• Two-sample t tests always ask whether the two means studied come from the same or a different population.
  
  
• The specific formula you use for a two-sample t test depends on whether there are equal numbers of subjects in each group, whether standard deviations are equal, and whether data are independent or dependent.
  
  
• The logic of two-sample t tests requires that we calculate the difference between each sample mean and test whether this difference is bigger than zero, given that they might be a little bit different on the basis of luck or chance.
  
  
• A dependent-means t test is required when the two samples are linked in some
  way=naturally in the population (e.g., mother-infant or brother-sister), matched or paired on background characteristics, or repeatedly studied over time (e.g., pretreatment or posttreatment).
  
  
• Dependent-means t tests have more power because they reduce the chance of randomly selecting very different groups of people.

Test your knowledge of this workshop by taking the Workshop Quiz.