# Statistics for argumentative data and politics

There is a drug trial underway and that it has shown, based on 200 participants, 100 randomly assigned to treatment and 100 to placebo, that the mean health outcomes of the 100 treated and the mean health outcomes of the 100 placebo patients do not overlap - in terms of a 90 percent confidence level. We see that drug results in healthier outcomes.

A Formulate the Null H of the study

B Would you be comfortable rejecting the null? List considerations based on the 90 per cent Confidence Level of your findings. Would a 99 per cent CI make more sense? Why or why not? Talk about the likelihood of wrongly rejecting a correct Null - how does that change in 90 to 99 standard and what it means for you.

C The company wants to increase the sample of the present drug trial, and hold back approval till results on a new, bigger sample (N=400) are in. What are potential benefits of this approach? Are there potential downsides? Talk about likelihood of rejecting Null and use some math.

D The company believes that they have an even more powerful drug under development, one that makes you even healthier on average relative to placebo. Talk about what that entails for the likelihood of rejecting the Null. Use some math, possibly pictures of CI.

E Would you rather take a risk of rejecting an H0 which is true or take the risk of maintaining an H0 which is false in case of rejecting a Null being a precondition for a drug to hit the market? Explain what one means and what the other means. Talk about upsides and downsides.

- under review
- 51 views
- $30.00

### Related Questions

- 2 conceptual college statistics questions - no equations
- Markov Chain
- Prove that $\lim_{n\rightarrow \infty} \int_{[0,1]^n}\frac{|x|}{\sqrt{n}}=\frac{1}{\sqrt{3}}$
- Calculating P values from data.
- Statistics/percentage question
- Questions for Statistics Project
- How would I perform an exploratory data analysis to visually assess the relationship between two sets of data?
- Sample size calculation for a cross sectional healthcare study