lavaan.mi (version 0.1-0)

parameterEstimates.mi: Pooled Parameter Estimates

Description

This function pools parameter estimates from a lavaan model fitted to multiple imputed data sets.

Usage

parameterEstimates.mi(
  object,
  se = TRUE,
  zstat = se,
  pvalue = zstat,
  ci = TRUE,
  level = 0.95,
  fmi = FALSE,
  standardized = FALSE,
  cov.std = TRUE,
  rsquare = FALSE,
  asymptotic = FALSE,
  scale.W = !asymptotic,
  omit.imps = c("no.conv", "no.se"),
  remove.system.eq = TRUE,
  remove.eq = TRUE,
  remove.ineq = TRUE,
  remove.def = FALSE,
  remove.nonfree = FALSE,
  remove.unused = FALSE,
  output = "data.frame",
  header = FALSE
)

Value

A data.frame, analogous to lavaan::parameterEstimates(), but estimates, SEs, and tests are pooled across imputations.

Arguments

object

An object of class lavaan.mi

se, zstat, pvalue, ci, level, standardized, cov.std, rsquare, remove.system.eq, remove.eq, remove.ineq, remove.def, remove.nonfree, remove.unused, output, header

See lavaan::parameterEstimates().

fmi

logical indicating whether to add 2 columns:

• the fraction of missing information ($fmi), which is the ratio of between-imputation variance to total (pooled) sampling variance

• the relative increase in variance ($riv), which is the ratio of between-imputation variance to within-imputation variance

Thus, RIV = FMI / (1 \(-\) FMI) and FMI = RIV / (1 + RIV). Ignored when se=FALSE.

asymptotic

logical. When FALSE, pooled Wald tests will be t statistics with associated degrees of freedom (df). When TRUE, the df are assumed to be sufficiently large for a t statistic to approximate a standard normal distribution, so it is printed as a z statistic.

scale.W

logical. If TRUE (default), the vcov method will calculate the pooled covariance matrix by scaling the within-imputation component by the ARIV (see Enders, 2010, p. 235, for definition and formula). Otherwise, the pooled matrix is calculated as the weighted sum of the within-imputation and between-imputation components (see Enders, 2010, ch. 8, for details).

omit.imps

character indicating criteria for excluding imputations from pooled results. See lavaan.mi for argument details.

Author

Terrence D. Jorgensen (University of Amsterdam; )

References

Enders, C. K. (2010). Applied missing data analysis. New York, NY: Guilford.

Rubin, D. B. (1987). Multiple imputation for nonresponse in surveys. New York, NY: Wiley. tools:::Rd_expr_doi("10.1002/9780470316696")

See Also

standardizedSolution.mi() to obtain inferential statistics for pooled standardized parameter estimates.

Examples

data(HS20imps) # import a list of 20 imputed data sets

## specify CFA model from lavaan's ?cfa help page
HS.model <- '
  visual  =~ x1 + x2 + x3
  textual =~ x4 + x5 + x6
  speed   =~ x7 + x8 + x9
'
## fit model to 20 imputed data sets
fit <- cfa.mi(HS.model, data = HS20imps)

## pooled estimates, with various optional features:

parameterEstimates.mi(fit, asymptotic = TRUE, rsquare = TRUE)
parameterEstimates.mi(fit, ci = FALSE, fmi = TRUE, output = "text")
parameterEstimates.mi(fit, standardized = "std.all", se = FALSE)