APST 207 Lecture Notes - Lecture 5: Kurtosis, Unimodality, Sampling Distribution
Document Summary
More descriptives : kurtosis, skew, and sampling distributions. Kurtosis: a measure of how peaked the data is. Skew: is the degree to which the distribution is asymmetrical. Positive skew: most of the value are low with high outliers. Mean greater than the median thus positive. Negative skew: most of the values are higher, but there are some low outliers. Most distributions are unimodal-they only have one peak or mode. Some distributions are bimodal- they have two peaks or modes. The shape of a distribution is determined by its kurtosis and skew. Kurtosis : how peaked the distribution is. Normal- what we hope for, kurtosis value of 0. Platykurtic- not very peaked, - kurtosis value. Positive skew- mostly low values, tail points toward higher values. Negative skew- mostly high values, tail points toward lower values. Modal points: how many peaks the data has.
