# Mathematics - Compare Clinical Studies

In a recent clinical study, Brand ABC was proved to be 1950 percent better than creatine." I have to turn this question into a math problem and I do not know how to go about doing this.

### Comments for Mathematics - Compare Clinical Studies

 Jul 31, 2011 Compare Clinical Studies by: Staff The question: In a recent clinical study, Brand ABC was proved to be 1950 percent better than creatine." I have to turn this question into a math problem and I do not know how to go about doing this. The answer: There are a number of possible reasons why Brand ABC was proved to be 1950 % better than Creatine. For example, the test methodology can be constructed in such a way that Brand ABC will look like a miracle brand compared to Creatine. This could be something as simple as lowering the temperature for the tests conducted with Creatine and raising the temperature for the tests conducted with Brand ABC. Or, the samples used in the “clinical study” can be carefully selected so positive results are over-reported for Brand ABC. You asked specifically how to turn in test results into a math problem. A possible math question could be: How does the margin of error for the clinical study vary if the number of samples used in the tests is 10, 100, or 1,000? All clinical studies have a margin or error which can be calculated. True Result = calculated/test result ± margin of error Margin of Error formula = (Z value)*(Standard Deviation)/( square root of sample size) As you can see, the numerator for the margin of error formula is divided by the square root of the sample size/or number of samples). This means that if the sample size is only 10, the numerator is divided by √10 (which is about 3.2). However when the sample size is increased to 100, the numerator is divided by √100 (which is 10). By increasing the sample size to 100, the margin of error drops to 1/3 of what it was when the sample size was only 10. The results become even more reliable (smaller margin of error) when the sample size is raised to 1000. By increasing the sample size to 1000, the numerator is divided by √1000 (which is 31.6). By increasing the sample size to 1000 the margin of error drops to 1/10 of what it was when a sample of 10 was used. To get a low margin or error, the number of samples must be large. Thanks for writing. Staff www.solving-math-problems.com