checkmate project page

Checkmate

This is an ongoing project out of the MIT Media Lab to study decision-making and impression formation in online dating. Please contact checkmate at media dot mit dot edu with questions or comments.

Research Problem:

Dating practices have been indelibly marked by data retrieval and communication opportunities available online. But how do people identify and pursue a romantic partner online (verses off)? How do people reason and make decisions based solely on the dubious information users volunteer themselves? We are examining this process, running studies on online dating impression formation and experimenting with new methods to get acquainted for online dating.

This project has three components:

Data analysis of online behavior. Unlike in other decision-making domains, online each behavior results in a trace: button clicks, pages visited. Even from anonymous histories, we can begin to reconstruct the decision-making processes and variables that inform such decisions. In this work we examine if general human attraction literature predicts online behavior.

Empirical work. Getting acquainted can be thought of as a process of reducing ambiguity where each individual strives to collect information in order to be able to describe and predict another’s behavior (Berger, 1979). Offline, we have a set of tools (such as observation, direct interaction and querying third parties) and social cues (gesture and intonation) for this process. In online communication, information flow differs; it is decreased per unit time (Walther, 1996) and lacks credibility. Meanwhile we have fewer ways to gather information, usually just direct email and instant messaging interaction limiting ability to reduce ambiguity. Our experiments explore the interaction of the resulting ambiguity and attractiveness in this domain. People state that they want more information but our results suggest that ambiguity is attractive in itself.

Game-based experiments. Our results demonstrated ambiguity itself implicates online impression formation; therefore varying the quantity of trait and attribute information will not help people form more accurate social evaluations of one another. Instead, we propose new types of online dating interaction to facilitate data gathering in ways that are analogous to offline contact (e.g. observation, collaborative activity as well as direct questioning). To structure this interaction we are creating a series of online dating games. We will use these simple games to experiment with what types of exchanges are useful in making accurate social judgments and which lead to feelings of intimacy and liking.