[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123746-en":3,"doc-seo-123746-105":30,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},123746,16904993612988,"Olivia Brown","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","Combining Natural Language and Machine Learning for Predicting Survey Responses of Social Constructs in a Dyad","Measuring social constructs such as engagement, rapport, and trust often depends on surveys and behavioral observation. The paper presents a predictive approach that combines psychology-informed language analysis features with machine learning to forecast participant survey responses in a training setting. Data were collected from 120 dyad transcripts within the SCOTTIE project, using trainer–subject interactions across teleconference, virtual reality, and face-to-face conditions. Results indicate low utterance counts and weak feature–response correlations, limiting the verbal expression of social behaviors, though the methodology is described as sound.","Combining Natural Language and Machine Learning for Predicting Survey Responses of Social Constructs in a Dyad *  \nBruno Abreu Calfa**, Mohammadamin Sanaei***, Peggy Wu**, Stephen Gilbert***, Andrew  \nRadlbeck**, Brett Israelsen**  \nAbstract—Measuring social constructs such as engagement, rapport, and trust often rely heavily on surveys and behavioral observations. This paper describes a method to use features identified by psychology-based language analysis, combined with machine learning, to predict participant survey responsesin a training context based on 120 dyad transcripts. The method analyzed data collected from subjects performing a circuit board training task within the project called SCOTTIE, Systematic Communication Objectives and Telecommunications Technology Investigations and Evaluations. In this study, the collected data showed low utterance count and a lack of correlation between features and survey responses, suggesting that the context in which the interactions occurred may limit opportunities for interlocutors to manifest social behaviors verbally, which in turn affected the ability to use language analysis to predict subject perceptions of the interaction. However, the methodology appears sound.  \nI. INTRODUCTION  \nOne modality in which humans exhibit social behaviors is language. Linguistic categories [1] have been used in diverse applications from measuring emotional expression [2], to evaluating team dynamics through discourse [3], and identifying correlations between written student selfintroductions with course performance [4] . This paper describes the use of Linguistic Inquiry and Word Count (LIWC) categories to examine transcripts between trainerstudent pairs in a project called Systematic Communication Objectives and Telecommunications Technology Investigations and Evaluations (SCOTTIE) . SCOTTIE’s goal is to investigate the impact of the interaction media on the effectiveness of achieving communication objectives. The definition of communication objectives is described in [5] . Briefly, the communication objectives of interest include copresence, engagement, virtual embodiment, rapport, perceived usability, trust, and mental workload.  \nII. METHOD  \nThe study protocol involved a scripted scenario where a trained confederate staff member provided instruction on a circuit board repair task to subjects. Subjects were assigned to one of three conditions. Trainer-subject interactions were either conducted through teleconference software (Zoom), in virtual reality (also called extended reality or XR), or Face-toFace (F2F) visits. All conditions used the same circuit board simulator testbed, where the trainer-subject pair used screen  \n* Research supported by the Advanced Research Projects Agency-Energy (ARPA-E), U.S. Department of Energy.  \n**Is with the Raytheon Technology,(e-mail in order of the writers: [bruno.abreucalfa@rtx.com](bruno.abreucalfa@rtx.com); [peggy.wu@rtx.com](peggy.wu@rtx.com); [andrew.radlbeck@rtx.com](andrew.radlbeck@rtx.com); [bris1087@colorado.edu](bris1087@colorado.edu)).  \nshare, controlled their own avatars in the virtual environment, or shared physical screen, for the Zoom, XR and F2F conditions respectively. The testbed and virtual environment, called Circuit World, is software created by the study staff as described in [6] . At the start of the trial, a research assistant explained the purpose of the study and obtained informed consent. The researcher then administered pre-trial surveys. Upon survey completion and other introductory materials, the trainer entered the session. The trainer provided subjects with approximately 15 minutes of instruction on how to repair a specific circuit board and invited subjects to ask questions. The trainer then left the session, and the researcher initiated the test portion of the session, cuing the testbed for the subject to repair a virtual circuit and complete a multiple-choice quiz based on knowledge conveyed during training. Subjects ","cbCaitwvfcDnVK7C","https://ap.wps.com/l/cbCaitwvfcDnVK7C","pdf",358367,1,4,"English","en",105,"# Abstract\n# Introduction\n# Method\n# Feature Extraction from Transcript Data","[{\"question\":\"How does the study predict survey responses about social constructs?\",\"answer\":\"It extracts language features from dyad transcripts using psychology-based linguistic analysis categories and combines them with machine learning to predict participants’ survey responses in a training context.\"},{\"question\":\"What interaction conditions and data were used?\",\"answer\":\"Trainer–subject interactions occurred in teleconference (Zoom), virtual reality (XR/extended reality), and face-to-face formats, using the same circuit board training testbed and totaling 120 dyad transcripts.\"},{\"question\":\"Why did the feature analysis perform poorly in predicting survey responses?\",\"answer\":\"The study observed low utterance counts and a lack of correlation between extracted features and survey responses, suggesting the interaction context limited opportunities for participants to express social behaviors verbally.\"}]","Combining Natural Language and Machine Learning for Predicting Survey Responses of Social Constructs in a Dyad | PDF",1785818302,10,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":28},"combining-natural-language-and-machine-learning-for-predicting-survey-responses-of-social-constructs-in-a-dyad","",{"@graph":36,"@context":84},[37,53,67],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":21},"https://docshare.wps.com/document/combining-natural-language-and-machine-learning-for-predicting-survey-responses-of-social-constructs-in-a-dyad/123746/",{"url":52,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-08-04",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"How does the study predict survey responses about social constructs?","Question",{"text":74,"@type":75},"It extracts language features from dyad transcripts using psychology-based linguistic analysis categories and combines them with machine learning to predict participants’ survey responses in a training context.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"What interaction conditions and data were used?",{"text":79,"@type":75},"Trainer–subject interactions occurred in teleconference (Zoom), virtual reality (XR/extended reality), and face-to-face formats, using the same circuit board training testbed and totaling 120 dyad transcripts.",{"name":81,"@type":72,"acceptedAnswer":82},"Why did the feature analysis perform poorly in predicting survey responses?",{"text":83,"@type":75},"The study observed low utterance counts and a lack of correlation between extracted features and survey responses, suggesting the interaction context limited opportunities for participants to express social behaviors verbally.","https://schema.org",{"og:url":52,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,127,130,133],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":46,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":29,"doc_module":4,"doc_module_name":46,"category_name":131,"show_sort_weight":29,"slug":132},"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":46,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]