Outlier

(redirected from Outlying Observation)
Also found in: Dictionary, Thesaurus, Medical, Encyclopedia.

Outlier

A data point significantly different from others in the same set. An outlier is generally due to statistical noise and not to any fundamental difficulty with the data set. Taking the mean and median values of a data set can help reduce the influence of outliers. They are also called outlying observations.
Mentioned in ?
References in periodicals archive ?
ANOVA revealed that when level-1 outliers were present, the interaction of method by level-2 sample size by magnitude of contamination was significantly related to [gamma].sub.00] estimation bias ([F.sub.54,54] = 1.554, p = 0.02, [[eta].sub.2] = 0.154), as was the main effect of percentage of outlying observations ([F.sub.1,9] = 6.533, p = 0.022, [[beta].sub.2] = 0.140).
Outlying observations were not found in relation to the measured parameters and the distribution normality was not impaired.
In particular, extreme points, influential and outlying observations, which might represent erroneous data or otherwise reflect unusual market conditions such as non-arm's length transactions, can seriously undermine the performance of parametric estimator.
If any data points appear aberrant, an evaluation is done using tools such as the Dixon, Grubbs or Cochran tests for statistically evaluating outlying observations. If a single outlier is found, it is rejected or included based on the results of this evaluation.
Consensus means were determined by a recursive testing procedure for outlying observations. The means and SDs for serum pools 1-20 were first calculated using all reported observations.
Two new organizations in the sample were so large that, in statistical terms, they were outlying observations. Therefore, consideration was given to excluding them from the sample.
The estimation of the regression function values for the ranges is an iterative process with minimization of the influence of the most outlying observations (Matlab, 2005).