Skip to main content
. 2024 Oct 3;3(10):e0000613. doi: 10.1371/journal.pdig.0000613

Table 2. Results of the mixed effect models estimating the fixed effects of group, campus familiarity, and sex on five performance measures.

(a) Wayfinding distance (b) Wayfinding duration (c) Movement speed (d) Number of map views (e) Number of orientation stops
Fixed effects
Est/Beta 95% CI p Est/Beta 95% CI p Est/Beta 95% CI p Est/Beta 95% CI p Est/Beta 95% CI p
(Intercept) 5.368 5.007; 5.729 < .001 5.120 4.771; 5.468 < .001 1.310 1.196; 1.424 < .001 -0.655 -0.384; 0.758 .195 0.187 -0.384; 0.758 .521
Group YA -0.127 -0.230; -0.024 .018 -0.283 -0.420; -0.146 < .001 0.223 0.123; 0.323 < .001 -1.243 -1.470; -0.369 .016 -0.919 -1.470; -0.369 .001
Group SCD 0.071 -0.038; 0.179 .208 0.128 -0.017; 0.273 .088 -0.048 -0.153; 0.056 .371 0.820 0.184; 1.157 .059 0.670 0.184; 1.157 .006
Familiarity -0.003 -0.009; 0.002 .219 -0.008 -0.015; -0.001 .038 0.006 0.001; 0.011 .031 -0.048 -0.033; 0.019 .050 -0.007 -0.033; 0.019 .595
Sex Female 0.102 0.016; 0.188 .023 0.139 0.025; 0.253 .020 -0.054 -0.137; 0.029 .203 0.183 -0.105; 0.707 .618 0.301 -0.105; 0.707 .146
Random effects
Variance SD Variance SD Variance SD Variance SD Variance SD
Participant 0.003 0.055 0.022 0.148 0.028 0.168 1.246 1.116 0.402 0.634
Track 0.154 0.393 0.130 0.361 0.002 0.130 0.003 0.063 0.000 0.000
Model fit
Delta AIC -9.25 -25.34 -23.32 -14.1 -25.94
(Marginal / Pseudo) R2 0.033 0.108 0.277 0.265 0.268

For all models, random intercepts were estimated per participant and track and confidence intervals were calculated using the Wald method. Model equations: performance measure ~ group + familiarity + sex + (1|participant) + (1|track). Performance measures were the (a) log-transformed wayfinding distance; (b) log-transformed wayfinding duration; (c) movement speed, (d) number of map views, and (e) number of orientation stops. P-values for fixed effects in the LME models (a-c) were calculated using the Satterthwaite’s approximation for degrees of freedom. For LME models (a-c) marginal R2 is reported. The GME models (d-e) were calculated using the maximum likelihood estimation and Pseudo R2 was calculated. For (d), a zero-inflated negative binomial distribution and for (e), a zero-inflated poisson distribution of the data was assessed and a log link function applied. Bold font indicates significant effects; italic font indicates statistical trends.