|
|
Post by Admin on Jan 24, 2016 22:40:26 GMT
 To assess the efficacy of a new bronchodilator drug X, the peak expiratory flow rate of a population of asthmatics is measured during an asthma attack under three conditions: after no intervention, after the administration of albuterol, or after the administration of drug X. The graph shows the mean peak flow and 95% confidence intervals for each condition. Which is the best statement regarding the results of the study? A. There is likely a significant difference between the mean peak expiratory flow rate of albuterol versus drug X B. There is likely a significant difference between the mean peak expiratory flow rate of albuterol versus no intervention C. There is likely a significant difference between the mean peak expiratory flow rate of drug X versus no intervention D. There is not likely a significant difference between the mean peak expiratory flow rate of albuterol versus drug X because the ratio of these two means is close to 1 E. There is not likely a significant difference between the means of any of the conditions because the confidence intervals are too wide |
|
|
|
Post by Admin on Jan 24, 2016 22:49:59 GMT
1. What are Confidence intervals (CIs)?
- They specify a lower and upper limit of variability around a specific probability in which data points are likely to fall.
2. What is the Frequently 95% CI is used?
- This means that in 95% of cases, the true value for the population will fall within the given CI.
- Generally, if the 95% CIs do not overlap, the means of the groups differ significantly.
- If the CI overlap considerably, it is less likely that there is a significant difference between the means of the two groups.
- However, one must be careful with this assumption because if two CI come very close but do not overlap, there may not be significant difference at the p = 0.05 level.
- Conversely, a small overlap of CI does not preclude the possibility of difference between the means, but the chance that this difference is significant at the p = 0.05 level is much
less likely.
- When assessing odds ratio or relative risk, if the 95% CI includes the value 1, the null hypothesis cannot be rejected. This means that the data do not confirm a significant difference between the odds of having a disease based on the odds ratio or the risk of getting a disease based on the relative risk versus the unexposed population. However, in this case, the CI refers to means of the population, not to odds ratio or relative risk, and these values cannot be determined simply by dividing the two means. Thus, it is unlikely that there is a significant difference between the means of albuterol and drug X because the CI of the means overlap.
|
|
