無条件潜在成長曲線モデル[Mplus]

無条件潜在成長曲線モデル(Latent Growth Curve Model: LGCM)

Higher-Order Growth Curves and Mixture Modeling with Mplus (Multivariate Applications Series)

https://www.routledge.com/Higher-Order-Growth-Curves-and-Mixture-Modeling-with-Mplus-A-Practical/Wickrama-Lee-ONeal-Lorenz/p/book/9781138925151

こちらのFigure 2.3の例。

モデル

IS = 内在化症状 Internalizing Symptoms
I = 第一レベル Initial Level
S = 線形勾配 Linear Slope
Q = 二次勾配 Quadratic Slope
MC = 結婚における不和 Marital Conflict
RV = 恋愛における暴力 Romantic Violence

コード

  TITLE: Unconditional LGCM
  DATA:FILE=example_ch_2.dat;
  VARIABLE:NAMES = IS1-IS5 MC1 RV5;
           USEVAR = IS1-IS5;
           MISSING ARE ALL (9.00);
  
  MODEL: 
  I S | IS1@0 IS2@1 IS3@3 IS4@4 IS5@6; 

  OUTPUT: SAMPSTAT STANDARDIZED;
  PLOT: SERIES=IS1-IS5(S); TYPE=PLOT3;

結果

MODEL RESULTS

                                                    Two-Tailed
                    Estimate       S.E.  Est./S.E.    P-Value

 S        WITH
    I                 -0.016      0.002     -6.587      0.000

 Means
    I                  1.490      0.023     65.804      0.000
    S                 -0.027      0.004     -6.458      0.000

 Variances
    I                  0.165      0.015     10.771      0.000
    S                  0.004      0.001      6.009      0.000

 Residual Variances
    IS1                0.066      0.010      6.348      0.000
    IS2                0.121      0.011     10.872      0.000
    IS3                0.171      0.014     12.356      0.000
    IS4                0.129      0.011     12.218      0.000
    IS5                0.060      0.010      5.702      0.000

余談

著者の名前がこれ。