Work in Progress

 Behavioral Economics: Well-Being and Data Science

Crisis Informatics represents a multidisciplinary field combining Computing and Social Science knowledge of disasters; its fundamental tenet is people use personal information and communication technology to respond to a disaster in creative ways to cope with uncertainty [Leysia Palen \& Kenneth M. Anderson (2016)]. There are many thematic applications of big crisis data analytics including (1) data-driven digital epidemiology (in which public health research is conducted using CDRs and social media)[M. Salathe et al. (2012)]; (2) population surveillance and urban analytics [M. N. K. Boulos (2011)](in which big crisis data is used for tracking the movement of a crisis-affected population as the crisis unfolds); (3) crisis informatics and sociology [L. Palen, S., et al. (2007)] (in which data, along with participatory mapping and crowdsourcing technology, is used for analyzing the sociological behavior of the affected community through behavioral inference and “reality mining”). 

The most important application of AI-based computational linguistics for big crisis data analytics is an automated analysis of social media using sentiment analysis and opinion mining [M. Imran, C et al. (2015)] While there have been many advances in Natural Language Processing (NLP) applying it to Crisis Informatics is still non-trivial [Junaid Qadir et al. (2016)]. The effects of violence in communities with long-temporal waves of violence have showed behavior of resilience [ Drury, J et al.(2009), Collins,R et al.(2004)],social support [Lin, YR \& Margolin D (2014)] and some other mental health problems [Lemyre, L et al. (2005), Jenkin, C.M (2006)]. 

This Crisis Informatics project carries out Data Science, Machine Learning and NLP (natural language processing) models to study behavioral responses to crises (i.e violence produced by Urban Warfare, Terrorism), and complex social interactions;  by exploring experimental measures of subjective well-being, risk perception and policy preferences. We look at the general research topics areas:

  • Choices, Values, and Frames in Risk Environments
  • Cognitive Bias, Social Identity and Preferences
  • Terrorism and Distress
  • Online Social Networking and Mental Health
  • Modeling Emotion-Based Decision-Making

Working Papers in Behavioral Economics

  • Stated Preferences and Well-Being in Risky Environments 
Current literature that studies the impact of terrorism on the psychological well-being of people suggests that it takes more than one agent (e.g. threat to life) to provoke psychopathology. By using an online experimental survey anchored on stated preferences this research measure people's well-being preferences in locations considered as risky environments. By implementing an online recruitment by using Facebook Ads, we tested the decision-making process in personal and policy hypothetical scenarios measuring 259 respondents' preferences. We found that the impact on the mental health of the civilian population by this phenomenon is one of the most significant. We found suggestive evidence of negative well-being aspects such as anger, anxiety, and depression. In addition, we also found that the behavior of respondents have low uncertainty regarding other's preferences and in this case, people state preferences with high intensity in policy

Submitted to Journal for Publication, Coming Soon (Draft Available Upon Request)

  • Risk Perception, Policy Preferences, and Pro-Social Behavior:  Experimental Evidence in Risky Environments
The study of risk perception and policy preferences in locations affected by violence reflects negative emotional expressions as violence increases and resilience in a steady wave of violence. Threat perceptions and emotions can jointly impact individuals’ attitudes towards risk. In this article, we implemented an online experiment to prove two main hypotheses. The first hypothesis lies in the appraisal-tendency theory to test the behavior of 111 respondents residents in violent and non-violent locations. We found that anger triggered in one situation evokes optimistic risk estimates and risk-seeking choices. Fear does the opposite, evoking pessimistic estimates and risk-averse choices. The second hypothesis lies in the perception of risk and risky decision making. In particular, risky decision making suggests that people are loss-averse—they dislike losses much more than they like equivalent valued gains (Kahneman and Tversky, 1979). By using an identifiable victim effect on an online modified dictator game, we provide evidence that suggests that people see saving a statistical life as a gain, but saving an identified victim was seen as avoiding a loss. Then this predicts that people put greater value on identified victims than on statistical ones.

Submitted to Journal for Publication, Coming soon (Draft Available Upon Request)

Working Papers in Social Networks 

  • Risky Environments, Happiness, and Networks in Blau Space
How does experiencing mass urban warfare, violence, or other traumatic events affect individuals' interaction? Social media has become central to the public's response to violence, particularly social networks support has an important role in helping to reduce negative emotional violence effects. By using a network instrument based on household data from INEGI Subjective Well-Being 2012 Survey in Mexico. The present study analyzes homophily differences by using log-linear models based on a self-reported level of happiness of 10,400 respondents. The analysis is made within and between groups by fitting layer effect parameters. There are main differences between male and female non-kin ties residents in violent and nonviolent locations. The study compares the pattern of ties among dissimilar alters for both sexes. The results of how personal characteristics relate to differences in strength of homophily related with people's level of happiness or life satisfaction indicates that residents in violent places are almost equally homophilous to affiliate or having support from a specific social network, although involvement in community activities exerts a stronger influence in women than men and the resulting tendency to form ties is based on that particular locus. 

Submitted to Journal for Publication, Coming soon (Draft Available Upon Request)

A fuzzy analysis is commonly used for handling various forms of uncertainty in the decision-making process and it is related to the design and control of complex systems which is the case of fuzzy clustering. This paper initially will aboard this issue by incorporating approaches of group decision making based on intuitionistic preference relations. The common cases of study are data that comes from subjective well-being decisions that are embedded in behavioral networks. By using fuzzy theory and optimization methods, this paper proposes to follow a two-step algorithm to analyze behavioral networks by (1) identify complete intuitionistic preferences and by (2) approximate a network feature matrix for a conventional fuzzy community detection which is an extension of a fuzzy k-means clustering procedure. [NARSC 2016] 

This paper introduces the concept of "Citizen Commission" as a structure of social capital and show methods for its mathematical analysis. The paper explains the network formation and how information transmission feeds back into the evolution of social links. The social network is analyzed in cooperative games and its role in public goods. The concepts of the equilibrium point and its different solutions are introduced by mathematical definition. And later in the paper, the process of social learning is discussed. As an illustration of the possibilities of this Citizen Commission for a real application in the economy is included. [ NARSC 2014 ]

Conference Papers 

  • Plascencia, Fernando J. (2016). "Fuzzy Analysis for Intuitionistic Preferences on Complex Networks". Presented at the  63rd North American Regional Science Council (NARSC). Social Interaction, Group Decision Making and Spatial Considerations.  Minneapolis, MN.
  • Plascencia, Fernando J. (2014). "Cooperative Games, Learning and Strategic Interaction in Social Networks". Presented at the  61st North American Regional Science Council (NARSC). Economic Development.  Washington, DC.