The dataset 'Individual Choices in Dynamic Networks' contains the outcome of an experiment on network formation, studying how social preferences play a role in networking relations actors make.
In total we ran 16 experimental sessions, each of them having between 9 and 15 subjects. Subjects were contacted using the Online Recruitment System for Economic Experiments (Greiner B (2004) The online recruitment system ORSEE 2.0 - A guide for the organization of experiments in economics. University of Cologne, Working Paper Series in Economics 10.)
A total of 227 subjects subscribed for one of the 16 sessions, of which 205 subjects participated. We allowed at most 16 registrations for each session and every registered subject who showed up in time participated.
Most subjects were students at Utrecht University from a wide range of disciplines, although non-student subjects also participated.
Units of analysis are:
- Groups of subjects
- Subject Decisions
Subjects were between 17 and 60 (mean age being 21.6), 68.3% female, and 78.5% Dutch.
The experiment was programmed using z-Tree (Fischbacher 2007).
For detailed information on the set-up of the experiment, the analysis and discussion, see the article 'VanDolderBuskens2014.pdf' in de folder Publication.
The set contains:
1) Experimental instructions in the folder titled "Instructions"
This folder contains the instructions in both Dutch and English and for the four
orderings that we employed in the experiment, making (2x4=) 8 sets of instructions
overall.
2) Ztree treatment files in the folder titled "Treatments"
This folder contains the ZTree files needed to run the experiment. Note that here,
as with the instructions, there are four orderings leading to four sets of treatment
files. The program itself is bilingual, so there is no need to have separate files
for the Dutch or English version.
3) Dataset.dta contains the data used in the paper.
We did not include a codebook, but rather described each variable in the variable
label in the Stata data file. The dataset contains many variables that are not
directly necessary for the analyses, but might be useful for other aims.
4) Analyses.do is the Stata dofile to generate the results presented in the paper.
We have added annotations to help the reader understand what happens at each step.
The set contains original raw quantitative data.
The experiment has been conducted with X subjects.
