Page 103 - NordicLightAndColour_2012

Basic HTML Version

NORDIC LIGHT & COLOUR
101
Table 2: Mean Illuminance values measured at two different points of the
RE. All values measured each time before the participants evaluated the
rooms.
Virtual Environments:
Tri-dimensional pictures (3D) taken with
two photographic cameras OLYMPUS SP-800UZ with a lens
30xwide. The cameras were placed horizontally at an observer’s
height (1,65 m.) when taking the pictures. One camera captured
a “left eye view” and the second camera captured the “right
eye view”. These pictures have a 4:3 format and 14 megapixels.
The images were opened in a PC using the Stereoscopic Player
software. They were then projected with the help of two High
Definition Projectors of more than 5000 ANSI Lumen on an An-
tipolarized Silver Screen, especial to have a good performance
of 3D imaging. Participants of the experiments made use of
circular polarized glasses to assess the scenes from a distance
not greater than 3 meters (See Figure 5).
Participants:
A total of 8 persons were participants of the
pilot experiment. The participants were recruited via email
and announcements at the Intranet side of NTNU. The sample
consisted of participants with and without architecture training.
They were from different nationalities and cultural background.
The gender of the sample was 4 female and 4 male partici-
pants. The age of the participants ranged between 26 and 62
years old (
M
= 30). Every participant was tested for their vision
prior the experiment to confirm that they did not have particu-
lar vision impairments which could compromise the collected
data. The two vision tests conducted were: a. Luminance Con-
trast Test (Spatial Contrast Sensitivity), via computer software
VigraC
and using the Michelson Plot curve to test 5 different
spatial frequencies and b. Stereoscopic Vision Test, making use
of the
Random Dot 2 – Stereo Acuity Test
. All of the participants
received a free movie ticket to see a movie of their choice for
their cooperation to the experiment. Every participant read a
“General Information Sheet for Participants” with full explana-
tion of the experimental session. After having read the infor-
mation sheet and heard the oral instruction, every participant
signed an individual consent form where they freely consent to
participate in the study.
Experimental Hypothesis:
In order to carry out an experiment
able to throw results that can be compared between the RE and
the VE, an
Experimental Hypothesis
was formulated. This way,
the different variables (independent, dependent and extrane-
ous) can be identified and controlled. By maintaining the same
settings and variables in both VE and RE regarding daylight
and colour, the scores from both environments can be easily
compared.
The Experimental Hypothesis has been defined as:
H
1
= High daylighting level and white wall surfaces can produce a
better visual evaluation of friendliness, complexity and spacious-
ness of a room than low daylighting level and black wall surfaces.
From this hypothesis, the independent and dependent variables
were deduced. Independent variables: Daylight (3 different
daylight levels) and Colour (White and Black wall surfaces). De-
pendent variables: Friendliness, Complexity and Spaciousness.
Making use of a Within-Participant Experimental Design, where
the participants evaluated both VE and RE, it was obtained a
better control over the difference between participants, making
it easier to compare difference between results.
Experimental Procedure:
The comparison between methods
(RE and VE) was carried out making use of a mixed method
approach. The Quantitative part used: Questionnaires with
Semantic Differential Scaling, luminance pictures and Illumi-
nance measurements of two points of the REs. The Qualitative
part used: Open-ended and In-depth interviews after the ques-
tionnaires were answered by the participants. The objective
was to complement the information gathered in the quantitative
part of the experiment.
The pilot experiment was carried out at midday when the
daylight level is at its highest point. The task introduction, along
with written and verbal instruction was given by an assistant
experimenter, who was not completely aware of the critical
aspects of the experiment, in order to avoid giving the partici-
pants too much information that could bias their scores. After
this part, the group of participants was divided in three sub-
groups; where the first group started evaluating the RE – White
Room, the second group started evaluating the RE – Black
Room and the third group started with the evaluation of the VEs
– stereoscopic pictures. The groups then were rotating between
the different stimuli presented until all the stimuli could be
evaluated by all of the participants. By the randomization of
stimuli presentation, bias connected to the Context effect or
sequential contraction bias was controlled. The participants
filled one written questionnaire for each presented stimulus. In
total, there were 12 different stimuli (six in RE and six in VE).