maid-Methods: Methods for 'maid' objects

In maid: Meta-Analysis of Individual Differences

Methods for maid objects

Description

Standard and plot methods are documented here for maid objects that are the output of the mabb function

Usage

## S3 method for class 'maid'
print(x, digits = 4, ...)
## S3 method for class 'maid'
summary(object, ci.level = 0.95, ...)
## S3 method for class 'summary.maid'
print(x, digits = 4, ...)
## S3 method for class 'maid'
logLik(object, ...)
## S3 method for class 'maid'
qtest(object, ...)
## S3 method for class 'maid'
plot(x, pooled = TRUE, 
    main = "Meta-analysis of Beta-binomial parameters", 
    ellipsecol = "grey50", pool_pch = 19, pool_cex = 1, 
    predellipse_lty = 2, predellipse_col = "black", 
    backtransform = TRUE, xlim = "auto", ylim = "auto", ...)
## S3 method for class 'maid'
alphaKR21(x, ci.level = 0.95, ...)
## S3 method for class 'alphaKR21.maid'
print(x, digits = 4, ...)

Arguments

x	a maid object (these are the output of the mabb function).
object	a maid object (these are the output of the mabb function).
digits	an integer specifying the number of digits to which printed results must be rounded.
ci.level	a numerical value between 0 and 1, specifying the confidence level for the computation of confidence intervals.
pooled	logical. Should the pooled value and its prediction ellipse be plotted?
main	caption of plot.
ellipsecol	color of confidence ellipses for study level estimates.
pool_pch	symbol for pooled value.
pool_cex	size of pooled value.
predellipse_lty	style of line of prediction ellipse.
predellipse_col	color of prediction ellipse.
backtransform	logical. Should logit(pi) and, if applicable, logit(rho) values be backtransformed prior to plotting? Defaults to TRUE
xlim	Either "auto" or a numeric vector of length 2. If the former, the function determines the plotting range for the x-axis.
ylim	Either "auto" or a numeric vector of length 2. If the former, the function determines the plotting range for the y-axis.
...	further arguments passed to or from other methods.

Details

All methods are identical to or derived from their counterparts in the mvmeta-package.

Value

The summary method function for maid objects produces a list of class summary.maid. The plot method function produces a plot and returns an invisible NULL. See the examples for plotting examples.

The alphaKR21 method for maid objects extracts the moment estimates of logit(\pi) and \rho (or logit(\rho) if logitrho is TRUE) on the study level and computes Wald confidence intervals. In addition, the function calculates the Kuder & Richardson (1937) formula 21 reliability estimates. The confidence interval for these estimates is based on the standard errors of the rho estimates. The alphaKR21 produces an object of class c("alphaKR21.maid", "data.frame") which is essentially a data.frame with columns logitpi, se.logitpi, ci.logitpi.lower, ci.logitpi.upper, rho, se.rho, ci.rho.lower, ci.rho.upper, alphaKR21, ci.alphaKR21.lower and ci.alphaKR21.upper. If logitrho is TRUE, the estimates are given for logit(\rho) and the respective columns are labeled logitrho, se.logitrho, ci.logitrho.lower, ci.logitrho.upper.

Author(s)

Philipp Doebler <[e-mail]>

See Also

mabb, summary.mvmeta

Examples

## Plotting a meta-analysis of individual differences

## Fit model
data(ICAR60) # load example data file
fit <- mabb(N = ICAR60$N, size = ICAR60$test_length, 
            obsmean = ICAR60$obsmean, obssd = ICAR60$obssd)
            
## print moment estimates and alpha KR21 estimates
alphaKR21(fit)

## change the default ci.level to .8 and print more digits
print(alphaKR21(fit, ci.level = .8), digits = 5)

## Produce default plot
plot(fit)
## Add a legend
legend("bottomright", 
       legend = c("study level estimate", "95% confidence ellipse", 
       "pooled value", "95% prediction ellipse"),
       lty = c(NA, 1, NA, 2), col = c("black", "grey50", "black"),  pch = c(1,NA, 19, NA))

## Variants of this plot (using all arguments and some ... methods)
plot(fit,
     pooled = TRUE, # TRUE is the default value
     main = "My meta-analysis",
     ellipsecol = "red", # color of the small ellipses for each study,
     pool_pch = 13, # symbol for pooled value
     pool_cex = 2, # size of symbol for pooled value
     predellipse_lty = 3, # style of line
     predellipse_col = "blue", # default is "black"
     backtransform = FALSE, # logit(pi) on x-axis
     xlim = "auto", # function computes xlim (default)
     ylim = c(.11, .15), # only plot values between .11 and .15 (+eps)
     # now some further argument passed on to plot.default() :
     pch = 2,  # symbol for study estimates
     cex = 0.5, # size of symbols for study estimates
     col = "green") # color of symbols for study estimates

maid documentation built on March 31, 2023, 3:07 p.m.