iboot-v3.2.1

• Added a new function, bootanovan, which is a wrapper for anovan (from the Statistics package in Octave, or the Statistics and Machine Learning toolbox in Matlab). The p-values are calculated from the bootstrap distribution of the ANOVA F-statistics.
• Corrected file version and dependencies in DESCRIPTION

iboot-v3.1.7

• Changed behaviour of NaN in group for bootnhst
• Improved maxq function used by bootnhst to cope better with unequal sample sizes

iboot-v3.1.5

• Changed order of input arguments for bootnhst

iboot-v3.1.4

• Updated bootnhst to compute standard errors using bootknife resampling

iboot-v3.1.3

Fixed bug which meant that the wrong standard error was reported in the stats.groups output of bootnhst

iboot-v3.1.2

Update to bootnhst (and maxq)

iboot-v3.1.1

• fixed broadcast warning in smoothmedian
• fixed MarkerFace warning in Octave

iboot-v3.1.0

• bootnhst can now compute (and plot) bootstrap-t confidence intervals for post hoc comparisons
• bootnhst now have bootfun 'robust' which uses the smoothed median
• bootnhst now prints group numbers allocated to elements of the GROUP argument entered by the user
• bootnhst (and maxq) now support group exclusion from hypothesis testing by assigning Inf or NaN in Group
• speed enhancement to smoothmedian function and added ability to compute standard errors

iboot-v3.0.3

• fixed minor bug that caused data and text to be displayed when using jack with octave parallel

iboot-v3.0.2

• fixed bug in bootnhst that could cause low power for posthoc tests - the test statistic is now studentized and posthoc tests now control FWER similarly to Tukey's Honest Signficance test and Dunnett's test for all pairwise comparisons or treatment versus control comparisons respectively. The sampling variance is estimated using Tukey's Jackknife, and the calculation of pooled standard error accomodates for unequal sample size.