acc.1test()
and acc.paired()
now allow the user (via the
argument method.ci
)to choose from a range of differnt types of confidence
intervals.tab.paired()
and tab.1test()
now allow for data where all
subjects are either diseased or nondiseased.Test 1
is consistently used as the reference test.pv.gs()
and pv.wgs()
, it is now diff.ppv <- ppv.2-ppv.1
(instead of
diff.ppv <- abs(ppv.1-ppv.2)
), and accordingly for negative predictive
values.pv.prev()
) to allow computation of positive and
negative predictive values for different theoretical prevalences.sesp.gen.mcnemar()
) implementing a generalized
McNemar's test for a joint comparison of sensitivity and specificity.man/dtcompair-package.rd
was deleted.pv.rpv()
now returns the full variance-covariance matrix (Sigma
).ellipse.pv.rpv()
generates the data to plot a joint confidence region for
rPPV and rNPV (depends on the ellipse
package) (as in Moskowitz and Pepe,
2006).sesp.rel()
calculates relative sensitivity and relative specificity (with
Wald CIs and p-value).tpffpf.rel()
calculates relative sensitivity (rTPF) and relative 'one minus
specificity' (rFPF) (with Wald CIs and p-value), but it does not calculate
their individual components (ie, TPFs and FPFs); this function is meant to be
used with paired screen-positive designs, where only rTPF and rFPF are
estimable form the data (see Cheng and Macaluso, 1997 or Alonzo, Pepe,
Moskowitz, 2002).sesp.diff.ci
(detected by F. Gimenez - many thanks!).sesp.exactbinom
(detected by J. Swiecicki - many thanks!).Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.