IEEE Consumer Electronics Magazine - July 2018 - 51

sEnsE of EmBarrassmEnt
Out of the 18 individuals who were assigned to group A, only
one female confirmed a sense of embarrassment. That number
increased by a rate of nine times for each of the other two
groups. Overall, 18 out of the 36 participants of groups B and
C stated feeling discomfort, with most of them communicating feelings of embarrassment, of being uncomfortable, or
awkwardness. Sixteen stated that the source of their discomfort was the salesman, without being directly asked to point to
the salesman as their source of discomfort. Two male participants, who both confirmed being present in the VE, identified
the female researcher, who was in the room with them during
the experiment, as the main source of their discomfort. The
same female from the first group was the only participant to
also name the researcher as an additional source of her feeling
embarrassed. Overall, out of the 19 participants who experienced a sense of embarrassment during the moment of the
VB's exposure, only two males indicated the researcher as the
sole reason for their embarrassment.
In total, during the moment of the VB's exposure, 19 out of
the 54 participants experienced negative emotions such as
embarrassment, discomfort, and awkwardness. At least eight
individuals expressed positive emotions, such as amusement, a
sense of comfort, or a sense of freedom. The remaining 27 used
terms such as normal or remained completely neutral, stating
that they had no emotion toward the state of the naked VB; 12
of those were in group A, where there was no other character
present. Additionally, 12 out of the 27 who did not express any
emotion toward the naked VB had chosen the DrM (five in
group A, five in group B, and two in group C).

statistical significancE of thE rEsults
The statistical analysis of the collected questionnaires' data
indicated that the statistical significance of our results is, in
most cases, above the typical significance level of 5%
( p 2 0.05); i.e., the results may not be regarded as statistically
significant (on average, we have a yield of p = 0.26 ). Indeed,
this was expected, primarily due to the size of our statistical
sample (54 participants divided into three groups).
Nevertheless, the validity of our statements is supported
by qualitative data compiled from the semistructured followup interviews, wherein several questions were intentionally
articulated to cross-check the respective questionnaires' findings (e.g., participants were encouraged to explain their
behavioral choices while in the VE).

conclusion
The most important outcome of this study is that the majority
of participants assigned the expected social context to the
clothing store. This conclusion is drawn based on the DrM versus CrM results, where, in total, 36 participants were placed
under DrM. The decision to choose the DrM was not solely
attributed to a sense of embarrassment or the participants'
intention to conceal the process of changing clothes. The "outof-habit" and the "because it is reasonable" responses also
assigned a specific social context to the VE and subsequently

In total, during the moment of
the VB's exposure, 19 out of the 54
participants experienced negative
emotions such as embarrassment,
discomfort, and awkwardness.
to the virtual experience itself. Based on these results, we considered that individuals were prone to handling (or challenging) information or events in VR according to their existing
social skills and knowledge.
Additionally, participants from groups B and C experienced negative emotions toward the naked VB, by a rate of
nine times more than group A, which indicates that users felt
more compelled to abide by a specific social context when
others were present or engaged in the interaction. Considering that the mere existence of the salesman did not affect the
levels of presence and BOI, his affective impact was associated with the level of social presence, whereas both the sense
of embarrassment as well as the increased levels of presence
during the exposure of the VB were related to other aspects
of the social context of the VE; namely, the public nature of
the clothing store. Participants' references to "other people/
customers" were also indicative of the social context's effect.
An unexpected outcome of this study relates to the juxtaposition of all of the relevant indicators between groups B
and C; participants did not demonstrate any acknowledgment
that the salesman was a real actor. A few even expressed surprise at how intelligent the avatar was for providing real-time
answers, yet none of the participants attempted to initiate an
actual conversation or test the realness of the salesman. Had
they tried to do so, the researcher was prepared to engage in
conversation. We assume that they considered the probability
of another user sharing the VR environment implausible, and
as such, it was more sensible to perceive all verbal statements
and movements as prerecorded or programmed.
The salesman is one of the components that defined the social
context of the VE. There was no need for participants to realize
the existence of a real actor for this component to be still applicable, as it was for group B. It would be unwise to speculate what
sorts of results group C would have produced, had those participants realized the experimental condition in full; after all, we
wished to observe whether participants would acknowledge the
existence of a real actor on their own, based on his realistic movements and naturally delivered verbal instructions. The fact that
they did not might be partially related to the task-oriented nature
of their experience, i.e., a different experimental setup might
encourage participants to pay closer attention to the nature of the
salesman. Nonetheless, this does raise the question as to how
easy it will be in future applications to determine whether an avatar is controlled by a real user, or an intelligent agent.
Various studies have produced interesting findings in regard
to the affective impact of BOI [9]-[11], VR as interactive media
July 2018

IEEE Consumer Electronics Magazine

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - July 2018