
摘" 要" 近年來, 研究者們將“治療聯盟” (Therapeutic Alliance, TA)的概念與在線自助干預(Internet-based Self-help Interventions, ISIs)相結合, 以解決ISIs中用戶參與度較低的問題。這種在數字環境中形成的TA, 被稱之為“數字治療聯盟” (Digital Therapeutic Alliance, DTA)。隨著人工智能的迅速發展, 聊天機器人可模擬人類指導, 相對于傳統ISIs程序更易于與用戶建立關系, 可通過友好、尊重、傾聽、鼓勵、真誠、理解、信任這幾個關系線索來促進DTA的發展, 為解決用戶低參與度的問題提供了一種新思路。未來的研究可從影響因素、ISIs技術迭代、測量規范、實驗操縱等方面對DTA作進一步的探索。
關鍵詞" 數字治療聯盟, 聊天機器人, 關系線索
分類號" B849
1" 引言
目前, 在線自助干預(Internet-based Self-help Interventions, ISIs)的可行、有效性已得到廣泛驗證(Izzaty et al., 2021; Johansson et al., 2021; Sun et al., 2021; Taylor et al., 2021; Weisel et al., 2019), 或可成為面對面治療的有力補充(Berry et al., 2019), 但高脫落率、低參與度仍是其眾所周知的困境, 對于無指導ISI, 此問題則更為凸顯(Linardon et al., 2019; Pratap et al., 2020; Zhang et al., 2021)。隨著人工智能(Artificial Intelligence, AI)技術的迅猛發展, 能模擬人類對話的聊天機器人(Chatbot)可讓無指導ISI在自動化后兼顧效率及成本效益(Luo et al., 2022)。具體而言, 在聊天機器人的設計中引入關系線索(Relational Cues), 如自我披露、真誠、理解、幽默等(Gallen et al., 2018), 可在認知、情感兩個維度上滿足用戶的需要(Abd- alrazaq et al., 2019; Provoost et al., 2017; Wiese et al., 2022), 并與用戶建立數字治療聯盟(Digital Therapeutic Alliance, DTA), 進而促進用戶的參與度及治療效果(Goldberg et al., 2021; Liu et al., 2022; Provoost, 2021)。縱觀既有研究, 國外相關成果盡管豐富但較為零散, 而國內關于ISIs的研究進展尚處初期階段(Henson et al., 2019; Grekin et al., 2019; Yao et al., 2020; Zhang et al., 2021)。鑒于此, 文章集中探討聊天機器人在無指導ISI中通過關系線索對DTA產生的可能影響, 以引發同行們對該領域的研究興趣, 為進一步的研究提供參考。
2" DTA的發展
治療聯盟(Therapeutic Alliance, TA), 也稱工作同盟(Working Alliance, WA), 是來訪者和咨詢師之間為實現治療目標而合作的關系的質量與強度(朱旭, 江光榮, 2011a)。上世紀70年代, Bordin (1979)將TA分為三個成分——情感紐帶、對治療任務達成共識、就治療目標達成一致, 并成為TA最流行之定義。而后, Horvath與Greenberg (1989)基于Bordin的定義編制了首個TA量表——工作同盟量表(Working Alliance Inventory, WAI)。近年來, 隨著WAI開始逐漸被用于數字心理健康的研究(Andersson et al., 2012), DTA一詞也由此誕生。在諸如以電子郵件、在線聊天、視頻會議、ISIs程序等干預形式所建立的TA均可稱為DTA (D’Alfonso et al., 2020; Henson et al., 2019; Lederman amp; D’Alfonso, 2021)。
DTA之所以得到發展, 可能的原因有三:第一, 新冠疫情加速了社會的數字化進程, 虛擬現實(Virtual Reality)、元宇宙(Metaverse)等概念也得到了發展(張夏恒, 李想, 2022)。從哲學的創世觀看, 人類雖生活在既定的宇宙中且被排除于創世者之外, 但人類一直有超越自然的夢想, 而數字化的發展則為人類提供了造世的機會(黃欣榮, 曹賢平, 2022)。因此, 未來現實生活的數字化將成為客觀趨勢, 人機關系也變得越來越重要。第二, ISIs正朝著效益最大化的方向發展, 但更高水平的自動化也伴隨著用戶參與度低、脫落率高的問題, 基于此, 在面對面心理咨詢/心理治療中占有重要地位的TA也自然受到研究者們的關注。第三, 基于自我決定理論(Self-determination Theory, SDT), 自主、勝任、關系這三個基本需求的滿足, 能促進個體的外在動機向內在動機轉化, 進而保證其心理健康成長(Deci amp; Ryan, 1985)。與之相應, ISIs程序可輔助用戶解決問題, 提高用戶的自主感、勝任感, 進而有助于培養TA中的情感紐帶。同時, TA中“在目標和任務上達成一致”與用戶使用ISIs程序時的目標確立及所獲得的階段性反饋有關。因此, DTA在可行性、有效性方面具有一定的理論支撐。
有一系列研究表明, 在ISIs中建立的DTA與面對面心理咨詢/心理治療中建立的TA水平相接近(Andersson et al., 2012; Heim et al., 2018; Klasen et al., 2013; Pihlaja et al., 2018; Tremain et al., 2020)。同時, DTA與參與度呈正相關(Baumel amp; Kane, 2018; Goldberg et al., 2021; Hargreaves et al., 2018; Heim et al., 2018; Perski et al., 2017; Rodrigues et al., 2021), 而參與度則是改善ISIs治療效果的關鍵因素(Arndt et al., 2020; Asaeikheybari et al., 2021; Fuhr et al., 2018; Puls et al., 2020)。另一項元分析還指出, DTA與治療效果也存在相關, 且總體效應量中等, 但實際的研究結果好壞參半(Probst et al., 2019), 這與測量工具的發展及選用不無關系。DTA測量需針對數字環境具體考量, 若簡單改編傳統WAI, 可能無法解釋數字干預中TA的復雜性。研究者們逐漸認識到這一點, 并著手將傳統WAI基于數字環境進行改編(Tremain et al., 2020)。例如, Berger等人(2014)在WAI-SR的基礎上進行改編, 使之適應有指導ISI。最初, Kiluk等人(2014)提出基于原版WAI的數字改編版(WAI-Tech), 用于無指導ISI中DTA的測量, 但其僅僅只將量表中的“咨詢師”換成了“應用程序”。隨著對DTA的進一步探究, Meyer等人(2015)在研究中發現, 被試和ISIs程序之間的TA并不等同于和人類咨詢師之間的TA。因此, 他們對幫助聯盟問卷(Helping Alliance Questionnaire, HAQ)進行了改編, 以評估被試在多大程度上認為程序有所助益。在實證研究中, 被試在干預后第3周的HAQ得分即可成功預測其3個月后的治療效果。最近, Berry等人(2018)考慮了無指導ISI的特點, 在阿格紐關系量表(Agnew Relationship Measure, ARM)的基礎上編制了移動版阿格紐關系量表(Mobile Agnew Relationship Measure, mARM)。隨后, Henson等人(2019)在WAI-SR的基礎上編制了D-WAI, 以專門評估無指導ISI的DTA。Gómez Penedo等人(2020)也為更好地測量有指導ISI中的DTA, 在Berger等(2014)的基礎上編制了WAI-I, 并在大樣本中驗證了此量表的可靠性。為進一步使DTA測量適應數字干預場景, D’Alfonso等人(2020)在mARM的基礎上, 將人機交互(Human- Computer Interaction, HCI)理論與TA理論結合, 并嘗試開發一種能更可靠地在無指導ISI中評估DTA的量表。時下, 聊天機器人技術正不斷地改變傳統無指導ISI程序的交互體驗, 其既提供了類似于人類的指導, 但又實現了全自動化。因此, 未來DTA測量的發展或將與新興AI技術的迭代趨勢相適應。
3" 關系線索或是影響DTA的重要因素
目前, ISIs多是基于認知行為療法(Cognitive Behavioral Therapy, CBT)進行設計, 且對壓力、抑郁、焦慮、煙癮、酒癮、失眠及創傷后應激障礙等問題均有顯著的療效(Weisel et al., 2019)。根據Bielinski和Berger (2020)的劃分, ISIs的常見類型有三:一是無指導干預(Unguided Interventions), 指在線干預的過程中無咨詢師介入, 用戶僅通過程序自助; 二是有指導干預(Guided Interventions), 指將用戶自助與定期、簡短的在線輔導(同步或異步)相結合; 三是混合干預(Blended Interventions), 指將在線干預與面對面心理咨詢/心理治療相結合, 以前者作為后者的補充。
在ISIs的情境中, 若有咨詢師的支持, TA則相對更容易建立。有研究指出, 有指導ISI與面對面治療所建立的TA水平并無顯著差異, TA不但預測了參與度, 也預測了治療效果(Anderson et al., 2012; Kaiser et al., 2021)。盡管有指導ISI的整體效果往往優于無指導ISI (Baumeister et al., 2014), 但也有研究表明, 在低強度的有指導ISI中, 被試在干預初期的情感紐帶得分較低且增速緩慢(Jasper et al., 2014), 而被試在參與度、治療效果上的得分也與無指導ISI上的對應得分無顯著差異(Chen et al., 2020), 這說明人類介入的缺乏限制了TA的發展。不過, Berry等人(2018)指出, 被試實際上也能與傳統的無指導ISI程序發展虛擬關系, 這有助于彌補缺乏人類指導帶來的影響。Holter等人(2020)以扎根理論(Grounded Theory)建立的人機關系模型也認為, 個體與無指導ISI程序能夠建立情感紐帶, 但前提是要使個體對程序的感知在社會行動者與無生命的程序之間交替轉換。基于此, 研究者們開始嘗試在無指導ISI程序中加入基于傳統編程的虛擬化身(Avatar), 以縮小其與有指導ISI效果的差距。例如, 在Heim等人(2018)的研究中, 被試的情感紐帶得分因虛擬化身的加入而穩定發展, 并與失眠改善相關。但是, 一些被試卻表示他們更想與人類咨詢師交流, 此意愿也預測了療效。類似的, Fenski等人(2021)指出, 若虛擬化身不能對被試的負面情緒準確識別并給予恰當回應, 則很有可能起到反作用。總的來說, 嵌入于無指導ISI程序中的虛擬化身有希望與人類建立類似于有指導ISI中的DTA, 且DTA可正向影響參與度及治療效果, 但如何設計虛擬化身以保障療效仍需進一步討論。
人類線索(Human Cues), 是計算機程序因模擬人類形象、言語、行為等條件而具有的特征, 能讓與之交互的個體產生往往只有與真人交互時才特有的感受(Rodrigues et al., 2021)。社會行動者范式(Computers as Social Actors, CASA)也指出, 人類往往會下意識地對計算機程序呈現的人類線索做出反應, 且無論這些線索有多么初級(Nass et al., 1994)。為將人類線索具體化, Gallen等人(2018)將其分為4類:視覺線索(Visual Cues), 如年齡、性別、外貌、表情、動作等; 言語線索(Verbal Cues), 如文字、語音、語調、語速等; 準非言語線索(Quasi-Nonverbal Cues), 如表情符號; 關系線索(Relational Cues), 如自我披露、理解、幽默等。這些線索都可能對情感紐帶及整個DTA的建立、發展造成影響, 可作為指導虛擬化身設計的起點。
由上文可知, 被試往往對虛擬化身存有更高的情感期待, 而這種情感期待是否能得到滿足也會影響情感紐帶的發展。若情感紐帶建立的不夠牢固, 則可能會限制DTA的發展, 進而導致較差的參與度及治療效果。然而, 要形成情感紐帶, 前提則是虛擬化身向被試傳遞了溫暖、安全和信任等關系線索(Negri et al., 2019)。在早期的研究中, 研究者就已發現在虛擬化身的對話設計中引入寒暄、幽默、同理心等關系線索對情感紐帶的影響相對目標、任務維度更大(Bickmore et al., 2005)。在最近的研究中, ter Stal等人(2020)也指出, 富有同理心的話語是影響人機關系的關鍵因素。因此, 賦予虛擬化身恰當的關系線索對其與用戶發展DTA可能有重要的作用。
4" 如何設計關系線索來促進DTA的發展
若關系線索對DTA的發展可能起到重要的作用, 那么, 如何設計關系線索, 并讓其更高效地介入自然也變得重要(Müssener, 2021)。此時, 基于AI自然語言處理(Natural Language Processing, NLP)技術的聊天機器人就展現出了優勢, 其不但能呈現豐富的人類線索, 還能基于用戶的行為數據進行持續的“學習” (Zhang et al., 2020), 并給予用戶個性化的反饋(Laranjo et al., 2018; Zhang et al., 2020), 比基于傳統編程的虛擬化身更高效、靈活且人性化, 無疑是更活躍的社會行動者(Alkhaldi et al., 2016; Ames et al., 2019; Hardeman et al., 2019; Tremain et al., 2020)。
自1966年世界上第一個真正意義上的聊天機器人ELIZA誕生以來(Weizenbaum, 1983), 聊天機器人的技術就一直在持續迭代, 并逐步融入到數字心理健康之中(Elmasri amp; Maeder, 2016; Fitzpatrick et al., 2017; Gaffney et al., 2014)。目前, 聊天機器人通常作為單獨的功能模塊嵌入ISIs程序之中, 以語音用戶界面(Voice User Interface, VUI)的形式為用戶提供幫助, 可替代人類咨詢師的指導而使程序成為一種新型的無指導ISI (如MYLO, Woebot), 或是配合人類咨詢師作為一個輔助的功能(如dll心聆“小天”)。此外, 根據回復生成機制, 聊天機器人可分為兩類:一是檢索式(Retrieval-based), 聊天機器人將從靜態的知識庫中檢索預定義的規則來進行回復; 二是生成式(Generation-based), 聊天機器人將通過學習及推理機制來動態生成回復(Song et al., 2018)。在形態方面, 聊天機器人還可大致分為兩類:一是具有虛擬化身, 這是一種將聊天機器人和交互式化身(計算機生成的數字角色, 其外觀可能為人類或卡通人物)結合在一起的程序形態, 可通過眼神、表情、動作、語音、文本等方式與人類交互(如Replika); 二是僅以語音、文本與人類交互(如Siri, 微軟“小冰”)。
近年來, 聊天機器人在ISIs中的應用逐漸增多, 有研究發現其不但比傳統的無指導ISI程序更能促進被試的參與度(Perski et al., 2019; Vaidyam et al., 2019), 且其與被試所建立的TA水平也與人類相當(Darcy et al., 2021)。盡管如此, 若要問聊天機器人為何有效, 研究者們卻知之甚少。本文假設, 可能的原因有四:第一, 心智感知理論(Mind Perception Theory; Waytz et al., 2010)指出, 個體可感知到其他對象具有心理能力, 并對其作擬人化的信息加工。因此, 聊天機器人的關系線索越豐富, 就越可能提升社會存在(Social Presence), 使個體產生與真實人類交互的感知(Lee et al., 2020; Sundar, 2008)。同時, 擬人化的聊天機器人通常比人類更可靠、易得, 個體與其交互也容易獲得更多安全感(Wanser et al., 2019), 從而更傾向與其合作(Wiese et al., 2022)。第二, 基于社會線索減少理論(Reduced Social Cues, RSC), 在網絡文本信息交互的過程中, 由于思想和情感必須轉化為文字以彌補非言語信息的缺乏(Kiesler et al., 1984)。因此, 個體在信息加工的過程中可能會產生網絡去抑制效應(Online Disinhibition Effect, ODE), 進而表現出不同于面對面交流時的行為, 包括放松、較少的約束感以及較開放的情感表達等(Suler, 2004), 這可能會使人機關系變得更為緊密、牢固。第三, 聊天機器人天然具有人類線索, 能擬人化地輔助個體解決問題, 滿足了SDT原則, 進而促進情感紐帶的發展。第四, 由人際投資模型(The Investment Model of Personal Relationships)可知, 聊天機器人提供的情感支持及有價值的信息, 可使個體的感知獲益及感知投入持續增加、認知成本及疑慮持續降低, 進而逐漸建立信任感, 提升對ISIs程序的參與度(Rusbult et al., 1994)。
歸納上述, 具備關系線索且更為靈活的聊天機器人更利于從認知及情感的角度切入, 在無指導ISI中促進DTA的快速發展, 解決用戶參與度低的問題。然而, 盡管已有少部分研究者針對此問題進行了探索, 但目前尚未有研究者歸納出確實、有效的關系線索以指導聊天機器人的設計。例如, Rodrigues等人(2021)發現, 與僅具有視覺線索的聊天機器人相比, 僅具有關系線索的聊天機器人更能與被試建立DTA, 且參與度也更高。但是, 此研究只探討了視覺線索與關系線索在DTA上的差異, 而并未檢驗不同關系線索對DTA的影響。為將聊天機器人的作用具體化, 應對可能更為關鍵的關系線索作深入的探討。因此, 下文將在前人研究的基礎之上提出幾種ISIs中可能會對DTA帶來積極影響的關系線索(Bordin, 1979; Horvath amp; Greenberg, 1989; Norcross, 2002), 以幫助聊天機器人發展人工智慧(Artificial Wisdom)。
4.1" 友好尊重
在面對面咨詢中, 溫暖、和諧、寬松、自由且安全的談話氛圍以及咨訪雙方的相互尊重都是TA的助長因素(Luborsky, 1976)。同樣, 在過往的ISIs研究中, 友好與尊重也被認為是程序中必備的基礎性設計, 其中ISIs程序所傳遞信息的語氣、語調都會對干預的可信度、參與度、有效性造成影響(Ames et al., 2019; Bock et al., 2015)。盡管被試的偏好各不相同, 但禮貌、尊重、友好、幽默、積極等友好的對話語氣相對更受其青睞, 反之, 被試普遍對壓力、教訓、羞辱等較為排斥(Ames et al., 2019; Müssener, 2021)。在使用聊天機器人時, 這種影響還可能被實時的對話交流強化。例如, 當聊天機器人直呼被試的名字, 并在適當的時候使用幽默, 也能增進雙方的友好關系(Bickmore et al., 2009)。原因在于, 聊天機器人的“人格”特征會影響被試情緒反應的強度(Medhi Thies et al., 2017), 若被試因聊天機器人的互動反饋而將其歸因為禮貌、友好、尊重的, 即便被試知曉這是虛擬交互, 也仍會將這種類社會互動(Parasocial Interaction)視作一種親密的社交互動(Horton amp; Wohl, 1956), 并將對應的社會化規范應用于與聊天機器人的交互中。而當聊天機器人具有虛擬形象時, 由于其呈現出更為豐富的非言語信息(如表情、姿勢、動作、唇同步等), 還可通過人工情緒傳染(Artificial Emotional Contagion)的機制(如模仿及情緒鏡像), 使被試更容易感受到其友好、尊重之特征(Nofz amp; Vendy, 2002)。例如, 在面對面咨詢的場景之中, 來訪者往往能夠敏感地捕捉咨詢師的微表情, 以評估咨詢師的價值觀及評判意圖(Datz et al., 2019)。而在ISIs的情境之中, 若賦予虛擬化身較高的模型面數(Tris)并精細化其骨骼(Bone)設計, 化身不但能模擬生動的宏表情, 甚至也可能模擬出積極的微表情以進一步促進真實且融洽的交談氛圍。
4.2" 傾聽鼓勵
咨詢師對咨詢工作的投入通常被認為是TA的預測因素, 通常包含了積極地傾聽與適時鼓勵(朱旭, 江光榮, 2011b)。而在無指導ISI的研究中, 研究者也發現無論被試傾訴的對象是真人還是聊天機器人, 情緒宣泄所達到的效果并無顯著差異(Ho et al., 2018)。不過, ISIs程序往往需要與服務器通訊, 因此以純文本聊天機器人進行傾聽與回應時, 其回復速度也會影響其社交吸引力(Lew amp; Walther, 2022), 而基于上下文的動態回復相對于過于即時或延遲的效果更好(Samsudin, 2020)。此外, 聊天機器人在動態回復時若能模擬“正在輸入”狀態, 用戶可更明顯地感知到其適時的停頓與猶豫, 并產生其正在“思考”的印象。對于鼓勵, Chikersal等人(2020)也指出, 在ISIs中, 那些帶來更積極影響的支持信息, 不但更為簡短, 而且也包含更多積極、肯定、鼓勵等詞匯。基于此, 聊天機器人除了直接鼓勵用戶之外, 在多輪對話中通過關鍵詞復述也能給予用戶間接鼓勵, 并提升其傾訴體驗。當聊天機器人具有虛擬形象時, 則可在用戶表達時, 針對對應的內容, 回應以適當的眼神凝視、積極的面部表情、點頭以及開放的手勢等作為言語鼓勵的補充, 這能有效地使用戶產生人際互動的感知(Cummins amp; Cui, 2014), 進而將關系線索歸因于聊天機器人(Hortensius amp; Cross, 2018), 并進一步提升傾聽、鼓勵的效果。然而, 鼓勵可能并不適用于所有群體(Arndt et al., 2020)。因此, 聊天機器人可先甄別出哪些人群更容易受鼓勵的積極影響后再做反應。
4.3" 真誠理解
心理咨詢的效果往往取決于咨訪關系的質量——若咨詢師善解人意、真誠一致并無條件積極地關注著來訪者, 咨詢效果則更好(Rogers, 1957)。而在無指導ISI的情境中, 若聊天機器人能經常對被試的話語進行關注, 并以誠實、謙虛的態度對被試進行請教, 被試對它的積極評價也會更多(Zhou et al., 2019), 這不僅會讓被試更具有參與感, 而且還會幫助聊天機器人“學習”新概念。此外, 若聊天機器人能主動向被試披露其“個人信息”, 也有可能讓被試感受到它的“真誠”, 進而作更多的自我暴露(Kang amp; Gratch, 2014)。對于理解, 共情技術在傳統咨詢中較為常用, 而在ISIs的情境中, 聊天機器人基于真實的咨詢語料來進行訓練, 亦可具備復述的能力來做到一定程度的“共情”。然而, 復述并非鸚鵡學舌即可, 而是需要用“自己的話”加上來訪者話中的重要詞語來提煉內容。目前, 基于先進的自然語言生成模型GPT-3 (General Pre-trained Transformer-3)就可使聊天機器人做到長話短說、取其精要(Sezgin et al., 2022)。但若要推測來訪者的言外之意來達到更高級的“共情”且能兼顧對話歷史來保持咨詢的連貫性, 則仍需技術的進一步迭代。此外, 聊天機器人還可基于個性化技術對用戶獨特的需求、偏好、情緒進行積極、精準的響應(Valentine et al., 2022), 進而使用戶體驗到一種區別于傳統咨詢的一種獨特“理解”。例如, Liu-Thompkins等人(2022)嘗試將一個個性化的系統框架融入于真實的營銷場景中, 使聊天機器人通過偏好分析、人格評估、目標推理三個步驟, 來具備換位思考的“共情”能力。有研究發現, 引入個性化的設計有利于被試與ISIs程序在治療任務及目標上達成一致(Penedo et al., 2020), 進而加強參與動機(Liu et al., 2013)及DTA (Oinas-Kukkonen amp; Harjumaa, 2009; Tremain et al., 2020; Valentine et al., 2022)。
4.4" 相互信賴
在一段咨訪關系中, 若來訪者認為咨詢師可信, 他們才會作更多自我暴露, 進而促進TA的發展(Bachelor, 2013; 朱旭, 江光榮, 2011b)。同樣,"可信度與DTA的質量也高度相關, ISIs程序的低可信度可能會導致被試參與度低甚至脫落(Mackie et al., 2017)。反之, 若被試覺得ISIs程序可信, 其繼續使用的意愿(Radomski et al., 2019)及其對被治愈的期望也會更高(Sauer-Zavala et al., 2018)。在無指導ISI中, 聊天機器人所營造的擬人的第一印象會影響其可信度(Kelders et al., 2012; Neuberg, 1989; Oinas-Kukkonen amp; Harjumaa, 2009), 而外在刺激特征則是最關鍵的預測因素(Kim et al., 2021; Richards et al., 2020; Tremain et al., 2020)。但不同于傳統咨詢的是, 在ISIs的情境中可讓用戶自主設計、搭配, 或基于用戶畫像來賦予聊天機器人特定的種族、形象、年齡、性別、個性、聲音(Brown et al., 2013), 并基于用戶反饋迭代、調整, 因此更具靈活性。此外, 類同于復述, 包含了情感詞語的情感反映技術也值得探究。在干預早期, 通過基于深度學習的情感預測技術(Kumar, 2021), 聊天機器人能以簡短的情感反映(如“我感到你現在很焦慮”)與用戶迅速建立信任。隨著DTA水平的逐步提升, 聊天機器人還可進一步將更為關鍵的情感反饋給用戶, 并通過詢問以澄清其情感體驗, 促使其作更深入的暴露。但由于情感通常以隱喻、明喻、舉例等方式表達, 因此, 聊天機器人在不明其意時可靈活運用主動提問來進行確認及“學習”, 以豐富知識圖譜(Yin et al., 2017)。最后, 并非所有用戶都對情感反映表示歡迎, 因此聊天機器人在作情感反映前, 需綜合評估個人知識圖譜、DTA水平及上下文情感詞出現的頻率、強度等因素來確定回復的時機及內容, 并結合用戶的后續反饋來習得其偏好。
綜上所述, 文章對現有研究結果進行了歸納, 并梳理了DTA的前因、后果。基于此, 文章將提出一個模型(見圖1)。并假設, DTA對治療效果有直接影響; DTA對參與度有直接影響; 參與度對治療效果有直接影響。同時, 友好尊重、傾聽鼓勵、真誠理解、相互信賴等關系線索或可對DTA造成影響, 進而帶來更優的參與度及治療效果。
5" 存在問題及未來展望
5.1" 需進一步探索DTA的影響因素
盡管人類線索的效用顯著(Rietz et al., 2019), 但目前關于聊天機器人的研究多集中于言語、視覺線索, 將關系線索與不同線索比較的研究相對較少(Bao et al., 2022; Grekin et al., 2019; ter Stal et al., 2020)。然而, 在傳統咨詢領域中, 研究者會將關系線索與其他變量對比, 以確定不同變量對療效的貢獻大小, 但文章僅涉及可能影響DTA的部分關系線索(Gallen et al., 2018)。因此, 究竟是哪一個變量在發揮關鍵作用, 仍不得而知(Heim et al., 2018)。在基于聊天機器人的無指導ISI程序中, 除人類線索之外還包括用戶體驗、AI對話水平、用戶期望等影響因素。首先, 基于Hentati等人(2021)的發現, 程序用戶界面(User Interface, UI)的易用與否對被試的參與度有顯著的影響。因此, 這一額外變量可能會導致研究者錯誤地評估聊天機器人的作用。其次, 有研究發現人類線索之間存在交互作用, 當聊天機器人呈現強視覺線索(人類照片)時, AI對話水平與被試態度無關, 但呈現弱視覺線索(氣泡圖)時, 強AI對話水平補償了弱視覺線索的低擬人化效果。最后, 此研究還指出身份線索設定了被試對聊天機器人性能的期望, 當聊天機器人被識別為人類時被試對其有更高的期望, 而低AI對話水平則會導致更多負面評價(Go amp; Sundar, 2019)。因此, 聊天機器人呈現的人類線索并非越多越好, 不同線索的影響不同且內在關系復雜。在未來的研究中, 研究者可評估更多的TA助長因素, 并將其它人類線索及變量與之對比或組合, 探索不同變量之間可能存在的交互作用。
5.2" ISIs需作進一步的技術迭代
時下, ISIs多是將傳統心理療法數字化, 而計算機科學領域仍有諸多成果可促進ISIs的技術迭代。首先, 可將其他成熟的結構化技術與ISIs程序結合, 使之進一步體系化。例如, 以說服性系統設計(Persuasive System Design, PSD)來構建ISIs程序, 程序將以更多地支持、提醒、安排來提高參與度(Baumel amp; Yom-Tov, 2018)。此外, 使用動機性訪談(Motivational Interviewing, MI)這種結構化的對話技術, 也可提升用戶改變的動機(Rollnick et al., 2010), 進而快速且有效地提升其參與度(Malins et al., 2020)。其次, 可使用較先進的算法模型來進一步提升ISIs程序的性能。例如, 以創新的BERT (Bidirectional Encoder Representation from Transformers)或GPT-3模型來替代依賴人力、拓展性較差的LIWC (Linguistic Inquiry and Word Count)模型(Tanana et al., 2021)。如此, 聊天機器人不但能動態評估用戶的情緒及DTA水平, 其情感識別/交互能力也能得到極大的加強(Rajagopal et al., 2021)。然而, 當聊天機器人的回復生成更具靈活性且富有創意時, 其生成內容的不確定性也是一把雙刃劍。因此, 在咨詢情境中將檢索式與生成式結合, 開發聯合型的聊天機器人, 或許更有利于實際的應用(Song et al., 2018)。最后, 未來也可將NLP及計算機視覺(Computational Vision, CV)相結合的多模態技術運用于ISIs中。例如, 通過深度學習模型對語音、語調、語速、宏表情、微表情、肢體動作、瞳孔擴張等因素進行綜合分析, 以進一步提升聊天機器人對用戶意圖、情緒推斷的準確度(Hu et al., 2018; Jonell, 2019; Kuo et al., 2021; Lee et al., 2020; Liu amp; Yang, 2021), 并提供諸如文字、圖像、選項、語音等交互形式, 以適應不同群體的習慣。此外, 還可結合虛擬現實(Virtual Reality, VR)技術來彌補虛擬與真實交互的差距, 強化沉浸感及社會存在(Donker et al., 2019; Miloff et al, . 2020)。
5.3" 開發適應ISIs的DTA測量工具并結合客觀數據進行報告
具備有效的測量工具, 是推進領域研究發展的重要條件, 但在現階段, 研究者們對于如何衡量DTA卻幾乎沒有共識(Gómez Penedo et al., 2020)。例如, 研究者們要么直接使用WAI量表, 要么只對WAI量表進行最小程度的微調(Ellis- Brush, 2020), 但簡單地將“咨詢師”替換為“應用程序”可能有失偏頗。一方面, 將ISIs程序定位為人時, 被試可能產生更高的預期并提高評價標準(Go amp; Sundar, 2019)。另一方面, 面對面治療中的重要因素在ISIs中可能并非同等重要, ISIs具有其自身的特殊性及復雜性(Clarke et al., 2016), 而在基于聊天機器人的ISIs中, 研究者不但要考量程序的交互體驗, 還需要對其中的人類線索進行評估。因此, 研究者未來可在傳統TA理論的基礎上, 還考慮如社會行動者范式(CASA)、恐怖谷效應(UVE)等HCI理論(Smelser amp; Baltes, 2001; Zhang et al., 2020), 針對ISIs情境及干預形式來設計專門的DTA量表(D’Alfonso et al., 2020; Heim et al., 2018)。此外, 越來越多關于TA的研究強調, 需要更準確地識別TA的建立以及破裂的發生(Colli et al., 2019), 但時下DTA的測量幾乎都依賴被試的自我報告(Berger, 2017), 而沒有結合行為、生理數據等進行更為客觀的量化分析。因此, 未來可結合更詳盡的客觀數據(Nof et al., 2021), 對被試的聲學特征、行為軌跡、文本及視聽數據進行建模, 并動態分析當下的DTA質量, 監測聊天機器人與被試的DTA在何時建立、破裂, 進而為聊天機器人的行為決策提供更優的指導。最后, 研究者還可綜合評估量化數據及咨詢師、觀察者的主觀數據, 以加強研究結果的嚴謹性。
5.4" 關注在ISIs中不同療法及不同群體于DTA上所呈現出的新問題
目前, DTA研究中所使用的ISIs程序多是基于CBT設計的, 盡管在線CBT的可行、有效性均得到驗證(Newby et al., 2017; Titov et al., 2015),"但仍有部分群體并未充分受益于此(Rozental et al., 2019; Sunderland et al., 2012)。因此, 在ISIs中仍要開發和測試更多的替代療法。例如, 正念干預(Mindfulness-based Interventions, MBIs)就被認為是CBT的有效替代(Li et al., 2021)。有研究表明, 被試不但在MBIs中的TA得分高于CBT (Jazaieri et al., 2018), 且狀態正念也與TA存在高度的相關(Johnson, 2018)。但是, MBIs與TA關系的研究仍然較少, 在ISIs環境中的類似證據則更是缺乏。因此, 未來的研究可在DTA的研究中使用正念減壓療法(Mindfulness-based Stress Reduction, MBSR)、正念認知療法(Mindfulness-based Cognitive Therapy, MBCT)、接納與承諾療法(Acceptance and Commitment Therapy, ACT)等MBIs, 并嘗試以聊天機器人模擬正念教練, 優化現有在線MBIs的體驗。此外, 基于一種療法的ISIs在不同心理問題(如抑郁、焦慮、恐懼、成癮等)、群體(如青少年、成年人、老年人或男性、女性等)中所建立的DTA水平可能存在差異, 但現有研究對此少有討論(Darcy et al., 2021; Ellis-Brush, 2020; Werz et al., 2021)。因此, 未來的研究在探討DTA與某一癥狀的關系時, 還可將被試劃分為更多的亞組, 以檢驗不同特征人群的結果差異, 進而加深對DTA作用機制的理解。
參考文獻
黃欣榮, 曹賢平. (2022). 元宇宙的技術本質與哲學意義. 新疆師范大學學報(哲學社會科學版), 43(3), 119–126.
張夏恒, 李想. (2022). 國外元宇宙領域研究現狀、熱點及啟示. 產業經濟評論, 10(2), 199–214.
朱旭, 江光榮. (2011a). 工作同盟的概念. 中國臨床心理學雜志, 19(2), 275–280.
朱旭, 江光榮. (2011b). 當事人眼里的工作同盟:質的分析. 心理學報, 43(4), 420–431.
Abd-alrazaq, A. A., Alajlani, M., Alalwan, A. A., Bewick, B. M., Gardner, P., amp; Househ, M. (2019). An overview of the features of chatbots in mental health: A scoping review. International Journal of Medical Informatics, 132, 103978. https://doi.org/10.1016/j.ijmedinf.2019.103978
Alkhaldi, G., Hamilton, F. L., Lau, R., Webster, R., Michie, S., amp; Murray, E. (2016). The effectiveness of prompts to promote engagement with digital interventions: A systematic review. Journal of Medical Internet Research, 18(1), e6. https://doi.org/10.2196/jmir.4790
Ames, H. M., Glenton, C., Lewin, S., Tamrat, T., Akama, E., amp; Leon, N. (2019). Clients’ perceptions and experiences of targeted digital communication accessible via mobile devices for reproductive, maternal, newborn, child, and adolescent health: A qualitative evidence synthesis. Cochrane Database Systematic Reviews, 10(10), CD013447. https://doi.org/10.1002/14651858.cd013447
Anderson, R. E. E., Spence, S. H., Donovan, C. L., March, S., Prosser, S., amp; Kenardy, J. (2012). Working alliance in online cognitive behavior therapy for anxiety disorders in youth: Comparison with clinic delivery and its role in predicting outcome. Journal of Medical Internet Research, 14(3), e88. https://doi.org/10.2196/jmir.1848
Andersson, G., Paxling, B., Wiwe, M., Vernmark, K., Felix, C. B., Lundborg, L., … Carlbring, P. (2012). Therapeutic alliance in guided internet-delivered cognitive behavioural treatment of depression, generalized anxiety disorder and social anxiety disorder. Behaviour Research and Therapy, 50(9), 544–550.
Arndt, A., Rubel, J., Berger, T., amp; Lutz, W. (2020). Outpatient and self-referred participants: Adherence to treatment components and outcome in an internet intervention targeting anxiety disorders. Internet Interventions, Advance online publication. https://doi.org/10.1016/j.invent.2020.100319
Asaeikheybari, G., Hooper, M. W., amp; Huang, M.C. (2021). A context-adaptive smoking cessation system using videos. Smart Health, 19, 100148. https://doi.org/10.1016/j.smhl. 2020.100148
Bachelor, A. (2013). Clients’ and therapists’ views of the therapeutic alliance: Similarities, differences and relationship to therapy outcome. Clinical Psychology and Psychotherapy, 20(2), 118–135.
Bao, Z., Chen, J., Luo, Y., amp; Reani, M. (2022, July). Convincing or odd: Anthropomorphic design cues in chatbots. Paper presented at the meeting of the pacific asia conference on information systems, Sydney, AUS.
Baumeister, H., Reichler, L., Munzinger, M., amp; Lin, J. (2014). The impact of guidance on Internet-based mental health interventions: A systematic review. Internet Interventions, 1(4), 205–215.
Baumel, A., amp; Kane, J. M. (2018). Examining predictors of real-world user engagement with self-guided eHealth interventions: Analysis of mobile apps and websites using a novel dataset. Journal of Medical Internet Research, 20(12), e11491. https://doi.org/10.2196/11491
Baumel, A., amp; Yom-Tov, E. (2018). Predicting user adherence to behavioral eHealth interventions in the real world: Examining which aspects of intervention design matter most. Translational Behavioral Medicine, 8(5), 793–798.
Berger, T. (2017). The therapeutic alliance in internet interventions: A narrative review and suggestions for future research. Psychotherapy Research, 27(5), 511–524.
Berger, T., Boettcher, J., amp; Caspar, F. (2014). Internet-based guided self-help for several anxiety disorders: A randomized controlled trial comparing a tailored with a standardized disorder-specific approach. Psychotherapy, 51(2), 207–219.
Berry, K., Salter, A., Morris, R., James, S., amp; Bucci, S. (2018). Assessing therapeutic alliance in the context of mHealth interventions for mental health problems: Development of the mobile agnew relationship measure (mARM) questionnaire. Journal of Medical Internet Research, 20(4), e90. https://doi.org/10.2196/jmir.8252
Berry, N., Lobban, F., amp; Bucci, S. (2019). A qualitative exploration of service user views about using digital health interventions for self-management in severe mental health problems. BMC Psychiatry, 19(1), 35.
Bickmore, T., Gruber, A., amp; Picard, R. (2005). Establishing the computer-patient working alliance in automated health behavior change interventions. Patient Education and Counseling, 59(1), 21–30.
Bickmore, T. W., Pfeifer, L. M., amp; Jack, B. W. (2009). Taking the time to care: Empowering low health literacy hospital patients with virtual nurse agents. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 1265–1274). New York, NY: Association for Computing Machinery.
Bielinski, L. L., amp; Berger, T. (2020). Internet interventions for mental health: Current state of research, lessons learned and future directions. Counseling Psychology and Psychotherapy, 28(3), 65–83.
Bock, B. C., Rosen, R. K., Barnett, N. P., Thind, H., Walaska, K., Foster, R., … Traficante, R. (2015). Translating behavioral interventions onto mhealth platforms: Developing text message interventions for smoking and alcohol. JMIR MHealth and UHealth, 3(1), e22. https://doi.org/10.2196/mhealth.3779
Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research amp; Practice, 16(3), 252–260.
Brown, A., Mountford, V. A., amp; Waller, G. (2013). Is the therapeutic alliance overvalued in the treatment of eating disorders? International Journal of Eating Disorders, 46(8), 779–782.
Chen, H., Rodriguez, M. A., Qian, M., Kishimoto, T., Lin, M., amp; Berger, T. (2020). Predictors of treatment outcomes and adherence in internet-based cognitive behavioral therapy for social anxiety in China. Behavioural and Cognitive Psychotherapy, 48(3), 291–303.
Chikersal, P., Belgrave, D., Doherty, G., Enrique, A., Palacios, J. E., Richards, D., amp; Thieme, A. (2020). Understanding client support strategies to improve clinical outcomes in an online mental health intervention. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems (pp. 1–16). New York, NY: Association for Computing Machinery.
Clarke, J., Proudfoot, J., Whitton, A., Birch, M. R., Boyd, M., Parker, G., … Fogarty, A. (2016). Therapeutic alliance with a fully automated mobile phone and web-based intervention: Secondary analysis of a randomized controlled trial. JMIR Mental Health, 3(1), e10. https://doi.org/10.2196/ mental.4656
Colli, A., Gentile, D., Condino, V., amp; Lingiardi, V. (2019). Assessing alliance ruptures and resolutions: Reliability and validity of the collaborative interactions scale-revised version. Psychotherapy Research, 29(3), 279–292.
Cummins, R. G., amp; Cui, B. (2014). Reconceptualizing address in television programming: The effect of address and affective empathy on viewer experience of parasocial interaction. Journal of Communication, 64(4), 723–742.
D’Alfonso, S., Lederman, R., Bucci, S., amp; Berry, K. (2020). The digital therapeutic alliance and human-computer interaction. JMIR Mental Health, 7(12), e21895. https:// doi.org/10.2196/21895
Darcy, A., Daniels, J., Salinger, D., Wicks, P., amp; Robinson, A. (2021). Evidence of human-level bonds established with a digital conversational agent : Cross-sectional, retrospective observational study. JMIR Formative Research, 5(5), e27868. https://doi.org/10.2196/27868
Datz, F., Wong, G., amp; L?ffler-Stastka, H. (2019). Interpretation and working through contemptuous facial micro-expressions benefits the patient-therapist relationship. International Journal of Environmental Research and Public Health, 16(24), 4901. https://doi.org/10.3390/ijerph16244901
Deci, E. L., amp; Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York, NY: Plenum.
Donker, T., Cornelisz, I., van Klaveren, C., van Straten, A., Carlbring, P., Cuijpers, P., amp; van Gelder, J. L. (2019). Effectiveness of self-guided app-based virtual reality cognitive behavior therapy for acrophobia: A randomized clinical trial. JAMA Psychiatry, 76(7), 682–690.
Ellis-Brush, K. (2020). Coaching in a digital age : Can a working alliance form between coachee and “coaching app”? (Unpublished doctorial dissertation). Oxford Brookes University.
Elmasri, D., amp; Maeder, A. (2016). A conversational agent for an online mental health intervention. In G. Ascoli, M. Hawrylycz, H. Ali, D. Khazanchi, amp; Y. Shi (Eds.), Lecture notes in computer science: Vol. 9919: Brain informatics and health (pp. 243–251). Berlin, Germany: Springer.
Fenski, F., Rozental, A., Heinrich, M., Knaevelsrud, C., Zagorscak, P., amp; Boettcher, J. (2021). Negative effects in internet-based interventions for depression: A qualitative content analysis. Internet Interventions, 26, 100469. https://doi.org/10.1016/j.invent.2021.100469
Fitzpatrick, K. K., Darcy, A., amp; Vierhile, M. (2017). Delivering cognitive behavior therapy to young adults with symptoms of depression and anxiety using a fully automated conversational agent (Woebot): A randomized controlled trial. JMIR Mental Health, 4(2), e19. https://doi.org/ 10.2196/mental.7785
Fuhr, K., Schr?der, J., Berger, T., Moritz, S., Meyer, B., Lutz, W., Hohagen, F., … Klein, J. P. (2018). The association between adherence and outcome in an Internet intervention for depression. Journal of Affective Disorders, 229, 443–449.
Gaffney, H., Mansell, W., Edwards, R., amp; Wright, J. (2014). Manage your life online (MYLO): A pilot trial of a conversational computer-based intervention for problem solving in a student sample. Behavioural and Cognitive Psychotherapy, 42(6), 731–746.
Gallen, S., Gallen, S., Zurich, E. T. H., amp; Zurich, E. T. H. (2018, June). The impact of interpersonal closeness cues in text-based healthcare chatbots on attachment bond and the desire to continue interacting: An experimental design. Paper presented at the 26th European Conference on Information Systems, Portsmouth, UK.
Go, E., amp; Sundar, S. S. (2019). Humanizing chatbots: The effects of visual, identity and conversational cues on humanness perceptions. Computers in Human Behavior, 97, 304–316.
Goldberg, S. B., Baldwin, S. A., Riordan, K. M., Torous, J., Dahl, C. J., Davidson, R. J., amp; Hirshberg, M. J. (2021). Alliance with an unguided smartphone app: Validation of the digital working alliance inventory. Assessment, 29(6), 1331–1345.
Gómez Penedo, J. M., Berger, T., Grosse Holtforth, M., Krieger, T., Schr?der, J., Hohagen, F., … Klein, J. P. (2020). The working alliance inventory for guided internet interventions (WAI-I). Journal of Clinical Psychology, 76(6), 973–986.
Grekin, E. R., Beatty, J. R., amp; Ondersma, S. J. (2019). Mobile health interventions: Exploring the use of common relationship factors. JMIR MHealth and UHealth, 7(4), e11245. https://doi.org/10.2196/11245
Hardeman, W., Houghton, J., Lane, K., Jones, A., amp; Naughton, F. (2019). A systematic review of just-in-time adaptive interventions (JITAIs) to promote physical activity. International Journal of Behavioral Nutrition and Physical Activity, 16(1), 31.
Hargreaves, A., Dillon, R., Castorina, M., Furey, E., Walsh, J., Fitzmaurice, B., … Donohoe, G. (2018). Predictors of adherence to low support, computerised, cognitive remediation training in psychosis. Psychosis, 10(4), 298–306.
Heim, E., R?tger, A., Lorenz, N., amp; Maercker, A. (2018). Working alliance with an avatar: How far can we go with internet interventions? Internet Interventions, 11, 41–46.
Henson, P., Wisniewski, H., Hollis, C., Keshavan, M., amp; Torous, J. (2019). Digital mental health apps and the therapeutic alliance: Initial review. BJPsych Open, 5(1), e15. https://doi.org/10.1192/bjo.2018.86
Hentati, A., Forsell, E., Ljótsson, B., Kaldo, V., Lindefors, N., amp; Kraepelien, M. (2021). The effect of user interface on treatment engagement in a self-guided digital problem- solving intervention: A randomized controlled trial. Internet Interventions, 26, 100448. https://doi.org/10.1016/ j.invent.2021.100448
Ho, A., Hancock, J., amp; Miner, A. S. (2018). Psychological, relational, and emotional effects of self-disclosure after conversations with a chatbot. Journal of Communication, 68(4), 712–733.
Holter, M. T. S., Ness, O., Johansen, A. B., amp; Brendryen, H. (2020). Making come-alive and keeping un-alive: How people relate to self-guided web-based health interventions. Qualitative Health Research, 30(6), 927–941.
Hortensius, R., amp; Cross, E. S. (2018). From automata to animate beings: The scope and limits of attributing socialness to artificial agents. Annals of the New York Academy of Sciences, 1426(1), 93–110.
Horton, D., amp; Wohl, R. R. (1956). Mass communication and para-social interaction: Observations on intimacy at a distance. Psychiatry, 19(3), 215–229.
Horvath, A. O., amp; Greenberg, L. S. (1989). Development and validation of the working alliance inventory. Journal of Counseling Psychology, 36(2), 223–233.
Hu, T., Xu, A., Liu, Z., You, Q., Guo, Y., Sinha, V., … Akkiraju, R. (2018, April). Touch your heart: A tone- aware chatbot for customer care on social media. Paper presented at the 2018 CHI Conference on Human Factors in Computing Systems, Montreal, Canada.
Izzaty, R. E., Astuti, B., amp; Cholimah, N. (2021). Evaluation of an online cognitive behavioural therapy, mindfulness meditation, and yoga (CBT-MY) intervention for posttraumatic stress disorder: A single arm clinical trial with psychometric and psychophysiological outcomes. Angewandte Chemie International Edition, 6(11), 951–952.
Jasper, K., Weise, C., Conrad, I., Andersson, G., Hiller, W., amp; Kleinst?uber, M. (2014). The working alliance in a randomized controlled trial comparing internet-based self-help and face-to-face cognitive behavior therapy for chronic tinnitus. Internet Interventions, 1(2), 49–57.
Jazaieri, H., Goldin, P. R., amp; Gross, J. J. (2018). The role of working alliance in CBT and MBSR for social anxiety disorder. Mindfulness, 9(5), 1381–1389.
Johansson, M., Berman, A. H., Sinadinovic, K., Lindner, P., Hermansson, U., amp; Andréasson, S. (2021). Effects of internet-based cognitive behavioral therapy for harmful alcohol use and alcohol dependence as self-help or with therapist guidance: Three-armed randomized trial. Journal of Medical Internet Research, 23(11), e29666. https://doi.org/ 10.2196/29666
Johnson, D. A. (2018). The relationship between state mindfulness and working alliance among counselors- in-training. Journal of Humanistic Counseling, 57(1), 31–50.
Jonell, P. (2019). Using social and physiological signals for user adaptation in conversational agents. In N. Agmon, M. E. Taylor, E. Elkind, amp; M. Veloso (Eds.), Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems (pp. 2420–2422). Richland, SC: International Foundation for Autonomous Agents and Multiagent Systems.
Kaiser, J., Hanschmidt, F., amp; Kersting, A. (2021). The association between therapeutic alliance and outcome in internet-based psychological interventions: A meta-analysis. Computers in Human Behavior, 114, 106512. https:// doi.org/10.1016/j.chb.2020.106512
Kang, S. H., amp; Gratch, J. (2014). Exploring users’ social responses to computer counseling interviewers’ behavior. Computers in Human Behavior, 34, 120–130.
Kelders, S. M., Kok, R. N., Ossebaard, H. C., amp; van Gemert-Pijnen, J. E. W. C. (2012). Persuasive system design does matter: A systematic review of adherence to web-based interventions. Journal of Medical Internet Research, 14(6), e152. https://doi.org/10.2196/jmir.2104
Kiesler, S., Siegel, J., amp; McGuire, T. W. (1984). Social psychological aspects of computer-mediated communication. American Psychologist, 39(10), 1123–1134.
Kiluk, B. D., Serafini, K., Frankforter, T., Nich, C., amp; Carroll, K. M. (2014). Only connect: The working alliance in computer-based cognitive behavioral therapy. Behaviour Research and Therapy, 63, 139–146.
Kim, H. C., Cha, M. C., amp; Ji, Y. G. (2021). The impact of an agent’s voice in psychological counseling: Session evaluation and counselor rating. Applied Sciences, 11(7), 2893. https://doi.org/10.3390/app11072893
Klasen, M., Knaevelsrud, C., amp; B?ttche, M. (2013). Die therapeutische beziehung in internetbasierten therapieverfahren: Ein überblick. Nervenarzt, 84(7), 823–831.
Kumar, T. S. (2021). Construction of hybrid deep learning model for predicting children behavior based on their emotional reaction. Journal of Information Technology, 3(1), 29–43.
Kuo, P. H., Lin, S. T., Hu, J., amp; Huang, C. J. (2021). Multi- sensor context-aware based chatbot model: An application of humanoid companion robot. Sensors, 21(15), 5132.
Laranjo, L., Dunn, A. G., Tong, H. L., Kocaballi, A. B., Chen, J., Bashir, R., … Coiera, E. (2018). Conversational agents in healthcare: A systematic review. Journal of the American Medical Informatics Association, 25(9), 1248–1258.
Lederman, R., amp; D’Alfonso, S. (2021). The Digital therapeutic alliance : Prospects and considerations. JMIR Mental Health, 8(7), e31385. https://doi.org/10.2196/31385
Lee, S., Lee, N., amp; Sah, Y. J. (2020). Perceiving a mind in a chatbot: Effect of mind perception and social cues on co-presence, closeness, and intention to use. International Journal of Human-Computer Interaction, 36(10), 930–940.
Lew, Z., amp; Walther, J. B. (2022). Social scripts and expectancy violations: Evaluating communication with human or ai chatbot interactants. Media Psychology, 25(6), 1–16.
Li, J., Cai, Z., Li, X., Du, R., Shi, Z., Hua, Q., … Zhan, X. (2021). Mindfulness-based therapy versus cognitive behavioral therapy for people with anxiety symptoms: A systematic review and meta-analysis of random controlled trials. Annals of Palliative Medicine, 10(7), 7596–7612.
Linardon, J., Cuijpers, P., Carlbring, P., Messer, M., amp; Fuller-Tyszkiewicz, M. (2019). The efficacy of app- supported smartphone interventions for mental health problems: A meta-analysis of randomized controlled trials. World Psychiatry, 18(3), 325–336.
Liu, H., Peng, H., Song, X., Xu, C., amp; Zhang, M. (2022). Using AI chatbots to provide self-help depression interventions for university students : A randomized trial of effectiveness. Internet Interventions, 27, 100495. https://doi.org/10.1016/j.invent.2022.100495
Liu-Thompkins, Y., Okazaki, S., amp; Li, H. (2022). Artificial empathy in marketing interactions: Bridging the human-AI gap in affective and social customer experience. Journal of the Academy of Marketing Science, 50, 1198–1218.
Liu, X., Zhang, J., amp; Guo, C. (2013). Full-text citation analysis : A new method to enhance scholarly networks. Journal of the American Society for Information Science and Technology, 64(9), 1852–1863.
Liu, Y., amp; Yang, R. (2021). Federated learning application on depression treatment robots (DTbot). In Proceeding of the 13th International Conference on Computer Research and Development (pp. 121–124). Piscataway, NJ: IEEE.
Luborsky, L. L. (1976). Helping alliance in psychotherapy. In J. L. Cleghhorn (Ed.), Successful Psychotherapy (pp. 92–116). New York, NY: Brunner/Mazel.
Luo, B., Lau, R. Y. K., Li, C., amp; Si, Y. (2022). A critical review of state-of-the-art chatbot designs and applications. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 12(1), e1434. https://doi.org/10.1002/ widm.1434
Mackie, C., Dunn, N., Maclean, S., Testa, V., Heisel, M., amp; Hatcher, S. (2017). A qualitative study of a blended therapy using problem solving therapy with a customised smartphone app in men who present to hospital with intentional self-harm. Evidence-Based Mental Health, 20(4), 118–122.
Malins, S., Biswas, S., Rathbone, J., Vogt, W., Pye, N., Levene, J., … Russell, J. (2020). Reducing dropout in acceptance and commitment therapy, mindfulness-based cognitive therapy, and problem-solving therapy for chronic pain and cancer patients using motivational interviewing. British Journal of Clinical Psychology, 59(3), 424–438.
Medhi Thies, I., Menon, N., Magapu, S., Subramony, M., amp; O’neill, J. (2017). How do you want your chatbot? An exploratory Wizard-of-Oz study with young, urban Indians. In R. Bernaupt, G. Dalvi, A. Joshi, K. Balkrishan, J. O'Neill, amp; M. Winckler (Eds.), Proceeding of the IFIP Conference on Human-Computer Interaction (pp. 441– 459). Berlin, Germany: Springer.
Meyer, B., Bierbrodt, J., Schr?der, J., Berger, T., Beevers, C. G., Weiss, M., … Klein, J. P. (2015). Effects of an Internet intervention (Deprexis) on severe depression symptoms: Randomized controlled trial. Internet Interventions, 2(1), 48–59.
Miloff, A., Carlbring, P., Hamilton, W., Andersson, G., Reuterski?ld, L., amp; Lindner, P. (2020). Measuring alliance toward embodied virtual therapists in the era of automated treatments with the Virtual Therapist Alliance Scale (VTAS): Development and psychometric evaluation. Journal of Medical Internet Research, 22(3), e16660.
Müssener, U. (2021). Digital encounters : Human interactions in mHealth behavior change interventions. Digital Health, 7, 1–6.
Nass, C., Steuer, J., amp; Tauber, E. R. (1994, April). Computers are social actors. Paper presented at the SIGCHI Conference on Human Factors in Computing Systems, Boston, MA.
Negri, A., Christian, C., Mariani, R., Belotti, L., Andreoli, G., amp; Danskin, K. (2019). Linguistic features of the therapeutic alliance in the first session: A psychotherapy process study. Research in Psychotherapy: Psychopathology, Process and Outcome, 22(1), 71–82.
Neuberg, S. L. (1989). The goal of forming accurate impressions during social interactions: Attenuating the impact of negative expectancies. Journal of Personality and Social Psychology, 56(3), 374–386.
Newby, J. M., Mewton, L., amp; Andrews, G. (2017). Transdiagnostic versus disorder-specific internet-delivered cognitive behaviour therapy for anxiety and depression in primary care. Journal of Anxiety Disorders, 46, 25–34.
Nof, A., Amir, O., Goldstein, P., amp; Zilcha-Mano, S. (2021). What do these sounds tell us about the therapeutic alliance: Acoustic markers as predictors of alliance. Clinical Psychology and Psychotherapy, 28(4), 807–817.
Nofz, M. P., amp; Vendy, P. (2002). When computers say it with feeling: Communication and synthetic emotions in Kubrick’s 2001: A space odyssey. Journal of Communication Inquiry, 26(1), 26–45.
Norcross, J. C. (2002). Psychotherapy relationships that work: Therapist contributions and responsiveness to patients. Oxford University Press.
Oinas-Kukkonen, H., amp; Harjumaa, M. (2009). Persuasive systems design: Key issues, process model, and system features. Communications of the Association for Information Systems, 24(1), 485–500.
Penedo, J. M. G., Babl, A. M., Holtforth, M., Hohagen, F., Krieger, T., Lutz, W., … Berger, T. (2020). The association of therapeutic alliance with long-term outcome in a guided internet intervention for depression: Secondary analysis from a randomized control trial. Journal of Medical Internet Research, 22(3), e15824. https://doi.org/ 10.2196/15824
Perski, O., Blandford, A., West, R., amp; Michie, S. (2017). Conceptualising engagement with digital behaviour change interventions: A systematic review using principles from critical interpretive synthesis. Translational Behavioral Medicine, 7(2), 254–267.
Perski, O., Crane, D., Beard, E., amp; Brown, J. (2019). Does the addition of a supportive chatbot promote user engagement with a smoking cessation app? An experimental study. Digit Health, 5, 1–13.
Pihlaja, S., Stenberg, J. H., Joutsenniemi, K., Mehik, H., Ritola, V., amp; Joffe, G. (2018). Therapeutic alliance in guided internet therapy programs for depression and anxiety disorders – A systematic review. Internet Interventions, 11, 1–10.
Pratap, A., Neto, E. C., Snyder, P., Stepnowsky, C., Elhadad, N., Grant, D., … Omberg, L. (2020). Indicators of retention in remote digital health studies: A cross-study evaluation of 100, 000 participants. Npj Digital Medicine, 3(1), 1–10.
Probst, G. H., Berger, T., amp; Flückiger, C. (2019). The alliance-outcome relation in internet-based interventions for psychological disorders: A correlational meta-analysis. Verhaltenstherapie, 29(3), 182–195.
Provoost, S. (2021). Embodied Conversational Agents in Internet-Based Cognitive Behavioral Therapy for Depression Bridging the Gap Between Unguided and Guided Interventions (Unpublished doctorial dissertation). Vrije Universiteit Amsterdam.
Provoost, S., Lau, H. M., Ruwaard, J., amp; Riper, H. (2017). Embodied conversational agents in clinical psychology: A scoping review. Journal of Medical Internet Research, 19(5), e151. https://doi.org/10.2196/jmir.6553
Puls, H. C., Schmidt, R., Herpertz, S., Zipfel, S., Tuschen- Caffier, B., Friederich, H. C., … Hilbert, A. (2020). Adherence as a predictor of dropout in internet-based guided self-help for adults with binge-eating disorder and overweight or obesity. International Journal of Eating Disorders, 53(4), 555–563.
Radomski, A. D., Wozney, L., McGrath, P., Huguet, A., Hartling, L., Dyson, M. P., … Newton, A. S. (2019). Design and delivery features that may improve the use of internet-based cognitive behavioral therapy for children and adolescents with anxiety: A realist literature synthesis with a persuasive systems design perspective. Journal of Medical Internet Research, 21(2), e11128. https://doi.org/ 10.2196/11128
Rajagopal, A, Nirmala, V., Andrew, J., amp; Arun, M. (2021). Novel AI to avert the mental health crisis in COVID- 19 : Novel application of GPT2 in cognitive behaviour therapy. Research Square, Advance online publication. https://doi.org/ 10.21203/rs.3.rs-382748/v3
Richards, D., Alsharbi, B., amp; Abdulrahman, A. (2020). Can I help you? Preferences of young adults for the age, gender and ethnicity of a virtual support person based on individual differences including personality and psychological state. In Proceedings of the Australasian Computer Science Week Multiconference (pp. 1–10). New York, NY: Association for Computing Machinery.
Rietz, T., Benke, I., amp; Maedche, A. (2019). The impact of anthropomorphic and functional chatbot design features in enterprise collaboration systems on user acceptance. In Proceedings of the 14th International Conference on Wirtschaftsinformatik (pp. 1642–1656). Siegen, Germany.
Rodrigues, T. R. C., Reijnders, T., de Buisonjé, D. D., Kowatsch, T., Janssen, V. R., Kraaijenhagen, R. A., … Evers, A. W. M. (2021). Human cues in self-help lifestyle interventions : An experimental field study (Preprint). Journal of Medical Internet Research, Advance online publication. https://doi.org/10.2196/preprints.30057
Rogers, C. R. (1957). The necessary and sufficient conditions of therapeutic personality change. Journal of Consulting Psychology, 21(2), 95–103.
Rollnick, S., Butler, C. C., Kinnersley, P., Gregory, J., amp; Mash, B. (2010). Motivational interviewing. BMJ, 340, 1242–1245.
Rozental, A., Andersson, G., amp; Carlbring, P. (2019). In the absence of effects: An individual patient data meta- analysis of non-response and its predictors in internet- based cognitive behavior therapy. Frontiers in Psychology, 10, 589.
Rusbult, C. E., Drigotas, S. M., amp; Verette, J. (1994). The investment model: An interdependence analysis of commitment processes and relationship maintenance phenomena. In D. J. Canary amp; L. Stafford (Eds.), Communication and relational maintenance (pp. 115–139). Academic Press.
Samsudin, C. M. (2020). The effect of chatbots response latency on users’ trust (Unpublished doctorial dissertation). University of Oklahoma.
Sauer-Zavala, S., Boswell, J. F., Bentley, K. H., Thompson-Hollands, J., Farchione, T. J., amp; Barlow, D. H. (2018). Expectancies, working alliance, and outcome in transdiagnostic and single diagnosis treatment for anxiety disorders: An investigation of mediation. Cognitive Therapy and Research, 42(2), 135–145.
Sezgin, E., Sirrianni, J., amp; Linwood, S. L. (2022). Operationalizing and implementing pretrained, large artificial intelligence linguistic models in the US health care system: Outlook of generative pretrained transformer 3 (GPT-3) as a service model. JMIR Medical Informatics, 10(2), e32875. https:// doi.org/10.2196/32875
Smelser, N. J., amp; Baltes, P. B. (Eds). (2001). International encyclopedia of the social amp; behavioral sciences (Vol. 11). Amsterdam, Netherlands: Elsevier.
Song, Y., Yan, R., Li, C.-T., Nie, J.-Y., Zhang, M., amp; Zhao, D. (2018). An ensemble of retrieval-based and generation- based human-computer conversation systems. In Proceedings of the International Conference of Learning Representation ICLR 2018 (pp. 4382–4388). ICLR.
Suler, J. (2004). The online disinhibition effect. Cyberpsychology amp; Behavior, 7(3), 321–326.
Sun, S., Lin, D., Goldberg, S., Shen, Z., Chen, P., Qiao, S., … Operario, D. (2021). A mindfulness-based mobile health (mHealth) intervention among psychologically distressed university students in quarantine during the COVID-19 pandemic: A randomized controlled trial. Journal of Counseling Psychology, 69(2), 157–171.
Sundar, S. S. (2008). The MAIN model: A heuristic approach to understanding technology effects on credibility. In M. J. Metzger, A. J. Flanagin, D. John, amp; C. T. MacArthur (Eds.), Digital media, youth, and credibility (pp. 73–100). Cambridge, MA: MIT press.
Sunderland, M., Wong, N., Hilvert-Bruce, Z., amp; Andrews, G. (2012). Investigating trajectories of change in psychological distress amongst patients with depression and generalised anxiety disorder treated with internet cognitive behavioural therapy. Behaviour Research and Therapy, 50(6), 374–380.
Tanana, M. J., Soma, C. S., Kuo, P. B., Bertagnolli, N. M., Dembe, A., Pace, B. T., … Imel, Z. E. (2021). How do you feel? Using natural language processing to automatically rate emotion in psychotherapy. Behavior Research Methods, 53(5), 2069–2082.
Taylor, H., Strauss, C., amp; Cavanagh, K. (2021). Can a little bit of mindfulness do you good? A systematic review and meta-analyses of unguided mindfulness-based self-help interventions. Clinical Psychology Review, 89, 102078. https://doi.org/10.1016/j.cpr.2021.102078
ter Stal, S., Kramer, L. L., Tabak, M., op den Akker, H., amp; Hermens, H. (2020). Design features of embodied conversational agents in eHealth: A literature review. International Journal of Human Computer Studies, 138, 102409. https://doi.org/10.1016/j.ijhcs.2020.102409
Titov, N., Dear, B. F., Staples, L. G., Terides, M. D., Karin, E., Sheehan, J., …Wootton, B. M. (2015). Disorder- specific versus transdiagnostic and clinician-guided versus self-guided treatment for major depressive disorder and comorbid anxiety disorders: A randomized controlled trial. Journal of Anxiety Disorders, 35, 88–102.
Tremain, H., McEnery, C., Fletcher, K., amp; Murray, G. (2020). The therapeutic alliance in digital mental health interventions for serious mental illnesses: Narrative review. JMIR Mental Health, 7(8), e17204. https://doi.org/ 10.2196/17204
Vaidyam, A. N., Wisniewski, H., Halamka, J. D., Kashavan, M. S., amp; Torous, J. B. (2019). Chatbots and conversational agents in mental health: A review of the psychiatric landscape. The Canadian Journal of Psychiatry, 64(7), 456–464.
Valentine, L., D’Alfonso, S., amp; Lederman, R. (2022). Recommender systems for mental health apps: Advantages and ethical challenges. AI amp; SOCIETY, Advance online publication. https://doi.org/10.1007/s00146-021-01322-w
Wanser, S. H., Vitale, K. R., Thielke, L. E., Brubaker, L., amp; Udell, M. A. R. (2019). Spotlight on the psychological basis of childhood pet attachment and its implications. Psychology Research and Behavior Management, 12, 469–479.
Waytz, A., Gray, K., Epley, N., amp; Wegner, D. M. (2010). Causes and consequences of mind perception. Trends in Cognitive Sciences, 14(8), 383–388.
Weisel, K. K., Fuhrmann, L. M., Berking, M., Baumeister, H., Cuijpers, P., amp; Ebert, D. D. (2019). Standalone smartphone apps for mental health: A systematic review and meta-analysis. npj Digital Medicine, 2(1), 118.
Weizenbaum, J. (1983). ELIZA: A computer program for the study of natural language communication between man and machine. Communications of the ACM, 26(1), 23–28.
Werz, J., Voderholzer, U., amp; Tuschen-Caffier, B. (2021). Alliance matters: But how much? A systematic review on therapeutic alliance and outcome in patients with anorexia nervosa and bulimia nervosa. Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity, 27, 1279–1295. https://doi.org/10.1007/s40519-021-01281-7
Wiese, E., Weis, P. P., Bigman, Y., Kapsaskis, K., amp; Gray, K. (2022). It’s a match: Task assignment in human–robot collaboration depends on mind perception. International Journal of Social Robotics, 14(1), 141–148.
Yao, H., Chen, J. H., amp; Xu, Y. F. (2020). Rethinking online mental health services in China during the COVID-19 epidemic. Asian Journal of Psychiatry, Advance online publication. https://doi.org/10.1016/j.ajp.2020.102015
Yin, Z., Chang, K., amp; Zhang, R. (2017). Deepprobe: Information directed sequence understanding and chatbot design via recurrent neural networks. In Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 2131–2139). New York, NY: Association for Computing Machinery.
Zhang, J., Oh, Y. J., Lange, P., Yu, Z., amp; Fukuoka, Y. (2020). Artificial intelligence chatbot behavior change model for designing artificial intelligence chatbots to promote physical activity and a healthy diet: Viewpoint. Journal of Medical Internet Research, 22(9), e22845. https://doi.org/ 10.2196/22845
Zhang, X., Lewis, S., Firth, J., Chen, X., amp; Bucci, S. (2021). Digital mental health in China: A systematic review. Psychological Medicine, 51(15), 2552–2570.
Zhou, M. X., Wang, C., Mark, G., Yang, H., amp; Xu, K. (2019). Building real-world chatbot interviewers: Lessons from a wizard-of-oz field study. In C. Trattner, D. Parra, amp; N. Riche (Eds.), Proceedings of ACM IUI 2019 Workshops (Vol. 2327, pp. 1–6). New York, NY: ACM.
Abstract: To address the issue of users’ poor engagement, researchers have recently integrated the therapeutic alliance (TA) concept with Internet-based self-help interventions (ISIs). Digital therapeutic alliance (DTA) are TAs established within a digital environment. A chatbot can replicate human guidance due to the rapid development of artificial intelligence, and it is easier to establish relationships with users than traditional ISIs. Furthermore, it may enhance DTA through amiability, respectfulness, attentiveness, encouragement, sincere comprehension, and mutual trust, which presents a novel solution to this issue. Future research can investigate DTA from the perspectives of affecting factors, technology iteration of ISIs, measurement specification, and experimental manipulation.
Keywords: digital therapeutic alliance, chatbot, relational cues