Experiment and Data Sources

Nonexperimental component examines:

  • Managers' labor flexibility practices and their relationship to store-level outcomes, such as turnover, and worker-level outcomes, such as worker commitment, engagement, and sales
  • Dimensions of work schedules (stability, predictability, flexibility) and their relationship to worker performance and well-being.

Experimental component includes:

  • Cluster-randomized experiment. Stores randomly assigned to intervention and control conditions
  • Intervention. Posting schedules for a full month at a time (usually schedules are posted the Tuesday or Wednesday before the workweek that begins on Sunday)
  • Experiment complete in Chicago area stores (N=21). Preparing to expand experiment to stores in the Northeast.

Multiple sources of data:

For 151 stores in Midwest and Northeast:

  • Corporate administrative data
    Monthly personnel records 2006 to present (N=1,179 hourly employees, 2008)
    Used to track turnover, retention, and churning
    Provides employee demographic information
  • Manager Survey (N=139; 92% response rate)
    Telephone survey that asked managers about their staffing and scheduling strategies

For 21 stores in Chicago area that participated in experiment

  • Weekly posted schedules (6-8 months)
  • Payroll hours, i.e., actual hours employees worked (6-8 months)
  • Weekly sales by associate and store, including sales-per-hour (6-8 months)
  • Employee Surveys (pre and post intervention)
    Telephone survey that gathered information on work attitudes and behaviors (e.g., engagement, commitment, organizational citizenship behavior) and worker well-being (e.g., work to family conflict, work to family interferences, stress, caregiving responsibilities)