Академический Документы
Профессиональный Документы
Культура Документы
CCHS 315
Due 11/26/13
smoking status. Women who were white and older were more likely to smoke, as well as
those in higher levels of poverty. Minority women were more likely to take up smoking
as they got older, whereas white women started smoking at a younger age but their use of
tobacco diminished as they got older.
The article closed by reviewing issues that are related to tobacco use such as lost
wages, health problems, and outcomes in pregnancy. They state that only 18 to 25% of
maternal smokers [are] successful in stopping smoking. The article ends with a call to
various disciplines to address this important issue.
3. What did this statistic show and/or prove in your article AND what are some
examples of this statistic in your article? This raw data was the material which the
researchers used to analyze and do calculations with. And example would be the
7,802 birth certificates available or the 500 smokers and 500 non-smokers records
selected for the study. The numbers themselves prove the outcomes of the study but
no one can see that until it is converted into rates, ratios, means, and so on.
4. Was there an associated P value? No
Statistic #5 Logistic regression
1. How many times was the statistic used in the article? Once
2. Why do you think this statistic was used in the article? Conditional multinomial
logistic regression was done using software to compare each of the case control pairs
(matched) which were used in the study. The dependent variable used for these
calculations was smoking status and the independent variables used have already
been discussed. Two different models were used.
3. What did this statistic show and/or prove in your article AND what are some
examples of this statistic in your article? Using logistic regression provided the
researchers with further statistics to analyze and interpret such as odds ratio, mean,
average, and percentage. These statistics were then analyzed and the results showed
that the majority of the participants were African American, not married, receiving
medical assistance, had less than a high school education, and were living in areas of
poverty. They also showed that white women were more likely to smoke, especially
as they got older and the poverty level increased.
4. Was there an associated P value? Yes
5. If yes, what was the P value number; state if this value would be considered highly
significant, significant, or not significant; and what did this significance or lack
of significance actually mean in reference to your specific article? p values were
calculated for each variable and ranged from <0.0005 to 0.606. Many of these were
not significant and indicated that arrest rate, education rate, and medical assistance
status were not closely correlated with maternal smoking. The p values that were
<0.0005 showed that the results of race, age, and poverty being significant
predictors of smoking status have a very low probability of being the result of
chance.
Statistic #6 - Mean
1. How many times was the statistic used in the article? 2
2. Why do you think this statistic was used in the article? Mean was used to discuss the
ages of participants in the study, which ranged from 13 to 45. They chose to include
teen moms in the study because teen pregnancy is so prevalent in Baltimore. Mean
was also used to discuss the arrest rate in each of the census blocks.
3. What did this statistic show and/or prove in your article AND what are some
examples of this statistic in your article? The mean age was found to be 24.8 years and
the mean arrest rate was 7.8%. These numbers gave the researchers further insight
into the participants being studied and their community.
4. Was there an associated P value? No
Statistic #7 Standard deviation
1. How many times was the statistic used in the article? 3
2. Why do you think this statistic was used in the article? Standard deviation was used
in connection to the mean age of participants, the average percent of women living
below the poverty level, and the mean arrest rate for the communities in question.
Since all of these values are a reflection of a spread of data points it is helpful to
know the standard deviation in order to determine if data points are grouped closely
to the mean/average or if they are quite spread out.
3. What did this statistic show and/or prove in your article AND what are some
examples of this statistic in your article? The standard deviation for age was 6.6 years,
showing that the ages were clustered relatively near to the mean between 18.2 and
31.4 years. The rates for arrest had a standard deviation of 21.8%, reflecting a
wider spread from the mean.
4. Was there an associated P value? No
Statistic #8 Confidence interval
1. How many times was the statistic used in the article? 14
2. Why do you think this statistic was used in the article? Confidence intervals were
calculated for all of the variables studied in this research. It was necessary to do so
in order to evaluate the reliability of the findings.
3. What did this statistic show and/or prove in your article AND what are some
examples of this statistic in your article? The confidence intervals for race, age,
poverty level, interaction of race by age, and interaction of race by poverty were all
95%. This shows those statistics indicating a correlation between an increase in
smoking and an increase in poverty level, etc. are fairly reliable.
4. Was there an associated P value? Yes
5. If yes, what was the P value number; state if this value would be considered highly
significant, significant, or not significant; and what did this significance or lack
of significance actually mean in reference to your specific article? p values ranged
from <0.0005 to 0.606. Many of these were not significant and indicated that arrest
rate, education rate, and medical assistance status were not closely correlated with
maternal smoking. The p values that were <0.0005 showed that the results of race,
age, and poverty being significant predictors of smoking status have a very low
probability of being the result of chance.