T-tests

• deductive reasoning

  logical approach where you progress from general ideas to specific conclusions
• t-tests are used to compare groups
• regression and correlation are for analysing a relationship
• type 1 error

  if null is rejected that is actually true within a population
• type 2 error


  if null is accepted when it should be rejected
• 1 sample t-test


  significant difference between the mean of 1 sample and a theoretical value
• 2 sample t-test


  significant difference between the means of 2 independent samples
• paired t-test


  significant difference between the means of the two dependent paired samples
• paired t-tests can be before and after or comparing two methods
• 2 sample t-test can be comparing rainfall in the north and south