In a study of the impact of smoking on birth weight, researchers analyze birth weights (in grams)...

Question:

In a study of the impact of smoking on birth weight, researchers analyze birth weights (in grams) for babies born to 189 women who gave birth in 1989 at a hospital in Massachusetts.

In the group, 74 of the women were categorized as "smokers" and 115 as "non-smokers."

The difference in the two sample mean birth weights (non-smokers minus smokers) is 281.7 grams and the 95% confidence interval is (76.5, 486.9)

Which gives the best interpretation of what we can conclude about the impact of smoking on birth weight?

A) We are 95% confident that on average, smoking causes lower birth weights of between 76.5 grams to 486.9 grams.

B) There is a 95% chance that if a woman smokes during pregnancy her baby will weigh between 76.5 grams to 486.9 grams less than if she did not smoke.

C) Smoking is associated with lower birth weights. When smokers are compared to non-smokers, we are 95% confident that the mean weight of babies of non-smokers is between 76.5 grams to 486.9 grams more than the mean weight of babies of smokers.

D) This study does not suggest that there is a difference in mean birth weights when we compare smokers to non-smokers.

Confidence Intervals

The confidence interval for a difference between two means gives a range in which the true population might reside. If this interval omits 0 as a possibility, than a statistically significant difference exists. While a statistically significant difference signifies an association, it does not necessarily mean one thing is causing the other.

Answer and Explanation:

C) Smoking is associated with lower birth weights. When smokers are compared to non-smokers, we are 95% confident that the mean weight of babies of non-smokers is between 76.5 grams to 486.9 grams more than the mean weight of babies of smokers.

The 95% confidence interval does not show causation, but just an association, and gives the possible range in which the true population mean might be for a certain level of confidence.


Learn more about this topic:

Loading...
Using the t Distribution to Find Confidence Intervals

from Statistics 101: Principles of Statistics

Chapter 9 / Lesson 6
6.2K

Related to this Question

Explore our homework questions and answers library