Statistical Tests
- Created by: mollyb
- Created on: 13-06-13 15:19
View mindmap
- Statistics
- Mean
- add the quantities together and divide by the number of quantities
- Most useful but requires calculation
- Distorted by extreme values and involves calculation
- Median
- the central value when values are put in order
- gives no idea of other values
- Mode
- the value that occurs most frequently
- gives no idea of other values
- Range
- difference between the highest and the lowest
- misleading if their are abnomalies
- Interquartile Range
- difference between the inner half around the mean
- quick and easy to calculate
- ignores values above and below this
- Standard Deviation
- shows a spread of all values around the mean
- uses a formula and calculation and distorted by extreme values
- larger the S.D the larger the variation
- uses when comparing similar means
- Mann Whitney U
- compares medians and ranks them to see if data sets differ
- Disadvantage: uses a formula and calculation making working out difficult and uses a significance tables
- Tests for a difference and starts by assuming that both sets of data are alike and then establishes whether its is safe to reject assumptions
- Chi Squared
- Difference
- compares observed and expected frequecies
- uses a formula, calculation + significance table
- how do you determine expected frequency?
- Correlation
- compares observed and the frequency expected given a certain hypothesis
- uses a formula, calculation and a significance table
- how do determine expected frequency?
- Difference
- Spearmans Rank
- measures the strength of a relationship between two sets of ranked data
- uses a formula and calculation
- only uses ranks of data and significance tables
- Skew
- the extent to which the mode and mean differ.
- can be positive or negative and can occur if distributions are not symmetrical
- if the mode is lower that the mean it is said to be negatively skewed and visa versa.
- Kurtosis
- used to show wave length/ amplitude but shows shape
- shows shapes such as wavelength and distributions can be flat or in between.
- Mean
- Tests for a difference and starts by assuming that both sets of data are alike and then establishes whether its is safe to reject assumptions
Comments
No comments have yet been made