A function for pooling imputed scale scores, SGPs and Baseline SGPs

Function produces summary tables and multiple imputation statistics for imputed score data and SGP analyses.

imputationSummary(
  data.to.summarize,
  institution.level = NULL,
  summary.level = NULL,
  standardize.scores = NULL)

Arguments

data.to.summarize	An dataset of imputed scale scores and SGPs (cohort and baseline referenced) calculated using those imputed scores.
institution.level	Institution ID for summary table aggregation. E.g., "SCHOOL_NUMBER" or "DISTRICT_NUMBER". NULL (default) aggregates across all students (entire state).
summary.level	Additional aggregation levels. E.g., "GRADE" and/or "CONTENT_AREA".
standardize.scores	If not NULL (default), then level by which Scale scores are standardized (by mean and SD of the observed scores). E.g., "GRADE" to standardize by grade for cross grade comparisons.

Details

Produces summary tables and multiple imputation statistics. Function assumes data contains variables named "SCALE_SCORE_OBSERVED", "SGP_OBSERVED", and "SGP_BASELINE_OBSERVED" for the unaltered (missing) data, and imputed/imputation derived variables with the tag _IMPUTED_ added. For example, "SCALE_SCORE_IMPUTED_1", "SGP_IMPUTED_1", and "SGP_BASELINE_IMPUTED_1". That is, the data is in a wide format (not stacked by imputation).

Value

Function returns a list including a summary table disaggregated by the requested factors and a table including the overall mean differences and correlations between the observed and (averaged/pooled) imputed values (i.e. mean SGPs and scale scores). The summary table includes several statistics derived for the evaluation of multiple imputation analyses.