dataFigure2 {superb}

Data for Figure 2

Description

The data, taken from (Cousineau 2017)7, is an example where the "stand-alone" 95\% confidence interval of the means returns a result in contradiction with the result of a statistical test. The paradoxical result is resolved by using adjusted confidence intervals, here the correlation- and different-adjusted confidence interval.

Usage

data(dataFigure2)

Format

An object of class data.frame.

Source

doi:10.5709/acp-0214-z

References

Cousineau D (2017). “Varieties of confidence intervals.” Advances in Cognitive Psychology, 13, 140 – 155. doi:10.5709/acp-0214-z.

Examples

library(ggplot2)
library(gridExtra)
data(dataFigure2)

options(superb.feedback = 'none') # shut down 'warnings' and 'design' interpretation messages

## realize the plot with unadjusted (left) and ajusted (right) 95% confidence intervals
plt2a <- superbPlot(dataFigure2, WSFactors = "Moment(2)", 
    adjustments=list(purpose = "difference"), 
    variables = c("pre","post"), plotStyle="bar" ) + 
  xlab("Group") + ylab("Score") + labs(title="Difference-adjusted 95% CI\n") +
  coord_cartesian( ylim = c(85,115) ) +
  geom_hline(yintercept = 100, colour = "black", linewidth = 0.5, linetype=2)
plt2b <- superbPlot(dataFigure2, WSFactors = "Moment(2)", 
    adjustments=list(purpose = "difference", decorrelation = "CA"), 
    variables = c("pre","post"), plotStyle="bar" ) + 
  xlab("Group") + ylab("Score") + labs(title="Correlation and difference-adjusted\n95% CI") +
  coord_cartesian( ylim = c(85,115) ) + 
  geom_hline(yintercept = 100, colour = "black", linewidth = 0.5, linetype=2)
plt2  <- grid.arrange(plt2a,plt2b,ncol=2)

## realise the correct t-test to see the discrepancy
t.test(dataFigure2$pre, dataFigure2$post, paired=TRUE)