Download Introduction to Chance Models (Section 1.1) Introduction A key step

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Document related concepts

Statistics wikipedia, lookup

History of statistics wikipedia, lookup

Transcript
10. Use the applet to determine what proportion of your 1000 simulated random assignments
were the results as (or more) extreme as the actual study (which, you’ll recall, saw a 0.467
difference in improvers). This is your p-value. Interpret this p-value, including what is meant
by the phrase “by chance alone” in this context.
11. Is your p-value small enough so that you would consider an difference in the observed
proportion of improvers of 0.467 or more surprising under the null model that dolphin
therapy is not effective? Does this give strong evidence that the dolphin therapy is effective
in treating depression in patients similar to those in the study?
12. Is it reasonable to draw a cause-and-effect conclusion between dolphin therapy and
improvement or not in depression symptoms? Explain why or why not.
13. Is it reasonable to generalize your conclusion to all people who suffer from depression?
Explain why or why not.
Further Exploration
What if the study had found the same overall proportion of improvers, but with the group
proportions closer together? More specifically, suppose that the actual data had turned out with
2 fewer improvers in the dolphin therapy group and 2 more improvers in the control group.
14. How will this affect our analysis, and the p-value, and the conclusion? First make a
conjecture, and explain your reasoning.
15. Then investigate this question with a simulation. [Hints: Think about how the two-way table
will change. Then use your earlier simulation results but explain what you are doing
differently now to find the approximate p-value.] Explain why your answers (about how the
strength of evidence against the null model changes when the group difference is not as
large) make intuitive sense.
June 27, 2014
MAA PREP workshop
68