kurtosis
- noun
- /kərˈtoʊsɪs/
- Specialized
- Preliminary tests of assumptions showed the data to be independent and normally distributed; kurtosis and skewness were at acceptable levels.
- kurtosis value
- skewness and kurtosis
- kurtosis statistic
And maybe the utility function should be adding some penalties for skewness or kurtosis, and that's being ignored here.
- And maybe the utility function should be adding some penalties for skewness or kurtosis, and that's being ignored here.
Examples
-
The values of skewness and kurtosis did not exceed the criteria of extreme skewness (>3) or kurtosis (>8).
Academic text (2013) -
In addition, preliminary statistical procedures were utilized, including tests for skewness and kurtosis.
Academic text (2013) -
With respect to the kurtosis, the reported value suggests a restricted range.
Academic text (2002) -
For instance, skewness and kurtosis values were less than 2 and fell within 1 SE.
Academic text (2006) -
Reliability, skewness, and kurtosis were also calculated for each age group.
Academic text (2015) -
Additionally, a similar pattern of distributional differences with greater kurtosis in the African American group was also observed.
Academic text (2001) -
The items display moderate skewness and kurtosis and possibly should be considered censored.
Academic text (1993) -
No outliers were found, scores were not significantly skewed, and there was no significant kurtosis.
Academic text (2004) -
We first inspected the data for univariate normality and found that skewness and kurtosis for all measures were within 1.0.
Academic text (2010) -
The kurtosis of the dataset indicated that the distribution had heavy tails.
Surface Forms
Morphology
kurtosis = kurt + osis
The element 'kurt' is a non-productive Greek combining form and '-osis' is a technical noun-forming suffix; an average B1 learner cannot derive the statistical meaning from these parts.
Etymology
The word kurtosis comes from the Greek root kurtos, meaning 'curved' or 'bulging'. So kurtosis tells you how 'pointed' or 'flat' a data curve is and how much it has rare very large or very small values on the sides.