The study compared different types of correlation coefficients when analyzing data with continuous latent variables. Polychoric correlation coefficients were found to be more accurate than Pearson and Spearman correlation coefficients in this scenario. The closer the real coefficients were to 1, the more deviation was observed in Pearson and Spearman correlation coefficients. Polychoric correlation coefficients were not affected by rank, while Pearson and Spearman correlation coefficients were influenced by unequally-spaced rank.