Variable names and data column names must not contain periods. Rename them consistently, for example x.1 to x_1, and check that the new names are unique. Study, group, and cluster values may contain periods.
It conducts univariate and multivariate meta-analysis with maximum likelihood estimation method. Mixed-effects meta-analysis can be conducted by including study characteristics as predictors. Equality constraints on intercepts, regression coefficients, and variance components can be easily imposed by setting the same labels on the parameter estimates.
meta(
y,
v,
x,
data,
intercept.constraints = NULL,
coef.constraints = NULL,
RE.constraints = NULL,
RE.startvalues = 0.1,
RE.lbound = 1e-10,
intervals.type = c("z", "LB"),
I2 = "I2q",
R2 = TRUE,
model.name = "Meta analysis with ML",
suppressWarnings = TRUE,
silent = TRUE,
run = TRUE,
...
)metaFIML(
y,
v,
x,
av,
data,
intercept.constraints = NULL,
coef.constraints = NULL,
RE.constraints = NULL,
RE.startvalues = 0.1,
RE.lbound = 1e-10,
intervals.type = c("z", "LB"),
R2 = TRUE,
model.name = "Meta analysis with FIML",
suppressWarnings = TRUE,
silent = TRUE,
run = TRUE,
...
)
An object of class meta with a list of
Object
returned by match.call
A data matrix of y, v and x
Number of effect sizes
Number of predictors
A vector indicating whether the predictors are missing.
Studies will be removed before the analysis if they are TRUE
Types of I2 calculated
Logical
A fitted
object returned from mxRun
A fitted
object without any predictor returned from mxRun
A vector of effect size for univariate meta-analysis or a \(k\) x \(p\) matrix of effect sizes for multivariate meta-analysis where \(k\) is the number of studies and \(p\) is the number of effect sizes.
A vector of the sampling variance of the effect size for univariate
meta-analysis or a \(k\) x \(p*\) matrix of the sampling
covariance matrix of the effect sizes for multivariate meta-analysis where
\(p* = p(p+1)/2 \). It is arranged by column major as used
by vech.
A predictor or a \(k\) x \(m\) matrix of predictors where \(m\) is the number of predictors.
An optional data frame containing the variables in the model.
A \(1\) x \(p\) matrix specifying
whether the intercepts of the effect sizes are fixed or free. If the input
is not a matrix, the input is converted into a \(1\) x \(p\)
matrix with t(as.matrix(intercept.constraints)). The default is that
the intercepts are free. When there is no predictor, these intercepts are
the same as the pooled effect sizes. The format of this matrix follows
as.mxMatrix. The intercepts can be constrained
equally by using the same labels.
A \(p\) x \(m\) matrix specifying how the
predictors predict the effect sizes. If the input is not a matrix, it is
converted into a matrix by as.matrix(). The default is that all
\(m\) predictors predict all \(p\) effect sizes. The format of
this matrix follows as.mxMatrix. The regression
coefficients can be constrained equally by using the same labels.
A \(p\) x \(p\) matrix specifying the
variance components of the random effects. If the input is not a matrix, it
is converted into a matrix by as.matrix(). The default is that all
covariance/variance components are free. The format of this matrix follows
as.mxMatrix. Elements of the variance components can
be constrained equally by using the same labels. If a zero matrix is
specified, it becomes a fixed-effects meta-analysis. Lower bounds keep the
variances non-negative, but with free covariances the estimated variance
component can still be not positive semidefinite; meta() warns in
this case, and a diagonal matrix, e.g.,
RE.constraints=Diag(paste0("0.1*Tau2_", 1:p, "_", 1:p)), may be
used instead.
A vector of \(p\) starting values on the diagonals of the variance component of the random effects. If only one scalar is given, it will be duplicated across the diagonals. Starting values for the off-diagonals of the variance component are all 0. A \(p\) x \(p\) symmetric matrix of starting values is also accepted.
A vector of \(p\) lower bounds on the diagonals of the
variance component of the random effects. If only one scalar is given, it
will be duplicated across the diagonals. Lower bounds for the off-diagonals
of the variance component are set at NA. A \(p\) x \(p\)
symmetric matrix of the lower bounds is also accepted.
Either z (default if missing) or LB. If
it is z, it calculates the 95% Wald confidence intervals (CIs) based
on the z statistic. If it is LB, it calculates the 95%
likelihood-based CIs on the parameter estimates. Note that the z values and
their associated p values are based on the z statistic. They are not related
to the likelihood-based CIs.
Possible options are "I2q", "I2hm" and
"I2am". They represent the I2 calculated by using a typical
within-study sampling variance from the Q statistic, the harmonic mean and
the arithmetic mean of the within-study sampling variances (Xiong, Miller, &
Morris, 2010). More than one options are possible. If
intervals.type="LB", 95% confidence intervals on the heterogeneity
indices will be constructed.
Logical. If TRUE and there are predictors, R2 is calculated
(Raudenbush, 2009).
A string for the model name in
mxModel.
Logical. If TRUE, warnings are suppressed.
The argument to be passed to mxRun.
Logical. An argument to be passed to
mxRun
Logical. If FALSE, only return the mx model without
running the analysis.
Further arguments to be passed to mxRun
An auxiliary variable or a \(k\) x \(m\) matrix of auxiliary variables where \(m\) is the number of auxiliary variables.
Mike W.-L. Cheung <[email protected]>
Cheung, M. W.-L. (2008). A model for integrating fixed-, random-, and mixed-effects meta-analyses into structural equation modeling. Psychological Methods, 13, 182-202.
Cheung, M. W.-L. (2009). Constructing approximate confidence intervals for parameters with structural equation models. Structural Equation Modeling, 16, 267-294.
Cheung, M. W.-L. (2013). Multivariate meta-analysis as structural equation models. Structural Equation Modeling, 20, 429-454.
Cheung, M. W.-L. (2015). Meta-analysis: A structural equation modeling approach. Chichester, West Sussex: John Wiley & Sons, Inc.
Hardy, R. J., & Thompson, S. G. (1996). A likelihood approach to meta-analysis with random effects. Statistics in Medicine, 15, 619-629.
Neale, M. C., & Miller, M. B. (1997). The use of likelihood-based confidence intervals in genetic models. Behavior Genetics, 27, 113-120.
Raudenbush, S. W. (2009). Analyzing effect sizes: random effects models. In H. M. Cooper, L. V. Hedges, & J. C. Valentine (Eds.), The handbook of research synthesis and meta-analysis (2nd ed., pp. 295-315). New York: Russell Sage Foundation.
Xiong, C., Miller, J. P., & Morris, J. C. (2010). Measuring study-specific heterogeneity in meta-analysis: application to an antecedent biomarker study of Alzheimer's disease. Statistics in Biopharmaceutical Research, 2(3), 300-309. doi:10.1198/sbr.2009.0067
reml, Hox02,
Berkey98, wvs94a