Monday, February 8, 2010

STATISTICS HELP-indicate whether the following statement is true or false and why??

';an outlier constitutes an obviously bad data value which must be perged from the data set before analysis can be undertaken.';STATISTICS HELP-indicate whether the following statement is true or false and why??
FALSE





A complete definition of an outlier has yet to be developed. The only way to remove a data point from a sample and not risk compromising the data is if you know for a fact the data point is a mistake due to a transcribing error. There are many convensions for outliers, but not one of them is a robust and 100% correct.





one method is any data point more than 1.5 inner quartile ranges above or below the 3rd or 1st quartile respectively can be called an outlier.





Some will use the idea that a data point more than 3 standard deviations away from the mean is an outlier, the 6sigma idea.





';outliers'; can be very helpful to have. they can give us insight into the behavior of a distribution in this extremes.





There is no formal definition for an outlier, only conventions.





For example. Say you have ten test scores. 90, 84, 87, 91, 88, 91, 90, 23, 86, 89. The value of 23 looks like an ';obviously bad data point'; but the student taking this test had not performed well for any number of reasons and this score is consistent with his/her other scores. The value is not an outlier and should be left in the data set even though it will weight in heavy for the calculations of mean and standard deviation.STATISTICS HELP-indicate whether the following statement is true or false and why??
True. If a value is so far removed from every other value being analyzed, there is an obvious flaw in that one particular value. If it used, it will skew your analysis, and cause the result to be unreliable.
I only did hear that word in biology-

No comments:

Post a Comment