It’s not Complicated: A Study of Non-Specialists Analyzing GSR Sensor Data to Detect UX Related Events

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

206 Downloads (Pure)

Resumé

Emotion is a key factor in understanding user experiences (UX) of interactive systems. An emerging trend within HCI is to apply physiological sensors for uncovering emotions. Previous studies rely on various sophisticated analysis techniques and specialized knowledge to interpret sensor data. While commendable for increasing accuracy at fine grained latencies (to detect events within seconds), this can be challenging for UX practitioners without specialized knowledge. This study contributes in two ways. Firstly by understanding the level of sensor accuracy in detecting UX related events. Secondly by applying a basic analysis approach where sensor data is interpreted by 21 non-specialist participants (no previous experience in doing this). Their performance is compared to random guessing. Findings show that sensor data analyzed by non-specialists are significantly more accurate in capturing subjectively reported UX events than random guessing. An accuracy level of 60-80% was obtained at granularities within 3.5-11 seconds of UX related events.
OriginalsprogEngelsk
TitelNordiCHI 2018 : Revisiting the Life Cycle - Proceedings of the 10th Nordic Conference on Human-Computer Interaction
Antal sider14
ForlagAssociation for Computing Machinery
Publikationsdato1 okt. 2018
Sider170-183
ISBN (Elektronisk)978-1-4503-6437-9
DOI
StatusUdgivet - 1 okt. 2018
BegivenhedNordiCHI 2018 - Oslo, Norge
Varighed: 1 okt. 20183 okt. 2018

Konference

KonferenceNordiCHI 2018
LandNorge
ByOslo
Periode01/10/201803/10/2018

Fingerprint

Sensors
Human computer interaction

Citer dette

Bruun, A. (2018). It’s not Complicated: A Study of Non-Specialists Analyzing GSR Sensor Data to Detect UX Related Events. I NordiCHI 2018: Revisiting the Life Cycle - Proceedings of the 10th Nordic Conference on Human-Computer Interaction (s. 170-183). Association for Computing Machinery. https://doi.org/10.1145/3240167.3240183
Bruun, Anders. / It’s not Complicated : A Study of Non-Specialists Analyzing GSR Sensor Data to Detect UX Related Events. NordiCHI 2018: Revisiting the Life Cycle - Proceedings of the 10th Nordic Conference on Human-Computer Interaction. Association for Computing Machinery, 2018. s. 170-183
@inproceedings{178a543cbdc64b2b9fcf4fa144b65ce3,
title = "It’s not Complicated: A Study of Non-Specialists Analyzing GSR Sensor Data to Detect UX Related Events",
abstract = "Emotion is a key factor in understanding user experiences (UX) of interactive systems. An emerging trend within HCI is to apply physiological sensors for uncovering emotions. Previous studies rely on various sophisticated analysis techniques and specialized knowledge to interpret sensor data. While commendable for increasing accuracy at fine grained latencies (to detect events within seconds), this can be challenging for UX practitioners without specialized knowledge. This study contributes in two ways. Firstly by understanding the level of sensor accuracy in detecting UX related events. Secondly by applying a basic analysis approach where sensor data is interpreted by 21 non-specialist participants (no previous experience in doing this). Their performance is compared to random guessing. Findings show that sensor data analyzed by non-specialists are significantly more accurate in capturing subjectively reported UX events than random guessing. An accuracy level of 60-80{\%} was obtained at granularities within 3.5-11 seconds of UX related events.",
keywords = "Emotion, GSR, Non-specialists, Orienting responses, Physiological sensors, Sensor Data Analysis, Subjective",
author = "Anders Bruun",
year = "2018",
month = "10",
day = "1",
doi = "10.1145/3240167.3240183",
language = "English",
pages = "170--183",
booktitle = "NordiCHI 2018",
publisher = "Association for Computing Machinery",
address = "United States",

}

Bruun, A 2018, It’s not Complicated: A Study of Non-Specialists Analyzing GSR Sensor Data to Detect UX Related Events. i NordiCHI 2018: Revisiting the Life Cycle - Proceedings of the 10th Nordic Conference on Human-Computer Interaction. Association for Computing Machinery, s. 170-183, NordiCHI 2018, Oslo, Norge, 01/10/2018. https://doi.org/10.1145/3240167.3240183

It’s not Complicated : A Study of Non-Specialists Analyzing GSR Sensor Data to Detect UX Related Events. / Bruun, Anders.

NordiCHI 2018: Revisiting the Life Cycle - Proceedings of the 10th Nordic Conference on Human-Computer Interaction. Association for Computing Machinery, 2018. s. 170-183.

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

TY - GEN

T1 - It’s not Complicated

T2 - A Study of Non-Specialists Analyzing GSR Sensor Data to Detect UX Related Events

AU - Bruun, Anders

PY - 2018/10/1

Y1 - 2018/10/1

N2 - Emotion is a key factor in understanding user experiences (UX) of interactive systems. An emerging trend within HCI is to apply physiological sensors for uncovering emotions. Previous studies rely on various sophisticated analysis techniques and specialized knowledge to interpret sensor data. While commendable for increasing accuracy at fine grained latencies (to detect events within seconds), this can be challenging for UX practitioners without specialized knowledge. This study contributes in two ways. Firstly by understanding the level of sensor accuracy in detecting UX related events. Secondly by applying a basic analysis approach where sensor data is interpreted by 21 non-specialist participants (no previous experience in doing this). Their performance is compared to random guessing. Findings show that sensor data analyzed by non-specialists are significantly more accurate in capturing subjectively reported UX events than random guessing. An accuracy level of 60-80% was obtained at granularities within 3.5-11 seconds of UX related events.

AB - Emotion is a key factor in understanding user experiences (UX) of interactive systems. An emerging trend within HCI is to apply physiological sensors for uncovering emotions. Previous studies rely on various sophisticated analysis techniques and specialized knowledge to interpret sensor data. While commendable for increasing accuracy at fine grained latencies (to detect events within seconds), this can be challenging for UX practitioners without specialized knowledge. This study contributes in two ways. Firstly by understanding the level of sensor accuracy in detecting UX related events. Secondly by applying a basic analysis approach where sensor data is interpreted by 21 non-specialist participants (no previous experience in doing this). Their performance is compared to random guessing. Findings show that sensor data analyzed by non-specialists are significantly more accurate in capturing subjectively reported UX events than random guessing. An accuracy level of 60-80% was obtained at granularities within 3.5-11 seconds of UX related events.

KW - Emotion

KW - GSR

KW - Non-specialists

KW - Orienting responses

KW - Physiological sensors

KW - Sensor Data Analysis

KW - Subjective

UR - http://www.scopus.com/inward/record.url?scp=85056580964&partnerID=8YFLogxK

U2 - 10.1145/3240167.3240183

DO - 10.1145/3240167.3240183

M3 - Article in proceeding

SP - 170

EP - 183

BT - NordiCHI 2018

PB - Association for Computing Machinery

ER -

Bruun A. It’s not Complicated: A Study of Non-Specialists Analyzing GSR Sensor Data to Detect UX Related Events. I NordiCHI 2018: Revisiting the Life Cycle - Proceedings of the 10th Nordic Conference on Human-Computer Interaction. Association for Computing Machinery. 2018. s. 170-183 https://doi.org/10.1145/3240167.3240183