In: Statistics and Probability
Prior research has shown a significant relationship between maternal smoking during pregnancy. Investigators at University Anywhere were interested in looking at whether or not there was a dose response between smoking and low birth weight. They collected data from 500 women receiving prenatal care at a local health center and categorized their smoking status at their first prenatal visit into the following categories:
(1) never smokers, (2) former smoker (quit smoking) (3) non-daily smoker (currently smokes but not every day) and (4) daily smoker. The table below provides a descriptive summary of data from the study:
Birth weight | ||
never smokers (n = 250) | 3000 | 647 |
Former Smoker (n=100) | 2800 | 593 |
Non-daily smoker (n=75) | 2750 | 495 |
Daily Smoker (n=75) | 2300 | 500 |
A. Describe what type of graph you would use to examine the relationship between maternal smoking and birth weight. Justify the graph you selected.
B. What statistical test can be performed to see if the mean birth weight varied across the smoking categories? Be sure to include a brief explanation/justification for the test that you selected.
C. Specify the null and alternative hypotheses for the test that you specify in part b.
D. Perform the test you suggested in part b, report a p-value
E. Interpret this result in a sentence (remember that your interpretation should be one that makes sense to the scientific community who will use these results, not just biostatisticians).
F. What else would you like to know in order to address this hypothesis? Describe how you would examine this.
Using SPSS:
First Enter the data:
A): Means plot graph use to examine the relationship between maternal smoking and birth weight.
plot interpretation: mean birth weight of never smoker is high. mean birth weight of daily smoker is low.
B): One Way ANOVA statistical test can be performed to see if the mean birth weight varied across the smoking categories. One way anova use when we have to check significant difference between more than two category.
So there are 4 categories in our problem, therefore we are using one way anova test.
C):
Null Hypothesis H0 : There is no significant difference between group.
Alternative Hypothesis H1 : There is a significant difference between any group.
SPSS Steps:
Analyze-->Compare Means-->One Way Anova.
D): Here p-value is 0.993 that is greater than 0.05, we do not reject the null hypothesis. It is mean there is no significant difference between group
E): Birth weight does not depend on non-smoker, former-moker, non-daily smoker and daily-smoker.