You may be referring to the statistical term 'outlier(s)'. Also, there is a rule in statistics called the '68-95-99 Rule'. It states that in a normally distributed dataset approximately 68% of the observations will be within plus/minus one standard deviation of the mean, 95% within plus/minus two standard deviations, and 99% within plus/minus three standard deviations. So if your data follow the classic bell-shaped curve, roughly 1% of the measures should fall beyond three standard deviations of the mean.
Measurements. Just because a particular result lies far from the mean doesn't make it any different. If it's noticeably far from the "crowd" of all the other measurements, it can be called an outlier. An outlier isn't necessarily bad, just different. It should be examined in detail to see if there's something odd about it, but not discarded out of hand.
99.7% of scores fall within -3 and plus 3 standard deviations around the mean in a normal distribution.
This is a question on a statistics puzzle. Yes, it is for a puzzle. What is the answer?
Chebyshev's inequality: The fraction of any data set lying within K standard deviations is always at least 1-1/K^2 where K is any positive number greater than 1. It does not assume that any distribution. Now, there is the empirical rule of bell shaped curves or the 68-95-99.7 rule, which states that for a bell shaped curve: 68% of all values should fall within 1 standard deviation, 95% of all values should fall within 2 standard deviations and 99.7% of all values should fall within 3 standard deviation. If we suspect that our data is not bell shaped, but right or left skewed, the above rule can not be applied. I note that one test of skewness is Pearson's index of skewness, I= 3(mean of data - median of data)/(std deviation) If I is greater or equal to 1000 or I is less than 1, the data can be considered significantly skewed. I hope this answers your question. I used the textbook Elementary Statistics by Triola for the information on Pearson's index. If this answer is insufficient, please resubmit and be a bit more definitive on what you mean by empirical rule.
Outliers.
It is one of the informal definitions for an outlier.
outliers
I believe outliers is the best answer to this question. The previous answer was average, which is the mean.
Usually they would be observations with very low probabilities of occurrence.
variances
They are observations with a low likelihood of occurrence. They may be called outliers but there is no agreed definition for outliers.
Outliers
Extreme values. They might also be called outliers but there is no agreed definition for the term "outlier".
You may be referring to the statistical term 'outlier(s)'. Also, there is a rule in statistics called the '68-95-99 Rule'. It states that in a normally distributed dataset approximately 68% of the observations will be within plus/minus one standard deviation of the mean, 95% within plus/minus two standard deviations, and 99% within plus/minus three standard deviations. So if your data follow the classic bell-shaped curve, roughly 1% of the measures should fall beyond three standard deviations of the mean.
It is a measurement which may, sometimes, be called an extreme observation or an outlier. However, there is no agreed definition for outliers.
Measurements. Just because a particular result lies far from the mean doesn't make it any different. If it's noticeably far from the "crowd" of all the other measurements, it can be called an outlier. An outlier isn't necessarily bad, just different. It should be examined in detail to see if there's something odd about it, but not discarded out of hand.