Inferential statistics

• Nominal level of data is the least info and category data
• Ordinal level of data is data that can be ordered in some way without knowing exact differences (likert scales)
• Interval level of data is the most information and it uses exact measurements with differences like psychological measure of stress
• Alternate hypothesis are written predicting that the result will be significant
• Null hypothesis predicts the results to not be significant (due to chance)
• The main inferential statistical tests are

  • Chi - squared
  • Wilcoxon signed ranks
  • Mann - whitney U
  • Spearman's rho correlation coefficient
• They choose a significance level when they begin their research to express the level of chance they are prepared to accept to be sufficiently satisfied
• The traditional level of significance is 5% p<0.05 which means that the results occurring due to chance is less than or equal to 5%
• IF the probability is higher than that then the null hypothesis will be accepted
• Every inferential statistical test has its own procedure to calculate the observed value (or calculated value) with a table of critical values
• Observed value is when statistical tests turn all of the known results into one value
• The critical value is precalculated
• If the observed value is bigger than the critical value is bigger the results are not significant
• p<0.05/the5% level is the results having occurred by chance is less than 5% its the most traditional and most frequently used and it offers balance between the likelihood of making type 1 and 2 errors
• p<0.1/10% is the results occurring due to chance is less than 10% this used when a researcher does not need a strict level of significance like new research 90% significant
• p<0.01 The probability of the results occurring by chance is less than 1% its very strict as any mistakes in the research could have serious consequences 99% significance rate its normally used will human health.
• Type 1 error is reject the null hypothesis when the results are actually due to chance (false positive)
• Type 2 is fail to reject the null hypothesis when the results were not due to chance (false negative)
• Type 2 error is more common in 1% level
• Type 1 error is more typical in the 10% level
• What are you testing for like difference or correlation e.g experiment or correlational study
• What level of data do you have like nominal or ordinal and above
• Is it related or unrelated sampling e.g repeated measures and matched groups are related and independent groups are unrelated
• Chi- squared

  • Difference
  • Nominal
  • Unrelated
• Spearman's RHO

  • Correlation
  • Ordinal
  • Related
• Wilcoxon Signed ranks

  • Difference
  • Ordinal
  • Related
• Mann-whitney u

  • Difference
  • Ordinal
  • Unrelated