PERFORMANCE SALIENCE
Description
This dataset contains observations collected from a controlled experiment designed to analyze the role of information salience in individual decision-making. The experiment consists of multiple rounds in which participants were exposed to varying levels of information salience and were required to make choices under different informational conditions. The dataset includes individual-level responses, treatment conditions, timing variables, and outcome measures capturing behavioral adjustments. The primary objective of this dataset is to facilitate the study of how salient information influences attention allocation, belief updating, and choice behavior. The data structure enables analysis of 3 groups, as well as identification of temporal patterns in responses to changing informational environments. This dataset is suitable for empirical research in behavioral economics, experimental economics, and related fields focusing on decision-making under varying informational constraints.
Files
Steps to reproduce
To reproduce the results associated with this dataset, follow the steps outlined below: 1. Download the dataset and the accompanying Stata script file ("StataResults.do" file) from the repository. 2. Ensure that Stata (version 15 or higher) is properly installed on your system. 3. Place the dataset file and the "StataResults.do" file in the same working directory, or update the file paths within the "StataResults.do" file accordingly. 4. Open Stata and set the working directory to the folder containing the files using the command: cd "path_to_your_directory" 5. Load the "StataResults.do" file into Stata by either: * Opening it through the Stata interface, or * Typing in the command window: do StataResults.do 6. Execute the "StataResults.do" file. This will automatically load the dataset, perform the necessary data cleaning and transformations, and generate the results, tables, or figures as specified in the script. 7. Verify that all outputs are correctly generated and saved in the designated folders. These steps ensure full reproducibility of the empirical analysis and allow users to replicate the results or adapt the code for further research.
Institutions
- Pontificia Universidad JaverianaBogota D.C., Bogotá
- Military University Nueva GranadaBogota D.C., Bogotá