# Statistical Skills

?
• Created by: 8cburton
• Created on: 02-06-15 16:53

## Measures of central tendancy- Mean, median and mod

• Simple and easy
• Mode can be used with non numerical data
• Median- very large and very small numbers do not affect result
• Mean- useful in making measurements more accurate

• Cant use discontinuous data
• Median and mode do not account for whole set of data
• Mean is easily disorted by very large/small anomalies
1 of 6

## Interquartile range

• Interquatile range is the spread of values around the median
• Find out the LQ and the HQ and the difference is the IQR

• Not affected by the outliers

DISVANTAGES

• Not all data considered
• Complicated to work out
2 of 6

## Standard deviation

• the average amount by which the values in a data set vary from the mean
• Calculate the mean and minus it from X
• Then divide by n and square root
• Low standard deviation means little range and therefore reliable mean

• More reliable measure of dispersal as it uses all the data

• Can be greatly affected by outliers
3 of 6

## Spearmans rank

• Formulate a null hypothesis.
• Individually rank the values of each variable. 1 = highest value.
• Find the difference between the two.
• Square the differences and sum the values.
• Input into the formula.

• Indicates the statistical significance of a result - rules out chance.
• Gives numerical value to the strength and direction of a correlation.

• Does not show if there is a casual link
• Too many tied ranks affect the validity of the test.
• Subject to human error.
• Only appropriate for data with 10-30 values with 2 variables that are believed to be related
4 of 6

## Mann Whitney U

• Select null hypothesis.
• Rank the data sets across the two columns. 1 = lowest value.
• Treat as two seperate columns. Add ranks in first column to get your R1 value then add ranks in the second column to get your R2  value.
• Input int the formula.
• Choose the smaller U value of either U1 or U2.
• Compare to the critical values table: less than the critical value means you should reject the null hypothesis at 95% confident. Greater than the critical value - accpet the nul.

Use:

• Used to show if there is a statistical difference between two sets of data e.g. size of rocks in upper course and lower course.

Cons:

• Does not explain cause and effect.
5 of 6

## Chi square

• Identify null hypothesis - no significant difference between observed an expected.
• Subtract observed frequencies from expected and square the result.
• Divide this by the expected value for that group.
• Compare with degrees of freedom: on the critical values chart, the degree will be one less than the total number of observed values.

• To assess the degree of difference between observed and theoretical data e.g. number of pebbles along a river.
• Statistical significance of results can be tested.