Models of colour semiotics

This thesis is concerned with the relationship between
colour and certain bi-­‐polar characteristics. This
topic has previously been studied through highly controlled
laboratory-based experiments but this thesis starts from
the question of whether experiments conducted in the
laboratory (which are necessarily constrained to have low
numbers of participants) can ever hope to capture the full
complexity of the relationships being studied, since there
are likely to be strong cultural and regional differences.
The key advance in this work therefore was to explore the
use of a web-­based experiment for collecting data on a large
scale and from all over the world. A laboratory-­‐based
experiment to explore colour semiotics was carried out and
broadly supports the earlier work carried out by Ou et
al. A novel paradigm for carrying out colour semiotic
experiments based on a large-­scale internet presentation
and distributed over large numbers of participants (over
2000 from 58 countries) was then conceived and implemented.
Comparison with the laboratory-­based experiment broadly
validated the use of this new paradigm. The large amount
of data collected allowed an analysis of gender and
cultural differences to be carried out and it was shown
that cultural and age may be significant factors but
that gender is probably not. The thesis has made a contribution in terms of collecting new data, generating
new models, and testing a web-­based paradigm for carrying
out colour-­‐based experiments. One application of the
colour-­‐semiotic models that has been developed at the end
of this thesis is in the design process and a potential
new software tool that could build a bridge between science
and design has been considered.