[CANCELLED] Analyzing Social Behavior in the Digital Age (2225.YR.011213.1)

General information

Type:	OPT	Curs:	2	Period:	S semester
ECTS Credits:	3 ECTS

Teaching Staff:

Group	Teacher	Department	Language
Year 2	Matteo Prato	Dirección General y Estrategia	ENG

"If I launch an ad with Colin Kaepernick, will my FB followers increase or decrease? "If I introduce a pet-friendly policy in my company, will my employees' productivity drop or augment??, "If I support a social initiative for philanthropy, will my firm's stock price go up or down? Whether to retain customers, motivate employees, or entice investors, a successful business rests on developing accurate representations of cause-effect relationships. For many years, these representations laid on the intuition that successful leaders derived from their consolidated experience in the field. As recent technological advancements are making possible to collect and process data on nearly every aspect of social behavior on a massive scale, managers face new opportunities to make compelling causal inferences. To tap into the potential afforded by big data, however, managers must develop analytical skills that allow them to distinguish robust causal patterns from misleading associations. The challenge is to do so without being statisticians or computer scientists. This course aims to offer an intuitive and non-technical approach to develop such skills.

At the end of this course, participants will become familiar with how to tackle business problems in a data-analytic fashion. They will develop working knowledge on:

- How to leverage data-analytic to develop novel managerial intuitions
- How to transform a managerial intuition (e.g., if "A" then "B") in a testable relationship between measurable constructs
- Why a simple correlation between "A" and "B" can be misleading, and what techniques can be employed to test whether "A" does affect "B"
- What can be done to understand the underlying mechanism through which (i.e., "why") "A" causes "B" and discard alternative interpretations
- How to represent data visually to gain insights and effectively convey arguments and results

The course will be strongly focused on uncovering social evaluation biases that underpin inequality, discrimination, and prejudices. Data-driven decision making has the potential to mitigate these social biases. Nevertheless, improper use of data-analytic tools can spread and even magnify them. The course will help you forming socially-conscious managerial decisions and understand the increasing social impact caused by the digitalization of human behavior and humans' interactions.

1. Introduction to Analytics for Social Data One of the main building blocks of any data-analytic project is finding associations between variables. In this section, we will: a) explore how the digital age offers unprecedented opportunities to measure social behavior b) review the basic principles of statistical analysis aimed at identifying relationships between variables c) discuss how to interpret regression results. d) examine how this technique can generate insights into concrete cases and uncover novel associations.

2. Building Models for Social Behavior In this section, we will have an overview of basic principles and underlying analytical thinking that inform the formulation of compelling models. We will discuss the fundamental differences between models aimed at improving predictions versus those aimed at drawing causal inferences. We will explore the trade-off between the two in the context of people analytics. People analytics is one of the emerging trends in business analytics. A wide range of employees' data is collected and analyzed to efficiently manage people within organizations, whether to optimize hiring, improve engagement, reduce attrition, or rationalize compensation. This section will review the journey (of trials and errors and learning) that a worldwide leading data-analytic company (i.e., Google) undertook when implementing a people analytic system. At the end of the class, each group will compete in developing an accurate HR model based on past employees' performance data.

3. Social Network Analysis Whether to analyze the "viral" spread of social trends among customers or the flow of information among employees, business leaders must develop an understanding of how networks operate. In this section, we will review how to visually map relationships between actors in a network structure and how to interpret the benefits and disadvantages of occupying different network positions. We will discuss how network analysis can be used to re-thinking the design of organizational structures.

4. Natural Language Processing Natural language processing (NLP) is one of the most rapidly advancing subjects in the data science field. This section will explore the fundamentals of topic modeling (i.e., how to identify relevant themes from a corpus) and sentiment analysis (i.e., hot to gauge the positive vs negative sentiment in a text). We will show how NLP results can be used to feed clustering models and identify social categories.

5. Social Experiments Experiments are the gold standard in the field of descriptive data analytics aimed at drawing causal inferences. In this section, we will develop working knowledge on how to run experiments to inform decision-making. We will review some of the best practices in the field and discuss ethical concerns.

The course format will be based on a mix of lectures and class discussions, case studies, students' presentations, and group exercises

All the readings for this course will be posted on the course's page on the portal. Readings might be added as the course progresses. The most important readings will be the course slides themselves.

Matteo Prato joined the Department of Strategy and General Management at ESADE Business School as Associate Professor in September 2019. Before joining ESADE, he was Associate Professor at USI (Switzerland). He holds a Ph.D. and Master of Research in Management from IESE Business School, where he also served as a post-doctoral researcher. He has been visiting researcher at Stanford University and Columbia University. His current research interests lie at the intersection of economic sociology and organization theory, with particular emphasis on the role that social structures (e.g., status hierarchies, social networks, and categorization systems) play in shaping market valuations and in enhancing actors' performance. His research has been published in the leading journals in the field of Management and Sociology such as Administrative Science Quarterly, Academy of Management Journal and Organization Science, and funded by the most prestigious institutions such as the US National Science Foundation (NSF) and the European Research Council (ERC)

Group	Teacher	Department
Year 2	Matteo Prato	Dirección General y Estrategia