Affectifi Inc.

Affectifi aims to use the power of Hollywood, emotion science and technology to improve the mental health and wellness of children, parents and educators.

To accomplish this, Affectifi has created ThinkHumanTV which integrates with major streaming platforms like Netflix, Disney+ and Peacock to deliver truly engaging and effective social emotional skills training, and employs an evidence-based methodology and curricula developed by its founding team of Columbia University cognitive scientists.

Ilya Lyashevsky is Affectifi's cofounder and CEO and a faculty member at Columbia Teachers College. The ThinkHumanTV platform and training methodology is based in part on his research. At Affectifi he is responsible for R&D, including product development and evaluation, as well as business development and strategy.

The company founders are faculty members at Columbia Teachers College with PhDs in cognitive science, and are experienced researchers and educators. Other senior staff hold masters degrees in education, psychology, or related fields. Thus from an academic perspective the team is well prepared to work on the LEAP Project. Moreover, the team is used to devoting a portion of its time to research as this has been an integral part of the company's approach to product development from the outset. Consequently, the time commitment required for LEAP would not be difficult to accommodate.

ThinkHumanTV integrates with major streaming platforms like Netflix and Disney+ to deliver truly engaging and effective social emotional skills training.

Most US educators now recognize social emotional skills as crucial to young people’s success and wellbeing. Research indicates that emotional flexibility and resilience enable better stress management, serve as protective factors against anxiety and depression, and improve academic performance. Children in underserved communities are in particular need of support in fostering socio-emotional skills given the additional challenges they may face compared to their more privileged peers. The need has become even more acute in the wake of the coronavirus pandemic. For instance, from 2009 to 2021 the share of American high school students who feel 'persistent feelings of sadness or hopelessness' rose from 26 to 44%, the highest level ever recorded. 

At the same time, solutions that purport to improve social and emotional skills feature content that usually can’t live up to emotionally compelling Hollywood productions while also lacking cultural responsiveness, leading to learner disengagement and poor outcomes. Most solutions also fail to adequately assess students’ social emotional skills, relying primarily on self-report measures that are inherently inaccurate. The above pain points have been a common theme in educator interviews, and have been reflected in the willingness of educators from public and private K12 schools to undertake paid pilots of the ThinkHumanTV platform, including ongoing work with the NYC Dept of Education. Similarly, over 95% of surveyed parents wanted a service like ThinkHumanTV for their kids.

Developed with support from the National Science Foundation, ThinkHumanTV integrates directly with major streaming sites (including free ones like Peacock to ensure equal access) to enable the improvement of essential social emotional skills such as perspective-taking and emotion regulation. Leveraging the streaming services’ vast and diverse catalogs enables ThinkHumanTV to deliver truly compelling and culturally responsive training content for all learners, coupled with an evidence-based, practice-driven emotion science curriculum that facilitates skill and knowledge transfer. 

The platform’s training methodology is based in part on the Affectifi founders’ doctoral and post-doctoral research at Columbia University, which has been published in the American Education Research Journal and presented at the SEL Exchange Conference, the Association for Psychological Science Conference, and elsewhere.

The platform is powered by an exercise generation engine which offers unique flexibility in that a user can train with any media narratives in any order while still seamlessly progressing through the training curriculum. This provides for choice and personalization and supports learning variability.

At the same time, ThinkHumanTV offers an unprecedented level of precision in assessing students’ performance on a range of socio-emotional skills, which are notoriously difficult to measure. It does so by employing activities that are based on specific emotion science principles, such as emotion valence, and that lend themselves to sufficiently objective evaluation.

In addition to being piloted in elementary school, ThinkHumanTV is designed to work well with middle and high school students.

Adolescent and young adult learners have been shown to be resistant to more prescriptive social-emotional learning programs that simply teach them the “right” way to behave in specific situations. When working with this age group, better outcomes have been achieved with interventions that provide a new conceptual framework, or mental model, of emotional functioning, which changes how learners view themselves, other people, and social situations. This is why ThinkHumanTV training focuses on teaching fundamental emotion science principles such as situational appraisal. Understanding such principles helps build a more accurate mental model of emotions, leading to more effective social problem-solving, better self-regulation, and increased empathy.

At the same time, ThinkHumanTV makes use of high quality popular media such as TV and film, which enables the teaching of psychological principles to be grounded in concrete, compelling and widely varied examples of human emotion and behavior. Use of such media is particularly important with adolescent learners who are more likely to tune out if presented with content perceived as 'cheesy.'

Foundational Research

Lyashevsky, I., Cesarano, M., Black, J. (2020). To Understand Is to Forgive: Learning a Simple Model of Appraisal Leads to Emotion Knowledge Transfer and Enhances Emotional Acceptance and Empathy. American Educational Research Journal, 57(2), 906–940.

Lyashevsky, I. and Berkowitz, M. (2020). Seeing the Deep Structure of Emotion: Applying the Triad Judgment Task to the Perception of Emotion Causes. Poster Presented at the Association for Psychological Science Conference. Online. June 1-Sep 1.

Lyashevsky, I., Cesarano, M., Black, J. (2019). Teaching to Transfer in the Social Emotional Learning Context: Is targeting deep structure the key? Poster presented at the annual SEL Exchange Conference. Chicago, IL. October 7-9.

Cesarano, M. (2018). Implicit Theories of Emotion and Social Judgment. Doctoral Dissertation. Columbia University Academic Commons.

Lyashevsky, I. (2018). Teaching to Transfer in the Social Emotional Learning Context: The Case for an Instructional Model of the Human Emotion System. Doctoral Dissertation. Columbia University Academic Commons.

Formative Research

Usability study. Graduate students. N=14. February, 2020.

Usability study. High school students. N=8. April, 2020.

Usability study. Graduate students. N=36. October, 2020.

Usability study. Graduate students. N=6. February 2021.

Alpha testing. Undergraduate students. N > 100. June-August 2021.

School pilot/usability study. High school students. N=11. Public school. May-June, 2022. At home use & online video discussion sessions.

Usability study. Community college students. N=42. Oct, 2022.

School pilot/usability study. High school students. N=41. Private school. In-class & at home use use. April-May 2022.

School pilot/usability study. Middle school classroom. Private school. In-class use. Ongoing.

School pilot/usability study. Middle school students. Public school. After school group use. Ongoing.

School pilot/usability study. High school students. N=25. Private school. In-class & at home use use. Ongoing.

Summative Research

Lyashevsky, I. A. & Cesarano, M. (2022). Assessing the Efficacy of the Affectifi Online Platform for Emotional Skill Development. Study findings presented at the Association for Psychological Science Conference. Chicago, IL. May 26-29, 2022.

Lyashevsky, I. A. & Cesarano, M. (2023). Building Social Emotional Skills with the Power of Hollywood. In preparation.

Foundational Research

The foundational research suggested that:

- emotion science concepts can be seen as the 'deep structure' of social emotional learning

- people's ability to perceive important underlying features (deep structures) of emotional situations can be enhanced with training

- teaching emotion science concepts can (1) facilitate emotional knowledge transfer, (2) enhance self- and other-awareness (3) that the methodology
works with adolescents and young adults, a population for whom SEL interventions have often proven ineffective; and (4) that it can be delivered via an online platform.

This foundational research provided the basis for designing the ThinkHumanTV online social emotional learning platform.

Formative Research

The formative research indicated that:

- Over 90% of users found the application easy to use and the content easy to understand

- Nearly 100% found the use of popular media to be an enjoyable way to learn about emotions

- Over 90% thought that using the application improved their understanding of emotions

- Over 85% said they would recommend the application to others

- Significant percentage of users wanted more supported media content

- Users wanted UI and UX improvements such as a bigger fonts (for in-class display), multiple profiles (for multiple class support), media content age rating, and others

- Users wanted open-response capability

- Users wanted to be able to use ThinkHumanTV with their TVs

We have worked to improve the platform based on the users' feedback by making the suggested enhancements, including increasing the media library (now nearly 60 titles), improving the UI/UX, preparing to deploy open-response exercises, and developing a mobile solution that will work with TVs.

Summative Research

A recent study at Columbia University showed that participants who spent two weeks using the platform were significantly less likely to use suppression as a regulation strategy (22% reduction) compared to a control group.

A second study at LaGuardia Community College showed significant improvement in emotional self-efficacy (i.e. confidence in managing one’s emotions) after completing the training (nearly 10% increase).

Our present goal is to obtain additional empirical evidence of the platform's efficacy with K12 students. Such evidence would ideally come from school-based use, as that is the primary context in which the application is being piloted.

In particular, we are interested in partnering with outside academic institutions on running school-based quasi-experimental or RCT studies that would look at some more long-term outcomes for students.

This would enable us to both make adjustments to the training experience to make it potentially more impactful for students, and to provide additional proof of the platform's value to stakeholder and decision-makers.

- Can we replicate previous findings and show significant improvements in certain social emotional competencies among K12 students following sustained use of the ThinkHumanTV application?

- Does sustained use of the ThinkHumanTV application lead to significant reduction in the incidence of anxiety or depression?

- Does sustained use of the ThinkHumanTV application lead to better grades, reduced absenteeism or higher graduation rates?

- Potential study design recommendations to address the research questions outlined above

- Identification of additional research questions of interest

- Potential outside research partners identified 

- Potential K12 partners identified

- Assessment of resources needed to carry out the study/ies

- Suggestions for potential studying funding sources (e.g. grants)

The goal would be to conduct K12 studies, e.g. pre-post or RCT, with one or more outside academic partners. 

The study results would be used to make potential adjustments to the ThinkHumanTV training program as well as to the design of future studies. If successful, the studies would also help demonstrate platform efficacy to end users and facilitate platform adoption. The team would seek to have the findings published in research journals, and shared with the academic and broader community via other channels such as social media.

Short-term Outcomes:

- Being prepared to conduct one or more K12 studies (pre-post or RCT)

- Having potential research and K12 partners lined up for this purpose

- Having potential research funding sources identified

Long-term outcomes:

- Additional evidence of platform efficacy

- Reusable study design and process to facilitate future data collection

- Ongoing relationships with research and K12 partners

- Increased platform adoption

- Larger number of students positively impacted