Document Type


Publication Date


Find this in a Library

A link to the resulting conference proceeding can be found at


In this document, we present all of the model predictions for an upcoming study to be to run in the Spring of 216. During this experiment, participants will sequentially play two games of strategic interaction for 50 rounds in one of four possible game orders, playing either iterated Prisoner’s Dilemma (PD) or iterated Chicken Game (CG) twice (PDPD or CGCG order) or playing each game once (PDCG or CGPD order). During each game, participants will play with a computerized confederate agent which uses a particular strategy, playing the first game of a condition using the T4T strategy and using the PT4T strategy during the second game (T4T_PT4T order) or vice versa (PT4T_T4T order). The trustworthiness of the confederate agent will remain constant over the course of both games, manipulated to be either high (HT) or low (LT) trustworthiness. Finally, the information that participants receive about confederate agents will be manipulated. Participants will be told that they will play both games with either the same participant (one-agent condition) or that they would be randomly paired with another participant to play with during the second game after the first game (two-agent condition).

This document is organized into 32 sections, one for each of the experimental conditions, each with six different figures. The first graph of a section shows the smoothed round-by-round predicted proportion of each of the five game outcomes during a condition. The proportion of each outcome was smoothed to remove the round-by-round variability in the game’s outcomes and to allow the reader observe the general behavioral trends predicted by the model. Following the first figure of a section, five additional figures are presented, showing the actual round-by-round predicted proportion of a single outcome over the course both games during condition, with 95% CI intervals.

Included in

Psychology Commons