## Abstract

Structural equation modeling (SEM) based on covariant (CBSEM), in defining a structural model which assumes that it is linear, can be modified to the form of nonlinear structural model that is called the spline truncated SEM ω_{η} = Γ_{ξκ}γ+ξ, for the γ is vector of parameter and a particular matrix Γ_{ξκ} which contains factor scores of exogenous latent variable and knot point. To test the hypothesis of H_{0}: υ′j γ = 0 against H_{0}: υ′j γ ≠ 0, for a vector line υ′_{j} in which the element j is 1 and the other is 0, by using likelihood ratio test (LRT), is obtained the statistical test of hypothesis namely (Formula presented) distributed by t with independent degree s_{2}. Area of rejection H_{0} for this hypothesis fulfills the equation of P(T < -K^{**})=α/2 or P(T > K^{**})=α/2 for a constant K**.

Original language | English |
---|---|

Pages (from-to) | 3775-3786 |

Number of pages | 12 |

Journal | Global Journal of Pure and Applied Mathematics |

Volume | 12 |

Issue number | 4 |

Publication status | Published - 2016 |

## Keywords

- Critical region
- Nonlinear SEM
- Partially statistical test
- Spline truncated