population means are grouped together based on statistical s

population means are grouped together based on statistical significance. The graphs at the right are from a study investigating hippocampal volume loss in 107 patients with drug-resistant epilepsy (Cook et al. 1993). The graphs depict the association between hippocampal volume loss (measured using MRI) as the volume of the smaller half of the hippocampus divided by the volume of the larger half, expressed as a percentage) and patient history. Patients were grouped on the basis of whether they had a record of childhood febrile seizures (CFS), childhood non-febrile seizures (no CFS) ind no childhood seizures. Which accompanying graph, the box plot (top) or the bar graph (bottom; indicating means and SEs), best depicts the patterns in the data? Why? Which statistical method would you recommend to test whether groups differed in hippocampal volume loss? Why?

Solution

a) CFs

because the data as you can see in the box plots are normally distributed because we have

data in all quartiels

b)

I would recomend T test or anova