Why are weights included in the Head Start Impact Study dataset?

The Head Start Impact Study (HSIS) used a multi-stage sampling process to select a representative group of Head Start children.  HSIS uses a sample not the universe of participants; it is not a simple random assignment; and not all programs, teachers, parents, and children participated.  Weights are used to compensate for not collecting data from the universe of participants, for adjusting for different probabilities of selection, differential nonresponse, and to reduce possible bias associated with nonresponse.

The HSIS uses complex survey data and the appropriate weights must be used for analysis.  Different cross-sectional weights were created for each instrument and each year.  Three sets of longitudinal weights are included in the weights file.  The first set applies to children with assessments at two or more time points and the second set applies to children with assessments at three or more time points.  The third set of longitudinal weights applies to children with a teacher survey and a teacher-child rating in both kindergarten and 1^st grade.

Careful consideration should be given to the choice of a weight for a specific analysis since it depends on the type of data analyzed. Each set of weights is appropriate for a different set of data or combination of sets of data. Due to non-response adjustments, sometimes the decision of which weight to use is affected by how much of the sample will be excluded if one weight were selected over the other. A general description of the types of weights to be used is provided in the Head Start Impact Study Data User’s Guide.  A more detailed description of the weights is provided in the Head Start Impact Study Technical Report.