section 3: data analysis

• What kind of test is used for nominal data and related designs (repeated measures or matched pairs)?


  A sign test
• What kind of test is used for nominal data and independent design?


  Chi squared
• What kind of test is used for ordinal data and related designs (repeated measures or matched pairs)?


  Wilcoxon
• What kind of test is used for ordinal data and independent design?

  Mann-Whitney
• What kind of test is used for ordinal data and a correlation?

  Spearman's Rho
• What kind of test is used for interval data and related designs (repeated measures and matched pairs)?


  Related t-test
• What kind of test is used for interval data and independent design?


  Unrelated t-test
• What kind of test is used for interval data and a correlation?

  Pearson's r
• Histograms are used for continuous data.
• A type 1 error is a false positive. This is where there is believed to be a real difference in the data produced where there is not, so the hypothesis is accepted in error.
• A type 2 error is a false negative. This is where a real difference in data is overlooked and the null hypothesis is accepted in error.
• A 5% level of significance is used in psychological research because it strikes a balance between the risk of making a type 1 and type 2 error.
• Nominal data is categorical. Ordinal and interval data are rated on a scale. Ordinal is subjective. Interval is objective.
• A sign test is done by seeing if the results increase, decrease or stay the same for each pair/individual across the categories. If it stays the same disregard this bit of data from the total. Add up all the increases (+) and all the decreases (-). Whichever is smallest is S.
• Concurrent validity is where there is a close agreement in data between a new test and an old test. There must be a coefficient between the two sets of data of +0.8 or more.
• Scattergraphs are used to display correlations. They indicate the relationship between co-variables.
• In a meta-analysis, combined data is often statistically tested to provide an overall conclusion.
• A symmetrical/normal distribution of data occurs when the mean is equal to the mode/median.
• A positive skew in distribution occurs when the mean is greater than the mode/median.
• A negative skew in distribution occurs when the mean is less than the mode/median.