Page 35 - FDMAsia MarApr 2025

TECHNOLOGY 33
www.fdmasia.com | FDM ASIA MAR/APR 2025

(5, 35) = 1.105, p = 0.035 < 0.05, ηp2 = 1.227), whereas form difference in subjective evaluation scores, suggesting whether
(p = 0.225 > 0.05) and interaction (p = 0.217 > 0.05) had no participants had a design-related education affected their
significant effect. preference for chairs.
No significant sex-based differences were found, consistent

Correlation Analysis with previous studies that reported no significant differences
Based on the above results, we chose only different metrics between men and women in terms of preferences.
to determine whether eye movement, physiological responses,
or EEG signals could reflect furniture aesthetics. Multimodal Measurement Results
This study conducted a correlation analysis between subjective Visual fixation is active when assessing internal mental
and objective data. Defining the nine-level evaluation scores representations, and eye movement indicators can be used
as Ordinal Data, all data were processed using Spearman’s as measures of choice and preference.
correlation inspection and the Kendall correlation coefficient test. Based on data analysis, it was concluded that among
The results suggest a weak positive correlation between the five eye movement parameters, only the average pupil

the subjective evaluation score, average pupil diameter diameter and first fixation duration were significant and
(Kendall’s tau-b = 0.044, p = 0.020 < 0.05; Spearman’s rho = 0.060, effective. The chair colour significantly affected the average
p = 0.023 < 0.05), and first fixation duration (Kendall’s tau-b = 0.045, pupil diameter, whereas the interaction between chair colour
p = 0.022 < 0.05; Spearman’s rho = 0.060, p = 0.023 < 0.05). and form significantly affected the first fixation duration.
However, there was no correlation between the subjective In previous studies, there was little research on the form
evaluation score and the SC signal (Kendall’s tau-b = 0.016, and colour of furniture. However, there are references in other
p = 0.404 > 0.05; Spearman’s rho = 0.022, p = 0.411 > 0.05) or design fields. Bradley et al. found that pupil sizes evoked by

the P300 amplitude (Kendall’s tau-b = 0.038, p = 0.327 > 0.05; International Affective Picture System images differed significantly.
Spearman’s rho = 0.057, p = 0.334 > 0.05). Guo, Ding, Liu, Liu, and Zhang reported that first fixation
The results were consistent, although the correlation duration and average pupil diameter could reflect the user
coefficients differed. experience of products when participants browse products freely.
It is possible that emotional responses observed through
Behavioural Result Discussion pupil size changes are also influenced by visual characteristics
Regarding chair colour and form, each element significantly such as colour and contrast.
affected the subjective evaluation scores; however, the Research suggests that these visual factors can affect
interaction between the two was not significantly different for

the behavioural results.
For the overall chair colour, the trend for all subjective
behavioural data means was warm > cool > neutral. These
results support the notion that long-wavelength colours (red
and yellow) are more arousing than other colours, leading to
a linear association between affective tone and wavelength.
In addition, the participants preferred complex to simple

forms, which agrees with the literature. Regarding participants’
sex and specialty, only educational majors showed a significant