sens_plot {ConfoundedMeta}  R Documentation 
Produces line plots (.type=="line"
) showing the bias factor on the relative risk (RR) scale vs. the proportion
of studies with true RRs above .q
(or below it for an apparently preventive relative risk).
The plot secondarily includes a Xaxis scaled based on the minimum strength of confounding
to produce the given bias factor. The shaded region represents a 95% pointwise confidence band.
Alternatively, produces distribution plots (.type=="dist"
) for a specific bias factor showing the observed and
true distributions of RRs with a red line marking exp(.q
).
sens_plot(.type, .q, .muB = NULL, .Bmin = log(1), .Bmax = log(5), .sigB = 0, .yr, .vyr = NULL, .t2, .vt2 = NULL, breaks.x1 = NULL, breaks.x2 = NULL, CI.level = 0.95)
.type 

.q 
True effect size that is the threshold for "scientific significance" 
.muB 
Single mean bias factor on log scale (only needed for distribution plot) 
.Bmin 
Lower limit of lower Xaxis on the log scale (only needed for line plot) 
.Bmax 
Upper limit of lower Xaxis on the log scale (only needed for line plot) 
.sigB 
Standard deviation of log bias factor across studies (length 1) 
.yr 
Pooled point estimate (on log scale) from confounded metaanalysis 
.vyr 
Estimated variance of pooled point estimate from confounded metaanalysis 
.t2 
Estimated heterogeneity (tau^2) from confounded metaanalysis 
.vt2 
Estimated variance of tau^2 from confounded metaanalysis 
breaks.x1 
Breaks for lower Xaxis (bias factor) on RR scale 
breaks.x2 
Breaks for upper Xaxis (confounding strength) on RR scale 
CI.level 
Poitnwise confidence level as a proportion 
Arguments .vyr
and .vt2
can be left NULL
, in which case no confidence
band will appear on the line plot.
# with variable bias and with confidence band sens_plot( .type="line", .q=log(1.1), .Bmin=log(1), .Bmax=log(4), .sigB=0.1, .yr=log(1.3), .vyr=0.005, .t2=0.4, .vt2=0.03 ) # with fixed bias and without confidence band sens_plot( .type="line", .q=log(1.1), .Bmin=log(1), .Bmax=log(4), .yr=log(1.3), .t2=0.4 ) # apparently preventive sens_plot( .type="line", .q=log(0.90), .Bmin=log(1), .Bmax=log(4), .yr=log(0.6), .vyr=0.005, .t2=0.4, .vt2=0.04 ) # distribution plot: apparently causative # commented out because takes 510 seconds to run # sens_plot( .type="dist", .q=log(1.1), .muB=log(2), # .yr=log(1.3), .t2=0.4 ) # distribution plot: apparently preventive # commented out because takes 510 seconds to run # sens_plot( .type="dist", .q=log(0.90), .muB=log(1.5), # .yr=log(0.7), .t2=0.2 )