Thought Questions (Think about, discuss briefly with partner and then move on)
- The null hypothesis in this study is that there is no treatment effect (genuine disadvantage) three days later from the sleep deprivation. Is it possible that through the random assignment process we could create groups where the skills on this task still tended to favor the unrestricted sleep group?
- Do you think it’s unlikely that the observed difference in group means would be this far from zero due to the random assignment process alone?
- How could we go about deciding whether the difference we are observing between two groups could have happened by the random assignment process alone?
