Mental Health Crisis: Can AI Tackle The Dearth of Counsellors?
India is the most depressed country in the world. Mental health is still greatly misunderstood, and the taboos surrounding it are as prevalent as ever before.
According to a study by WHO, at least 6.5 per cent of the Indian population suffers from some form of the serious mental disorder, with no major difference between rural and urban populations. WHO reports that 500 Child Care Institutions working towards rehabilitation of girls from trafficking background in India face challenges in getting access to counselling services mainly because there are only 3 psychiatrists and less than 1 psychologist for every million people in India.
The severe lack of mental health professionals has made the available therapists an extremely expensive, often unaffordable treatment for general public. So much so, mental health counsellors are slowly being seen as a “rich people’s” perks.
The inaccessibility to mental health resources often means that people only approach mental health experts when the problem has escalated from treatable to merely controllable. Children are perhaps the most worse-off. India has one of the highest cases of child abuse in the world, and the Child Care centers in the world majorly focus on physical impact of abuse, and not the mental health aspect of it.
Is Artifical Intelligence the Answer?
EmancipAction, an NGO working in this field, has partnered with IBM to use their AI to tap “potential” lay-counsellors (people not formally trained in the field in university setting), but those who have the qualities required to help those in need with some basic-level counselling. This AI looks at identifying qualities like empathy and sympathy using a series of tests, including audio and MCQs. Infusing Watson Personality Insights API that can assess and objectively identify those with better emotional intelligence suited for the role of an adolescent counsellor. For the last 7 months, EmancipAction is using this model to identify and shortlist candidates to hire as lay counsellors from its pool of applicants.
The AI Test for “Empathy”
The evaluation process consists of collating basic information about the candidate, (Resume/LinkedIn Profile), followed by answering an MCQ and finally an audio interview. Apart from measuring emotional intelligence quotient, personality traits and gender bias are the two main things that the model tries to decipher. This evaluation process specifically identifies key qualities, namely – Counselling Awareness, Adolescence Concerns, Team work, facilitation/communication skills, curiosity and commitment to work. The test starts off with an “introduce” yourself plate:
And moves on to an audio test that asks a number of open ended questions that you answer into a microphone:
The test then moves on to the MCQ section, where you answer questions with choices ranging from
I. The Tech-End: How Does It Work?
How do you track something like empathy? Gargi Dasgupta, Director - IBM Research India tells us: “We are trying to identify personality traits from their speech and text and also emotional intelligence traits such as curiosity, self and social emotional awareness and management, biases and other skills. The candidates - Emotional Quotient (EQ) such as Emotional Awareness, Emotional management (self and social), relationship management, Awareness of Gender Responsibilities, past relationship with Parents are measured by asking the applicants to take the Standard emotional quotient assessment ( which is part of the MCQs).”
What About Cultural Barriers?
Gargi says, “The AI model is India specific as we have selected culturally appropriate MCQs in the Indian context to assess the social, emotional and relationship strengths. Right now, the technology piece currently works only on English, but for instances where the audio recordings were a mix of Hindi, English, Konkani or Marathi etc., these were transcribed/translated to English for the textual processing by Watson Personality Insights.”
Apart From Hiring, What Can We Hope to Achieve With AI?
“The scope of this work was limited to identifying emotional state and emotional capabilities of people with a hiring perspective. But parts of this work can be used to build systems detecting mental health illness such as depression and Post-traumatic stress disorder. We also aim to incorporate video analysis as studies have shown that body language, gestures, eye-gaze and facial expressions to be important cues to asses Emotional Intelligence.”
II. What after the lay-counsellor is hired?
We spoke to EmancipAction, the NGO working with IBM. When asked about what happens once these individuals are hired by the tech end, they said: “The shortlisted candidates undergo a two-week training program. We call our trainers ‘Facilitators’, as we believe the role is to facilitate growth in others and not to teach. The participants who are hired after the training is completed, are called ‘Mental Health Workers’.”
The training starts with sessions on ‘Personal Growth’, the foundation of the training. It is based on the belief that one must explore one’s own thoughts, feelings and values to develop an understanding and acceptance of self. By reflecting and analyzing their experiences in training, the participants not only develop important skills necessary for working with vulnerable populations, but facilitator can also assess the participants’ capabilities based on their participation, interactions with others, and their self-exploration and understanding.
The training also focuses on ‘person-centered counseling’, which is a way to provide support, help in exploration, help in finding solutions, rather than trying to “fix” someone. Once the participants understand their role, the focus is on introducing basic counselling skills such as active listening, empathy etc., that are the building blocks to be practiced in the training. Practice is crucial and hence the focus is on role plays to develop these fundamental skills. Finally, the concepts of vicarious trauma, stress and caring for oneself are introduced at the end of the training.
What is the future of mental health?
What’s Not Working Out?
1. Identification of severe problems:
The NGO speaks of its main concern: Can lay-counsellors identify children with extreme problems? “It is very important for our Mental Health Workers to be well versed in identifying children with more serious problems, and we do a tremendous amount of supervision and training for them to be equipped in this regard. We do not take on any individual counselling but we do refer those with severe symptoms to a mental health professional.”
2. Of Taboos: It is well known that while more mental health experts will definitely be a boon to the rising problem of mental health within the country, how will AI actually drive people to take this help? EmancipAction representatives agree. “The taboo for a child to approach a counsellor is very high and the understanding of mental health and its impact is very low. Hence, most often, it is only the children who display outward behavioural issues are sent to counsellors by the caregivers in the CCIs.”
(Make sure you don't miss fresh news updates from us. Click here to stay updated)