Monday, November 16, 2009

Statistics?

This ones kind of a problem i was given to do at my 2nd year at uni in my statistics module. I was given 30 mesurments of the heights of men from Hungary and 30 Heights of woomen from Peru. And basicly from here we were told to proceed with any necesary tests. I thought of doing a boundry test on the sample to erase any outliers, then variance test and then a mean test. any sujestions? Am i going the right way??

Statistics?
Sounds about right...but don't forget to state that you are assessing a hypothesis that men from hungary have the same height as women from peru, based on the sample data you have available.
Reply:With 30 height-data-points in each sample you expect a normal distribution. Weeding out outliers may be biased. Histogram the heights and try a Kolmogorov-Smirnov test.
Reply:You can do Average, Median, Standard deviation test.


The teacher wants to know what can you deduce from the measurements. The median will tell you if the men from Hungary are really tall than women form Peru. In median the outlier are sorted of weighted out


No comments:

Post a Comment