After the initial item pool was generated, then the items were purified.
This purification step is designed to remove the potential for measurement error from the new construct to improve their reliability.
Collecting data from an initial sample of respondents helps to address these issues.
Specifically, a manual factor technique was used to establish tentative scale reliability and validity,
as well as to assess potential problems with the unidimensionality of the constructs.
The manual sorting procedures was conducted iteratively,
using independent panels of expert judges for each round.
The judges had recent industry experience with the implementation and use of ERP software in a business environment.