Table 2.

Bayesian Linear Mixed Model Characteristics Using Varied Random Effect Distributions

	Population Intercept (β0) (dB), Mean ± SE	Eye-Specific Intercepts (dB), Median (IQR) [Range]	Population Slope (β1) (dB/y), Mean ± SE	Eye-Specific Slope (dB/y), Median (IQR) [Range]	WAIC
Gaussian	–3.93 ± 0.06	–2.24 (–5.37 to –0.66) [–26.38 to 5.07]	–0.15 ± 0.01	–0.09 (–0.26 to 0.01) [–2.89 to 1.55]	218,562
Student's t	–1.87 ± 0.04	−2.15 (–5.15 to –0.79) [–27.75 to 5.87]	–0.07 ± 0.01	–0.07 (–0.19 to –0.01) [–6.93 to 6.01]	215,711
Log-gamma	–3.80 ± 0.06	–2.10 (–5.19 to –0.66) [–27.47 to 3.40]	–0.29 ± 0.01	–0.07 (–0.29 to 0.07) [–6.35 to 1.86]	214,188

Eye-specific intercepts and slopes are estimates derived from the posterior distributions and include both fixed and random effect components. Boldface indicates lowest WAIC, demonstrating best model fit.