豊田＠早大心理です
吉長 裕司<yosinagy (at) msr.mes.co.jp> さんは書きました：
>吉長＠岡山県立大学です。
>Ａｍｏｓの適合度指標の中にＮＦＩという指標がありますが、
>この指標の使い方がよくわかりません。
>詳しく解説した文献等がありましたら、ご教示いただければ幸いです。
オリジナルペーパーは
Bentler, P. M., & Bonett, D. G. (1980).
Significance tests and goodness of fit in the
analysis of covariance structures.
Psychological Bulletin, 88, 588-606.
です．邦書では
AMOS,EQS,LISRELによるグラフィカル多変量解析 ー目で見る共分散構造分析ー
狩野裕著 １９９７ 現代数学社 p１４９
が，分かり易いので参照なさってください．
Yasuyuki Fukukawa <fukukawa (at) nils.go.jp> さんは書きました：
>長寿医療研究センターの福川と申します．
>２値データの因子分析をしたいのですが，既存の統計ソフトで対応できるのでし
>ょうか．プログラムの組み方など，具体的な方法に関しても，どなたかご存知の
>方がいらしたら教えていただきたく思います．私は現在， SAS，AMOSなどが使え
>る環境下にあります．よろしくお願いいたします．
既存の統計ソフトで対応できますが，AMOS,SASでは実行できません．
LISREL,MPULSで実行するのがベストです(EQSはお勧めしません)．具
体的な方法は，拙著，共分散構造分析ー構造方程式モデリングー[入
門編]朝倉書店 1998 第１２章を御参照下さい．プログラムはLISREL
に関しては付録に掲載されています．MPLUSに関しては下記１２章の
プログラムを参照して下さい．
MPLUSは最近リリースされたソフトです．たいへん良くできたソフトで
す．
http://www.statmodel.com/
でデモバージョンを配っており，上記[入門編]の実習はほとんどデモバ
ージョンで学習できますから，学生さんが多いゼミでも出費がかかり
ません．そこで上記[入門編]に登場するモデルをMPLUSで書いてみました
ので，学習にご利用ください．データはFPR1511で，既に，配っていま
すので省略します．このため今回は10Kで，前回の半分です．データは
ブランクで区切るのがデフォルトです．詳しくは上記サイトのサンプル
プログラムの入力データを参照して下さい．営利目的でなく学習に使用
する限りにおいて再配布自由（奨励）とします．
---------ここから最後まで
TITLE: 第３章 単回帰重回帰組み合わせ 趣味の印象
DATA: FILE IS h:\Mplus\DAT\ND01.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=30;
VARIABLE: NAMES ARE Z2 Z3 Z1;
MODEL: Z2 ON Z1;
Z3 ON Z1 Z2;
OUTPUT: SAMP STAND RES;
TITLE: 第４章 確認的因子分析モデル趣味の印象
DATA: FILE IS h:\Mplus\DAT\self.dat;
VARIABLE: NAMES ARE V1-V6;
MODEL: F1 BY V1* V2*(1); F2 BY V3* V4*(2); F3 BY V5* V6*(3);
F1 (at) 1 F2 (at) 1 F3 (at) 1; V1* V2*(4); V3* V4*(5); V5* V6*(6);
OUTPUT: STANDARDIZED
TITLE: 第４章 因子数１の因子分析 記憶テスト
DATA: FILE IS h:\Mplus\DAT\ND02.dat;
TYPE IS COVARIANCE;
NOBSERVATIONS=130;
VARIABLE: NAMES ARE V1-V3;
MODEL: F1 BY V1* V2* V3*; F1 (at) 1;
OUTPUT: SAMP STAND RES;
TITLE: 第４章 ５つの学力テストモデル
DATA: FILE IS h:\Mplus\DAT\ND03.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V1-V5;
MODEL: F1 BY V1* V3* V4*; F2 BY V2* V3* V4* V5*;
F1 (at) 1 F2 (at) 1; F1 WITH F2; V1 WITH V2;
OUTPUT: SAMP STAND RES;
TITLE: 第５章 連結モデル
DATA: FILE IS h:\Mplus\DAT\ND04.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=432;
VARIABLE: NAMES ARE Z2 Z3 Z1;
MODEL: Z2 ON Z1;
Z3 ON Z2;
OUTPUT: SAMP STAND RES;
TITLE: 第５章 逐次モデル 主観客観的社会地位
DATA: FILE IS h:\Mplus\DAT\ND05.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=432;
VARIABLE: NAMES ARE Z4 Z3 Z2 Z1;
MODEL: Z3 ON Z1 Z2;
Z4 ON Z2 Z3;
OUTPUT: SAMP STAND RES;
TITLE: 第５章 完全逐次モデル 主観客観的社会地位
DATA: FILE IS h:\Mplus\DAT\ND05.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=432;
VARIABLE: NAMES ARE Z4 Z3 Z2 Z1;
MODEL: Z2 ON Z1;
Z3 ON Z1 Z2;
Z4 ON Z1 Z2 Z3;
OUTPUT: SAMP STAND RES;
TITLE: 第５章 多変量回帰分析 主観客観的社会地位
DATA: FILE IS h:\Mplus\DAT\ND05.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=432;
VARIABLE: NAMES ARE Z4 Z3 Z2 Z1;
MODEL: Z3 ON Z1 Z2;
Z4 ON Z1 Z2;
Z3 WITH Z4
OUTPUT: SAMP STAND RES;
TITLE: 第５章 非逐次モデル 主観客観的社会地位
DATA: FILE IS h:\Mplus\DAT\ND06.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=432;
VARIABLE: NAMES ARE Z5 Z4 Z3 Z2 Z1;
MODEL: Z3 ON Z1 Z4;
Z4 ON Z2 Z3;
Z5 ON Z3 Z4;
Z3 WITH Z4 Z5;
Z4 WITH Z5;
OUTPUT: SAMP STAND RES;
TITLE: 第５章 外生的観測変数と誤差変数の相関モデル
DATA: FILE IS h:\Mplus\DAT\ND06.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=432;
VARIABLE: NAMES ARE Z5 Z4 Z3 Z2 Z1;
MODEL: Z3 ON Z1;
Z4 ON Z2;
Z5 ON Z3 Z4;
Z3 WITH Z2;
Z4 WITH Z1;
OUTPUT: SAMP STAND RES;
TITLE: 第５章 階数条件を使えない非逐次モデル
DATA: FILE IS h:\Mplus\DAT\ND06.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=432;
VARIABLE: NAMES ARE Z5 Z4 Z3 Z2 Z1;
MODEL: Z3 ON Z1;
Z4 ON Z2 Z3 Z5;
Z5 ON Z3 Z4;
OUTPUT: SAMP STAND RES;
TITLE: 第６章 趣味に対する印象 −分析２−
DATA: FILE IS h:\Mplus\DAT\self.dat;
TYPE IS IND;
NOBSERVATIONS=30;
VARIABLE: NAMES ARE X1-X6;
MODEL: F1 BY X1 (at) 1 X2 (at) 1;
F2 BY X3 (at) 1 X4 (at) 1;
F3 BY X5 (at) 1 X6 (at) 1;
F2 ON F1; F3 ON F2 F1;
F1* ;
X1 X2(1);
X3 X4(2);
X5 X6(3);
OUTPUT: SAMP STAND RES;
TITLE: 第６章 ２次因子分析
DATA: FILE IS h:\Mplus\DAT\ND07.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V1-V6;
MODEL: F1 BY V1 V2;
F2 BY V3 V4;
F3 BY V5 V6;
F4 BY F1 F2 F3;
OUTPUT: SAMP STAND RES;
TITLE: 第６章 MIMICモデル
DATA: FILE IS h:\Mplus\DAT\ND08.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V4 V5 V6 V1 V2 V3;
MODEL: F1 ON V1 V2 V3;
F1 BY V4 V5 V6;
OUTPUT: SAMP STAND RES;
TITLE: 第６章 縦断的モデル１
DATA: FILE IS h:\Mplus\DAT\ND09.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V1-V6;
MODEL: F1 BY V1 V2;
F2 BY V3 V4;
F3 BY V5 V6;
F3 ON F2;
F2 ON F1;
OUTPUT: SAMP STAND RES;
TITLE: 第６章 縦断的モデル１
DATA: FILE IS h:\Mplus\DAT\ND09.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V1-V6;
MODEL: F1 BY V1 (at) 1 V2 (at) 1;
F2 BY V3 (at) 1 V4 (at) 1;
F3 BY V5 (at) 1 V6 (at) 1;
F4 BY V1 V3 V5;
F5 BY V2 V4 V6;
F3 ON F2;
F2 ON F1;
F4 WITH F1 (at) 0 F5 (at) 0;
F1 WITH F5 (at) 0;
OUTPUT: SAMP STAND RES;
TITLE: 第６章 シンプレックス構造分析
DATA: FILE IS h:\Mplus\DAT\ND10.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V1-V5;
MODEL: F1 BY V1 (at) 1;
F2 BY V2 (at) 1;
F3 BY V3 (at) 1;
F4 BY V4 (at) 1;
F5 BY V5 (at) 1;
F5 ON F4;
F4 ON F3;
F3 ON F2;
F2 ON F1;
V1-V5(1);
OUTPUT: SAMP STAND RES;
TITLE: 第６章 図６．１２ 問題２
DATA: FILE IS h:\Mplus\DAT\ND11.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V1-V6;
MODEL: F1 BY V1*0.9 V2*-.8;
F2 BY V3*0.9 V4*-.8;
F3 BY V5*-.8 V6 (at) 1;
F3 ON F2*-.6 F1*-.4;
F1 (at) 1 F2 (at) 1;
OUTPUT: SAMP STAND RES;
TITLE: 第８章 Harman H. H.の不適解
DATA: FILE IS h:\Mplus\DAT\ND12.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V1-V5;
MODEL: F1 BY V1-V5;
OUTPUT: SAMP STAND RES;
TITLE: 第８章 問題
DATA: FILE IS h:\Mplus\DAT\ND13.dat;
TYPE IS CORRELATION;
NOBSERVATIONS=100;
ANALYSIS: ESTIMATOR=ML;ITERATIONS=500;
CONVERGENCE=0.000005;
VARIABLE: NAMES ARE V1-V4;
MODEL: F1 BY V1*0.8 V2*0.7 V3*0.6 V4*0.5;
OUTPUT: SAMP STAND RES TECH5;
TITLE: 第12章 カテゴリカル因子分析
DATA: FILE IS h:\Mplus\DAT\D1201.dat;
TYPE IS IND;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE V1-V5 U3-U5;
USEVARIABLES ARE V1 V2 U3-U5;
CATEGORICAL ARE U3-U5;
MODEL: F1 BY V1* V2* U3* U4* U5*;
F1 (at) 1;
OUTPUT: SAMP STAND RES TECH5;
TITLE: 第12章 2母数正規累積モデル
DATA: FILE IS h:\Mplus\DAT\IRT.dat;
FORMAT IS (9F1.0);
TYPE IS IND;
NOBSERVATIONS=226;
ANALYSIS: ESTIMATOR=WLS;
VARIABLE: NAMES ARE U1-U9;
CATEGORICAL ARE U1-U9;
MODEL: F1 BY U1* U2* U3* U4* U5* U6*
U7* U8* U9*;
F1 (at) 1;
OUTPUT: SAMP STAND RES TECH5;
TITLE: 第12章 プロビット回帰分析
DATA: FILE IS h:\Mplus\DAT\D1201.dat;
TYPE IS IND;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE X1-X5 U3-U5;
USEVARIABLES ARE Z1 Z2 U3;
CATEGORICAL ARE U3;
DEFINE: Z1=X1/100;Z2=X2/100;
MODEL: F1 BY Z1 (at) 1; F2 BY Z2 (at) 1;
F3 ON F1* F2*;
F3 BY U3;
Z1 (at) 0 Z2 (at) 0 F3 (at) 0;
OUTPUT: SAMP STAND RES TECH5;
TITLE: 第12章 ロジスティック回帰分析
DATA: FILE IS h:\Mplus\DAT\D1201.dat;
TYPE IS IND;
NOBSERVATIONS=100;
VARIABLE: NAMES ARE X1-X5 U3-U5;
USEVARIABLES ARE Z1 Z2 U3;
CATEGORICAL ARE U3;
DEFINE: Z1=X1/100;Z2=X2/100;
ANALYSIS: TYPE=LOG;
MODEL: U3 ON Z1 Z2;
TITLE: 第13章 趣味の印象の因子の平均とＳＤ
DATA: FILE IS h:\Mplus\DAT\SELF.dat;
TYPE IS IND;
NOBSERVATIONS=30;
VARIABLE: NAMES ARE X1-X6;
ANALYSIS: TYPE=MEANSTRUCTURE;
MODEL: F1 BY X1*(1);F1 BY X2*(2);
F2 BY X3*(1);F2 BY X4*(2);
F3 BY X5*(1);F3 BY X6*(2);
F1 (at) 1 F2* F3*;
F1 WITH F2 F3;
F2 WITH F3;
[F1 (at) 0 F2* F3*];
[X1 X3 X5](3);[X2 X4 X6](4);
OUTPUT: SAMP STAND RES TECH5;
TITLE: 第13章 知能検査の縦断データの分析
DATA: FILE IS h:\Mplus\DAT\ND14.dat;
TYPE IS MEANS STD CORR;
NOBSERVATIONS=1472;
VARIABLE: NAMES ARE X1-X8;
ANALYSIS: TYPE=MEANSTRUCTURE;
MODEL: F1 BY X1*(1);F1 BY X2*(2);F1 BY X3*(3);F1 BY X4*(4);
F2 BY X5*(1);F2 BY X6*(2);F2 BY X7*(3);F2 BY X8*(4);
F2 ON F1;
F1 (at) 1 F2*;
[F1 (at) 0 F2* ];
X1 X2 X3 X4 PWITH X5 X6 X7 X8;
X1 X5(5);X2 X6(6);X3 X7(7);X4 X8(8);
[X1 X5](9);[X2 X6](10);[X3 X7](11);[X4 X8](12);
OUTPUT: SAMP STAND;
TITLE: 第14章 偏差値模擬試験
DATA: FILE IS h:\Mplus\DAT\ND15.dat;
TYPE IS MEANS STD CORR;
NOBSERVATIONS=316 609;
NGROUPS=2;
VARIABLE: NAMES ARE X1-X6;
ANALYSIS: TYPE=MEANSTRUCTURE;
MODEL: F1 BY X1* X2* X3*;
F2 BY X4* X5* X6*;
MODEL g1:F1 (at) 1 F2 (at) 1;X1-X6;
[F1 (at) 0 F2 (at) 0];
OUTPUT: SAMP STAND;
TITLE: 第14章 不完全データの分析
DATA: FILE IS h:\Mplus\DAT\ND16.dat;
TYPE IS MEANS STD CORR;
NOBSERVATIONS=170 200 270 130;
NGROUPS=4;
VARIABLE: NAMES ARE X1-X6;
ANALYSIS: TYPE=MEANSTRUCTURE;
MODEL:F1 BY X1-X3;F2 BY X4-X6;[F1 (at) 0 F2 (at) 0];
F1 (at) 1 F2 (at) 1;F1 WITH F2*0.6;
MODEL g1:F1 BY X1*0.9(13);F1 BY X2*1.1(14);
F1 BY X3*1.2(15);F2 BY X4*1.4(16); F2 BY X5*1.5(17);
F2 BY X6*1.6(18);F1 WITH F2*(37);X1*1.1(1);X2*0.8(2);
X3*0.5(3);X4*2.0(4);X5*1.4(5);X6*1.3(6);[X1*5.8](25);
[X2*4.8](26);[X3*3.8](27);[X4*5.9](28); [X5*4.9](29);
[X6*3.8](30);
MODEL g2:F1 BY X1*(16); F1 BY X2*(17);F1 BY X3*(18);
F2 BY X4*1.2(19); F2 BY X5*1.2(20);F2 BY X6*1.2(21);
F1 WITH F2*(37); X1*(4); X2*(5); X3*(6); X4*1.1(7);
X5*0.8(8); X6*0.7(9);[X1*](28);[X2*](29); [X3*](30);
[X4*5.9](31);[X5*5.0](32);[X6*3.9](33);
MODEL g3:F1 BY X1*(19); F1 BY X2*(20); F1 BY X3*(21);
F2 BY X4*1.3(22); F2 BY X5*1.6(23);F2 BY X6*1.7(24);
F1 WITH F2*(37); X1*(7); X2*(8); X3*(9); X4*2.0(10);
X5*1.7(11);X6*1.1(12);[X1*](31);[X2*](32);[X3*](33);
[X4*6.0](34);[X5*4.8](35);[X6*3.9](36);
MODEL g4:F1 BY X1*(13);F1 BY X2*(14);F1 BY X3*(15);
F2 BY X4*(22);F2 BY X5*(23);F2 BY X6*(24);
F1 WITH F2*(37);X1*(1);X2*(2);X3*(3);X4*(10);X5*(11);
X6*(12);[X1*](25);[X2*](26);[X3*](27);[X4*](34);
[X5*](35);[X6*](36);
OUTPUT: SAMP STAND;
TITLE: 第14章 行動遺伝学ACEモデル(MZ,DZの順に入力)
DATA: FILE IS h:\Mplus\DAT\ND17.dat;
TYPE IS COVARIANCE;
NOBSERVATIONS=171 101;
NGROUPS=2;
VARIABLE: NAMES ARE V1-V2;
MODEL:F1 BY V1*(1); F2 BY V2*(1); F3 BY V1*(2);
F3 BY V2*(2);F4 BY V1*(3);F5 BY V2*(3);V1 (at) 0 V2 (at) 0;
F1 (at) 1 F2 (at) 1 F3 (at) 1 F4 (at) 1 F5 (at) 1; F1 F2 WITH F3 (at) 0 F4 (at) 0 F5 (at) 0;
F3 WITH F4 (at) 0 F5 (at) 0; F4 WITH F5 (at) 0; F1 WITH F2;
MODEL g1:F1 WITH F2 (at) 1;
MODEL g2:F1 WITH F2 (at) 0.5;
OUTPUT: SAMP STAND;
--
--------------------------------------------------------------------------
TOYODA Hideki Ph.D., Associate Professor, Department of Psychology
TEL +81-3-5286-3567 School of Lieterature, Waseda University
toyoda (at) mn.waseda.ac.jp 1-24-1 Toyama Shinjyuku-ku, Tokyo 162-8644 Japan
-----------------------------------------------------------------------
ここは心理学研究の基礎メーリングリストに投稿された過去の記事を掲載しているページです。