AI News, Could robots be counselors? Early research shows positive user experience

Could robots be counselors? Early research shows positive user experience

Many participants in the University of Plymouth study praised the 'non-judgemental' nature of the humanoid NAO robot as it delivered its session -- with one even saying they preferred it to a human.

The role of the interviewer in MI is mainly to evoke a conversation about change and commitment, and the robot was programmed with a set script designed to elicit ideas and conversation on how someone could increase their physical activity.

Participants found it especially useful to hear themselves talking about their behaviour aloud, and liked the fact that the robot didn't interrupt, which suggests that this new intervention has a potential advantage over other technology-delivered adaptations of MI.

JMIR Publications

Lifestyle factors such as physical inactivity impose a considerable burden on society’s health care resources and individuals’

Participants in qualitative studies focusing on weight management say they want motivational support to make lifestyle changes [2,3], but public health budgets constrain society’s ability to offer face-to-face counseling [4].

This collaborative stance is considered important, because people are likely to react to directive, advice-giving, (doctor-patient) counseling styles by trying to justify their current behavior [10,11].

The aim of MI is to encourage the client to voice their own arguments for change, as hearing oneself arguing for change increases belief that change is important and will happen [12].

Two streams of development dominated early robotics: remote navigation for observing hard-to-reach environments and manipulation for replacing human manual work in industries.

They have proven acceptable and effective for helping children with type 1 diabetes to learn about their condition and how to manage it [17] and are being trialed as therapeutic aids for children with autism spectrum disorders, with results showing therapeutic outcomes similar to those of one-to-one therapy [18,19].

Robots have also become personal trainers, instructing and motivating the completion of exercises such as spinning, rowing, and bodyweights [20] or engaging elderly users in physical exercises [21].

They have served as weight loss coaches, stimulating tracking of calorie consumption and exercise, and being twice as effective as a stand-alone computer or paper log [22].

Interfaces have generally relied on participants entering text or selecting preprogrammed options, making the intervention less person-centered than is ideal and removing the benefits central to MI of hearing oneself argue for change.

They attributed the lack of benefit of MI to a lack of fluency in the dialogue between the robot and the participant, with errors in speech recognition and incongruous nonverbal behaviors destroying the illusion of a meaningful two-way conversation.

A complete motivational interview, with personally tailored questions and reflections upon the client’s answers, still poses substantial challenges to robot speech recognition and artificial intelligence.

In contrast to previous attempts to automate MI, apart from Kanaoka and Mutlu’s study, the focus of the interview was on encouraging participants to talk to the robot about their motivation for change, using open questions designed to draw attention to the discrepancy between the participant’s current behaviors and core values.

However, if this approach succeeds in encouraging participants to talk freely about their concerns and their plans, we contend that it would present a substantial step forward in the use of technology to deliver motivational support.

The questions covered MI elements such as advantages and disadvantages of the status quo, optimism about change, intention to change, evocation of ideas about change, hypothetical change, setting goals, and arriving at a plan [32].

Because this personalized reflection is not possible in a prescripted interview, we sought to amplify emotion using open questions to encourage the participant to think deeply about their incentives.

To help readers understand the strengths and weaknesses of the robot’s script, two of the authors trained in MI characterized it using Shingleton and Palfai’s [24] schema for rating technology-delivered adaptations of MI, which was published after we developed the robot interview.

We used an anonymous, computerized questionnaire rather than a semistructured interview because we wanted participants to feel as free as possible to give an honest account of their experiences and not feel socially pressured into praising the robot.

Was it easy or difficult to use?”), how they felt about listening to themselves discussing their goals aloud (because this is a core component of MI), and whether they perceived an impact of the interview on their motivation (“Did this interview with the robot affect your motivation?

Eleven participants were aged 18 to 25 years, 4 participants 26 to 33 years, 1 participant 34 to 42 years, 2 participants 43 to 60 years, and 2 participants above 61 years.

In session I, participants answered the robot’s questions out loud in a simulated conversation with the robot, with the participants touching the robot’s head sensor to advance to the next question.

One week later, in lab session II, participants returned to the lab and evaluated the intervention through a computerized evaluative questionnaire with open-ended questions and typed answers.

answers to the evaluative questionnaire were content-analyzed utilizing a three step methodology recommended by Boyatzis [35]: (1) sampling and design, (2) developing themes and codes, and (3) validating and applying the codes.

Then, two new raters, with no involvement in the study, applied the adjusted coding scheme to five further randomly selected units of analysis by deciding if each item in the code was mentioned or not.

The interview evaluation theme incorporated answers to most of the questions and covered specific feelings experienced during the interview (for example, feeling relaxed, engaged, or self-conscious) and usability of the interface.

The overall evaluation theme covered impressions of the intervention as a whole and suggestions for improvements, particularly but not solely covering responses to the questions about the best and worst aspects of the intervention.

The theme on motivation covered ideas that participants used spontaneously, whereas the PA theme covered impressions of whether the interview affected motivation and activity in the week after the interview.

evaluations of the interview clustered around four subthemes: how they felt about the interaction with the robot, their evaluation of the script, usability of the interface, and their experiences of hearing themselves speaking aloud to the robot.

Although the novelty of being in proximity to a robot contributed to the initial awkwardness, it also added to the enjoyment of the experience, as illustrated below: For some participants, the lack of a personal response prevented them feeling connected with the robot: However, this participant [P2] later identified advantages of the robot over a human interviewer: Others also drew comparisons with talking to a human, and some preferred it because they felt they could talk without being judged: Most participants found the questions clear and easy to understand.

and motivational books: Participants wrote about challenges (2.2) that make it hard to keep themselves motivated, including health problems, bad weather, winter, laziness or being tired, and social distractions: There was mixed success in terms of whether participants achieved their goal for the week after the robot interview.

Common themes were the problem of not being able to replay a question that had not been understood, needing some time to get used to the robot, and wanting a more natural way of progressing to the next question: We developed a technology-delivered adaptation of MI using a humanoid robot.

We tried to avoid these problems by using the robot to deliver a series of open questions and requiring the participant to press the robot’s head sensor when they had finished talking and were ready to advance to the next question.

This social support could include the ideas suggested by participants, such as reminding them of their plan, providing encouragement, or using imagery to strengthen motivation, for example, by guiding visualization of the goal and how good it will feel to succeed [40].

Providing a longer introduction before beginning the motivational interview could help address some of the drawbacks identified by participants, including discomfort at being close to the robot and having to touch it, and difficulty understanding its speech.

Without more sophisticated speech recognition and branching logic, the robot is unable to reflect the participant’s meaning, affirm their choices and autonomy, or summarize what they have said (although we included suggestions, by the robot, that the participant summarize their plan).

However, a skilled MI practitioner would elicit the participant’s appraisal of their plan, rather than directly advising against it, and this approach could be reproduced in the robot, as we did in this intervention through asking questions that probed an issue deeply before moving to the next.

The NAO robot’s head, hands, and feet sensors also provide opportunities to follow different paths through the prescripted interview—for example, participants could choose information about diet by pressing a hand sensor or about exercise by pressing a foot sensor.

Further research should test the robot interview with different populations, including those who wish to start being physically active and those who wish to increase their activity, and measure their pre- and postintervention motivation and behavior.

To maximize the potential for observing benefits over meaningful timescales, we suggest that a series of interactions be designed to incorporate reminders and follow-up sessions so that the robot provides ongoing support for behavior change.

Combining analyses of change and sustain talk with quantitative data on behavior change could reveal whether a robot-led motivational interview affected motivation and behavior via the same mechanisms as human-led MI.

Because participants enjoyed the interaction and liked the novelty, a robot-delivered interview may help engage people to discuss sensitive issues and to get a feel for what counseling would be like, encouraging self-help or help-seeking earlier in the time course of a problem.

Given that our adult participants were concerned about being judged by another adult, the robot could be particularly important for encouraging children and adolescents to discuss mental health issues, as they may be more susceptible to fears of being judged or misunderstood by an adult.

The generic nature of the interview means it can easily be modified for a wide variety of target behaviors, potentially providing motivational support for the very large number of people who struggle with conditions such as addiction or obesity but do not meet the criteria for accessing professional support.

Concern about being judged by a human interviewer came across strongly in praise for the nonjudgmental nature of the robot, suggesting that robots may be particularly helpful for eliciting talk about sensitive issues.

CBC Conf 18: Prize winning oral abstracts

The conference International Advisory Board have selected 2 prize winning abstracts judged on their importance, strength of methodology, originality, clarity and translation potential.

Self-efficacy was promoted by asking about past successes and exploring how they could be applied to the current situation, eliciting awareness of increased confidence using 0-100 self-report ratings.  Interviews focused on participants' goal of increasing physical activity, and were recorded.

Digital wellbeing interventions amplify this complexity due to the topic's multi-dimensional nature combined with a lack of consensus concerning the exact way to define wellbeing.  Is it defined by the quality of life, wellness, mental health, emotional wellbeing, the absence of mental illness, satisfaction, happiness, or some combination thereof?  These terms are commonly used in promotional materials, websites, and academic and industry conferences focusing on wellbeing.  Knowing the appropriate definition of wellbeing is a critical foundation upon which intervention components are defined, measurement tools are selected, and analytics strategies are created to evaluate the impact of the intervention.

This session will describe the practice- and evidence-based processes used to create a working definition of wellbeing, build a framework to develop digital wellbeing interventions, and formalize an evaluation strategy for these interventions.  This includes a review of the various factors that influence wellbeing and the emerging evidence describing the nature of the relationship between health and wellbeing.

This model is supported by a comprehensive and robust body of wellbeing evidence and thus can be considered reasonable justification of resource investment related to the development or provision of wellbeing interventions.  The second logic model considers wellbeing as a determinant of other health-related behavior change interventions.  Can an improvement in wellbeing enable improvements in other health promoting behaviors?

The evidence to support this model is nascent and emerging.  However, should testing of this model confirm that improvements in wellbeing may indeed contribute to other health behavior change, the impact of wellbeing at a population health level could exceed current estimates.

JMIR Publications

In the United States, heavy drinking among college students is a major public health concern that results in negative consequences for both drinking and nondrinking students [1].

However, meta-analyses exploring the effects of computer-delivered and face-to-face interventions across studies show that face-to-face interventions may produce longer-lasting effects than computer-delivered interventions [8,9], and that face-to-face interventions may outperform computer-delivered interventions in their impact on drinking quantity, peak blood alcohol content, and alcohol-related problems [10-16].

This change talk is elicited in face-to-face MI interventions through open-ended questions and reflective listening techniques (including simple reflections, paraphrased reflections, double-sided reflections, and summarizations) that allow clients to hear their own change talk.

MI process research shows that clients are, in fact, significantly more likely to engage in change talk directly following simple reflections, complex reflections, and open questions posed by the interventionist [21].

Furthermore, existing computer-delivered interventions rely on a personal computer (PC) keyboard, mouse, or touchscreen to capture participant’s responses that lack the capacity to allow users to speak aloud and to hear their own change talk, which may be an important factor in the success of MI interventions.

In a voice-based system, users could respond to open questions about their behaviors and attitudes with natural language, and the computer-delivered intervention could use reflective listening techniques to encourage deeper reflection and highlight discrepancies between current behavior and desired goals.

Although it is relatively straightforward to program open-ended prompts for a computer to deliver using speech software and although natural language recognition programs are becoming increasingly sophisticated [22-24], understanding the meaning of the users’

We also conducted a follow-up assessment with participants 1 month after the initial interaction with the voice-based computer-delivered intervention in order to test our primary hypotheses that participants receiving the voice-based computer-delivered intervention would report a significant increase in perceived importance of changing their drinking and report significant reductions in drinking and alcohol-related problems, consistent with the literature on computer-delivered interventions in college student populations.

In order to gauge in a preliminary manner how the voice-based computer-delivered intervention might differ in its effect from traditional text-based computer-delivered intervention, we randomized one-third of participants to a text-based computer-delivered intervention, which matched the voice-based computer-delivered intervention in content, but relied on mouse and keyboard entries of participant responses and provided only text-based responses from the computer.

Given the literature cited previously regarding the importance of change talk and our supposition that a voice-based computer-delivered intervention may increase processing of change talk through verbalization, we hypothesized that the voice-based computer-delivered intervention, compared to the text-based computer-delivered intervention, would result in greater increases in perceived importance of changing drinking and greater reductions in drinking behavior and related problems.

Eligible participants were enrolled in undergraduate or graduate programs in the Northeastern United States, were 18 years of age or older, and endorsed at least one episode of heavy drinking (≥5 drinks in a single sitting for men, ≥4 drinks for women) in the past 30 days.

An initial sample size of 60, assuming an 85% follow-up rate, provided power of .94 to detect a medium effect size of d=0.50 and power of .80 to detect a somewhat smaller effect size of d=0.40;

We also wanted to acquire a large enough body of verbal participant responses and human-controller response selections to facilitate future machine-learning approaches to approximate the decisions that human controllers made (results not described here).

Results of machine-learning approaches could be used as the initial seeds for developing a fully automated voice-based computer-delivered intervention, and we decided that having 60 completed sessions should provide a minimal level of data to initiate that work.

After baseline assessments were completed, participants were randomized in a 2:1 ratio using the urn procedure [29,30]—to ensure equal balancing on gender and number of heavy drinking days—to one of two experimental interventions: (1) a human-controlled voice-based computer-delivered intervention with computer-generated voice communication or (2) a computer-based text-and-click entry interface comparison condition.

Alcohol use over the past 30 days was assessed using an online timeline follow-back measure [31], which assessed the number of standard drinks (12 oz beer, 5 oz wine, 1-1.5 oz liquor as a “shot”

To assess the extent to which the system approximated MI counseling characteristics, we administered two brief surveys, specifically designed for this project, to all participants that contained (1) five items (7-point Likert scale: 1=“not at all”

Other relevant measures were assessed from within the intervention as participants completed it, including whether participants (1) set a goal for reducing their drinking and/or (2) agreed to receive further information on changing their drinking.

The research assistant listened to the questions that the system asked and based on a participant’s responses selected appropriate paraphrases of content or prompts to the participant for further information from a pre-established list of possible responses.

The positive and negative consequences of drinking were then verbally reflected to the participant via computerized voice, with the phrases strung together to create a double-sided reflection: “On the one hand you like that drinking..., but on the other hand, you do not like that...”

ratings of the characteristics of the voice-based computer-delivered intervention to determine how well the system met the objective of reflecting positive therapist traits (eg, how supportive was the system) and MI-based therapy traits (eg, how well did the system help you talk about your own reasons for change).

We then conducted paired t tests to test the hypothesis that participants receiving voice-based computer-delivered intervention would show significant increases from baseline to the 1-month follow-up in perceived importance of changing drinking and confidence in their ability to change drinking and would show significant decreases in drinking and alcohol-related problems.

To test our secondary hypotheses regarding differences between the voice-based and text-based computer-delivered interventions, we conducted linear regressions to test the effects of experimental condition on self-rated importance of changing drinking and confidence in ability to change drinking, as well as number of drinks consumed per week at the 1-month follow-up.

Paired t tests showed significant main effects of time indicating reductions in drinks consumed per week (t51=–3.56, P=.001), number of heavy drinking days (t51=–4.53, P<.001), and reported problems with alcohol use (t51=–3.60, P=.001) from baseline to the 1-month follow-up assessment in the voice-based computer-delivered intervention condition.

Covarying baseline alcohol-related problems, participants randomized to the voice-based computer-delivered intervention reported 40% fewer alcohol-related problems at follow-up compared to participants in the text-based condition (incident rate ratio [IRR]=0.60, 95% CI 0.44-0.83, P=.002).

Experimental condition did not significantly predict number of drinks consumed per week (B=–0.12, 95% CI –0.41 to 0.17, P=.41), number of heavy drinking days (IRR 1.07, 95% CI 0.75-1.53, P=.72), or rated importance of changing drinking (B=0.76, 95% CI –0.51 to 2.03, P=.24) at the 1-month follow-up, covarying for the respective dependent variable at baseline.

Participants receiving the voice-based computer-delivered intervention agreed that the system demonstrated MI-consistent behavior (eg, helped me talk about reasons for change, asked me about my ideas before presenting its own), and displayed at least moderate levels of general therapist traits (eg, was understanding, was engaging).

Although no significant differences on the total score for either scale were observed between conditions, several ratings on the individual-item level that might have been expected to be greater for the voice-based computer-delivered intervention were observed to be numerically lower than the text-based computer-delivered intervention;

Participants in the voice-based computer-delivered intervention condition reported significant decreases in number of drinks consumed and number of heavy drinking days, and significant increases in perceived importance of changing drinking, but confidence in their ability to change drinking, which was high at baseline, did not increase significantly.

First, the sample consisted of college-aged participants who met criteria for heavy drinking, but whose overall levels of drinking were relatively low compared to other intervention studies with college students (eg, [39]);

Finally, an emerging line of experimental research has shown that compared to screen avatars, embodied robots (ie, robots that have a physical form and are in the room with participants) elicit greater engagement and compliance from people who are following directions from the automated system [42,43].

Sandy Jen: "Rethinking Home Care" | Talks at Google

Sandy Jen is co-founder and CTO of Honor, a comprehensive home care startup helping older adults continue to live comfortably and safely in their own homes ...

ROUNDTABLE : CLIF HIGH, PALADIN & KERRY : CRYPTO CURRENCIES & THE FUTURE

I talk with Clif High and Paladin about Crypto Currencies and the Future. CLIF HIGH is the originator of the Webbot system now focusing on crypto currencies.

Connecting Emotions, Brain, and Behavior with Wearables — Dr. Rosalind Picard

NIMH Director's Innovation Speaker Series March 3, 2016 Dr. Rosalind Picard is Director of Affective Computing Research at the MIT Media Lab, Faculty Chair ...

Superintelligence: Science or Fiction? | Elon Musk & Other Great Minds

Elon Musk, Stuart Russell, Ray Kurzweil, Demis Hassabis, Sam Harris, Nick Bostrom, David Chalmers, Bart Selman, and Jaan Tallinn discuss with Max Tegmark ...

Dissecting the Android Bouncer - Duo Tech Talk - December 2013

Google's Android Bouncer was intended to shore up weaknesses of the Android Market by testing applications within a dynamic analysis environment to ...

Behrend Speaker Series: Andrew Byrnes

Andrew Byrnes, the former chief of staff at the U.S. Patent and Trademark Office, discussed his career path and innovation on Tuesday, April 3, 2018, when he ...

Похудение с помощью кефирной диеты? Реальный опыт похудения диетой на кефире

reassured screamed liter favoring traction wondered reconsider realizing plow nap brain's ebb manifests CVD HDL ..

re:publica 2016 - Opening Keynote: The last kilometer, the last chance

Find out more at: During the period from 2016 – 2025, the "other half" of the human ..

#Рост_мышц и боль. Биохимия роста мышц, дома, питание, на турнике, галилео без жира

screamed liter favoring traction wondered reconsider realizing plow nap brain's ebb manifests CVD HDL minutiae ..

С какого возраста можно начинать заниматься бодибилдингом (спортом)?

reassured screamed liter favoring traction wondered reconsider realizing plow nap brain's ebb manifests CVD HDL ..