Learn R Programming

LCPA (version 1.0.0)

LRT.test.VLMR: Lo-Mendell-Rubin likelihood ratio test

Description

Implements the ad-hoc adjusted likelihood ratio test (LRT) described in Formula 15 of Lo, Mendell, & Rubin (2001), also known as VLMR LRT (Vuong-Lo-Mendell-Rubin Adjusted LRT). This method is typically used to compare models with L-1 and L classes. If the difference in the number of classes exceeds 1, conclusions should be interpreted with extreme caution.

Usage

LRT.test.VLMR(object1, object2)

Value

An object of class "htest" containing:

  • statistic: VLMR adjusted test statistic

  • parameter: Degrees of freedom (\(df = npar_2 - npar_1\))

  • p.value: P-value from \(\chi^2_df\) distribution

  • method: Name of the test

  • data.name: Model comparison description

Arguments

object1

Fitted model object with fewer parameters (i.e., fewer npar, small model).

object2

Fitted model object with more parameters (i.e., more npar, large model).

Details

Note that since the small model may be nested within the large model, the result of LRT.test.VLMR may not be accurate and is provided for reference only. More reliable conclusions should be based on a combination of fit indices (i.e., get.fit.index), classification accuracy measures (i.e., get.entropy, get.AvePP), and a bootstrapped likelihood-ratio test (i.e., BLRT, LRT.test.Bootstrap, which is very time-consuming). Above all and the most important criterion, is that the better model is the one that aligns with theoretical expectations and offers clear interpretability.

The LRT.test.VLMR statistic is defined as: $$VLMR = \frac{LRT}{c} \quad \text{where} \quad c = 1 + \frac{1}{df \cdot \log(N)}$$ where:

  • \(LRT\) is the standard likelihood ratio statistic. see LRT.test

  • \(df = npar_2 - npar_1\) is the difference in number of free parameters between models.

  • \(N\) is the sample size.

Under the null hypothesis (H_0: small model is true), VLMR asymptotically follows a chi-square distribution with \(df\) degrees of freedom.

References

Lo, Y., Mendell, N. R., & Rubin, D. B. (2001). Testing the number of components in a normal mixture. Biometrika, 88(3), 767-778. https://doi.org/10.1093/biomet/88.3.767

Nylund-Gibson, K., & Choi, A. Y. (2018). Ten frequently asked questions about latent class analysis. Translational Issues in Psychological Science, 4(4), 440-461. https://doi.org/10.1037/tps0000176