what is the null hypothesis and can you give me some simple examples?
In: 8
You may have question about the world around you. Does drinking beer cause you to have a beer belly?
You can design a study to determine Truth in the Universe. Randomly select 100 volunteers for your study; 50 will be in group A and will eat a specific diet and have a specific exercise regimen. You will choose some measure of “beer belly,” and measure it before and after your study period, say 6 weeks. A common measure for this is hip to waist ratio.
50 will be in group B, and will have the same diet and exercise regimen as group A, except some of the calories in group A’s diet will be removed and replaced with, say, 3 cans of Miller Genuine Draft each day, and this will go on for 6 weeks.
You can then compare how much the average hip to waist (HtW) ratio changes in group A, and compare it to the changes in group B.
***The null hypothesis states that there will be no difference in the change of HtW between the two groups***, and you will choose an appropriate statistical test to analyze this hypothesis. If there IS a difference in the change of HtW ratio in the two groups, your statistical test will tell you if the difference is statistically significant. If it is, then you get to decide if the difference is clinically significant.
Spoiler alert; the studies that have been done to examine this question have not been randomized double-blind controlled trials, and the answer is generally that beer consumption is not particularly associated with the presence or absence of a beer belly once you control for exercise and diet…
Latest Answers