Probability + significance

• What do psychologists want

  psychologists want to be confident the difference is caused by the IV rather than chance/that the IV is the only thing that causes an effect on the DV -> a significant result is one where there is a less than 5% probability that the results (difference or correlation between your variables) are due to chance (P less than equal to 0.05 / 5%)
• How can psychologists be sure the change in DV is caused by IV


  conduct a statistical test to look at the probability of results being due to chance. if they are due to chance = not significant, accept null hypothesis. if not due to chance + IV is having an effect on the DV = results are significant + accept alternate hypothesis
• What is probability


  how likely something is to happen + is expressed as a number between 0 and 1 or a percentage. by assessing probability we can determine the significance of results
• What is significance


  where there is a low probability that the difference or correlation between the variables is due to chance

  significance level = how confident a psychologist wants to be about whether their results are significant e.g want to be more confident = strict level such as P less than equal to 0.01, less confident make it more lenient such as P less than equal to 0.5
• Why may psychologists choose a smaller level of significance e.g. p<0.01 or p<0.001

  if errors could have dangerous implications i.e looking at effectiveness of drugs or in replication research where we want to be certain the results are due to what we are testing + nothing else OR in replications of previous research BUT by using such a strict level of significance + probability we may make a type 2 error
• 2 types of error

  • TYPE 1: we think that there was a significant effect when really results were due to chance, may accept alternate when it should have been rejected + null accepted. produced by lenient probability (e.g P<0.1)
  • TYPE 2: we think that there was no significant effect when there was, may accept null when should have rejected + accepted alternate. produced by strict probability (e.g P<0.01)
• type 1 vs type 2

  accept null = 2Cs = type 2

  reject null = 1 C = type 1