Project Essays Example

This analysis is considering the incidents of births with low infant weight through the use of health records from the North Carolina state center for health and environmental statistics. I will be interested in obtaining summary statistics for the variables contained in the dataset. Of particular interest is the birth weight of infants born by smoking mothers in comparison to birth weight of infants born to non-smoking mothers.
There are several variables under consideration and it would be of interest to explore these variables and there distributions first (univariate analysis) before we can do a comparison of the relationship between these variables (multivariate analysis). It will be important to compute the correlation between infant weight and the variable “smoke” which indicates whether the mother is a smoker or not.
In computing the summary statistics, there are variables which if their statistics are computed will have no meaning. These variables are classified as nominal variables and the values assigned to them do not measure a quantity, for instance for sex a male infant could be assigned the value 1 and a female infant the value 0. It is also possible to interchange the values as long as it is defined before analysis is done. Thus for these variables we shall consider histograms and other forms of representation that we deem appropriate.
An example of nominal variables is the variable “plural” which measures the number of children in a single birth. It would be appropriate to observe these variables through Histograms. Histograms will give a clear indication of the percentage of number of children in a single birth. Other variables which are similar to this variable include: Sex, Marital, Racemom, Hispmom, Smoke, Mature, and Premie. There are a number of factors also which might contribute to a difference in the infant weights and of interest are variables like Racemom, Mage, Hispmom and Premie which from our knowledge might have an effect on the infant’s weight at birth. For instance, Racemom will give an indication of the race of the mother. It is of interest to know that there are diverse cultures and eating habits which in most cases and determined by race and as such this might have an impact on the weight of the infant.
Histograms of the variables Fage, Mage, Weeks, Gained and Tpounds were created with normality curves. It is established that the mean of the variable tpounds is 7.1. This could be interpreted to mean that the mean weight of infants born is 7.1 pounds. The median of the variable tpound is 7.3125pounds which indicates that the distribution, although it appears to be normally distributed, is skewed. If an in-depth analysis of the variable was being conducted, it would have been appropriate to conduct a parametric test to establish if the distribution of the variable is indeed similar to that of a normally distributed random variable.
The mean of the variable Fage is 30.256, and its median is 30, the variable Mage has a mean of 27, and a median of 27. The mean and the median of the variable Mage are equal, this indicates that there is a high probability that the variable is normally distributed. The mean of the variable week is 38.335, and the median is 39. The mean of the variable “gained” is 30.326, and the median is 30. The mean is the best measure of central tendency as it considers all the values of a variable. Its disadvantage, however, is that it is affected by outliers. Since these variables are not most likely to have extreme outliers, the most appropriate measure is the mean.
The percentage of smokers is slightly over 100 which represent an approximate 11% of our sample. We are interested in establishing the correlation between the variable smoke and tpounds and from there we can make a conclusion as to whether there is a statistical correlation between smoking and the weight of the infant. The correlation between smoking and weight infant is -0.07, the negative indicates that they move in opposite directions that is an increase in one variable leads to a decrease in the other variable. The p-value is 0.028 which is less than 0.05, and this shows that the test is significant, and we reject the hypothesis that there is no correlation between smoking and infant weight.
The boxplots on the weight of infants for both smoking and non-smoking mothers does not show a difference. Most of the statistics calculated are computed at 95% confidence interval, as such the weight of the infant could fall anywhere in between these intervals. Thus since these intervals overlap for both smoking and non-smoking mothers then there is the possibility that the mean weight of infants for both smoking and non-smoking mothers are equal.
The average weight of infants born is 7.0997 pounds, the weight of the heaviest baby is 11.75pounds. Although for our friend the baby is heavy with a weight of 10.3 pounds, the weight is less than the maximum value (11.75 pounds) of the variable and as such there is no need for her to worry.
The mean weight gained by the mothers was 30.326 pounds, and the median was 30 pounds. The weight gained by our friend is comparatively equal to this and as such there is no need to be depressed.
Variables of interest that are not included in the analysis include the occupation of the mother, the amount of income, economic stability, eating habits and the environment which could include the presence of conflicts in the family. These variables play a role in both the health of the mother and weight of the infant.
Variables that are of interest regarding the infants weight will include the weight of the mothers, the nutritional status of the mother and the financial status. It is expected that there is a correlation between the mother's financial stability and her nutritional status. The residence of the mother also may be considered a variable of interest since it is expected that certain foods with specified nutritional quantities differ with regions. Questions of interest that the mother should be asked before delivery is if the baby was due or it’s a premature labor, it should also be established if there are previous births and the nature of these previous births whether normal or through caesarean. These questions are important in anticipating possible complications during delivery and preparing for them.

Appendix:

Descriptive Statistics: fage, mage, weeks, visits, gained, tpounds
Variable N N* Mean SE Mean StDev Minimum Q1 Median Q3
fage 829 171 30.256 0.235 6.764 14.000 25.000 30.000 35.000
mage 1000 0 27.000 0.196 6.214 13.000 22.000 27.000 32.000
weeks 998 2 38.335 0.0928 2.932 20.000 37.000 39.000 40.000
visits 991 9 12.105 0.126 3.955 0.000 10.000 12.000 15.000
gained 973 27 30.326 0.457 14.241 0.000 20.000 30.000 38.000
tpounds 1000 0 7.0997 0.0477 1.5089 1.0000 6.3750 7.3125 8.0625
Variable Maximum
fage 55.000
mage 50.000
Descriptive Statistics: tpounds
Variable smoke Mean SE Mean StDev Variance Minimum Q1 Median
tpounds 0 7.1430 0.0514 1.5188 2.3066 1.0000 6.4375 7.3125
1 6.827 0.123 1.386 1.921 1.688 6.047 7.063
* 3.6250 * * * 3.6250 * 3.6250
Variable smoke Q3 Maximum
tpounds 0 8.0625 11.7500
1 7.766 9.188
* * 3.6250