Statistical Tests Flashcards
How to calculate a binomial sign test
Step 1-is the difference positive, if it is give a +, if it’s negative give -
Step 2-if participant had no difference leave them out of test
Step 3-calculate the number of times the least frequency sign (- or +) appears, this is the calculated value of X
Step 4-find critical value
Step 5-if X value is less than critical, reject null
How to calculate Chi-square test
Step 1-calculate expected frequencies for each cell separately using formula fe=(RxC)➗T (R is total of row cells, C is total of column cells, T is overall total)
Step 2-calculate difference between observed and expected, O - E
Step 3-square each difference
Step 4-divide the square by the expected for that cell
Step 5-add up each one to give X squared value
Step 6-degrees of freedom is (row-1)(colour-1)
Step 7-if X squared value equal to or bigger than critical value to show a significance
How to calculate wilcoxon sign ranks test
Step 1-ignore any participants with same value in both conditions
Step 2-calculate difference, A-B
Step 3-rank the differences ignoring if it’s a positive or negative number
Step 4-add together the ranks of the numbers with the least frequent sign, this is the calculated T value
Step 5-T value must be less than critical for there to be significance
How to calculate Mann Whitney U test
Step 1-rank all the data together but use different colours (even if two separate groups of participants)
Step 2-calculate U value for each group of participants NaNb + (Na(Na+1)➗2) -total Ra and NaNb + (Nb(Nb+1)➗2) - totalRb
Na=total participants in group A
Total Ra=total sum of ranks for group A
Step 3-choose lower U value, this is the calculated value of U
Step 4-U value must be less than or equal to critical value for there to be significant
How to calculate spearman rho
Step 1-rank scores for column 1
Step 2-rank scores for column 2
Step 3-calculate difference of ranks
Step 4-square the difference
Step 5-add all of squares together
Step 6-sub in values into 1-((6x total d^2)➗N(N^2 -1)) this is the Rs value
Step 7-if Rs is greater than critical then there is significance
