Ive got to do my maths statistics coursework for next Friday and as the teacher went through what we had to do before the summer I cannot seem to remember what skewness and normal distribution are and how I would work them out.
Ive been searching through Google but all of the sites are far too complex.
Would greatly appreciate it if someone could explain these terms to me.
Turn on thread page Beta
What are normal distribution and skewness? watch
- Thread Starter
- 13-10-2006 22:15
- 13-10-2006 23:08
The normal distribution is a continuous random variable whose p.d.f. looks like a bell-shaped curve. It is useful in stats because many experiments produce data which is approximately normally distributed. Also it can be used to approximate the binomial and poisson distributions. If you haven't encountered such distributions before, think of your entire school being set a test, and the results collected. Most people will get a score near the average (maybe this is 70%), and very few people will get scores in the extremes (e.g. <50% and >90%). The 'normal distribution' is a theoretical variable which can be used to model the distribution of these scores. To be honest, if you don't really understand it, it's probably not worth putting it in coursework.
Skew is much simpler. When you draw the frequency graph of your data, if the 'tail' is on the right there is +ve skew. If the tail is on the left, there is a -ve skew. If the graph is largely symmetrical (like the normal distribution curve), there is zero skew. Skew is especially evident when summarising data with box plots.
- 14-10-2006 00:10
for a box plot skew is -ve when Q2-Q1 < Q3-Q2 and +ve when Q2-Q1 > Q3-Q2
alternatively can calculate skew with formula 3(mean-median)/ standard deviation