? ?c M
￥??
1. Và Q?M
Và Q?M
B?μ ?/ 4?¨
1 à Q??¨
?
^ μ ￥B?` V? b
2 à Q￥?ì??¨
à Q￥?ì??ü
^à Q ￥?ì?? b
3 ê4
à Q￥·S
9ü
^M
à Q￥ =d?ì??-a1??￥yTü
^¨
I
1·S
M
?V
UM? =d￥
ù5
b
à Q￥·S V[
^ "￥ bB B b
μ￥μ2 8" b
4 ??
à Q￥ 8?·S
à Q￥
31B1 ￥ f ?/1
8?Q?à Q =d￥ o
pü3
1üB* 8"8?1B? b
8?￥'é
ü
^ b
2. à Q￥ ?ù5
9￥e5
^1D ¨EBáà
^M ?￥ b?1 ?B? 3
B?
??Dè
? 8MQ b
tY e
|
ì ?^Tp? ?? SD?Z b
V
s ?ìa?
s ?F??
s ?/
?
s ? GDP
s ?5
s ??
s ? ,0 b
??T1? [5 ?ù? b1o?
^ ?7
^ =d[# =d?M￥il b
?
1c M
′DM
′￥?
1. M
′DM
′￥ |Z
T
1 |M
′
? ￥ZE ?Y
|M
′v·μ
v ?ZE
B
^1¨
? b
=
^1- b?
^ ?5 ? SD =￥
? SDù?￥K×1￥' ? ?-B b
2
? ?Báù5
?] Mk
? ￥?Báù5
^ènù5-B b 'é
^ M
￥ ￥μs b
3M
′?
￥à
?
'à Q ?
ü
^D
Z41¤ B?Hq/￥ b
?v?
M
′￥ìas ?óM
?′ V[s1
A Hq/?
D B ?
?Z
T b
?v?′?D$?′"￥1"M
￥?
9 V[s1
v ??
Aó ￥?′Z
T b
B ￥ 1?′ Z
T b
?
2T
^M
￥′
? b
?
ü?μμ b
G
Lty
N aNV a
3?s
!!? B?Hq/￥?
`￥M
′
Z4?? <B? μ
tvμ b
G LEóù?`?′?$?
k?8￥ 1?′?
^? ??
s￥?? a
?V
? b C?v ê b
2. M
?
ZE
1?
￥ ??
A. d? ?￥?
B. ? ?￥?
C. d? ? 1? ?￥?
D. ? ? 1d? ?￥?
2? ??
￥1o-B /?ùù5
Constructing Questions
?
1 ??
￥ ??? B aC
?
lá1? ??
￥ ?Y?A?/?ù￥ù
5
Questions b
?/ù5
Constructing Questions
^?
? ?M
′￥dè×1￥B? b
Và Q 1
ˉ?1
íM
1
?B1
íM
1ù5??
^B? 8?￥V? b
oμ?ùù5/?8? ??
? ??
s?? a
?￥V?? ? 7
S b
? f ?/+Y
^?M
=d￥ˉ???ú
H31/ ?ù5
B?ù5"
? a set of questions ??
??M
b
è ?M
o f??? p￥?
ü31YV ?/μ1 o f???￥VC p ??
At any time during the past week
Was your appetite so poor that you did nor like eating?
Did you feel so tired and worn out that you could not enjoy anything?
Did you feel depressed?
Did you feel unhappy about the way your life is going?
Did you feel discouraged and worried about your future?
Did you feel lonely?
?
M
¨￥ù5
question ?1μ ?/ ?? ??
A. +?
Yq
? ￥ù5
recalling question b
B. s?￥ù5
fixed-choices question b
C. ù5
open-ended question b
?ùù5
questions?? ?
P¨ ?9èè
P¨ b
3?
M
￥ ? ??ùù5
?
M
¨￥ù5
question ?1μ ?/ ?? ??
A. +?
Yq
? ￥ù5
recalling question b
B. s?￥ù5
fixed-choices question b
C. ù5
open-ended question b
-
?ù5
?¤?￥s?K d9) ?
^K ￥ b
4? ??
￥Sù5
?
￥ ù5
^? ??
￥K×1y
í-B b
?
￥Sμ
?
B
^ ￥S b
=
^ S b
5? ??
￥ ?'
ù5
?'
9
^? ??
? ?? $?? a
?￥??￥×1y
í-B b
6¨ - |￥8?·S
Index or scale
1
1ù?￥ZLù???A1 a9???
I
11-Vh 7
S?1-
!9B*ù
5 ??
B?M
à Q7??? , - |ù?￥8?·S
Index or scale b
?
^B?'e5 b
7?
y
? ??
7y
^
d? ??
?
<l￥?
y ?
8
? ￥ ??
u5 ?d9 vP. 28-30
? ?
9?
￥ ê
T=
S=
R=
3. ù B?×1￥?
y
1ù ￥F?
A B
2ù5￥ ??
A 7b?ù5D>?ù5
Bík?ù5Ddík?ù5
C?êD êù5
D ???ù5Dd ???ù5
E?êD êù5￥
? ) ?
??êù51ü
j1B?M
b
? êù51ü T1B?M
b
3
V￥?1 ??
A
V￥à Q
V
Scale +Y
a?? ?M
￥?
b
B ?? ?s
V
Y
scale V ? b
C Likert
V
?T
V
#? ù5ü?1
V
itemized rating scale b
D Item analysis[ "s
ù?￥
^ ú X+
V?￥0ù5
item ￥sO
qù5 b
E1??
VD ??
V
4ù
!9?￥ ??×1ù5
Aù ù5
! ?￥?Be5 o ?¤ p￥e5 b
B?êù5 !ês???
^B?s? bW￥
^ ￥ b
C ê5￥!ês?A? ￥ b
Dù5￥?
??!ês??? ?μ b
Eù5
!9￥¨?
??1 b
F ù5￥?
??1
??ù?1￥ K? b
G ù ù5?1D
K? b
Hù5￥?]￥ V ??á 2T b
I?¤???
Lís7?A?
3￥
? V[YV ¤
M?￥
? b
J ù ￥é
Kù5￥
¨?
Lù
!9B?1YV ??? b
l2 ù ? ?/ ? ? ￥ù5?
V?
ík? ?
? F
?
HW v V[?1
4. ù )
? ? ès
??c ?
￥?Dr
X=T+S+R
7bù5 99
^
$ >?ù5
7
yp?
ù5
s?￥
V?n?
μ
4[?1
>?ù5
dík? b
7bù5 ?
ù
4??
ù5
s?μB
?
￥ V?n
?
V[?1
>?ù5
5
?
s? s?μ?
<
?^?1
ù5 >?ù5
￥ V?n
?
êù5
!ês?
? Mo ?
!ês?s?
??
?ê4Y ?
??s
?
^B?
M
>
?[
V
?)s
V
??s
V
d ?êù5 d ??
?
??ù
5
!ês?
?
Likert
V
??
?
?êù
5
^B? !ês? ]B? s[
ù M
Mo ?
V ?
V
?lμs
V
??sé b
?
?0
5
Stapel
V
ù5
F?
Guttman BV
V BM
°¤ ??
!ês?
^??￥
?)
¨?
V
???
??$ ??
? 1?
`
?êù
5
V
1?
^B?
1? ??
V
!ês?
M
1.?
￥?
1?
￥ ? ￥?l
?
￥?
^·?
2 T ￥
9 ? ? b
2? ￥ ?
?9
?
VarX
STVar
r
xx
)( +
=
A ?
VarX
VarR
r
xx
?=1
3 ? ?
? ￥ 8ZE
A. ×? ?
Test-retest reliability
×? B? 1W? ? b
()( )
∑∑
∑
??
??
i
i
i
i
i
ii
yyxx
yyxx
22
)()(
S
Schutt￥ è0
One example of how test-retest reliability may be assessed is a study by Linda
Sobell and her associates (1988) of alcohol abusers’ past drinking behavior (using the
Lifetime Drinking History questionnaire) and life changes (using the Recent Life
Changes questionnaire). All 69 subjects in the study were patients in an addition
treatment program. They had not been drinking prior to the interview (determined by
a breath test). The two questionnaires were administered by different interviewers
about two or three weeks apart, both times asking the subjects to recall events eight
years prior to the interviews. Reliability was high: 92% of the subjects reported the
same life events both times and at least 81% of the subjects were classified
consistently at both interviews as having had an alcohol problem or not. When asked
about their inconsistent answers, subjects reported that in the earlier interview they
had simply dated an event incorrectly, misunderstood the question, evaluated the
importance of an event differently, or forgotten an event. Answers to past drinking
questions were less reliable when they were very specific, apparently because the
questions exceeded subjects’ capacities to remember accurately.
B. ˉ' ?
Alternate-form reliability, or Equivalent form reliability
n?M
x
1
x
2
l x
n
?
H ]
Hs
6? n?M
y
1
y
2
l y
n
1
px
i
D y
i
? b
?ì?
^·ù 5 ￥?é¤ í 4¤ í ?
T¤ í b
C. |??
Split-halves reliability
?ü]B ? vù5 ￥ lù5
M
s? ù5 a ù5
? ?a
p?
?ù5
? ? s?￥M1 "
? r
xx
b
D. =?Bá ?
Internal consistent reliability or inter-item reliability
a. 'à Q
Schutt 1999 P.88 ? Tù??¨ ×0ù5
item ??
]B?à
Q *
1
e
ìü?1? o ù5 =? ? p ù5 =?Bá?ù5 b
b. =?Bá ?￥9
?
Cronbach %"
? (r% "
? )￥ 9
?
T1
r% = )(
2
1
2
1
1 s
s
k
k
k
i
i∑
=
?
?
T? s
2
i
^?s?M
X
i
￥BF0ù5
?ês?′￥Zμ b
s
2
^?s?ís￥9Zμ
k
^
?1?
￥M
￥?
? b
E. ?s ??
Inter-observer reliability
?
^ YVE
?N | t | a
Y a té? ?N
H ?
￥×??
V
L? ￥· S b
S
()( )
∑∑
∑
??
??
i
i
i
i
i
ii
yyxx
yyxx
22
)()(
2.?
￥r
validity
1?
￥ r ￥?l
?
￥r
^·?
￥2T ¤ í
? 1?
￥M
￥ ￥? ?
?
￥? ?? b
2r ￥ ?
?9
?
Var X=Var(T)+ Var(S)+Var(R)
?
￥rü?l1
r =
VarX
TVar )(
3 ? ?
r ￥ 8ZE
A. V
?r
V
?r
B. = ? r
Content Validity
= ?rQ?￥
^ ?ùù5
^? -?
$ ?
à Q ￥ Z
? $
? ? ?
?ù5 b
C. 5 r
Criterion Validity
?V B??
?¤ ?￥ s′
M
′ ?D B? ￥°¤?
′ 1? ?
D B? X ü $ ￡
ü
^μr ￥ a ]B ?M
￥?
2 T
V [
^ ?
￥ ?1
?i O1?2T
^
Hü????
?¤?￥M
′
^5μr￥ μ
5r b
?? ￥°¤?
[ "
·Sü?15rS b
?
^B?? ??
s′
·?ù5￥? ??
s?? a
?2T￥1?` b
??r
Predictive Validity ? rS
^? ?′
H ?TB?? ??
￥2T D| ?
??
5′
rS′
^Bá￥ *
1??? ??
ü μ??r b
? ??
￥μ1·S?1??·S b
i?r
Concurrent Validity ?? ??
￥2TD]
H?
￥5s′
á MM1
Hü???? ??
μi?r b
D. Xyr
Construct Validity
1 ? o?
B?à Q
M
A ￥BF p￥μr? ?s??
H
?? ?$
?Tá
ì
??à Q
M
AD
6?B?C`
M
B
^ ￥
?X`
M
B ￥BF?
·S?
àμ$ 1rS bN
H ?T?
A D B
?à Q￥
F 2T
^úM1￥5??
A ￥?BF·S
ù5
^yμ
r￥?
a
^ μXr￥ b
Measurement validity can be established by showing that a measure is related to
a variety of other measures as specified in a theory. This validation approach, known
as Construct validity, is commonly used in social research when no clear criterion
exists for validation purposes.
? SDù???
àμ b#￥
Hyr
^üè¨?￥ b
E. = rD? r
?$ 7y ?TB??
￥ = ?r a5r a2r?
^z￥5
????
￥ r
^z￥ b
?B? = rú￥?
y ?
r9ú
H5????
y ￥ r
^ú￥ b
6. ?
￥?Dr1"￥ è0
?￥ è0 b
)
? ù mú
*￥?
D m ??￥?
?Bú
*￥?
??
￡ ü￥?
?
Bc "D?°
?μ??r￥?
y
ù ??aü1é ?"D?°¨
b
?°ZE? V[s?
v ? B
^ ZE=
^ ZE b
6.1 ?° a"?°D"ZE
6.1.1 ?°D"
?°ü
^
l" ? ￥μ1·S￥·S′
M
′ b
"?°
sampling ' ?°V
?
3￥ ?
3 f ? b
?°"'?8
"
?¤?￥?8
?¤?￥
? ?1 b
"?°￥?1il
^ ?'
HW Z
? V[÷1
V[÷1 b
6.1.2 "ZE
?v ￥e5V98? |?8￥TE?1 " b
è¨￥dà
q"μ "
" a "
" a " b
6.1.3 è¨￥8?"ZE
1sa
s ?"
stratified sampling
^·?v98? ü98?￥
?8s1 b ?a é?e?
?" b
sa"1 p ￥μ 1lBt bB?sa"￥ μ v? μ b
1?
?s ￥E
2?1s ￥E
3Kas ￥E
'Zμs ￥E
2? ?"
cluster sampling
^·|98?￥ò??8?v s
1 ?a[ 1?ê é?
e?
?" b¤" ?￥
é? b
B? ? ?"?
P ￥μsD
lBt7 μ?vμsií^?y
1 ?° b
3"d"
systematic sampling ??1 "
^?v ü98?
￥?8 ?a¨ ZE ?? $?°￥?8 ?v |
?8 b
6.2 è¨￥?°ZE
?°`￥ ??1G ù? # ZE
?°ZE
survey methods ?1μ ?/+ ?
1è?`
telephone interview ZE
2ˉq`
mail interviewZE
3 |?`
personal interview ZE
4è0ˉq
e-mail interview `
5
eè0Z
T￥`
?i 1 ?]￥ ZE??]￥ù5
^??M]￥
?i 2 ?
? '?` ?1 |¤ ￥ a V b
43ZE
observation methods vá V[s1
? bB
^1ù ?43￥Z
E=
^°¤ Y1 ?43￥ZE b
