Beef1297 / ReadingPapers

Haptics, Kinesthetic sensation, perception
0 stars 0 forks source link

Evaluating a Personalizable, Inconspicuous Vibrotactile (PIV) Breathing Pacer for In-the-Moment Affect Regulation #18

Open Beef1297 opened 4 years ago

Beef1297 commented 4 years ago

CHI2020, Honorable mention

Authors

Miri, Pardis, Emily Jusuf, Andero Uusberg, Horia Margarit, Robert Flory, Katherine Isbister, Keith Marzullo, and James J. Gross. 2020. “Evaluating a Personalizable, Inconspicuous Vibrotactile(PIV) Breathing Pacer for In-the-Moment Affect Regulation.” In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, 1–12. CHI ’20. New York, NY, USA: Association for Computing Machinery.

Author Keywords

Haptic; Vibrotactile; Anxiety; Affect Regulation; Affect; Respiration; Slow-paced breathing; Pacer; Wearable; Linear Mixed Model; XGBoost; Shapley Values; Machine Learning;

どんなもの?

認知的なストレスを与えたときの,心労や不安感を軽減する PIV (a personalized, breathing pacer) システムの評価

個人差を考慮した振動による感情をコントロールするシステムと user-technology engagement (ストレスへの効果を考慮してテクノロジーを使う場面を決める) の組み合わせは初めて

Q

振動提示方法はどうやって分けたの? (treatment group 内でランダム?)

先行研究と比べてどこがすごい?

Vibrotactile Affect regulation の比較

image

Doppel や EmotionCheck とは異なり PIV は affect regulation を explicit にする. つまり,知覚,認知し,行動を起こさせる.(slow-paced breathing を意図的に導く)

state, trait, personality の3点セットで実験しているところ つまり,実験手法がすごいところ

技術や手法のキモはどこ?

PIV はどのぐらい効果があるのか? (axiety, positive affect, negative affect を調査) treatment group と control group に分けて 二回ストレスを提示 (treatment group の二回目に提示)

Stressor がるとき,PIV-user engagement に対して, noticing, differentiating, synchronizing の難しさを調査. (そもそも,気づけるのかどうかや振動に呼吸を合わせられるのかなど)

個人差から anxiety がどれぐらい低下するのかを予測する. どういう人に PIV は効果があるのかを調べる 被験者の変化を変数として,Contribution を求めて,高い値から,効果がありそうな人というのを特定する

どうやって有効だと検証した?

stressor: vocabulary-based CRA stressor task; native English speakers を集めた (97人) (43: treatment group, 53 control group)

image

Meditation: 自分なりのゆっくりの呼吸にする. 同時に呼吸数 (respiratory rate)を測る Personalization: 個人に合わせて,周波数や振幅をチューニングする

Trait affect regulation measures, personality measures ⇒ 実験前のオンラインのアンケートを利用 (Anxiety Sensitivity Index など

State affect measures: 各実験セクションの合間に STAI-6 に回答

Post-stressor1 と Post-stressor2 との間で treatment group は有意に下がっている control group とで変化の仕方が異なる. (treatment group の効果量 Cohen's d = 0.33 だった)

image

image

Stressor and PIV-User Engagement

V-Breathing Practice では Control と Treatment のデータがあるが Post-stressor 2 では Treatment グループのデータだけ stressor があると Synchronization が難しくなっている

image

Predicting Anxiety Drop from Individual Differences

26 種類の individual difference measures (実験前と中に得たデータ) から学習モデルを作った. (各特徴の変化に対して,axiety のレベルがどう変化するか?というモデルで合ってる?)

それぞれ,Shapley value (シャープレイ値) を求めた (feature's contritubtion to model's predicted change in anxiety from Post-stressor 1 to Post-stresor2)

image

その後で,高い貢献を持っている feature にに対して,PIV の効果によって anxiety が減少したのかを検定 (treatment groupt と control group の比較かな?) ⇒ つまり,PIV の効果があった!だけではなく,それは個人差によるものではない?も検討している. (Big Five Inventory や 実験中に「こういう反応」をする人のように)

実験中に再評価(?)をした人は,PIV の使用に関わらずanxiety が減少していた. Big Five Inventory :パーソナリティ特性の分類テスト

議論はある?

(cognitive) reapprasal: changing the way one thinks about potentially emotion-eliciting events

reappraisal になれていない人 には PIV が効果あり データをそれぞれの特徴で抽出 image

implicit と explicit の差はあるのか?結果より,engagement が難しいと 効果が低くなってしまう.interface として簡単にしたらいいのか,implicit の方がいいのかは分からないので調査を行う必要がある

次に読むべき論文は?

Beef1297 commented 4 years ago

abdomen: 腹部, 下腹 prevalence: 流行, the fact or condition of being prevalent; commonness. paramount: 最高の trait: 習性,癖

Beef1297 commented 4 years ago

Cohen's d: 効果量 群間差についての効果量であるd族の一つ。平均値の差をプールされた標本標準偏差で割って標準化したもの。効果量を扱う場合、CohenのdよりもHedgesのgのほうが推奨されることが多い。 https://bellcurve.jp/statistics/glossary/2244.html#:~:text=Cohen's%20d,%E3%81%95%E3%82%8C%E3%82%8B%E3%81%93%E3%81%A8%E3%81%8C%E5%A4%9A%E3%81%84%E3%80%82

image

http://www.mizumot.com/method/mizumoto-takeuchi.pdf https://qiita.com/fhiyo/items/9cb2b05b36566ffe0eff

Beef1297 commented 4 years ago

Cognitive Reappraisal https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4168764/