PSYC 3000 Lecture Notes - Lecture 4: Error Bar, Standard Deviation, Interquartile Range
Document Summary
Boxplots tend to boil things down a bit too much and detail/information is lost because of this. Histograms are able to pick up bimodality if the binning is made small enough. Boxplots do give an idea of symmetry, outliers. Standard deviation tells us something about the variability in data. Standard error tells us how well the sample mean reflects the true population mean. If the standard error is small, then the sample mean is accurate. Bigger the sample size, the smaller the sample error is. Sample size will not affect variation in the population. The error bar can show +/- 1, 2 or 3 se or sd. Reason this graph is so popular is because you compare two means and using standard errors you can see if the difference between the two is significant. The means must be 1. 96 standard errors apart to be significant.