t
he sur
viv
al
st
rat
egi
esofpeopl
eli
vi
ngwi
thHI
V/AI
DSi
nur
bani
nfor
mal
set
tl
ement
sofNai
robi
,
Keny
a.
2.The pur
pose oft
his st
udy i
sto est
abl
i
sh t
he f
act
orst
hatcont
ri
but
eto
mal
nut
ri
ti
onofchi
l
dreni
nKeny
a
3.Thepur
poseoft
hisst
udyi
stof
indoutt
hecat
tl
egr
azi
ngpr
act
icesandl
and
degr
adat
ioni
nMoy
aleDi
str
ict

Page11of11
I
I(b)
.Thespeci
fi
cobj
ect
ives
Thespeci
fi
cobj
ect
ivesgi
vet
hespeci
fi
csoft
her
esear
chdet
ail
sbyr
efer
ri
ngt
ospeci
fi
c
v
ari
abl
eindi
cat
orsandmustonl
yuseoneact
ivev
erbatat
ime.Exampl
esofspeci
fi
c
obj
ect
ivescoul
dbest
atedasf
oll
ows:
Thespeci
fi
cobj
ect
ivesoft
hisst
udyar
e:
1.Toi
nvest
igat
ethesour
cesofi
ncomeamongpeasantf
armer
sinBar
ingoDi
str
ict
2.Todet
ermi
net
heeducat
ional
att
ainmentofar
ti
sani
nGi
kombaMar
ket
.
3.Tof
indoutt
hecausesofdomest
icv
iol
encei
nNj
uki
riSubl
ocat
ion,
EmbuDi
str
ict

I
nst
ati
ngbot
hthegener
alandt
hespeci
fi
cobj
ect
ivest
her
esear
chershoul
dbecar
eful
ont
hev
erbsused.Hencehe/
sheshoul
dnotusev
alue-
laden,bi
asedorsubj
ect
ive
wor
ds/
ver
bs;r
atheruseneut
ralv
erbsasshowni
nthef
oll
owi
ngt
abl
eadopt
edf
rom
Mugenda&Mugenda(
1999)
.

Tabl
e1:
Biased&neut
ral
wor
ds/
ver
bs
Bi
asedv
erbs/
wor
ds Neut
ral
wor
dsv
erbs
Toshow Todet
ermi
ne
Topr
ove Tocompar
e
Toconf
ir
m Toi
nvest
igat
e
Tov
eri
fy Todi
ff
erent
iat
e
Tocheck Toexpl
ore
Todemonst
rat
e Tof
indout
Toi
ndi
cat
e Toexami
ne
Tov
ali
dat
e Toi
nqui
re
Toexpl
ain Toest
abl
i
sh
Toi
l
lust
rat
e Tot
est
Sour
ce:
Mugenda&Mugenda,
1999:
21

F.Resear
chi
nst
rument
ati
on/
tool
sandmet
hods

Page12of12
Dat
aCol
l
ect
ionTechni
ques
I
ntr
oduct
ion
Dat
acol
l
ect
ioni
sacr
uci
alst
agei
nthepl
anni
ngandi
mpl
ement
ati
onofast
udy
.Ift
he
dat
acol
l
ect
ionhasbeensuper
fi
cial
,bi
asedori
ncompl
ete,dat
aanal
ysi
sbecomes
di
ff
icul
t,andt
hequal
i
tyoft
her
esear
chr
epor
twi
l
lbeofpoorqual
i
ty.Ther
efor
e,we
shoul
dconcent
rat
eal
lpossi
bleef
for
tsondev
elopi
ngappr
opr
iat
etool
s,andshoul
dtest
t
hem sev
eral
timesbef
oreunder
taki
ngt
heact
ual
study
.Dependi
ngont
het
ypeofst
udy
,
di
ff
erentdat
a-col
l
ect
iont
echni
quesmaybeused.I
nmanyst
udi
esweusual
l
ycombi
nea
numberofdi
ff
erentt
echni
ques and l
ookatpr
obl
ems f
rom di
ff
erentper
spect
ives
(
tri
angul
ati
on)
.Dat
a-col
l
ect
iont
echni
quesal
l
owust
osy
stemat
ical
l
ycol
l
ecti
nfor
mat
ion
aboutourobj
ect
sofst
udy(
peopl
e,obj
ect
s,phenomena)andaboutt
heset
ti
ngsi
n
whi
cht
heyoccur
.Int
hecol
l
ect
ionofdat
awehav
etobesy
stemat
ic.I
fdat
aar
e
col
l
ect
ed haphazar
dly
,itwi
l
lbe di
ff
icul
tto answerourr
esear
ch quest
ions i
na
concl
usi
veway

Sour
ces/
met
hodsofdat
acol
l
ect
ion
1.Usi
ngav
ail
abl
einf
ormat
ion/
document
aryr
evi
ew
I
tinv
olv
es t
he use of t
ext
s and document
s as sour
ce mat
eri
als-
gov
ernment
publ
i
cat
ions, newspaper
s, j
our
nal
s, census publ
i
cat
ions, f
il
ms, v
ideo, per
sonal
phot
ogr
aphsandot
herwr
it
ten,
visualandpi
ctor
ialsour
cesi
npaper
,el
ect
roni
c,orot
her
`
har
dcopy
’for
m.Usual
l
yther
eisal
argeamountofdat
athathasal
readybeencol
l
ect
ed
byot
her
s,al
thoughi
tmaynotnecessar
il
yhav
ebeenanal
yzedorpubl
i
shed.Locat
ing
t
hesesour
cesand r
etr
iev
ing t
hei
nfor
mat
ioni
sagood st
art
ing poi
nti
nanydat
a
col
l
ect
ionef
for
t.Forexampl
e,anal
ysi
soft
hei
nfor
mat
ionr
out
inel
ycol
l
ect
edbyheal
th
f
aci
l
iti
essuch asKeny
a Demogr
aphi
cand Heal
th Sur
vey
scan bev
eryusef
ulf
or
i
dent
if
yingpr
obl
emsi
ncer
tai
nint
erv
ent
ionsorar
eas.Anal
ysi
sofheal
thi
nfor
mat
ion
dat
a,censusdat
a,unpubl
i
shedr
epor
tsandpubl
i
cat
ionsi
nar
chi
vesandl
i
brar
iesori
n
of
fi
cesmaybeast
udyi
nit
sel
f.Usual
l
y,howev
er,
itf
ormspar
tofast
udyi
nwhi
chot
her
dat
acol
l
ect
iont
echni
quesar
eal
soused.

Page13of13
Adv
ant
ages
 Pr
ovi
desusef
uli
nsi
ght
s(hi
stor
ical
l
y)oft
hepastl
eadi
ngt
othepr
esent
 I
tisi
mpor
tantf
orhi
ghl
i
ght
ingwhathasbeendoneandpoi
ntst
owhatshoul
dbe
done-
poi
ntsatr
esear
chgaps.
 Si
nceandwhenev
err
ecor
dsar
eav
ail
abl
e,t
hei
tbecomeseasyt
ogeti
nfor
mat
ion
f
rom secondar
ysour
ces
 Document
aryr
evi
ewsar
ever
sat
il
e.Theycanbeconduct
edf
oral
mostanyt
opi
c
andcanpr
ovi
dei
nfor
mat
ionei
theratt
heov
erv
iewl
evel
ori
n-dept
h.
 Document
aryr
evi
ewsar
erel
ati
vel
yinexpensi
veandef
fi
cient
-al
argeamountof
dat
acanbecol
l
ect
edqui
ckl
yatmi
nimal
cost
.
 Noschedul
i
ngorcoor
dinat
ioni
sinv
olv
ed.Thecooper
ati
onofot
her
sisnot
r
equi
red.
 Theonl
yresour
cesneededar
eagoodl
i
brar
yoronl
i
nedat
abaseanda
compet
entr
evi
ewer
.
 Document
aryr
evi
ewscanbeanexcel
l
entf
ir
stst
epi
napr
ojectorst
udybecause
t
heypr
ovi
deaconcept
ual
framewor
kforf
urt
herpl
anni
ngandst
udy
.

Di
sadv
ant
ages
 Maynothav
e been col
l
ect
ed sci
ent
if
ical
l
yand hence maygi
vet
he wr
ong
i
mpr
essi
onoft
hef
act
s
 Maynotbeupt
odat
eandt
her
efor
emeani
ngal
otofgaps
 Somet
imesi
tmaynotbeav
ail
abl
e/accessi
ngi
tmaybedi
ff
icul
t
 An ef
fect
ivedocument
aryr
evi
ew r
equi
resahi
gh l
evelofski
l
lini
dent
if
ying
r
esour
ces,anal
yzi
ngt
hesour
cest
oident
if
yrel
evanti
nfor
mat
ion,andwr
it
inga
meani
ngf
ulsummar
y.
 Li
mit
edt
ocol
l
ect
ingi
nfor
mat
ionaboutwhathashappenedi
nthepast
,and
usual
l
ywi
thi
nor
gani
zat
ionsot
hert
hant
her
esear
cher
’sownwor
kpl
ace.
 Theycannotpr
ovi
dedat
aaboutcur
rentact
ual
behav
ior
-al
way
shi
stor
ical

2.KeyI
nfor
mant
s/i
ntensi
vei
nter
viewi
ng
Page14of14
Theuseofkeyi
nfor
mant
sisanot
heri
mpor
tantt
echni
quet
ogai
naccesst
oav
ail
abl
e
i
nfor
mat
ion.Thekeyi
nfor
manti
nter
viewr
efer
stoget
ti
ngi
nfor
mat
ionf
rom ani
ndi
vi
dual
whoi
sconsi
der
edt
obepar
ti
cul
arl
yknowl
edgeabl
eaboutt
het
opi
cofi
nter
est
.Thi
s
per
soni
scal
l
ed“
keyi
nfor
mant
”.Thesemi
‐st
ruct
uredi
nter
viewi
susual
l
yconduct
edi
na
f
acet
ofaceset
ti
ngwhi
chper
mit
sther
esear
chert
oseeknewi
nsi
ght
s,askquest
ions,
andassessphenomenai
ndi
ff
erentper
spect
ives.TheKI
Iisusedwhenwr
it
tenr
ecor
ds
orpubl
i
shed document
s ar
eli
mit
ed ordo notexi
st,when i
nfor
mat
ion di
ff
erent
per
spect
ivesi
sneeded,
andwhent
her
ear
ekeyi
nfor
mant
swhoar
eaccessi
bleandhav
e
i
n‐dept
hknowl
edgeaboutat
opi
c.Keyi
nfor
mant
scoul
dbeknowl
edgeabl
ecommuni
ty
l
eader
sorheal
thst
affatv
ari
ousl
evel
sandoneort
woWel
lInf
ormedmember
soft
he
t
argetgr
oup(
e.g.
,adol
escent
sont
hei
rsexual
behav
ior
).

Theycanbei
nvol
vedi
nvar
iousst
agesoft
her
esear
ch,f
rom t
hest
atementoft
he
pr
obl
em t
oanal
ysi
soft
hedat
aanddev
elopmentofr
ecommendat
ions.Ot
hersour
ces
ofav
ail
abl
edat
aar
enewspaper
sandpubl
i
shedcasehi
stor
ies,e.
g.,pat
ient
ssuf
fer
ing
f
rom ser
iousdi
seases,
ort
hei
rrel
ati
ves,
tel
l
ingt
hei
rexper
iencesandhowt
heycope.I
n
or
dert
oret
ri
evet
hedat
afr
om av
ail
abl
esour
ces,t
her
esear
cherwi
l
lhav
etodesi
gnan
i
nst
rumenti
n such a wayt
hathe/
sheonl
ycol
l
ect
sther
elev
anti
nfor
mat
ion.The
adv
ant
age ofusi
ng ex
ist
ing dat
aist
hatcol
l
ect
ion i
sinexpensi
ve.Howev
er,i
tis
somet
imesdi
ff
icul
ttogai
naccesst
other
ecor
dsorr
epor
tsr
equi
red,
andt
hedat
amay
notal
way
sbecompl
ete,
preci
seenough,
wel
lor
gani
zedorupt
odat
e.

Adv
ant
ages
 Theypr
ovi
demuchmor
edet
ail
edi
nfor
mat
iont
hanwhati
sav
ail
abl
ethr
ough
ot
herdat
acol
l
ect
ionmet
hods,
suchassur
vey
s.
 Theypr
ovi
deamor
erel
axedat
mospher
einwhi
cht
ocol
l
ecti
nfor
mat
ion-
peopl
e
mayf
eelmor
ecomf
ort
abl
ehav
ingaconv
ersat
ionwi
thy
ouaboutt
hei
rpr
ogr
am
asopposedt
ofi
l
li
ngoutasur
vey
.
 I
n-dept
hint
erv
iewspr
ovi
dev
aluabl
einf
ormat
ionf
orpr
ogr
ams,par
ti
cul
arl
ywhen
suppl
ement
ingot
hermet
hodsofdat
acol
l
ect
ion.

Page15of15
 Det
ail
edandr
ichdat
acanbegat
her
edi
nar
elat
ivel
yeasyandi
nexpensi
veway
-
l
esscost
ly
 Al
l
owsi
nter
viewert
oest
abl
i
shr
appor
twi
tht
her
espondentandcl
ari
fyquest
ions-
al
l
owspr
obi
ng
 Pr
ovi
des an oppor
tuni
tyt
o bui
l
d orst
rengt
hen r
elat
ionshi
ps wi
thi
mpor
tant
communi
tyi
nfor
mant
sandst
akehol
der
s
 Canr
aiseawar
eness,i
nter
est
,andent
husi
asm ar
oundani
ssueamongpot
ent
ial
r
espondent
sandi
nter
viewer
sal
i
ke
 Cancont
acti
nfor
mant
slat
ert
ocl
ari
fyi
ssuesasneededsi
ncet
heyar
enotv
ery
many
.
 Gener
all
yeasyt
ofi
ndpeopl
ewi
tht
henecessar
yski
l
lst
oconductkeyi
nfor
mant
i
nter
viewsbecausemostsoci
alsci
ent
ist
spossesst
hepr
ofessi
onalt
rai
ningand
exper
iencer
equi
red
 Canbeusedwi
thal
lagegr
oups,
incl
udi
ngel
der
lyandchi
l
dren

Di
sadv
ant
ages
 Pr
onet
obi
as:Becausepr
ogr
am st
affmi
ghtwantt
o“pr
ove”t
hatapr
ogr
am i
s
wor
king,t
hei
rint
erv
iewr
esponsesmi
ghtbebi
ased.Responsesf
rom communi
ty
member
sandpr
ogr
am par
ti
cipant
scoul
dal
sobebi
asedduet
othei
rst
akei
nthe
i
ssueathand.
 Canbet
ime-
int
ensi
ve:I
nter
viewscan beat
ime-
int
ensi
veev
aluat
ion act
ivi
ty
becauseoft
het
imei
ttakest
oconducti
nter
views,t
ranscr
ibet
hem,andanal
yze
t
her
esul
ts.
 I
nter
viewermustbeappr
opr
iat
elyt
rai
nedi
nint
erv
iewi
ngt
echni
ques:Topr
ovi
de
t
hemostdet
ail
edandr
ichdat
afr
om ani
nter
viewee,t
hei
nter
viewermustmake
t
hatper
soncomf
ort
abl
eandappeari
nter
est
edi
nwhatt
heyar
esay
ing.They
mustal
sobesur
etouseef
fect
ivei
nter
viewt
echni
ques,
suchasav
oidi
ngy
es/
no
and l
eadi
ng quest
ions,usi
ng appr
opr
iat
e bodyl
anguage,and keepi
ng t
hei
r
per
sonal
opi
nionsi
ncheck.
 Notgener
ali
zabl
e:When i
n-dept
hint
erv
iews ar
e conduct
ed,gener
ali
zat
ions
Page16of16
aboutt
her
esul
tsar
eusual
l
ynotabl
etobemadebecausesmal
lsampl
esar
e
chosenandr
andom sampl
i
ngmet
hodsar
enotused.
 Sel
ect
ingt
he“
ri
ght
”keyi
nfor
mant
smaybedi
ff
icul
tsot
heyr
epr
esentdi
ver
se
backgr
oundsandv
iewpoi
nts
 Maybechal
l
engi
ngt
oreachandschedul
eint
erv
iewswi
thbusyand/
orhar
d-t
o-
r
eachr
espondent
s
 Pr
ovi
deonl
yav
eryl
i
mit
edbasi
sforquant
if
icat
ion;t
heyar
erar
elyappr
opr
iat
e
when quant
it
ati
ve dat
a ar
e needed.Deci
sion maker
s of
ten expectpr
eci
se,
st
ati
sti
cal
dat
awi
thwhi
cht
odesi
gnapr
ojectormoni
tori
tspr
ogr
ess
 Fi
ndi
ngscanbebi
asedi
fthei
nfor
mant
sar
enotcar
eful
l
ysel
ect
ed.Especi
all
y
wheni
nter
viewer
sar
enotf
ami
l
iarwi
tht
hel
ocalcondi
ti
ons,t
heymaysel
ect
i
nfor
mant
sont
hebasi
soft
hei
rSEXSorf
luencyi
nani
nter
nat
ionall
anguage
r
athert
hanont
hei
rknowl
edgeoft
hel
ocal
sit
uat
ion.
 Whenonl
yaf
ew peopl
e(f
ewert
han15)ar
eint
erv
iewed,i
tmaybedi
ff
icul
tto
demonst
rat
ethe v
ali
dit
y oft
he f
indi
ngs.I
tis noteasy t
o pr
ovet
hatt
he
i
nter
viewees ar
e,i
nfact
,knowl
edgeabl
e and i
nfor
med and t
hatt
hey ar
e
r
epr
esent
ati
veof
fthei
rpeer
sint
hei
rinf
ormat
ionandr
ecommendat
ions.
 Ot
hercommuni
tymember
swhoar
enotbei
ngusedaskeyi
nfor
mant
smay
becomej
eal
ousandr
esentbei
ngl
eftout
.
 Mayhav
etobecombi
nedwi
thot
hermet
hods,
becauser
epr
esent
ati
venessoft
he
t
otal
communi
tyi
sdi
ff
icul
ttoachi
eve
 Mayov
erl
ookt
heper
spect
ivesofcommuni
tymember
swhoar
elessv
isi
ble
 Takest
imet
osel
ectgoodi
nfor
mant
sandbui
l
dtr
ust
 Ther
ecoul
dbel
anguagebar
ri
erbet
weent
hei
nter
vieweeandt
her
esear
cher

Sour
cesofi
nfor
mat
ion/
keyi
nfor
mant
s
I
n-dept
hint
erv
iewst
ypi
cal
l
yrel
yon mul
ti
plesour
cesofi
nfor
mat
ion t
o pr
ovi
deas
compl
eteapi
ctur
easpossi
ble.I
nfor
mat
ionsour
cescoul
dincl
ude:
 Pol
i
cyMaker
s,Pr
ogr
am Par
ti
cipant
s/Cl
i
ent
s,Pr
ojectSt
aff
,Communi
tyMember
s,
Pr
ofessi
onal
Staf
f
Page17of17
3.Obser
vat
ion
Whati
sobser
vat
ioni
nresear
ch?
Obser
vat
ion i
sat
echni
que t
hati
nvol
ves sy
stemat
ical
l
y sel
ect
ing,wat
chi
ng and
r
ecor
ding behav
iorand char
act
eri
sti
cs ofl
i
ving bei
ngs,obj
ect
s orphenomena.
Obser
vat
ionofhumanbehav
iori
samuch-
useddat
acol
l
ect
iont
echni
que.I
tcanbe
under
takeni
ndi
ff
erentway
s:
I
.Par
ti
cipantobser
vat
ion:Theobser
vert
akespar
tint
hesi
tuat
ionheorsheobser
ves.
Forexampl
e,adoct
orhospi
tal
i
zedwi
thabr
okenhi
p,whonow obser
veshospi
tal
pr
ocedur
es‘
fr
om wi
thi
n’t
hehospi
tal
.
I
I.Non-
par
ti
ci
pant obser
vat
ion:The obser
ver wat
ches t
he si
tuat
ion,openl
y or
conceal
ed,butdoesnotpar
ti
cipat
e. Thi
sismor
eusedbysecur
it
yint
ell
i
gence
of
fi
cer
sort
heCr
imi
nal
Inv
est
igat
ionDepar
tment
.

Obser
vat
ionscangi
veaddi
ti
onal
,mor
eaccur
atei
nfor
mat
iononbehav
iorofpeopl
ethan
i
nter
viewsorquest
ionnai
res.Theycanal
socheckont
hei
nfor
mat
ioncol
l
ect
edt
hrough
i
nter
viewsespeci
all
yonsensi
ti
vet
opi
cssuchasal
coholordr
uguse,orst
igmat
izi
ng
di
seases.Forexampl
e,whet
hercommuni
tymember
sshar
edr
inksorf
oodwi
thpat
ient
s
suf
fer
ingf
rom f
ear
eddi
seases(
lepr
osy
,TB,
AIDS)ar
eessent
ialobser
vat
ionsi
nast
udy
onst
igma.Obser
vat
ionsofhumanbehav
iorcanf
orm par
tofanyt
ypeofst
udy
,butas
t
heyar
eti
meconsumi
ngt
heyar
emostof
tenusedi
nsmal
l
-scal
est
udi
es.Obser
vat
ions
canal
sobemadeonobj
ect
s.Forexampl
e,t
hepr
esenceorabsenceofal
atr
ineandi
ts
st
ateofcl
eanl
i
nessmaybeobser
ved.Her
eobser
vat
ionwoul
dbet
hemaj
orr
esear
ch
t
echni
que.I
fobser
vat
ions ar
e made usi
ng a def
ined scal
ethey may be cal
l
ed
measur
ement
s.Measur
ement
s usual
l
yrequi
re addi
ti
onalt
ool
s.Forexampl
e,i
n
obser
vingl
engt
hofar
oadwemayuseki
l
omet
ersormi
l
es.I
naddi
ti
on,i
nnut
ri
ti
onal
sur
vei
l
lancewemeasur
ewei
ghtandhei
ghtbyusi
ngwei
ghi
ngscal
esandt
her
momet
ers
f
ormeasur
ingbodyt
emper
atur
e.Thus,obser
vat
iondoesnotj
ustmeanl
i
ter
all
yusi
ng
y
ourey
es.

Page18of18
Ty
pesofobser
vat
ion
Par
ti
cipantobser
vat
ion
I
tissomet
imescal
l
edaf
orm ofsubj
ect
ivesoci
ology
,notbecauset
her
esear
cherai
ms
t
oimposet
hei
rbel
i
efsont
her
espondent
,butbecauset
heai
mist
ounder
standt
he
soci
alwor
ldf
rom t
hesubj
ect
'spoi
nt-
of-
view.Thi
smet
hodi
nvol
vest
her
esear
cher
"
get
ti
ngt
oknow"t
hepeopl
ethey
'r
est
udy
ingbyent
eri
ngt
hei
rwor
ldandpar
ti
cipat
ing-
ei
theropenl
yandsecr
etl
yint
hatwor
ld.Thi
smeansy
ouputy
our
sel
fint
heshoesoft
he
peopl
eyou'
rest
udy
ingi
nanat
tem ptt
oexper
ienceev
ent
sint
hewayt
heyexper
ience
t
hem.

Soci
ologi
stswhousepar
ti
cipantobser
vat
ionai
mtodi
scov
ert
henat
ureofsoci
alr
eal
i
ty
byunder
standi
ngt
heact
or'
sper
cept
ion/under
standi
ng/i
nter
pret
ati
onoft
hesoci
al
wor
ld.Fort
hisr
eason,Par
ti
cipantObser
vat
ion i
s somet
imes cal
l
ed a nat
ural
i
sti
c
met
hodandt
endst
obeassoci
atedwi
tht
heI
nter
act
ioni
storSoci
alAct
ionper
spect
ive.
The met
hodol
ogy i
s,as y
ou woul
d expect
,pr
imar
il
yint
erpr
eti
ve.The par
ti
cipant
obser
ver
,ther
efor
e,expl
oit
sthehumanabi
l
ityt
oempat
hise-t
hemai
nobj
ect
ivebei
ng
t
opar
ti
cipat
eina soci
algr
oupwhi
l
e,att
hesamet
ime,empl
oyi
ngt
hei
nsi
ght
sand
under
standi
ng ofa soci
ologi
calobser
ver
.The poi
nt,t
her
efor
e,i
sto obser
ve and
exper
ience t
he wor
ld as a par
ti
cipant
,whi
l
eret
aini
ng an obser
ver
's ey
efor
under
standi
ng,
anal
ysi
sandexpl
anat
ion.

Non-
par
ti
cipantobser
vat
ion
Ar
esear
cht
echni
quewher
ebyt
her
esear
cherwat
chest
hesubj
ect
sofhi
sorherst
udy
,
wi
tht
hei
rknowl
edge,butwi
thoutt
aki
nganact
ivepar
tint
hesi
tuat
ionunderscr
uti
ny.
Thi
sappr
oachi
ssomet
imescr
it
ici
zedont
hegr
oundst
hatt
hev
eryf
actoft
hei
rbei
ng
obser
vedmayl
eadpeopl
etobehav
edi
ff
erent
ly,t
husi
nval
i
dat
ingt
hedat
aobt
ained,as
f
orexampl
eint
hef
amouscaseoft
heso-
cal
l
edHawt
hor
neef
fect
.Non-
par
ti
cipant
,or
di
rect
,obser
vat
ioni
swher
edat
aar
ecol
l
ect
edbyobser
vingbehav
iorwi
thouti
nter
act
ing
wi
tht
he par
ti
ci
pant
s.I
n non-
par
ti
cipantobser
vat
ion,t
he r
esear
cherdoes notget
i
nvol
vedi
nthedai
l
yact
ivi
ti
esoft
hest
udypopul
ati
on,butmai
ntai
nsasoci
aldi
stance.

Page19of19
Hemayormaynothowev
er,
makei
tknownt
othet
argetpopul
ati
ont
hathei
sst
udy
ing
t
hem.

Di
rect
/Obt
rusi
veObser
vat
ion
I
ndi
rectobser
vat
ions,
peopl
eknowt
haty
ouar
ewat
chi
ngt
hem.Theonl
ydangeri
sthat
t er
heyar eact
ingt
oyou.Asst
atedear
li
er,
ther
eisaconcer
nthati
ndi
vi
dual
swi
l
lchange
t
hei
ract
ionsr
athert
hanshowi
ngy
ouwhatt
hey
'r
eREALLYl
i
ke.Thi
sisnotnecessar
il
y
bad,howev
er.Forexampl
e,t
he cont
ri
ved behav
iormayr
evealaspect
s ofsoci
al
desi
rabi
l
ity
,howt
heyf
eelaboutshar
ingt
hei
rfeel
i
ngsi
nfr
ontofot
her
s,orpr
ivacyi
na
r
elat
ionshi
p.Ev
ent
hemostcont
ri
vedbehav
iori
sdi
ff
icul
ttomai
ntai
nov
ert
ime.Al
ong
t
erm obser
vat
ionalst
udywi
l
lof
tencat
chagl
i
mpseoft
henat
uralbehav
ior
.Ot
her
pr
obl
emsconcer
nthegener
ali
zabi
l
ityoff
indi
ngs.Thesampl
eofi
ndi
vi
dual
smaynotbe
r
epr
esent
ati
veoft
hepopul
ati
onort
hebehav
ior
sobser
vedar
enotr
epr
esent
ati
veoft
he
i
ndi
vi
dual(
youcaughtt
heper
sononabadday
).Agai
n,l
ong-
ter
m obser
vat
ionalst
udi
es
wi
l
lof
tenov
ercomet
hepr
obl
em ofext
ernalv
ali
dit
y.Whataboutet
hicalpr
obl
emsy
ou
say
?Et
hical
l
y,peopl
eseey
ou,t
heyknowy
ouar
ewat
chi
ngt
hem andt
heycanasky
ou
t
ost
op

I
ndi
rect
/Unobt
rusi
veObser
vat
ion
Unobt
rusi
vemeasur
esi
nvol
vesanymet
hodf
orst
udy
ingbehav
iourwher
eindi
vi
dual
sdo
NOT know t
heyar
ebei
ng obser
ved (
don'
tyou hat
etot
hinkt
hatt
hiscoul
d hav
e
happenedt
oyou!
).Her
e,t
her
eisnott
heconcer
nthatt
heobser
vermaychanget
he
subj
ect
'sbehav
iour
.Whenconduct
ingunobt
rusi
veobser
vat
ions,i
ssuesofv
ali
dit
yneed
t
obeconsi
der
ed.Numer
ousobser
vat
ionsofar
epr
esent
ati
vesampl
eneedt
otake
pl
acei
nor
dert
ogener
ali
zet
hef
indi
ngs.Thi
sisespeci
all
ydi
ff
icul
twhenl
ooki
ngata
par
ti
cul
argr
oup.Many gr
oups possess uni
que char
act
eri
sti
cs whi
ch make t
hem
i
nter
est
ingst
udi
es.Hence,of
tensuchf
indi
ngsar
enotst
rongi
next
ernalv
ali
dit
y.Al
so,
r
epl
i
cat
ion i
s di
ff
icul
t when usi
ng non-
conv
ent
ionalmeasur
es (
non-
conv
ent
ional

Page20of20
meani
ngunobt
rusi
veobser
vat
ion)
.Obser
vat
ionsofv
eryspeci
fi
cbehav
iour
sar
edi
ff
icul
t
t
orepl
i
cat
einst
udi
esespeci
all
yift
her
esear
cheri
sagr
ouppar
ti
cipant(
we'
l
ltal
kmor
e
aboutt
hisl
ater
).Themai
npr
obl
em wi
thunobt
rusi
vemeasur
es,howev
er,i
set
hical
.
I
ssuesi
nvol
vi
ngi
nfor
medconsentandi
nvasi
onofpr
ivacyar
epar
amounther
e.An
i
nst
it
uti
onalr
evi
ewboar
dmayf
rownupony
ourst
udyi
fiti
snotr
eal
l
ynecessar
yfory
ou
nott
oinf
ormy
oursubj
ect
s.

Adv
ant
agesofobser
vat
ion
 Excel
l
entappr
oacht
ogat
heri
n-dept
hat
ti
tudes,bel
i
efs,andanecdot
aldat
afr
om
al
argegr
oupofpat
ronsatonet
ime.
 Gr
oupdy
nami
csmi
ghtgener
atemor
eideast
hani
ndi
vi
dual
int
erv
iews.
 Canbeef
fect
ivel
yusedt
ofocusondet
ail
sregar
dingi
ssuesf
oundt
hrough
sur
vey
sorot
herdat
acol
l
ect
ionmet
hods.
 Par
ti
cipant
sar
enotr
equi
redt
oreadorwr
it
e.
 Techni
que r
eli
es on or
alcommuni
cat
ion r
esear
cher can hav
efi
rst
hand
exper
ienceoft
hei
rst
udy
.
 Theabi
l
ityt
orecor
dandr
epor
tal
lfi
ndi
ngst
hatar
etr
uet
othet
opi
cathand
 Suppl
ement
sot
herdat
acol
l
ect
ionmet
hods.
 Si
mpl
eandl
esscost
lyt
ouse.
 Ver
if
iesdat
afr
om ot
hersour
ces

Di
sadv
ant
ages
 Requi
resst
afft
imet
oobser
veandr
ecor
dobser
vat
ions.
 Cannotaskquest
ionsofpar
ti
cipant
sdur
ingobser
vat
ion.
 Mi
ghtwantt
ousef
oll
ow-
upi
nter
viewst
over
if
yobser
vat
ions
 Thev
iewer
'sorr
esear
cher
'sownper
cept
ionofwhatI
Swi
l
lgi
vebi
ast
otheend
r
ecor
dedr
esul
t.
 Ther
e maybe t
oo f
ew t
ri
als/
studi
es/
orobj
ect
s obser
ved t
o make an end
concl
usi
ont
othest
udy

Page21of21
 Ti
meconsumi
ng.
 Maybi
aswor
kerper
for
mance/
ther
ecoul
dbeal
otofwi
ndow-
dressi
ngont
he
par
toft
heobser
ved.
 Smal
lsampl
esi
ze.
 Requi
resski
l
ledobser
ver
.
 Val
i
dit
y&r
eli
abi
l
itymaybepr
obl
emat
ic.
 Notusef
ulf
orj
obsconsi
sti
ngofmost
lyment
alt
asks.

4.I
nter
viewi
ng
Ani
nter
viewi
sadat
a-col
l
ect
iont
echni
quet
hati
nvol
vesor
alquest
ioni
ngofr
espondent
s,
ei
theri
ndi
vi
dual
l
yorasagr
oup.Answer
stot
hequest
ionsposeddur
ingani
nter
view
canber
ecor
dedbywr
it
ingt
hem down(
eit
herdur
ingt
hei
nter
viewi
tsel
fori
mmedi
atel
y
af
tert
hei
nter
view)orbyt
ape-
recor
dingt
her
esponses,orbyacombi
nat
ionofbot
h.
I
nter
viewscanbeconduct
edwi
thv
ary
ingdegr
eesoff
lexi
bil
i
ty.Thet
woext
remes,hi
gh
andl
owdegr
eeoff
lexi
bil
i
ty,
aredescr
ibedbel
ow:

 Hi
ghdegr
eeoff
lexi
bil
i
ty:
Forexampl
e,whenst
udy
ingsensi
ti
vei
ssuessuchast
eenagepr
egnancyandabor
ti
ons,
t
hei
nvest
igat
ormayuseal
i
stoft
opi
csr
athert
hanf
ixedquest
ions.Thesemayi
ncl
ude
howt
eenager
sst
art
edsexuali
nter
cour
se,
ther
esponsi
bil
i
tygi
rl
sandt
hei
rpar
tner
stake
t
opr
eventpr
egnancy(
ifatal
l
),andt
heact
ionst
heyt
akei
ntheev
entofunwant
ed
pr
egnanci
es.Thei
nvest
igat
orshoul
dhav
eanaddi
ti
onall
i
stofquest
ionsr
eadywhen
t
her
espondentf
all
ssi
l
ent
.Forexampl
e,whenaskedaboutabor
ti
onmet
hodsused,
who
madet
hedeci
si
onandwhopai
d,t
hesequenceoft
opi
csshoul
dbedet
ermi
nedbyt
he
f
lowofdi
scussi
on.I
tisof
tenpossi
blet
ocomebackt
oaquest
iondi
scussedear
li
eri
na
l
aterst
ageoft
hei
nter
view.Af
lexi
blemet
hodofi
nter
viewi
ngi
susef
uli
far
esear
cher
hasl
i
ttl
eunder
standi
ngoft
hepr
obl
em orsi
tuat
ionhei
sinv
est
igat
ing,ori
fthet
opi
cis
sensi
ti
ve.I
tisf
requent
lyappl
i
edi
nexpl
orat
oryst
udi
es.Thei
nst
rumentusedmaybe
cal
l
edani
nter
viewgui
deori
nter
viewschedul
e.

Page22of22
 Lowdegr
eeoff
lexi
bil
i
ty:
Lessf
lex
ibl
emet
hodsofi
nter
viewi
ng ar
eusef
ulwhen t
her
esear
cheri
srel
ati
vel
y
knowl
edgeabl
eaboutexpect
edanswer
sorwhent
henumberofr
espondent
sbei
ng
i
nter
viewedi
srel
ati
vel
ylar
ge.Thenquest
ionnai
resmaybeusedwi
thaf
ixedl
i
stof
quest
ionsi
nast
andar
dsequence,whi
chhav
emai
nlycl
osed-
endedquest
ionswi
thpr
e-
cat
egor
izedanswer
s.Forexampl
e,af
teranumberofobser
vat
ionsont
he(
hygi
eni
c)
behav
iourofwomendr
awi
ngwat
eratawel
landsomekeyi
nfor
manti
nter
viewsont
he
useandmai
ntenanceoft
hewel
l
s,onemayconductal
argersur
veyonwat
eruseand
sat
isf
act
ionwi
tht
hequant
it
yandqual
i
tyoft
hewat
er.I
nreal
i
tyt
her
eisof
tenami
xtur
e
ofopenandpr
e-cat
egor
izedanswer
s

Adv
ant
agesofi
nter
viewi
ng
 Usef
ulf
ordet
ermi
ningi
ftheappl
i
canthasr
equi
sit
ecommuni
cat
iveorsoci
al
ski
l
lswhi
chmaybenecessar
yfort
hej
ob
 I
nter
viewercanobt
ainsuppl
ement
aryi
nfor
mat
ion
 Al
l
owst
heappl
i
cantt
oaskquest
ionst
hatmayr
evealaddi
ti
onali
nfor
mat
ion
usef
ulf
ormaki
ngasel
ect
iondeci
sion.
 Thei
nter
viewmaybemodi
fi
edasneededt
ogat
heri
mpor
tanti
nfor
mat
ion

Di
sadv
ant
agesofi
nter
viewi
ng
 Subj
ect
iveev
aluat
ionsar
emade
 Deci
sionst
endt
obemadewi
thi
nthef
ir
stf
ewmi
nut
esoft
hei
nter
viewwi
tht
he
r
emai
nderoft
hei
nter
viewusedt
oval
i
dat
eorj
ust
if
ytheor
igi
nal
deci
sion
 I
nter
viewer
sfor
m st
ereot
ypes concer
ning t
he char
act
eri
sti
cs r
equi
red f
or
successont
hej
ob
 Negat
ivei
nfor
mat
ionseemst
obegi
venmor
ewei
ght
 Notmuchev
idenceofv
ali
dit
yoft
hesel
ect
ionpr
ocedur
e

5.Admi
nist
eri
ngwr
it
tenquest
ionnai
res/
sur
veyr
esear
ch
Asur
veyi
sasy
stemat
icmet
hodofcol
l
ect
ingdat
afr
om apopul
ati
onofi
nter
est
.A

Page23of23
sy
stemat
icmet
hodf
orgat
her
ingi
nfor
mat
ionf
rom (
asampl
eof
)indi
vi
dual
sfort
he
pur
posesofdescr
ibi
ngt
heat
tri
but
esoft
hel
argerpopul
ati
onofwhi
cht
hei
ndi
vi
dual
s
ar
emember
s.Theat
tri
but
esat
temptt
odescr
ibebasi
cchar
act
eri
sti
csorexper
iencesof
l
argeandsmal
lpopul
ati
onsi
nourwor
ld.
Itt
endt
obequant
it
ati
vei
nnat
ureandai
mst
o
col
l
ecti
nfor
mat
ion f
rom a sampl
e oft
he popul
ati
on such t
hatt
he r
esul
ts ar
e
r
epr
esent
ati
veoft
hepopul
ati
onwi
thi
nacer
tai
ndegr
eeofer
ror
.Thepur
poseofa
sur
veyi
stocol
l
ectquant
it
ati
vei
nfor
mat
ion,
usual
l
ythr
ought
heuseofast
ruct
uredand
st
andar
dizedquest
ionnai
re

Feat
uresofasur
vey
 I
nfor
mat
ioni
sgat
her
edbyaski
ngpeopl
equest
ions.
 I
nfor
mat
ioni
scol
l
ect
edei
therbyhav
ingi
nter
viewer
saskquest
ionsandr
ecor
d
answer
sorbyhav
ingpeopl
ereadorhearquest
ionsandr
ecor
dthei
ranswer
s.
 I
nfor
mat
ioni
scol
l
ect
edf
rom onl
yasubsetoft
hepopul
ati
ont
obedescr
ibed(
a
sampl
e)r
athert
hanf
rom al
lmember
s.

Adv
ant
ages
 Cancompl
etest
ruct
uredquest
ionswi
thmanyst
akehol
der
swi
thi
nar
elat
ivel
y
shor
tti
mef
rame.
 Canbecompl
etedbyt
elephone,
mai
l
,fax,
ori
n-per
son.
 I
tisquant
if
iabl
eandgener
ali
zabl
etoanent
ir
epopul
ati
oni
fthepopul
ati
oni
s
sampl
edappr
opr
iat
ely
.
 St
andar
dized,
str
uct
uredquest
ionnai
remi
nimi
zesi
nter
viewerbi
as.
 Tr
emendousv
olumeofi
nfor
mat
ioncanbecol
l
ect
edi
nshor
tper
iodoft
ime.
 Cant
akel
esst
imet
oanal
yset
hanqual
i
tat
ivedat
a.
 I
tis an ef
fi
ci
entway ofcol
l
ect
ing i
nfor
mat
ion f
rom a l
arge numberof
r
espondent
s.
 Ver
ylar
gesampl
esar
epossi
ble.
 St
ati
sti
calt
echni
quescanbeusedt
odet
ermi
nev
ali
dit
y,r
eli
abi
l
ity
,andst
ati
sti
cal
si
gni
fi
cance.
Page24of24
 Sur
vey
sar
efl
exi
blei
nthesenset
hata wi
der
angeofi
nfor
mat
ion can be
col
l
ect
ed-
canbeusedt
ost
udyat
ti
tudes,
val
ues,
bel
i
efs,
andpastbehav
ior
s.
 Becauset
heyar
est
andar
dized,t
heyar
erel
ati
vel
yfr
eef
rom sev
eralt
ypesof
er
ror
s.
 Theyar
erel
ati
vel
yeasyt
oadmi
nist
er.
 Ther
eisaneconomyi
ndat
acol
l
ect
ionduet
othef
ocuspr
ovi
dedbyst
andar
dized
quest
ions.Onl
yquest
ionsofi
nter
estt
other
esear
cherar
easked,r
ecor
ded,
codi
fi
ed,
andanal
yzed.Ti
meandmoneyi
snotspentont
angent
ial
quest
ions

Di
sadv
ant
ages
 Dependonsubj
ect
s’mot
ivat
ion,honest
y,memor
y,andabi
l
ityt
orespond.
Subj
ect
smaynotbeawar
eoft
hei
rreasonsf
oranygi
venact
ion.Theymay
hav
efor
got
tent
hei
rreasons.Theymaynotbemot
ivat
edt
ogi
veaccur
ate
answer
s;i
nfact
,theymaybe mot
ivat
ed t
o gi
ve answer
sthatpr
esent
t
hemsel
vesi
naf
avor
abl
eli
ght
.
 St
ruct
uredsur
vey
s,par
ti
cul
arl
ythosewi
thcl
osedendedquest
ionsmayhav
e
l
owv
ali
dit
ywhenr
esear
chi
ngaf
fect
ivev
ari
abl
es.
 Al
thought
hechosensur
veyi
ndi
vi
dual
sar
eof
tenar
andom sampl
e,er
ror
sdue
t
onon-
responsemayexi
st.Thati
s,peopl
ewhochooset
orespondont
he
sur
veymaybedi
ff
erentf
rom t
hosewhodonotr
espond,t
husbi
asi
ngt
he
est
imat
es.
 Sur
veyquest
ionanswer
-choi
cescoul
dleadt
ovaguedat
aset
sbecauseat
t
imest
heyar
erel
ati
veonl
ytoaper
sonalabst
ractnot
ionconcer
ning"
str
engt
h
ofchoi
ce"
.Fori
nst
ancet
hechoi
ce"
moder
atel
yagr
ee"maymeandi
ff
erent
t
hingst
odi
ff
erentsubj
ect
s,andt
oany
onei
nter
pret
ingt
hedat
aforcor
rel
ati
on.
 Ev
eny
esornoanswer
sar
epr
obl
emat
icbecausesubj
ect
smayf
ori
nst
ance
put"
no"i
fthechoi
ce"
onl
yonce"i
snotav
ail
abl
e
 Mor
edi
ff
icul
tto col
l
ecta compr
ehensi
veunder
standi
ng ofr
espondent
s’
per
spect
ive(
in-
dept
hinf
ormat
ion)compar
edt
oin-
dept
hint
erv
iewsorf
ocus
gr
oups.

Page25of25
 Canbev
eryexpensi
ve.
 Requi
ressomest
ati
sti
calknowl
edge,
sampl
i
ngandot
herspeci
ali
zedski
l
lst
o
pr
ocessandi
nter
pretr
esul
ts.

St
epsi
ncar
ryi
ngoutasur
vey
1.Cl
ari
fypur
pose
 WhyconductaSur
vey
?
 Whoar
ethest
akehol
der
s?
 Whoi
sthepopul
ati
onofi
nter
est
?
 Whati
ssuesneedt
obeexpl
ored?
2.AssessResour
ces
 Whatext
ernal
resour
ceswi
l
lyouneed?
 Whi
chi
n-houser
esour
cescany
oumakeuseof
?
3.Deci
deonMet
hods
 Sel
ectt
hemet
hodwhi
chi
smostappr
opr
iat
e
4Wr
it
eQuest
ionnai
res
 Deci
deonwhatquest
ionst
oask
 Sett
het
ypesofr
esponsef
ormat
s
 Sett
hel
ayoutoft
hequest
ionnai
re
5Pi
l
ott
est
/Rev
isequest
ionnai
re
 Pi
l
ott
estt
hequest
ionnai
re
 Rev
iset
hequest
ionnai
re
6Pr
epar
eSampl
e
 Deci
deont
hesampl
edesi
gn
 I
dent
if
ysour
cesofsampl
e

Page26of26
7Tr
aini
nter
viewer
s
8Col
l
ectdat
a
9Pr
ocessdat
a
 Codet
hedat
a
 Dat
aent
ert
hei
nfor
mat
ion
10Anal
yset
heResul
ts
11I
nter
pretandDi
ssemi
nat
eResul
ts
12TakeAct
ionont
her
esul
ts/
recommendat
ions

UseofSur
vey
s
Sur
vey
sar
epr
imar
il
yusedt
ocol
l
ectquant
it
ati
vei
nfor
mat
ionont
heper
cept
ionsand
opi
nionsofasampl
eofpeopl
ewhi
chadequat
elyr
epr
esentt
hepopul
ati
onofi
nter
est
.
Theyar
eappr
opr
iat
efor
:
 Measur
ing peopl
e’
s per
cept
ions,opi
nions,knowl
edge,at
ti
tudes,behav
iour
al
i
ntent
ions,
andbehav
iourusi
ngpr
imar
il
ycl
osed-
endedquest
ions.
 Conduct
inganeedsassessment
.
Sur
v sshoul
ey dnotbeusedt
o:
 Repl
acequal
i
tat
ivet
echni
ques
 Expl
oreopi
nionsi
n-dept
h

Modesofadmi
nist
eri
ngsur
vey
s
Awr
it
tenquest
ionnai
re(
alsor
efer
redt
oassel
f-
admi
nist
eredquest
ionnai
re)i
sadat
a
col
l
ect
iont
ooli
nwhi
chwr
it
tenquest
ionsar
epr
esent
edt
hatar
etobeanswer
edbyt
he
r
espondent
sinwr
it
tenf
orm.Awr
it
tenquest
ionnai
recanbeadmi
nist
eredi
ndi
ff
erent
way
s,suchasby
:
 Sendi
ng quest
ionnai
res bymai
lwi
th cl
eari
nst
ruct
ions on how t
o answert
he
quest
ionsandaski
ngf
ormai
l
edr
esponses;
 Gat
her
ingal
lorpar
toft
her
espondent
sinonepl
aceatonet
ime,gi
vi
ngor
alor
wr
it
teni
nst
ruct
ions,
andl
ett
ingt
her
espondent
sfi
l
loutt
hequest
ionnai
res;
or
 Hand-
del
i
ver
ingquest
ionnai
rest
orespondent
sandcol
l
ect
ingt
hem l
ater
.

Page27of27
 Aski
ngt
hequest
ionsonphone

Thequest
ionscanbeei
theropen-
endedorcl
osed(
wit
hpr
e-cat
egor
izedanswer
s).
Whent
hequest
ionnai
rei
sadmi
nist
eredbyt
her
esear
chert
other
espondentonaf
ace
t
ofacebasi
s,i
tiscal
l
ed ani
nter
view schedul
e,wher
easwheni
tisgi
vent
othe
r
espondentt
ofi
l
lal
onei
tiscal
l
edaquest
ionnai
re.

6.Focusgr
oupdi
scussi
ons(
FGD)
I
tisaspeci
alt
ypeofgr
oupi
nter
msofpur
pose,si
ze,composi
ti
on,andpr
ocedur
es.A
f
ocusgr
oupi
sty
pical
l
ycomposedofsev
ent
otwel
vepar
ti
cipant
swhoar
esel
ect
ed
becauset
heyhav
ecer
tai
nchar
act
eri
sti
csi
ncommont
hatr
elat
etot
het
opi
coft
he
f
ocusgr
oup.Af
ocusgr
oupdi
scussi
onal
l
owst
hegr
oupmember
stof
reel
ydi
scussa
cer
tai
nsubj
ectwi
tht
hegui
danceofaf
aci
l
itat
ororr
epor
ter
.Thegr
oupmember
sar
e
nor
mal
l
ypeopl
ewhoar
efami
l
iari
nthear
eat
heyar
erequi
redt
odi
scuss.Forexampl
e,
whendoi
ngast
udyony
out
hrepr
oduct
iveheal
th,onemaybr
ingt
oget
heragr
oupof
y
out
horr
epr
oduct
iveheal
thser
vicepr
ovi
der
stodi
scusst
hepar
ti
cul
ari
ssueunder
st
udy
.Iti
simpor
tantt
hatf
ocusgr
oupdi
scussant
sar
eatorr
elat
ivel
ythesamel
eveli
n
t
ermsofage,qual
i
ficat
ion/
occupat
ionand beknowl
edgeabl
eint
hear
eat
heyar
e
di
scussi
ng.I
tisal
soi
mpor
tantt
hatast
heydi
scusst
heyshoul
dsi
tinasemi
-ci
rcl
e(t
he
UN Sy
stem)
.Thi
sisi
ntended t
o br
ing a f
eel
i
ng ofequal
i
tyamong t
hem so t
hat
di
scussi
onsar
eseent
obedonebypeopl
ewhoar
eequal
.Gr
oupmember
sinf
luence
eachot
herbyr
espondi
ngt
oideas,
opi
nionandcomment
sint
hedi
scussi
on.

Adv
ant
ages
 Excel
l
entappr
oacht
ogat
heri
n-dept
hat
ti
tudes,bel
i
efsandot
herdat
afr
om a
l
argegr
oupsonr
espondent
satonet
ime
 Gr
oupdy
nami
csmi
ghtgener
atemor
eideast
hani
ndi
vi
dual
int
erv
iews.
 Canbeef
fect
ivel
yusedt
ofocusondet
ail
sregar
dingi
ssuesf
oundt
hrough
sur
vey
sorot
herdat
acol
l
ect
ionmet
hods.
 Par
ti
cipant
sar
enotr
equi
redt
oreadorwr
it
e.

Page28of28
 Techni
quer
eli
esonor
alcommuni
cat
ion

Di
sadv
ant
ages
 Requi
resst
afft
imet
osetupandf
aci
l
itat
efocusgr
oup.
 Requi
resst
afft
imet
oident
if
yandschedul
epar
ti
cipant
sforf
ocusgr
oup.
 Requi
resst
rongf
aci
l
itat
ort
ogui
dedi
scussi
onandensur
epar
ti
cipat
ionbyal
l
member
s.
 Usual
l
yrequi
res speci
alequi
pmentt
orecor
d and t
ranscr
ibe f
ocus gr
oup
di
scussi
on
G.Hi
stor
ical
Resear
ch
Whati
shi
stor
icalr
esear
ch?
Thesy
stemat
iccol
l
ect
ionandev
aluat
ionofdat
arel
atedt
opastoccur
rencesi
nor
dert
o
descr
ibecauses,ef
fect
s,andt
rendsoft
hoseev
ent
sthatmayhel
pexpl
ainpr
esent
ev
ent
s and ant
ici
pat
efut
ure ev
ent
s.Dat
ais of
ten ar
chi
val
-i
ncl
udi
ng newspaper
cl
i
ppi
ngs,
phot
ogr
aphs,
etc.
-andmayi
ncl
udei
nter
views

Adv
ant
ages
1.Someadv
ant
agesofor
alt
radi
ti
oni
sthatt
hel
i
ter
atur
est
ayswi
thi
nyou.
2.Youdon'
thav
etogol
ooksomet
hingupi
fyouhav
eaquest
ionabouty
ourbel
i
efs

Di
sadv
ant
ages
1.Duet
olongt
imedat
acol
l
ect
ion,t
het
heor
yorst
atementdoesnotappl
yincur
rent
condi
ti
ons
2.Thel
i
ter
atur
emaygetr
evi
sedf
rom per
sont
oper
son.t
heaccount
sar
enotas
accur
ateast
hosewr
it
tendown

Combi
ningDat
aCol
l
ect
ionTechni
ques
Thecombi
nat
ionanduseofdi
ff
erentmet
hodsofdat
acol
l
ect
ioni
scal
l
edt
ri
angul
ati
on.
Tr
iangul
ati
on al
so i
mpl
i
esget
ti
ng i
nfor
mat
ion f
rom di
ff
erentsour
cesand usi
ng a
mul
ti
disci
pli
nar
yteam ofr
esear
cher
s.I
tiscl
eart
hatdi
ff
erentdat
acol
l
ect
iont
echni
ques

Page29of29
cancompl
ementeachot
her
.Aski
l
ful
useofdi
ff
erentt
echni
quescanr
educet
hechance
ofbi
asand gi
ve a mor
e compr
ehensi
ve under
standi
ng oft
he t
opi
cunderst
udy
.
Resear
cher
sof
tenuseacombi
nat
ionofbot
hqual
i
tat
iveandquant
it
ati
vemet
hodsof
dat
acol
l
ect
ionsuchasl
oosel
yst
ruct
uredi
nter
viewsusi
ngopen-
endedquest
ions,
focus
gr
oupdi
scussi
onsandpar
ti
cipantobser
vat
ion.Thef
oregoi
ngar
ecal
l
edqual
i
tat
ive
r
esear
cht
echni
quesandpr
oducequal
i
tat
ivedat
athati
sof
tenr
ecor
dedi
nnar
rat
ive
f
orm.Qual
i
tat
iveResear
chTechni
quesi
nvol
vet
hei
dent
if
icat
ionandexpl
orat
ionofa
numberofof
tenmut
ual
l
yrel
atedv
ari
abl
est
hatgi
vei
nsi
ghti
nhumanbehav
iorsuchas
mot
ivat
ions,opi
nionsandat
ti
tudes;t
henat
ureandcausesofcer
tai
npr
obl
emsandt
he
consequencesoft
hepr
obl
emsont
hoset
heyaf
fect
.Int
hiscase,
‘Why
’,‘
What
’and‘
How’
ar
eimpor
tantquest
ions.

Ont
heot
herhand,
str
uct
uredquest
ionnai
rest
hatenabl
ether
esear
chert
oquant
if
ypr
e-
orpost
-cat
egor
izedanswer
stoquest
ionsar
eanexampl
eofquant
it
ati
ver
esear
ch
t
echni
ques.Theanswer
stoquest
ionscanbecount
edandexpr
essednumer
ical
l
y.
Quant
it
ati
ver
esear
ch t
echni
ques ar
e used t
o quant
if
ythe si
ze,di
str
ibut
ion and
associ
ati
onofcer
tai
nvar
iabl
esi
nast
udypopul
ati
on.‘
How many
?’‘
How of
ten?
’and

How si
gni
fi
cant
?’ar
eimpor
tantquest
ions.Bot
hqual
i
tat
iveandquant
it
ati
ver
esear
ch
t
echni
quesar
eof
tenusedwi
thi
nasi
ngl
est
udy
.Forexampl
e,I
thasbeenobser
vedi
n
count
ryXt
hatchi
l
drenbet
ween1and2-
1/2y
ear
s,whohav
eal
readyst
art
edt
oeat
i
ndependent
ly,
hav
eunsat
isf
act
oryf
oodi
ntakeoncet
heyf
all
il
l.
i
. I
ntheabov
est
udy
,thef
ir
stqual
i
tat
ivepar
toft
hest
udywoul
dbeusedt
ofocus
t
hesur
veyont
hemostr
elev
anti
ssues(
mot
her
s’f
eedi
ngbehav
ior
sandr
easons
f
ort
hesebehav
ior
s)andt
ohel
pphr
aset
hequest
ionsi
nanopt
imalwayi
nor
der
t
oobt
aint
hei
nfor
mat
iont
hati
sneeded.
i
i
. Thesecond,quant
it
ati
vepar
toft
hest
udywoul
dbeusedt
ofi
ndoutwhat
pr
opor
ti
onoft
hemot
her
sfol
l
ow v
ari
ouspr
act
icesandt
her
easonsf
ort
hei
r
behav
ior
sandwhet
hercer
tai
ncat
egor
iesofchi
l
drenar
emor
eatr
iskt
hanot
her
s.
i
i
i. Thet
hir
d,qual
i
tat
ivepar
toft
hest
udywoul
dpr
ovi
def
eedbackont
hemaj
or
f
indi
ngsoft
hesur
vey
.Dot
heconcl
usi
onsmakesenset
owomeni
nthest
udy

Page30of30
ar
ea?Hav
ecer
tai
naspect
sbeenov
erl
ookedwheni
nter
pret
ingt
hedat
a?What
r
emedi
alact
ioni
sfeasi
blet
oimpr
ovepr
act
icesr
elat
edt
ofeedi
ngsi
ckchi
l
dren?

I
tisal
socommont
ocol
l
ectqual
i
tat
iveandquant
it
ati
vedat
ainasi
ngl
equest
ionnai
re.
Resear
cher
scol
l
ect
ingbot
hty
pesofdat
ahav
etomakesur
etheyhav
ebot
hcl
osed-
endedandopen-
endedquest
ions.
Bi
asi
nInf
ormat
ionCol
l
ect
ion
BI
ASi
ninf
ormat
ioncol
l
ect
ioni
sadi
stor
ti
oni
nthecol
l
ect
eddat
asot
hati
tdoesnot
r
epr
esentr
eal
i
ty.Possi
blesour
cesofbi
asdur
ingdat
acol
l
ect
ion:
i
. Def
ect
ivei
nst
rument
s,suchasquest
ionnai
reswi
th:
 Fi
xedorcl
osedquest
ionsont
opi
csaboutwhi
chl
i
ttl
eisknown(
oft
en
aski
ngt
he‘
wrongt
hings’
 Open-
endedquest
ionswi
thoutgui
del
i
nesonhow t
oask(
ort
oanswer
)
t
hem
 Vaguel
yphr
asedquest
ions
 Leadi
ng quest
ions’t
hatcause t
he r
espondentt
o bel
i
eve one answer
woul
dbepr
efer
redov
eranot
her
;
 Quest
ionspl
acedi
nani
l
logi
cal
order
.
i
i
. Wei
ghi
ngscal
esorot
hermeasur
ingequi
pmentt
hatar
enotst
andar
dizedor
def
ect
ive.Theset
wosour
cesofbi
ascanbepr
event
edbycar
eful
l
ypl
anni
ngt
he
dat
acol
l
ect
ionpr
ocessandbypr
e-t
est
ingt
hedat
acol
l
ect
iont
ool
s.
i
i
i. Obser
verbi
as:
Caneasi
l
yoccurwhenconduct
ingobser
vat
ionsorut
il
izi
ngl
oosel
y
st
ruct
uredgr
oupori
ndi
vi
duali
nter
views.Ther
eisar
iskt
hatt
hedat
acol
l
ect
or
wi
l
lonl
ysee orheart
hings i
n whi
ch he orshe i
sint
erest
ed orwi
l
lmi
ss
i
nfor
mat
iont
hati
scr
it
icalt
other
esear
ch.Obser
vat
ionpr
otocol
sandgui
del
i
nes
f
orconduct
ingl
oosel
yst
ruct
uredi
nter
viewsshoul
dbepr
epar
ed,
andt
rai
ningand
pr
act
ice shoul
d be pr
ovi
ded t
o dat
a col
l
ect
orsi
n usi
ng bot
hthese t
ool
s.
Mor
eov
eri
tishi
ghl
yrecommendedt
hatdat
acol
l
ect
orswor
kinpai
rswhenusi
ng
f
lexi
bler
esear
cht
echni
quesanddi
scussandi
nter
prett
hedat
aimmedi
atel
yaf
ter
col
l
ect
ingi
t.Anot
herpossi
bil
i
ty-commonl
yusedbyant
hropol
ogi
sts-i
susi
nga

Page31of31
t
aper
ecor
derandt
ranscr
ibi
ngt
het
apewor
dbywor
d.
i
v. Ef
fectoft
hei
nter
viewont
hei
nfor
mant
:Thi
sisapossi
blef
act
ori
nal
lint
erv
iew
si
tuat
ions.Thei
nfor
mantmaymi
str
ustt
hei
ntent
ionoft
hei
nter
viewanddodge
cer
tai
n quest
ionsorgi
vemi
sleadi
ng answer
s.Forexampl
e:i
n a sur
veyon
al
cohol
i
sm y
ouaskschoolchi
l
dren:‘
Doesy
ourf
athersomet
imesgetdr
unk?

Manywi
l
lpr
obabl
ydenyt
hathedoes,
eveni
fiti
str
ue.Suchbi
ascanber
educed
byadequat
elyi
ntr
oduci
ngt
hepur
poseoft
hest
udyt
oinf
ormant
s,byphr
asi
ng
quest
ionsonsensi
ti
vei
ssuesi
naposi
ti
veway
,byt
aki
ngsuf
fi
cientt
imef
ort
he
i
nter
view,
andbyassur
ingi
nfor
mant
sthatt
hedat
acol
l
ect
edwi
l
lbeconf
ident
ial
.
I
nter
viewer
swhohav
eani
nter
esti
nthest
udysuchasheal
thwor
ker
sst
udy
ing
t
heut
il
izat
ionofheal
thser
vicesmayal
soi
nfl
uencet
heout
come.
v
. I
nfor
mat
ionbi
as:Somet
imest
hei
nfor
mat
ioni
tsel
fhasweaknesses.Medi
cal
r
ecor
dsmayhav
emanybl
anksorbeunr
eadabl
e.Thi
stel
l
ssomet
hingaboutt
he
qual
i
tyoft
hedat
aandhast
ober
ecor
ded.Forexampl
e,i
naTBdef
aul
terst
udy
t
heper
cent
ageofdef
aul
ter
swi
thani
ncompl
eteormi
ssi
ngaddr
essshoul
dbe
cal
cul
ated.Anot
hercommoni
nfor
mat
ionbi
asi
sduet
ogapsi
npeopl
e’
smemor
y;
t
hisi
scal
l
edmemor
yorr
ecal
lbi
as.Amot
hermaynotr
ememberal
ldet
ail
sofher
chi
l
d’sv
acci
nat
iondat
eort
ypeofv
acci
nat
ionort
het
reat
mentshegav
etwoor
t
hreemont
hsbef
ore.Forsuchcommondi
seasesi
tisadv
isabl
etol
i
mitt
he
per
iodofr
ecal
l
,aski
ng,
forexampl
e,‘
Hasy
ourchi
l
dhadanyv
acci
nat
ionov
ert
he
pastt
woweeks?

Not
e:Al
lthesepot
ent
ial
biaseswi
l
lthr
eat
ent
hev
ali
dit
yandr
eli
abi
l
ityofy
ourf
indi
ngs.

3.Resear
chDesi
gn
Whati
sresear
chdesi
gn?
SeePower
Poi
ntpr
esent
ati
onsepar
atel
y
Ty
pesofr
esear
chdesi
gns
SeePower
Poi
ntpr
esent
ati
onsepar
atel
y
4.Measur
ementi
nSoci
alResear
ch
Cover
edi
nBAS204:Soci
alSt
ati
sti
cs

Page32of32
I
dent
if
ythe st
rengt
hs and weaknesses of any t
wo dat
a col
l
ect
ion
met
hodol
ogi
esy
ouhav
est
udi
edi
nthi
scour
se(
20mks

5.Sampl
i
ngandSampl
i
ngPr
ocedur
es
i
.Whati
ssampl
i
ng?
Sampl
i
ngi
sthepr
ocessofsel
ect
inganumberofst
udyuni
tsf
rom adef
inedst
udy
popul
ati
on.Somest
udi
esi
nvol
veonl
ysmal
lnumber
sofpeopl
eandt
husal
loft
hem can
bei
ncl
uded.Howev
er,of
tenr
esear
chf
ocusesonsuchal
argepopul
ati
ont
hati
tisnot
possi
blet
oincl
udeal
lthemember
soft
hepopul
ati
oni
nthei
nvest
igat
ion.Wet
henhav
e
t
odr
aw asampl
efr
om t
het
otalpopul
ati
on.I
nsuchcaseswemustconsi
dert
he
f
oll
owi
ngquest
ions:
1.Whati
sthegr
oupofpeopl
e(STUDYPOPULATI
ON)wear
eint
erest
edi
nfr
om
whi
chwewantt
odr
awasampl
e?
2.Howmanypeopl
edoweneedi
noursampl
e?
3.Howwi
l
lthesepeopl
ebesel
ect
ed?

Thest
udypopul
ati
onhast
obecl
ear
lydef
ined,f
orexampl
e,accor
dingt
oage,sex,and
r
esi
dence.Ot
her
wise we cannotdo t
he sampl
i
ng.Apar
tfr
om per
sons,a st
udy
popul
ati
onmayconsi
stofv
il
lages,i
nst
it
uti
ons,r
ecor
ds,et
c.Eachst
udypopul
ati
on
consi
stsofst
udyuni
ts.Thewaywedef
ineourst
udypopul
ati
onandourst
udyuni
t
dependsont
hepr
obl
em wewantt
oinv
est
igat
eandont
heobj
ect
ivesoft
hest
udy

i
i
.Whysampl
e?
1.Lesscost
ly-
ascompar
edt
ost
udy
ingt
hewhol
epopul
ati
onsi
ncet
hecostofr
esear
ch
i
sdi
rect
lypr
opor
ti
onalt
othenumberofhour
sspentondat
acol
l
ect
ion.Apopul
ati
on
maybespr
eadi
nal
argear
ear
equi
ri
ngal
otoft
rav
elandr
elat
edexpenses.Such

Page33of33
expensesar
eli
kel
ytober
educedbyr
educi
ngt
henumberofr
espondent
stobest
udi
ed.
2.Less t
ime consumi
ng-
itr
equi
res a l
otoft
ime t
oint
erv
iew f
orexampl
e 2000
r
espondent
sascompar
edt
oasi
tuat
ionwher
ether
esear
cherhast
oint
erv
iew onl
ya
sampl
eofl
etsay200,
whi
chi
s10%oft
hewhol
epopul
ati
on.
3.I
ncaseofi
nfi
nit
epopul
ati
on-
sampl
i
ngi
stheonl
ypr
act
icalwayofcol
l
ect
ingdat
a
when t
hepopul
ati
on i
sinf
ini
teorext
remel
ylar
ge,t
husmaki
ng ast
udyofal
lit
s
el
ement
simpossi
ble.
4.I
ncasesampl
i
ngnecessi
tat
esdest
ruct
ionofobj
ect
s-f
orexampl
e,whent
est
ingt
he
br
eakabi
l
ityofwat
ergl
asses,onemayber
equi
redt
odest
royal
lthegl
assesi
fno
sampl
eissel
ect
ed.I
twoul
dther
efor
ebeadv
isabl
eto br
eakj
usta f
ew because
dest
ruct
ionofal
lthem woul
dmeanwast
ageofr
esour
cesandespeci
all
ywhent
heyar
e
madeofexact
lyt
hesamemat
eri
als.

i
i
i.Repr
esent
ati
venessi
nsampl
i
ng
I
fresear
cher
swantt
odr
awconcl
usi
onswhi
char
eval
i
dfort
hewhol
est
udypopul
ati
on,
t
heyshoul
dtakecar
etodr
aw asampl
einsuchawayt
hati
tisr
epr
esent
ati
veoft
hat
popul
ati
on.A r
epr
esent
ati
ve sampl
e has al
lthe i
mpor
tantchar
act
eri
sti
cs oft
he
popul
ati
onf
rom whi
chi
tisdr
awn.Toensur
ear
epr
esent
ati
vesampl
e,oneshoul
dusea
compl
eteandcor
rectsampl
i
ngf
rame.Thi
sist
hecompl
etel
i
stofal
luni
tsf
rom whi
ch
t
hesampl
eist
obedr
awn.Forexampl
e,t
ouseat
elephonedi
rect
oryf
orsel
ect
ing
sampl
eiswr
ongbecausei
twoul
dexcl
udeal
lpeopl
ewi
thoutt
elephonel
i
nes.Assuch,
anadequat
esampl
i
ngf
rameshoul
dnotexcl
udeanyel
ementoft
hepopul
ati
onunder
i
nvest
igat
ion.

Forexampl
e,i
fyoui
ntendt
oint
erv
iew 100mot
her
sinor
dert
oobt
ainacompl
ete
pi
ctur
eoft
heweani
ngpr
act
icesi
nNar
okDi
str
icty
ouwoul
dhav
etosel
ectt
hese
mot
her
sfr
om ar
epr
esent
ati
vesampl
eofdi
vi
sions.I
twoul
dbeunwi
set
osel
ectt
hem
f
rom onl
yoneort
wodi
vi
sionsast
hismi
ghtgi
vey
ouadi
stor
ted(
biased)pi
ctur
e.I
t
woul
dal
sobeunwi
set
oonl
yint
erv
iew mot
her
swhoat
tendt
heunder
-f
ivescl
i
nic,as
t
hosewhodonotat
tendt
hiscl
i
nicmayweant
hei
rchi
l
drendi
ff
erent
ly.Whenusi
ng

Page34of34
qual
i
tat
iver
esear
chappr
oaches,howev
er,r
epr
esent
ati
venessoft
hesampl
eisnota
pr
imar
yconcer
n.

I
nexpl
orat
oryst
udi
eswhi
chai
m atget
ti
ngar
oughi
mpr
essi
onofhowcer
tai
nvar
iabl
es
mani
festt
hemsel
vesi
nast
udypopul
ati
onorati
dent
if
yingandexpl
ori
ngt
husf
ar
unknownv
ari
abl
es,
youmayt
ryt
osel
ectst
udyuni
tswhi
chgi
vey
out
her
ichestpossi
ble
i
nfor
mat
ion:y
ougof
ori
nfor
mat
ionr
ichcases.Forexampl
e,keyi
nfor
mant
sshoul
d
nev
erbechosenatr
andom,butpur
posi
vel
yfr
om amongt
hosewhohav
ethebest
possi
ble knowl
edge,exper
ience orov
erv
iew wi
thr
espectt
otopi
c ofy
ourst
udy
.
Mor
eov
ert
heyshoul
dbewi
l
li
ngt
oshar
ethi
sinf
ormat
ionwi
thy
oui
nasmuchast
hey
ar
eWel
lI
nfor
medPer
sons.

i
v.Ty
pesofSampl
i
ng
(
a).Pr
obabi
l
ity
/random sampl
i
ng.
Pr
obabi
l
itysampl
i
ngi
nvol
vesusi
ngr
andom sel
ect
ionpr
ocedur
est
oensur
ethateach
uni
toft
hesampl
eischosenont
hebasi
sofchance.Hencer
andom sampl
i
ngoccur
s
whent
hepr
obabi
l
ityofi
ncl
udi
ngeachel
ementoft
hepopul
ati
oncanbedet
ermi
ned.Al
l
uni
tsoft
hest
udypopul
ati
onshoul
dhav
eanequal
,oratl
eastaknownchanceofbei
ng
i
ncl
udedi
nthesampl
e.Pr
obabi
l
itysampl
i
ngr
equi
rest
hatal
i
sti
ngofal
lst
udyuni
ts
exi
stsorcanbecompi
l
ed.Thi
sli
sti
ngi
scal
l
edt
hesampl
i
ngf
rame.Thef
oll
owi
ng
pr
obabi
l
itysampl
i
ngmet
hodswi
l
lbedi
scussed:
1.Si
mpl
erandom sampl
i
ng
2.Sy
stemat
icsampl
i
ng
3.St
rat
if
iedsampl
i
ng
4.Cl
ust
ersampl
i
ng
5.Mul
ti
stagesampl
i
ng

(
1)Si
mpl
erandom sampl
i
ng
Thi
sist
hesi
mpl
estf
orm ofpr
obabi
l
itysampl
i
ng.Tosel
ectasi
mpl
erandom sampl
e
y
ouneedt
o:

Page35of35
 Makeorsear
chf
oranexi
sti
ngnumber
edl
i
stofal
ltheuni
tsi
nthepopul
ati
on
f
rom whi
chy
ouwantt
odr
awasampl
e
 Deci
deont
hesi
zeoft
hesampl
e
 Sel
ectt
her
equi
rednumberofsampl
i
nguni
ts,usi
nga‘
l
ott
ery
’met
hodorat
abl
e
ofr
andom number
s
Forexampl
e,asi
mpl
erandom sampl
eof50st
udent
sist
obesel
ect
edf
rom aschoolof250
st
udent
s.Usi
ngal
i
stofal
l250st
udent
s,eachst
udenti
sgi
venanumber(
1to250.These
number
sar
ewr
it
tenonsmal
lpi
ecesofpaper
.Al
lthe250paper
sar
eputi
nabox,af
terwhi
ch
t
heboxi
sshakenv
igor
ousl
y,t
oensur
erandomi
zat
ion.Then,
50paper
sar
etakenoutoft
hebox,
andt
henumber
sar
erecor
ded.Thest
udent
sbel
ongi
ngt
othesenumber
stakenf
rom t
hebox
wi
l
lconst
it
utet
hesampl
e.

(
2)Sy
stemat
ic/
int
erv
alsampl
i
ng
I
nSy
stemat
icSampl
i
ngi
ndi
vi
dual
sar
echosenatr
egul
ari
nter
val
s(f
orexampl
eev
ery
f
if
th)f
rom t
hesampl
i
ngf
rame.I
deal
l
ywer
andoml
ysel
ectanumbert
otel
luswher
eto
st
artsel
ect
ingi
ndi
vi
dual
sfr
om t
hel
i
st.Forexampl
e,asy
stemat
icsampl
eof100
st
udent
sist
obesel
ect
edf
rom 1200st
udent
sofaschool
.Thesampl
i
ngf
ract
ioni
s:

Popul
ati
on(
N)di
vi
dedbysampl
esi
ze(
n)
1200/
100=12(
sampl
i
ngi
nter
val
)

Thesampl
i
ngi
nter
vali
sther
efor
e12.Ther
esear
chershoul
dnumberpi
ecesofpaper
f
rom onet
otwel
ve.Thenumberoft
hef
ir
stst
udentt
obei
ncl
udedi
nthesampl
eis
chosenr
andoml
y,f
orexampl
ebybl
i
ndl
ypi
cki
ngoneoutoft
wel
vepi
ecesofpaper
,
number
ed1t
o12.I
fnumber6i
spi
cked,t
henev
eryt
wel
ft
hst
udentwi
l
lbei
ncl
udedi
n
t
he sampl
e,st
art
ing wi
th st
udentnumber6,unt
il100 st
udent
s ar
e sel
ect
ed:t
he
number
ssel
ect
edwoul
dbe6,
18,
30,
42,
54,
66,
78,
etc.

Sy
stemat
icsampl
i
ngi
susual
l
ylesst
imeconsumi
ngandeasi
ert
oper
for
mthansi
mpl
e
r
andom sampl
i
ng.Howev
er,
ther
eisar
iskofbi
as,
ast
hesampl
i
ngi
nter
val
maycoi
nci
de
wi
thasy
stemat
icv
ari
ati
oni
nthesampl
i
ngf
rame.Fori
nst
ance,i
fwewantt
osel
ecta
Page36of36
r
andom sampl
eofday
sonwhi
cht
ocountadul
tcl
assat
tendance,
syst
emat
icsampl
i
ng
wi
thasampl
i
ngi
nter
val
of7day
swoul
dbei
nappr
opr
iat
e,asal
lst
udyday
swoul
dfal
lon
t
hesamedayoft
heweek(
e.g.
,Tuesday
sonl
y,whi
chmi
ghtbeamar
ketdayandmany
adul
tscoul
dbei
nthemar
ketr
athert
hani
nadul
tli
ter
acycl
asses)
.

(
3)St
rat
if
iedsampl
i
ng
Thesi
mpl
erandom sampl
i
ngmet
hoddescr
ibedabov
ehasasdi
sadv
ant
aget
hatsmal
l
gr
oupsi
nwhi
cht
her
esear
cheri
sint
erest
edmayhar
dlyappeari
nthesampl
e.I
fiti
s
i
mpor
tantt
hatt
hesampl
eincl
udesr
epr
esent
ati
vest
udyuni
tsofsmal
lgr
oupswi
th
speci
fi
cchar
act
eri
sti
cs(
forexampl
e,r
esi
dent
sfr
om ur
banandr
uralar
eas,ordi
ff
erent
r
eli
giousoret
hni
cgr
oups)
,thent
hesampl
i
ngf
ramemustbedi
vi
dedi
ntogr
oups,or
st
rat
a,accor
dingt
othesechar
act
eri
sti
cs.Random orsy
stemat
icsampl
esofapr
e-
det
ermi
nedsi
zewi
l
lthenhav
etobeobt
ainedf
rom eachgr
oup(
str
atum)
.Thi
siscal
l
ed
St
rat
if
iedSampl
i
ng.

St
rat
if
ied sampl
i
ng i
sonl
ypossi
ble when we know whatpr
opor
ti
on oft
he st
udy
popul
ati
onbel
ongst
oeachgr
oupwear
eint
erest
edi
n.Anadv
ant
ageofst
rat
if
ied
sampl
i
ngi
sthatwecant
akear
elat
ivel
ylar
gesampl
efr
om asmal
lgr
oupi
nourst
udy
popul
ati
on.Thi
sal
l
owsust
ogetasampl
ethati
sbi
genought
oenabl
eust
odr
awv
ali
d
concl
usi
onsaboutar
elat
ivel
ysmal
lgr
oupwi
thouthav
ingt
ocol
l
ectanunnecessar
il
y
l
arge(
andhenceexpensi
ve)sampl
eoft
heot
her
,lar
gergr
oups.

Forexampl
e,i
fwewer
etodoast
udyGTI
-Embust
udent
s’awar
enessofHI
V/AI
DS,we
wi
l
ldi
vi
det
hest
udent
sint
odi
ff
erentst
rat
asuchasage,gender
,ar
eaofor
igi
n,cour
se
bei
ngt
aken,y
earofst
udy
,et
c.t
hen,i
nli
newi
tht
henumberofst
udent
sineachst
rat
a,
wedr
awpr
opor
ti
onat
esampl
efr
om eachst
rat
a

(
4)Cl
ust
ersampl
i
ng
I
tmaybedi
ff
icul
tori
mpossi
blet
otakeasi
mpl
erandom sampl
eoft
heuni
tsoft
he
st
udypopul
ati
on atr
andom,becauseacompl
etesampl
i
ng f
ramedoesnotexi
st.

Page37of37
Logi
sti
caldi
ff
icul
ti
es may al
so di
scour
age r
andom sampl
i
ng t
echni
ques (
e.g.
,
i
nter
viewi
ngpeopl
ewhoar
escat
ter
edov
eral
argear
eamaybet
oot
ime-
consumi
ngf
or
exampl
enomadi
cpast
oral
i
sts)
.Howev
er,whenal
i
stofgr
oupi
ngsofst
udyuni
tsi
s
av
ail
abl
e(e.
g.,v
il
lagesorschool
s)orcanbeeasi
l
ycompi
l
ed,anumberoft
hese
gr
oupi
ngscanber
andoml
ysel
ect
ed.Thesel
ect
ionofgr
oupsofst
udyuni
ts(
clust
ers)
i
nst
eadoft
hesel
ect
ionofst
udyuni
tsi
ndi
vi
dual
l
yiscal
l
edCl
ust
erSampl
i
ng.

Cl
ust
ersar
eof
tengeogr
aphi
cuni
ts(
e.g.
,di
str
ict
s,v
il
lages)oror
gani
zat
ionaluni
ts(
e.g.
,
cl
i
nics,t
rai
ning gr
oups)
.Forexampl
e,i
n a st
udyoft
heknowl
edge,at
ti
tudesand
pr
act
ices(
KAP)r
elat
edt
ofami
l
ypl
anni
ngi
nMuni
cipal
i
tyLocat
ion,EmbuDi
str
ict
,al
i
st
i
smadeofal
lthev
il
lagesi
nthemuni
cipal
i
ty.Usi
ngt
hisl
i
st,
arandom sampl
eofv
il
lages
i
schosenandal
lst
udyuni
tsi
nthesel
ect
edv
il
lagesar
eint
erv
iewed.

(
5)Mul
ti
-st
agesampl
i
ng
I
nver
ylar
geanddi
ver
sepopul
ati
onssampl
i
ngmaybedonei
ntwoormor
est
ages.Thi
s
i
sof
tent
hecasei
ncommuni
ty-
basedst
udi
es,i
nwhi
chpeopl
ear
etobei
nter
viewed
f
rom di
ff
erentv
il
lages,andt
hev
il
lageshav
etobechosenf
rom di
ff
erentar
eas.For
exampl
e,i
nast
udyofut
il
izat
ionofpi
tlat
ri
nesi
nEmbuDi
str
ict
,150homest
eadsar
eto
bev
isi
tedf
ori
nter
viewswi
thf
ami
l
ymember
saswel
lasf
orobser
vat
ionsont
ypesand
cl
eanl
i
nessofl
atr
ines.Thedi
str
icti
scomposedof6war
dsandeachwar
dhasbet
ween
6and9v
il
lages.Thef
oll
owi
ngf
our
-st
agesampl
i
ngpr
ocedur
ecoul
dbeper
for
med:
 Sel
ect3war
dsoutoft
he6bysi
mpl
erandom sampl
i
ng.
 Foreachwar
dsel
ect5v
il
lagesbysi
mpl
erandom sampl
i
ng(
15v
il
lagesi
ntot
al)
.
 Foreachv
il
lagesel
ect10househol
ds.Si
ncesi
mpl
ychoosi
nghousehol
dsi
nthe
cent
reoft
hev
il
lagewoul
dpr
oduceabi
asedsampl
e,t
hef
oll
owi
ngsampl
i
ng
pr
ocedur
eispr
oposed:
 Got
othecent
reoft
hev
il
lage.
 Chooseadi
rect
ioni
nar
andom way
:spi
nabot
tl
eont
hegr
oundand
chooset
hedi
rect
iont
hebot
tl
enecki
ndi
cat
es.
 Wal
kint
hechosendi
rect
ionandsel
ectev
eryhousehol
dunt
ily
ouhav
ethe

Page38of38
10y
ouneed.
 I
fyour
eacht
heboundar
yoft
hev
il
lageandy
oust
il
ldonothav
e10
househol
ds,r
etur
ntot
he cent
re oft
he v
il
lage,wal
kint
he opposi
te
di
rect
ionandcont
inuet
osel
ecty
oursampl
eint
hesamewayunt
ily
ou
hav
e10.
 I
fther
eisnobodyi
nachosenhousehol
d,t
aket
henextnear
estone.
 Deci
debef
orehandwhom t
oint
erv
iew(
forexampl
etheheadoft
hehousehol
d,i
f
pr
esent
,ort
heol
destadul
twhol
i
vest
her
eandwhoi
sav
ail
abl
e).

Amul
ti
-st
agesampl
i
ngpr
ocedur
eiscar
ri
edouti
nphasesandi
tusual
l
yinv
olv
esmor
e
t
hanonesampl
i
ngmet
hod.Themai
nadv
ant
agesofcl
ust
erandmul
ti
-st
agesampl
i
ng
ar
ethat
:
1.Asampl
i
ngf
rameofi
ndi
vi
dualuni
tsi
snotr
equi
redf
ort
hewhol
epopul
ati
on.
Exi
sti
ngsampl
i
ngf
ramesofcl
ust
ersar
esuf
fi
cient
.Onl
ywi
thi
nthecl
ust
erst
hat
ar
efi
nal
l
ysel
ect
edi
sther
eaneedt
oli
standsampl
ethei
ndi
vi
dualuni
ts(
ifnot
usi
ngt
hebot
tl
espi
nni
ngmet
hod)
.
2.Thesampl
eiseasi
ert
osel
ectt
hanasi
mpl
erandom sampl
eofsi
mil
arsi
ze,
becauset
hei
ndi
vi
dualuni
tsi
nthesampl
ear
ephy
sical
l
ytoget
heri
ngr
oups,
i
nst
eadofscat
ter
edal
lov
ert
hest
udypopul
ati
on.

Themai
ndi
sadv
ant
ageoft
hist
ypeofsampl
i
ngi
sthat
,Compar
edt
osi
mpl
erandom
sampl
i
ng,t
her
eisal
argerpr
obabi
l
ityt
hatt
hef
inalsampl
ewi
l
lnotber
epr
esent
ati
veof
t
he t
otalst
udypopul
ati
on.The l
i
kel
i
hood oft
he sampl
e notbei
ng r
epr
esent
ati
ve
dependsmai
nlyont
henumberofcl
ust
erst
hati
ssel
ect
edi
nthef
ir
stst
age.Thel
arger
t
he numberofcl
ust
ers,t
he gr
eat
eri
sthe l
i
kel
i
hood t
hatt
he sampl
e wi
l
lbe
r
epr
esent
ati
ve.Fur
ther
,thesampl
i
ng uni
tsatcommuni
tyl
evelshoul
d besel
ect
ed
r
andoml
y(av
oidconv
eni
encesampl
i
ng!
).

(
b).Non-
probabi
l
ity
/non-
random sampl
i
ng.
Thi
sref
erst
othecasewher
ethepr
obabi
l
ityofi
ncl
udi
ngeachel
ementoft
hepopul
ati
on

Page39of39
i
nthesampl
eisunknown.Thi
sist
hesampl
i
ngusedwhent
her
esear
chi
snoti
nter
est
ed
i
nsel
ect
ingasampl
ethati
srepr
esent
ati
veoft
hepopul
ati
on.I
tismost
lyusedi
n
qual
i
tat
ive st
udi
es because t
he f
ocus i
s on i
n-dept
hinf
ormat
ion and notmaki
ng
i
nfer
encesorgener
ali
zat
ions.Someoft
henon-
probabi
l
itysampl
i
ngpr
ocedur
esi
ncl
ude:

(
1).Pur
posi
ve/
judgment
alsampl
i
ng
Thi
stechni
queal
l
owsar
esear
chert
ousecasest
hathav
ether
equi
redi
nfor
mat
ionwi
th
r
espectt
otheobj
ect
ivesoft
hest
udy
.Casesar
ether
efor
ehandpi
ckedbecauset
heyar
e
i
nfor
mat
iveorpossesst
her
equi
redchar
act
eri
sti
cs.Ar
esear
cherwhousest
hist
ypeof
sampl
i
ngmustspeci
fyt
hecr
it
eri
aforchoosi
ngpar
ti
cul
arcasese.
g.acer
tai
nager
ange,
gender
,rel
i
giousaf
fi
li
ati
onoreducat
ionall
evel
.Somer
esear
cher
smayusepur
posi
ve
sampl
i
ngasamul
ti
-st
agesampl
i
ngpr
ocedur
e.Forexampl
e,i
tcanbeusedt
opi
ckt
he
l
ocat
ionwher
ethest
udysubj
ect
sar
etobef
oundsuchassampl
i
ngadi
str
ictofi
nter
est
.
Pur
posi
vesampl
esi
ze,whi
chmayormaynotbef
ixedbef
oredat
acol
l
ect
iondepends
ont
her
esour
cesandt
het
imeav
ail
abl
etot
her
esear
cherandt
hest
udyobj
ect
ives.I
n
pur
posi
ve sampl
i
ng,sampl
e si
ze i
s of
ten det
ermi
ned on t
he basi
s oft
heor
eti
cal
sat
urat
ion(
thepoi
ntwher
emor
edat
abr
ingsnomor
eaddi
ti
onalknowl
edgeconcer
ning
t
hest
udyquest
ion)
.

(
2).Snowbal
lsampl
i
ng
Al
socal
l
edmudbal
lsampl
i
ngorchai
nref
err
alsampl
i
ngi
sthet
echni
quewher
eini
ti
al
subj
ect
swi
tht
hedesi
red char
act
eri
sti
csar
eident
if
ied usi
ng pur
posef
ulsampl
i
ng.
Those who hav
e been i
dent
if
ied name ot
her
s who t
heyknow hav
ethe r
equi
red
char
act
eri
sti
csunt
ilt
her
esear
cherget
sthenumberofcases/
respondent
sher
equi
res
f
orhi
s/herst
udy
.Assuch,par
ti
cipant
sori
nfor
mant
swi
thwhom cont
acthasal
ready
beenmadeuset
hei
rnet
wor
kst
oref
ert
her
esear
chert
o ot
herpeopl
ewho coul
d
pot
ent
ial
l
ypar
ti
ci
pat
einorcont
ri
but
etot
hest
udy
.Thi
smet
hodi
snor
mal
l
yusedwhen
t
her
esear
cherdoesnothav
eadequat
einf
ormat
ionconcer
ningt
hepopul
ati
ont
hathas
t
hechar
act
eri
sti
csher
equi
res.Forexampl
e,i
far
esear
cherwant
stost
udypeopl
ewho
hav
ebeenr
etr
enchedi
nanor
gani
zat
ion,
heneedst
oident
if
yoneoft
her
etr
encheswho

Page40of40
t
heni
dent
if
iesanot
herandt
het
rendcont
inuesunt
ilt
her
esear
cherget
sther
equi
red
sampl
esi
ze.

(
3).Quot
asampl
i
ng
Si
mil
art
ost
rat
if
iedr
andom sampl
i
ngandt
heobj
ect
ivei
stoi
ncl
udev
ari
ousgr
oupsor
quot
asoft
hest
udypopul
ati
onbasedonsomecr
it
eri
aorcer
tai
nchar
act
eri
sti
cs.Such
char
act
eri
sti
cscoul
dincl
udeage,pl
aceofr
esi
dence,gendercl
ass,pr
ofessi
on,mar
it
al
st
atus,HI
Vst
atus,et
c.Ther
esear
cherdi
vi
dest
hest
udypopul
ati
onbasedoncer
tai
n
cr
it
eri
aand t
hen pur
posi
vel
ypi
cksr
espondent
sfr
om each cat
egor
ybecauset
hey
bel
ongt
othatcat
egor
y.Thecr
it
eri
ather
esear
cherchoosesal
l
owshi
m/hert
ofocuson
t
hepeopl
ehe/
shet
hinksar
emostl
i
kel
ytoexper
ience,
knowaboutorhav
einsi
ght
sint
o
t
her
esear
cht
opi
c.Thesel
ect
ionofact
ualpar
ti
cipant
sisnotr
andom si
ncest
udy
subj
ect
sar
epi
ckedast
heyf
iti
ntoi
dent
if
iedquot
as.Theov
eral
lsampl
ehencewi
l
l
somehowbeacci
dent
all
ysel
ect
ed.

(
4).Conv
eni
ent
/acci
dent
al/
avai
l
abi
l
itysampl
i
ng
Al
socal
l
edv
olunt
eersampl
i
ng,
iti
nvol
vessel
ect
ingcasesoruni
tsofobser
vat
ion/
study
ast
heybecomeav
ail
abl
etot
her
esear
cher
.Thi
sismai
nlydonebyj
our
nal
i
stsespeci
all
y
whent
heywantt
ogett
hegener
alf
eel
i
ngoft
hepubl
i
caboutani
ssue.Theycangot
o
t
hest
reetandaskpedest
ri
anst
ogi
vet
hei
ropi
nionaboutapubl
i
cpol
i
cyi
ssuessuchas
t
hePr
esi
dent
’sMadar
akaDaySpeechort
heFi
nanceMi
nist
er’
sBudgetSpeech.The
r
esear
cherhence sel
ect
sar
espondentwho i
s conv
eni
ent
ly av
ail
abl
e such as a
nei
ghbour
,col
l
eagueorf
ri
end.Themai
nfeat
ureoft
hissampl
i
ngi
sthatr
espondent
s
ar
ereadi
l
yav
ail
abl
eandaccessi
ble.Resear
cher
swhouset
hiski
ndofsampl
i
ngcannot
hav
eanybasi
stoar
guet
hati
tisr
epr
esent
ati
veoft
hepopul
ati
on.

3.Bi
asi
nsampl
i
ng
Bi
asi
n sampl
i
ng i
sa sy
stemat
icer
rori
n sampl
i
ng pr
ocedur
es,whi
ch l
eadst
oa
di
stor
ti
oni
nther
esul
tsoft
hest
udy
.Thi
shast
hepot
ent
ial
ofmi
srepr
esent
ingf
act
sand
t
her
efor
edi
stor
ti
ngr
eal
i
ty.Theuseoff
aul
tydat
acol
l
ect
iont
ool
swoul
dleadt
obi
ased

Page41of41
r
esul
ts.Bi
as can al
so be i
ntr
oduced as a consequence ofi
mpr
opersampl
i
ng
pr
ocedur
es,
whi
chr
esul
tint
hesampl
enotbei
ngr
epr
esent
ati
veoft
hest
udypopul
ati
on.
Forexampl
e,ast
udywasconduct
edt
odet
ermi
net
heheal
thneedsofar
uralpopul
ati
on
i
n or
dert
o pl
an pr
imar
yheal
th car
e act
ivi
ti
es.Howev
er,a nomadi
ctr
ibe,whi
ch
r
epr
esent
edonet
hir
doft
het
otalpopul
ati
on,wasl
eftoutoft
hest
udy
.Asar
esul
tthe
st
udydi
dnotgi
veanaccur
atepi
ctur
eoft
heheal
thneedsoft
het
otal
popul
ati
on.

Ther
ear
esev
eralpossi
blesour
cesofbi
ast
hatmayar
isewhensampl
i
ngandt
hemost
wel
lknownsour
cei
snon-
response.Non-
responsecanoccuri
nanyi
nter
viewsi
tuat
ion,
but i
tis most
ly encount
ered i
n l
arge-
scal
e sur
vey
s wi
th sel
f-
admi
nist
ered
quest
ionnai
res.Respondent
smayr
efuseorf
orgett
ofi
l
lint
hequest
ionnai
re.The
pr
obl
em l
i
esi
nthef
actt
hatnon-
respondent
sinasampl
emayexhi
bitchar
act
eri
sti
cs
t
hatdi
ff
ersy
stemat
ical
l
yfr
om t
hechar
act
eri
sti
csofr
espondent
s.Ther
ear
esev
eral
way
stodeal
wit
hthi
spr
obl
em andr
educet
hepossi
bil
i
tyofbi
as:
1.Dat
acol
l
ect
iont
ool
sshoul
dbepr
e-t
est
edandi
fnecessar
y,adj
ust
ment
smadet
o
ensur
ebet
terco-
oper
ati
on.
2.I
fnon-
responsei
sduet
oabsenceoft
hesubj
ect
s,f
oll
ow-
upofnon-
respondent
s
maybeconsi
der
ed.
3.I
fnon-
responsei
sduet
oref
usalt
oco-
oper
ate,anext
ra,separ
atest
udyofnon-
r
espondent
smaybeconsi
der
edi
nor
dert
oident
if
ytowhatext
entt
heydi
ff
er
f
rom r
espondent
s.
4.Anot
herst
rat
egyi
stoi
ncl
udeaddi
ti
onalpeopl
eint
hesampl
e,sot
hatnon-
r
espondent
swhower
eabsentdur
ingdat
acol
l
ect
ioncanber
epl
aced.Howev
er,
t
hiscanonl
ybej
ust
if
iedi
fthei
rabsencewasv
eryunl
i
kel
ytober
elat
edt
othe
t
opi
cbei
ngst
udi
ed.

NB:Thebi
ggert
henon-
responser
ate,t
hemor
enecessar
yitbecomest
otaker
emedi
al
act
ion.I
tisi
mpor
tanti
nanyst
udyt
oment
iont
henon-
responser
ateandt
ohonest
ly
di
scusswhet
herandhow t
henon-
responsemi
ghthav
einf
luencedt
her
esul
ts.Ot
her
sour
cesofbi
asi
nsampl
i
ngmaybel
essobv
ious,
butatl
eastasser
iousandi
ncl
udet
he

Page42of42
f
oll
owi
ng:
1.St
udy
ingv
olunt
eer
sonl
y.Thef
actt
hatv
olunt
eer
sar
emot
ivat
edt
opar
ti
cipat
ein
t
hest
udymaymeant
hatt
heyar
eal
sodi
ff
erentf
rom t
hest
udypopul
ati
onont
he
f
act
orsbei
ngst
udi
ed.Ther
efor
eiti
sbet
tert
oav
oidusi
ngnon-
random sel
ect
ion
pr
ocedur
est
hati
ntr
oducesuchanel
ementofchoi
ce.
2.Sampl
i
ngoft
hepeopl
eyoumeeti
nthei
rhomesorf
armsonl
y(conv
eni
ence)
.
Thesear
eli
kel
ytodi
ff
ersy
stemat
ical
l
yfr
om peopl
ewhoar
eel
sewher
efor
exampl
ethosedoi
ngwhi
tecol
l
arj
obsandt
her
efor
eleadi
ngt
ounr
epr
esent
ati
ve
sampl
eanddat
a.
3.Mi
ssi
ngcasesofshor
tdur
ati
on.I
nst
udi
esoft
hepr
eval
enceofdi
sease,
casesof
shor
tdur
ati
onar
emor
eli
kel
ytobemi
ssed.Thi
smaymeanmi
ssi
ngf
atalcases,
caseswi
thshor
til
l
nessepi
sodesandmi
l
dcases.
4.Seasonalbi
as.I
tmaybet
hatt
hepr
obl
em underst
udy
,forexampl
e,mal
nut
ri
ti
on
orpr
ost
it
uti
onexhi
bit
sdi
ff
erentchar
act
eri
sti
csi
ndi
ff
erentseasonsoft
hey
ear
.
Fort
hisr
eason,dat
ashoul
dbecol
l
ect
edont
hepr
eval
enceanddi
str
ibut
ionof
t
hephenomenonunder
studyi
nanar
eadur
ingal
lseasonsr
athert
hanj
ustatone
poi
nti
nti
me.When i
nvest
igat
ing heal
th ser
vices’per
for
mance,one has t
o
consi
dert
hef
actt
hatt
owar
dst
heendoft
hef
inanci
aly
earshor
tagesmayoccur
i
ncer
tai
nbudgeti
temswhi
chmayaf
fectt
hequal
i
tyofser
vicesdel
i
ver
ed.
5.Tar
macbi
as.St
udyar
easar
eof
tensel
ect
edbecauset
heyar
eeasi
l
yaccessi
ble
bycar
.Howev
er,t
hesear
easar
eli
kel
ytobesy
stemat
ical
l
ydi
ff
erentf
rom mor
e
i
naccessi
ble ar
eas and t
he i
nfor
mat
ion col
l
ect
ed i
s mor
eli
kel
yto be
unr
epr
esent
ati
veoft
hepopul
ati
onf
rom whi
chi
tisdr
awn.

6.Et
hical
&Pol
i
tical
Consi
der
ati
onsi
nResear
ch

Wi
thr
elev
antexampl
es,
wri
teacr
it
ical
essayont
heet
hicali
ssuesi
n
soci
alsci
encer
esear
ch.

Page43of43
Br
ief
lyexpl
ain whyt
he r
esear
cher
s at
tach i
mpor
tance t
o et
hical
i
ssues

Ther
ear
evar
iouset
hicali
ssuest
otakei
ntoaccountwhi
l
eunder
taki
ngr
esear
chand
t
hesear
elar
gel
yai
medatav
oidi
ngsi
tuat
ionswher
easr
esear
cher
sinf
ri
ngeonot
her
peopl
e’
sri
ght
s.Thesei
ncl
ude:
 Vol
unt
arypar
ti
cipat
ion.I
nter
views/
subj
ect
s/par
ti
cipant
sshoul
dnotbecoer
cedi
nto
par
ti
cipat
ioni
nanyst
udy
.Rat
her
,par
ti
cipat
ionshoul
dbev
olunt
ary
.Inot
herwor
ds,
coer
cionofsubj
ect
scompr
omi
sest
hequal
i
tyofdat
agi
vent
hatt
heyar
enott
hey
ar
efor
ced i
nto par
ti
cipat
ion;t
hey ar
e mor
eli
kel
yto gi
vef
orce ordi
stor
ted
i
nfor
mat
ion.Thi
sshoul
dbeav
oidedatal
lcostsi
ncet
heov
eral
lfi
ndi
ngsshoul
d
r
efl
ectt
her
eal
i
ty.Anexampl
eofcoer
cedi
nfor
mat
ionwoul
dbepol
i
cei
nter
rogat
ion
ofsuspect
sthr
ought
ort
ure.Anaddi
ti
on,
theyshoul
dnotbebr
ibedorpr
omi
sedgi
ft
s
soast
ogi
vet
hei
nfor
mat
iont
hatot
her
wiset
heywoul
dnothav
egi
ven.Thebot
tom
l
i
ne her
e shoul
d be t
hat r
esear
cher
s st
ri
ke a good r
appor
t wi
th
r
espondent
s/par
ti
cipant
s/i
nter
vieweesso t
hatt
heyar
eateasewi
th hi
m/heri
n
gi
vi
ngt
her
equi
redi
nfor
mat
ion.
 Nohar
mtoi
nter
viewees/
par
ti
cipant
s/r
espondent
s.Lett
hest
udynotdamaget
he
r
eput
ati
onoft
hei
nter
viewee,hi
s/heregoorpeopl
e.Thi
scal
l
suponr
esear
cher
sto
be awar
e ofnegat
ivef
eel
i
ngs f
rom r
espondent
s/subj
ect
s.Forexampl
e when
i
nter
viewi
ngpeopl
eli
vi
ngoraf
fect
edbyHI
V/AI
DS,
peopl
ewhohav
epr
evi
ousl
ybeen
af
fect
edadv
ersel
ybywarorwhohadl
ostl
ovedonesorsubst
ant
ialamountof
pr
oper
ty,peopl
eonr
epr
oduct
iveheal
th,t
her
esear
cheri
srequi
redt
omakesur
e
he/
shedoesnotar
ousenegat
ivef
eel
i
ngst
hatmayst
iremot
ionsanddi
sor
ientt
he
t
empooft
hei
nter
view.
 I
dent
it
ydi
scl
osur
e.Bef
orecommencementoft
hei
nter
view,expl
aincl
ear
lyt
othe
i
nter
viewee who y
ou ar
e,whaty
ou ar
e doi
ng,f
orwhom,t
he pur
pose oft
he
i
nfor
mat
ioncol
l
ect
edandhowi
twi
l
lbeused.I
ntr
oducey
our
sel
fandmakesur
ethe
r
espondent
sunder
standy
ouwel
ltoav
oidmi
sunder
standi
ngst
hatwoul
drequi
re
i
mmat
uret
ermi
nat
ionoft
hei
nter
viewandt
huswast
ageoft
ime.
Page44of44
 I
nfor
medconsentf
rom t
hei
nter
viewee.Resear
cher
sshoul
dendeav
ourt
ogetwher
e
necessar
yawr
it
tenconsentoft
hei
nter
vieweebef
oreunder
taki
ngt
hest
udy
.For
exampl
e,whi
l
eint
erv
iewi
ngpeopl
eli
vi
ngwi
thHI
V/AI
DS,
aft
ert
hev
erbal
consent
,iti
s
necessar
ytogetawr
it
tenconsentacknowl
edgi
ngt
hatt
heywer
enotcoer
cedand
t
hatt
heygav
ethei
nfor
mat
ionv
olunt
ari
l
y.Thi
sisspeci
fi
cal
l
ytoav
oidf
utur
elawsui
ts.
 Conf
ident
ial
i
ty/
pri
vacy
. Makei
tast
andar
drequi
rementt
okeepconf
ident
ialal
l
i
nfor
mat
ioncol
l
ect
ed,onl
ydi
scl
osi
ngi
ttor
elev
antandaut
hor
izedpeopl
eincl
udi
ng
f
ell
ow r
esear
cher
s,butonl
yifnecessar
y.Use pseudo/
fict
it
ious names whi
l
e
r
efer
ri
ngt
othei
nter
vieweest
o concealt
hei
rident
it
yandmakeev
eryef
for
tto
expunget
henamesandot
hermat
eri
alst
hatmayi
dent
if
ythei
nter
viewees.
 Honestr
epor
ti
ng.Pr
ovi
deanhonestaccountofhowt
hest
udywascar
ri
edout
.Poi
nt
outwher
ether
esear
chdesi
gnhadt
obechangedandsaywhyi
thadt
obechanged
andacknowl
edget
hesour
cesoff
undi
ng-
justhi
ghl
i
ghtt
hemaj
orchanges.
 Pl
agi
ari
sm/
academi
cfr
aud.I
ninst
ances wher
eyou use ot
herpeopl
e’
s wor
k,
acknowl
edget
hei
rsour
cesaspercopy
rightr
equi
rement
sandav
oidpl
agi
ari
sm
(
steal
i
ngot
herpeopl
eswor
k)andmaki
ngi
tyourown.Thi
sisi
l
legal
andpuni
shabl
e.
 Aut
hor
izat
ion. Wher
e r
equi
red (
and t
his i
s of
ten t
he pr
act
ice)
, get
aut
hor
izat
ion/
li
censef
rom r
elev
antaut
hor
it
iesandst
ickt
othel
egalr
equi
rement
sof
t
hecount
ryy
ouunder
taki
ng t
hest
udy
.Forexampl
e,i
nKeny
a,r
esear
cher
sar
e
r
equi
redt
opr
ovi
det
otheaut
hor
izi
ngof
fi
cecopi
esoft
her
esear
chf
indi
ngs,
inmost
casest
heMi
nist
ryofEducat
ion

Readi
ngMat
eri
als
Bl
ess,
C;C.Hi
gson-
Smi
th&A.Kagee( .Fundament
2006) alsofSoci
alResear
ch
Met
hods:AnAf
ri
canPer
spect
iv 4thed)
e( .Kenwy
n:Jut
a&Co.l
td.
Br
ynar
d,P.A&S.X.Hanekom ( .I
2006) nt
roduct
iont
oResear
chi
nManagement
-
Rel
atedFi
el 2nded)
ds( .Pr
etor
ia:
VanSchai
kPubl
i
sher
s.
Gupt
a,C.B&V.Gupt
a( .AnI
2001) ntr
oduct
iont
oSt
ati
sti
calMet
hods.NewDel
hi:
Vikas
Publ
i
shi
ngHousePVTLt
d.
Mugenda,
O.M &A.G.Mugenda( .Resear
1999) chMet
hods:Quant
it
ati
veand

Page45of45
Qual
i
tat
ive.Nai
robi
:ACTSPr
ess.
Mwanj
e,J.
I( .Qual
2001) i
tat
iveResear
chPr
ocess:Soci
alSci
enceResear
ch
Met
hodol
ogySer
iesModul
e2.Addi
sAbaba:
Organi
zat
ionf
orSoci
alSci
ence
Resear
chi
nEast
ernandSout
her
nAf
ri
ca(
OSSREA)
.
Mwanj
e,J.
I&B.Got
u( .I
2001) ssuesi
nSoci
alResear
ch:Soci
alSci
enceResear
ch
Met
hodol
ogySer
iesModul
e3.Addi
sAbaba:Or
gani
zat
ionf
orSoci
alSci
ence
Resear
chi
nEast
ernandSout
her
nAf
ri
ca(
OSSREA)
.
Ngechu,
M( .Under
2006) standi
ngt
heResear
chPr
ocessandMet
hods:AnI
ntr
oduct
ion.
Nai
robi
:Mar
yNgechu.
Ony
ango,
J.P&A.
M.Pl
ews( .AText
1987) bookofBasi
cSt
ati
sti
cs.Nai
robi
:EastAf
ri
can
Educat
ional
Publ
i
sher
s.
Schut
t,R.
K( .I
1996) nv
est
igat
ingt
heSoci
alWor
ld:ThePr
ocessandPr
act
iceofSoci
al
Resear
ch.ThousandOaks:Pi
neFor
gePr
ess.
Si
ngl
eton,
R(j
nr)
;B.C.St
rai
ts;
M.M.St
rai
ts&R.J.McAl
l
ist
er( .Appr
1988) oachest
o
Soci
alResear
ch.NewYor
k:Oxf
ordUni
ver
sit
yPr
ess.

Page46of46

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy