Refine
Year of publication
Document type
- Working Paper (23) (remove)
Has full text
- Yes (23) (remove)
Keywords
- Mediennutzung (3)
- Computermathematik (2)
- Online-Ressource (2)
- ACCESS (1)
- Angewandte Statistik (1)
- Automatisierung (1)
- Beer (1)
- Behälterberechnung (1)
- Behältervolumen (1)
- Biomedizinische Statistik (1)
Course of studies
Biomedizinische Statistik
(2024)
Die klassische konfirmatorische Statistik, auch frequentistische Statistik genannt, setzt voraus, dass man theoretisch unendlich viele Stichproben ziehen kann, und dass dann die aus den Stichproben berechnete Prüf- oder Testgröße unter der Nullhypothese H0 eine bestimmte Verteilung annimmt. Meistens sind die Testgrößen so konstruiert, dass bei Ziehung der Stichproben aus immer derselben Grundgesamtheit (es gilt die Nullhypothese H0) eine Verteilung der Testgröße um den Wert null herum entsteht, z.B. in Form einer Glockenkurve, d.h., kleine Werte überwiegen. Große Werte der Prüf- oder Testgröße kommen mit geringer Wahrscheinlichkeit vor und signalisieren einen möglichen Ausnahmefall. Statt anzunehmen, dass einer der seltenen Fälle einer großen Prüfgröße eingetroffen ist, nimmt man lieber an, dass sich die Grundgesamtheiten unterscheiden (Alternativhypothese HA).
Our report focuses on the possible relationships between the occurrence of stressors and the use of VR. The issue was methodologically addressed by developing a general taxonomy of possible stressors in VR through related work and firstly carrying out a series of tests isolating only the baseline of headset-related stress.
Twelve young adults were tested in a pilot study by a specific five minute VR item for qualitative evaluations of the consumed content, their experience with the equipment, and their general experience running through our study. Same tests were carried out with a comparison group in a boxed TV-environment. For both groups, biomedical parameters were derived that indicate the stress level of the subjects (ECG, EDA and Cortisol test/salivette). Additionally the subjects were screened by questionnaires (VRSQ, NASA TLX-raw, semantic differentials) regarding changes in comfort. VR test subjects were screened by the headset's sensor data too (heart rate, pupil diameter, cognitive load).
Qualitative questions could not reveal any major irritations regarding the content of our relaxation scene. The headset’s sensor data did not indicate any cause of stress in subjects, but it is not possible to take baseline measurements without having the subject wear the VR headset. All other data show experimental validity and robustness.
Although knowing that our samples are small and the power of our tests may not be sufficient to detect a precise difference between the means, we decided to perform an experimental statistical analysis of the pilot study subject data (e.g f-tests, equal/unequal, significance level: 0.05).
According to the results and limitations of this analysis, young adults are not likely to experience stress from short-term use of VR headsets where all other VR stressors have been eliminated. Ten hypotheses in this regard have been rejected.