noun: Misspelling of jackknife. [(statistics) A resampling method that applies estimators to all subsamples that each omit a single different group (possibly of a single datapoint) of the original sample to provide a sample distribution of the estimate.]
