Question

Smoking during pregnancy can cause a baby to be born too early or to have low birth weight. Design a study that would test this statement.? Describe the problem? What is the sample size? Variable should be define? Parametric or nonparametric? Make the conclusion?

Answer 1

Various studies suggest that smoking or alcohol leads to retarded fetus growth. However, We need to focus on the effects of smoking (by Mothers during pregnancy) on a baby being born too early or to have a low birth rate.

To derive any conclusions, a data set is required. The data set should involve factor values of mothers who smoke during pregnancy as a "yes" or "no" and some observations for the baby after delivery, which are "the weight at the time of birth" and "total months of pregnancy in which they were born".

This study will conclude an important result which is of interest to so many people around the globe. So, the sample size should be significantly larger to avoid any changes in research. The sample should be greater than equal to 10,000 data points, although the larger it gets the better it is for research.

The correlation between variables can be studied in two ways;

1. Response Variable: The weight at the time of birth

Explanatory Variable: If the mother smokes during her pregnancy

2. Response Variable: Total months of pregnancy in which the baby was born

Explanatory Variable: If the mother smokes during her pregnancy

This way, we can know separately if smoking affects the two claims hypothesized.

Given that testing is done from the non-parametric sample since we do not know previously if it follows any pre-defined statistical distribution or not. Clearly, one can directly compare the effect by using a bar-graph.

If the weight of babies (given that mother smoked) is significantly less than the weights of babies whose mothers did not smoke then it will be clear that smoking does affect the weight of a newborn baby. Similarly, it can be concluded for the other variable.