Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

BEYOND THE NUMBERS 3.18 | LEARNING OUTCOMES 4 TO7 Statistical Significance in th

ID: 3366424 • Letter: B

Question

BEYOND THE NUMBERS 3.18 | LEARNING OUTCOMES 4 TO7 Statistical Significance in the Media-Part II Section Number: Name: To be graded, all assignments must be completed and submitted on the original book poge. EXHIBIT 1 Prescription to Pass Drug Safety Title: Trial Intensifies Concerns about Safety of Vytorin Author: Alex Berenson Source: New York Times, July 22, 2008, http://www.nytimes.com/2008/07/22/ business/22drug.html The following is an extract from a New York Times article on Vytorin: In a clinical trial, the cholesterol-lowering drug Vytorin did not help people with heart-valve disease avoid further heart problems but did appear to increase their risk of cancer, scientists reported Monday. Vytorin and Zetia, a companion drug, are prescribed each month to almost three million people worldwide and are among the world's top-selling medicines. In the Seas trial, which involved nearly 1,900 patients whose heart valves were partially blocked, participants were given either Vytorin or a placebo pill that contained no medicine. Scientists hoped that the trial would show that patients taking Vytorin would have a lower risk of needing valve replacement surgery or having heart failure. But the drug did not show those benefits. "No significant difference was observed between the treatment groups for the combined primary endpoint," Dr. Terje Pedersen, the principal investigator for the study and a professor medicine at Uleval University Hospital in Norway, said. The primary endpoint is the result that scientists hope to prove when they conduct a clinical trial. However, patients taking Vytorin in the Seas trial did have a sharply higher risk of developing and dying from cancer. In the trial 102 patients taking Vytorin developed cancer, compared with 67 taking the placebo. Of those, 39 people taking Vytorin died from their cancer, compared with 23 taking placebo. The absolute numbers of cancer cases were relatively small, But they reached statistical significance, meaning the odds were less than 5% that they were the result of chance.

Explanation / Answer

Answers:

Question 1

In the given article regarding the clinical trial, the phrase ‘significant difference’ is used for the difference between the treatment groups for the combined primary endpoint. The phrase ‘statistical significance’ is used for denoting the significance of the result or claim in a statistical manner. The phrase ‘significant difference’ indicates the difference between two treatments or variables if it is reasonable. This significant difference does not consider small difference up to specific level. After this specific level of difference, it is considered as significant. The phrase ‘statistical significant’ is used when we do not consider small differences which are not significant. The word ‘small’ is comparative and it would be different for different scenarios. In the given research study, results are statistically significant at 5% level of significance, this means the probability of getting odds is less than 5%. That is, there is a 5% chance of getting difference between the treatment groups for the combined primary endpoint.

Question 2

Yes, the phrase ‘statistically significant’ is being used in a meaningful way, because significant difference is considered while using this statement. In an actual study, the percentage of the albatross pairs as same sex may be vary from ½ or 0.5, but it would be close to ½, but not exact ½ . So reader’s comment about “1/2 of the albatross pairs as same sex” being statistically significant. If there is considerable difference between the percentages of same sex, that is, it varies more from ½, and then we can say that this difference is statistically significant. The statistical significant difference is based on the significance level. The extent of difference is being measured by using statistical level of significance. Differences would be different for different levels of significance. Readers claim will be valid for some level of significance, but it would be invalid at other level of significance.