CES-D抑うつ尺度の心理測定法的特性 - jartest.jp

1 ( 2004, )1 35 CES-D ( 2004, )2 35 ( ) 10 ( 2004, )3 35 ( 2004, )4 35 NIMH-DIS(Diagnostic Interview Schedule) CIDI(Composite International Diagnostic Interview) ( ) (( )) /Inventory CESCES--DD(Center for Epidemiologic StudiesDepression Scale) GHQ(General Health Questionnaire) STAI(State-Trait Anxiety Inventory) ( ) ( 2004, )5 35 negativenegative positivepositive ( ) ( , ) ( ) ( , ) ( yea-saying ) : ( 2004, )6 35 CES-D NIMH CES (Center for Epidemiologic Studies) BDI SDS high riskhigh risk 16 16 ((symptomsymptom--present) present) ( , , ) 44 ((symptomsymptom--absent) absent) ( ) 30 ( 2004, )7 35 CES-D.

2 ( ) : ( ) ( 2004, )8 35 CES-D : ( ): ( 2004, )9 35 More About the CES-D : : (1 )(1 2 )(3 4 )(5 7 ) 0123 3210 : 0 60, (cut-off): 15/16 ( 2004, )10 35 CES-D (N 2,100) ( ), ( ) (Iwata , 1989) ( ) ( ), ( ) (Aneshensel ,1981) ( ) ( ) (Vega , 1987;Mendes de Leon & Markides, 1988) score ( ) ( ) ( 2004, )11 35 CES-D 1) ?

3 Overwork? ?2) ? ? ? ( 2004, )12 35 1) CES-D back-translation OK2) Differential Item FunctioningDifferential Item Functioning ( : DIF) ? ( 2004, )13 35 54 ( )44 ( ) 314 ( )324 ( ) : 51 ( )51 ( )55-63 93 ( )93 ( )45-54 98 ( )98 ( )35-44 126 ( )126 ( ) : 25-34 368 368 Iwata (2002) ( 2004, )14 35 IRT-DIF Study Differential Item Functioning 2PL Model: P( )=1/(1+ ( -b)) IRT likelihood ratio DIF detection method Likert 2 : 0-1-2-3 0-1-1-1 (1 )(1 2 )(3 4 )(5 7 )01111110 Negative Positive ( 2004, )15 35 Jpn vs USAD iscrimination: vs (NS)Difficulty: vs (p<.

4 0001) bJPbUSItem Characteristic CurveDifferential Item Functioning (DIF) (Latent Trait: Depressive Level)Probability of Endorsementa: Slope (Discrimination)b: Threshold (Difficulty)Probability of EndorsementP( )#8 (POS)"(Not) Hopeful"Depressive Level ( : latent trait)2 Parameter Logistic Model:P( )=1/(1+ ( -b)) ( 2004, )16 35 The Purified (anchor) SubtestMantel-Haenszel (MH-DIF) ( ) non-MH-DIF 5 7 subset subset IRT G2free(a, b ) 2 a, b G2constrained G2constrained G2free subset invariance G2 non-MH-DIF G2 STOP ( step )Iwata (2002) ( 2004, )17 35 DIF DetectionPurified subset (constrained) 1 : G2free aaaa : G2a bbbb : G2bTest H0: aaaaUS= aaaaJpnG2a G2free (slope ) bbbb aaaa : G2bSlope(a) DIF not b b b b : : : : Non-DIFa:a:a:a: Non-DIFb:b:b:b: DIFnot b b b b : : : : DIFa:a:a:a: DIFb:b:b:b: Non-DIFTest H0: bbbbUS= bbbbJpnG2b-G2aTest H0.

5 BbbbUS= bbbbJpnG2b-G2freeIwata (2002) ( 2004, )18 35 CES-D -2PL-IRT - DIF Scale/Items 3 6 **9 **10 * 14 17 purified18 ** 15 purified19 Iwata (2002) ( 2004, )19 35 CES-D -2PL-IRT - 1 **2 **5 7 **13 purified20 ** 4 **8 **12 **16 ** DIF Scale/Items Iwata (2002) ( 2004, )20 35 JapaneseUSJpn vs USDiscrimination: vs : vs Characteristic Level (Latent Trait)Probability of EndorsementDepressed Affect Depressed Affect Depressed Affect Depressed Affect ---- on average on average on average on average vs : vs : vs ( )Iwata (2002) ( 2004, )21 35 Item Characteristic Level (Latent Trait)Probability of EndorsementInterpersonal Relations Interpersonal Relations Interpersonal Relations Interpersonal Relations ---- on average on average on average on average vs : vs.

6 Vs ( )Iwata (2002) ( 2004, )22 35 Item Characteristic Level (Latent Trait)Probability of EndorsementSomatic & Retarded Activities Somatic & Retarded Activities Somatic & Retarded Activities Somatic & Retarded Activities ---- on average on average on average on average vs : vs : vs ( )Iwata (2002) ( 2004, )23 35 Item Characteristic Level (Latent Trait)Probability of Endorsement((((Lack of) Positive Affect Lack of) Positive Affect Lack of) Positive Affect Lack of) Positive Affect ---- on average on average on average on average vs : vs : vs ( )Iwata (2002) ( 2004, )24 35 IRT-DIF ( ) ( 2004, )25 35 1) CES-D 2) comparability ( 2004, )26 35 vs (%) 4 582258ns23ns8 531451ns24**12 612264ns21ns16 511562ns31** 4 9243** 29 **8 10344** 21 **12 13338** 17 **16 10246** 28 ** (n=528) (n=85)Scale/Items Iwata et al.

7 (1998)Demographically-matched samples (controls:depressed = 3:1) ( 2004, )27 35 Negative (Depressed) Affect Items0204060 Rarely or None ofthe TimeSome or a Little ofthe TimeOccasionally or aModerate Amountof the TimeMost or All of theTime ControlsPatientsPositive Affect Items0204060 Most or All of theTimeOccasionally or aModerate Amountof the TimeSome or a Little ofthe TimeRarely or None ofthe Time ControlsPatients vs Negatively Revised Items0204060 Rarely or None ofthe TimeSome or a Little ofthe TimeOccasionally or aModerate Amountof the TimeMost or All of theTime ControlsPatientsSomatic & Retarded Activities Items0204060 Rarely or None ofthe TimeSome or a Little ofthe TimeOccasionally or aModerate Amountof the TimeMost or All of theTime ControlsPatientsIwata et al.

8 (1998) ( 2004, )28 35 1) Iwata (Kim et al., 2000) 2) westernizing ( 2004, )29 35 Negative (Depressed) Affect Items020406080 Rarely or None ofthe TimeSome or a Littleof the TimeOccasionally or aModerate Amountof the TimeMost or All of theTime Issei-NiseiSansei-Somatic & Retarded Activities Items0204060 Rarely or None ofthe TimeSome or a Littleof the TimeOccasionally or aModerate Amountof the TimeMost or All of theTime Positive Affect Items0204060 Most or All of theTimeOccasionally or aModerate Amountof the TimeSome or a Littleof the TimeRarely or None ofthe Time Issei-NiseiSansei-Negatively Revised Items020406080 Rarely or None ofthe TimeSome or a Little ofthe TimeOccasionally or aModerate Amountof the TimeMost or All of theTime 1 2 vs 3 4 ( ) ( 2004, )30 35 ( ) ( )

9 ** **(15/16)( ) ( ) ( ) ( ) ** (12/13)( ) ( ) ( ) ( )** p < .001 ( )(95% CI)Iwata & Buka (2002) : DIF ( 2004, )31 35 (CES-D) 4 2 7 7 r > .4r .1 USAUSA 9 + 4 5 + 2 JPNJPN Iwata & Roberts (1996) ( 2004, )32 35 (CES-D) r .0 r .24 10 + 2 4 + 4 12 + + 2 3 2 Iwata & Roberts (1996) ( 2004, )33 35 Collectivistic Society Confucian Ethics (similar in East Asia) Self Criticism in Japan vs Self Enhancement in the .. any other factors ( 2004, )34 35 STAI (J Clin Psychol (Iwata et al, 2000)) USAUSA JPNJPN r 0 & ( 2004, )35 35 / Americanizing (100 )