Follower behavior renders leader behavior endogenous: The simultaneity problem, estimation challenges, and solutions
Description
This page contains all data needed to replicate the results reported in the article "Follower behavior renders leader behavior endogenous: The simultaneity problem, estimation challenges, and solutions ". First, we mathematically demonstrate the problem of simultaneity bias using a simulated dataset and show how the identification problem can be analytically solved by means of an instrumental variables-approach(Study 1). Second, we present data from naturally occurring interactions between leaders and followers that make the simultaneous influence processes particularly salient (n = 108; Study 2). Third, we present how simultaneity can be resolved from a research design perspective using experimental approaches (n = 31; Study 3). The file "data_study1" contains the simulated data (in .csv format). The file "data_study2" contains the frequencies of behavior codes (in .csv format) derived from dyadic conversations between leaders and followers from Study 2. Data was analyzed using lag sequential analysis implemented in the Generalized Sequential Querier (GSEQ, Bakeman & Quera, 1995). The file "data_study3" contains the frequencies of behavior codes (in .csv format) derived from an experiment involving dyadic conversations between leaders and followers from Study 3. The results from this data are reported in Table 1 and 2 of the manuscript. The point-by-point estimation procedure is provided in the manuscript.