simsem (version 0.2-8)

miPool: Function to pool imputed results

Description

The function takes a list of imputed results (parameters and standard errors) and returns pooled parameter estimates and standard errors.

Usage

miPool(Result.l)

Arguments

Result.l

List of SimModelOut used for pooling based on Rubin's rule.

Value

• MIpool returns a list with pooled estimates, standard errors, fit indices and fraction missing information
• Estimates
  Pooled parameter estimates.
• SE
  Pooled standard errors.
• FMI.1
  Fraction of missing information for each parameter.
• FMI.2
  Fraction of missing information for each parameter.

Details

All parameter estimates are pooled by Rubin's (1987) rule. The chi-square statistics are pooled by the procedure proposed by Li, Meng, Raghunathan, and Rubin (1991; Equations 2.1, 2.2, 2.16, and 2.17). The resulting value from the pooled chi-square is F-statistic. The F-statistics is multiplied by the numerator degree of feedom to provide the value equivalent to chi-square value. This chi-square value will be used to find other chi-squared related fit indices: RMSEA, CFI, and TLI. SRMR, AIC, and BIC are pooled by Rubin's (1987) rule. The function for pooling chi-squared statistics is adapted from Craig Enders' SAS code from "http://psychology.clas.asu.edu/files/CombiningLikelihoodRatioChi-SquareStatisticsFromaMIAnalysis.sas".

References

Rubin, D.B. (1987) Multiple Imputation for Nonresponse in Surveys. J. Wiley & Sons, New York. Li, K. H., Meng, X. L., Raghunathan, T. E., & Rubin, D. B. (1991). Significance levels from repeated p-values with multiply-imputed data. Statistica Sinica, 1, 65-92.

See Also

• miPoolVectorfor pooling multiple imputation results that providing in matrix or vector formats
• miPoolChifor pooling multiple imputated chi-square statistics
• runMIfor imputing missing values by multiple imputation and analyzing the imputed datasets.

Examples

# No Example