statistics

• measures of central tendency produce a single number which represents the middle value.
• the mean takes into account all scores so is more representative but it is affected by outliers
• the mode is less affected by outliers but data could have two modes
• the median isn't affected by outliers but it doesn't represent a typical average
• measures of dispersion show how spread out the results within a data set are
• the range is easy to calculate but it isn't representative
• the variance considers the difference between each data point and the mean. takes into account every score so isn't mislead by outliers
• standard deviation converts the variance into the same units as the mean
• a histogram uses ordinal or interval continuous data
• normal distribution curve has the mean, mode and median all together in the middle.
• a positive skewed curve is a side way P shape. most scores are low
• a negative skewed curve is a side way b shape. most scores are high
• for parametric tests the data must be equal interval and normal distribution.
• chi squared and spearman Rho value must be more to be significant.
• wilcoxon, binomial and mann whitney U value must be less to be significant.
• mann whitney U can have different sample sizes
• wilcoxon works out the difference between scores then ranks the differences. then add up the + and - ranks.
• mann whitney U ranks both groups as 1
• spearman rho ranks the data separate and then works out the difference between the ranks
• chi squared labels the cells and works out the total row x total column and degrees of freedom (n row - 1) x (n column -1)