inferential statistics

assumptions of parametric tests

* populations drawn from should be normally distributed
* variances of populations should be approximately equal
* at least interval or ratio data
* no extreme scores

non-parametric test

* Mann Whitney U Test
* Chi Square
* Binomial Sign Test
* Wilcoxon Signed Ranks test
* Correlations Spearman’s Rho

observed value

number produced after the various steps and calculations for a statistical test have been carried out

critical value

a value taken from a statistical test, which must be reached in order for results to be significant

checklist for Mann Whitney U Test

* dv produces at least ordinal level of data
* independent measures design
* exploring a difference between each condition

checklist for Wilcoxon Signed Ranks Test

* dv produces at least ordinal level of data
* repeated measures design
* exploring a difference between each condition

checklist for Chi Square

* dv produces nominal data
* independent measures design
* exploring a difference between each condition

checklist for Binomial Sign Test

* dv produces nominal level of data
* repeated measures design
* exploring a difference between each condition

checklist for Spearman’s Rho

* at least ordinal level of data
* correlational design
* relationship between co-variables

steps for calculations - Mann Whitney Test

1. put the scores in order
2. Rank them
3. separate each rank in the ratings they belong to
4. add up each ranks on their group separately
5. choose the lowest group value
6. formula time
7. the result from the formula is the observed value
8. at the critical value tables match the number of p’s in each group together

steps for calculations - Wilcoxon Signed Ranks Test

1. find the difference between each score (ignore when the difference is 0)
2. rank the differences
3. count how many positive figures and how many negative figures there are
4. make calculations using the less frequent sign. Add ranks of those who belonged to the less frequent sign. That’s your observed value
5. n value is the number of participants but we ignore the ones with difference 0
6. match the n value to the table of critical value.

steps for calculations - Chi Square

* add totals of each column
* calculate the expected frequency for each cell (raw total x column total)/overall total
* formula time for each cell
* calculate degrees of freedom (number of rows - 1) x (number of columns -1)

steps for calculations - Binomial Sign Test

* add positive and negative signs
* count each positive and negative sign assigned to participant’s scores
* smallest is the observed value

template for Mann Whitney, Wilcoxon, Binomial, Chi Square

For a test of difference at significance level of (0.05/0.01), when n = (…), the critical value is (…) and the observed value is (…), this (non-parametric test) is (significant/not significant). Therefore we will accept the (null/alternative) and reject the (null/alternative)

\
This means that (there will be a/there is no) significant difference between (…context…)

template for Spearman’s Rho

For exploring a relationship at significance level of (0.05/0.01), when n = (…), the critical value is (…) and the observed value is (…), this (non-parametric test) is (significant/not significant). Therefore we will accept the (null/alternative) and reject the (null/alternative)