Using Artificial Intelligence in R to Help Selecting Sampling Weights for Early Childhood Longitudinal Study–Kindergarten Data Analysis

• In Event: Methodological Innovations and Advanced Applications Using National and International Databases

Abstract

Sampling weight selection is an important step when analyzing longitudinal data but potentially overwhelming due to substantial number of available weights and complex rules. This study demonstrates an innovative artificial intelligence (AI) function developed by the authors in EdSurvey R package for ECLS-K:2011 data, which assists analysts selecting sampling weights for their data analysis.

Authors

• Huade Huo, American Institutes for Research
  Presenting Author

• Paul D. Bailey, American Institutes for Research
  Non-Presenting Author

• Ting Zhang, American Institutes for Research
  Non-Presenting Author

• Emmanuel Sikali, U.S. Department of Education
  Non-Presenting Author