factors affecting choice of statistical test

• -          If you had interval data and we're carrying out a test of association or correlation you would use person's r test
• -          if you had interval data and we're carrying out the test of difference using a related design (repeated measures / matched pairs) you would use a related t test
• -          if you had interval data and you were carrying out a test of difference using an unrelated design (independent groups) you would use a unrelated t- test
• -          if you had ordinal data and we're doing a test of association or correlation you would use a spearman’s rho test
• -          If you had ordinal data and were testing for difference using a related design (repeated measures slash matched pairs) you would sue a Wilcoxon text
• -          If you have ordinal data and are testing for difference an are using a unrelated design “ independent groups) you would use a Mann Whitney test
• -          If you have nominal data + and are testing for associations or correlations you should use a chi square 
• -          If you have nominal data + used a related design (matched pair / repeated measure design) and are testing for difference you should sue the sign test
• -          Statistical tests are complex mathematical formulas that have been designed for a specific purpose therefore the features of s study will decide which statistical test is suited to use for its analysis
• -          When making a decision on what statistical test to use you should consider if you are looking for a difference or a relationship within the results and what the experimental design of a study is – independent / repeated / matched and finally what type of data have you recorder – nominal / ordinal / interval
• -          If you have nominal data and have used a unrelated design (independent group design) and are testing for a difference you should use a chi square