Causing Fear, Suspense, and Anxiety Using Sound Design in Computer Games - Game Sound Technology and Player Interaction: Concepts and Developments - page 184

causing Fear Findings fear or anxiety. Furthermore, timed sound effects-

-synchronized game sound and visual threat--will
not promote anxiety because the threat becomes a
clear stimulus. Thus, timing should not promote
anxiety. Fear and anxiety were measured in the
experiment discussed below, rather than suspense,
because fear and anxiety are considered separate
emotions, whereas suspense is the overlap of these
emotions, which could confound the interpretation
of the results. Our hypothesis was tested using
the methodology described in the next section.

Results showed a statistically significant (p < 0.05)
and large difference (η2 > 0.14) in fear due to the
volume of sound effects between low volume
sound and high volume sound, as well as between
medium volume sound and high volume sound.
No meaningful qualitative data was gathered for
volume related fear responses.
For timing, results showed a statistically sig-
nificant and very large (see eta square, Cohen,
1988) increase between timed and untimed sound
effects. Qualitative data showed that timed sound
effects enhanced the fear of many players when
accompanied along with a visual gameplay ele-
ment, such as the presence of an in-game enemy,
though the sound effect by itself may not have
substantially elicited fear.
Findings for fear related to sourced sound ef-
fects appeared to be considerable but they were
not statistically significant. Several participants
verbally reported that the acousmatic sound ef-
fects, such as a breaking window or footsteps on
the ceiling, evoked fear. In particular, participants
reported that the acousmatic sound effect required
them to be attentive to possible threats.

EXPErIMENtAL EVIDENcE

The hypothesis, as stated in the previous section,
was tested using a survival horror game level in
Gears of War (2007), which was created in Unreal
Editor 3. During each test subject's play-through
the participant heard one randomly selected alter-
native (wolf howl, gunfire, or wretch growl) for
the volume test, one randomly selected alternative
(thunder, boomer growl, or creaking door) for the
timing test, and one randomly selected alterna-
tive (locust growl, glass shattering, or footsteps)
for the source test. Both quantitative data using
7-point self report surveys and qualitative data
were gathered and analyzed. Although there can
be issues with “after-the-fact narration” (Nacke
& Grimshaw, 2011) by participants completing
self report surveys and interviews, the use of
these indirect measures are common approaches
to data gathering in research of emotions. Thirty-
four participants in the U.S.A., ten females and
twenty-four males participated in the study. The
average participants' age was 26 years old. The
average playing time per week was about eleven
hours, and fifteen participants (approximately
44%) liked playing survival horror games. (For
a full exposition of the methodology and results,
see Amdel-Meguid, 2009).

causing Anxiety Findings

Results showed a statistically significant and large
difference in anxiety due to the volume of sound
effects between low volume sound and medium
volume sound, as well as between low volume
sound and high volume sound. No meaningful
qualitative data was gathered for volume related
anxiety responses.
There was not a statistically significant differ-
ence between timed and untimed sound effects
for anxiety. Qualitative data showed that untimed
sound effects caught some players off guard,
because they could not determine whether the
sounds were meant to signal danger, or if they
were benign sounds.