Skip to main content
. 2023 Sep 27;14:1222608. doi: 10.3389/fpsyg.2023.1222608

Table 5.

Generalized linear mixed model (Model 3) was fitted to the (ordinal) personal resonance ratings of the visual artworks and complex images (N = 30), with “viewing condition” as a fixed effect and “Image” as a random effects grouping factor.

Model 3: Effects of different viewing conditions on personal resonance with visual artworks and complex images in healthy adults
Effect Df ChiSq p VS-MPR *
Viewing condition 2 0.401 0.818 1.000
Estimated marginal means
95% CI
Viewing condition Estimate SE Lower Upper
Contextual information 2.586 0.160 2.272 2.900
External perspective 2.720 0.160 2.406 3.034
Internal perspective 2.678 0.160 2.364 2.992

Two observations were removed due to missing values.

Generalized linear mixed model with Gaussian family and identity link function.

Model terms tested with likelihood ratio tests Method.

The following variable is used as a random effects grouping factor: “Image” (N = 30).

Type III sum of squares.

*Vovk–Sellke maximum p-ratio: Based on a two-sided p-value, the maximum possible odds in favor of H1 over H0 equal 1/(-e p log(p)) for p ≤ 0.37 (Sellke et al., 2001).

The estimated marginal means below provide estimates for each factor level/design cell and their differences.

No effect of viewing condition on resonance ratings was found, and a control analysis showed that there were neither effects of age nor sex; hence, the presented results were averaged over all research participants (N = 36). Total observations: 1,078. A rating of 1 indicated a strong resonance with the artwork or complex image, a rating of 3 indicated a neutral response and a rating of 5 indicated very little to no resonance with the artwork or complex image. The estimated marginal means suggest that participants on average felt some resonance with the artworks and complex images that were selected as stimuli in this study but not particularly strongly. The different viewing conditions had no influence on this.

One participant from the young adults cohort was excluded from the analysis because their resonance ratings were much lower than average across all three viewing conditions (>2 std dev), possibly suggestive of an indiscriminate low engagement with and/or dislike of the image selection.