Diet and Exercise Habits
Generated Jun 14, 2013 by smaskell1
Analysis of Diet and Exercise Habits
Introduction
My group and I designed a survey regarding diet and exercise habits of our adult friends, acquaintances and coworkers. We surveyed adults aged 20 – 60 years old from Facebook, email contacts and at coworkers in the workplace. We have a convenience sample with a voluntary response of participants.
The questions asked were:
1. How many days a week to you exercise?
2. How many soft drinks do you drink per week?
3. In the past six months, have you lost weight? Yes/No
4. Of these healthy choices, which one do you have the most difficulty including in your diet. Vegetables/Fruits/Whole Grains/Lean Meats (Protein)/Dairy
Looking at a Categorical Variable
The following are responses to the question, “What healthy foods do you have the most difficulty including in your diet?”
<result 1>
This pie chart shows that dairy, followed closely by vegetables are the foods our participant have the most difficulty including in their diet, with whole grains being the lowest at 20% of participants.
In order to compare how healthy food choices are to whether the participant has lost weight in the last six months is reflected in the bar graph below:
<result 2>
Among those who have not lost weight, dairy seems to be there most difficult food choice. And, surprisingly, those who have lost weight seem to have the most difficulty including vegetables.
Looking at a Numerical Variable
These responses are to the question “How many days a weeks do you exercise?”. They are displayed in a histogram, boxplot and summary statistics:
<result 3>
<result 4>
<result 5>
The histogram shows a mostly rightskewed data with most of the data at less than 4 days a week. The most frequent number of days exercised by participants is 0 days, but followed closely by 3 days.
The center of the skewed data seems most reflected by the mean of close to 2.5 hours, whereas, the median, which is 2 days, is affected by the extreme value of zero. The midrange (0 + 7)/2 is 3.5 days , which appears slightly high for this data.
The IQR, which gives the range in the middle of 3 days a week, which appears close to center. Range rule of thumb: range/4 = 1.75 is accurate since the standard deviation is 1.99. Since there is only a range of 7 days, and the standard deviation is almost 2, most of the data fall within this range.
The boxplot reveals no outliers, mostly due to the fact there were only up 7 days of the week to choose from and no one chose from outside of the parameters. It is plausible for someone to exercise anywhere from 0 to 7 days a week.
As we can see from looking at the histogram that 0 and 3 seem to be the most popular numbers of days to exercise and after 3 days, it steadily decreases. Seven days a week is unlikely for most, but is possible, especially what one person may consider exercising? Someone could take a daily walk, and listed that as their seven days, since it was not specified as to what the survey considered exercise.
Looking for a Relationship Between Two Numerical Variables:
To determine if there is a relationship between: Number of Sodas Consumed in a Week vs. Number of Days a Week Participant Exercises. Here is the scatterplot comparing the data:
<result 6>
This scatterplot shows a very slight positive correlation between numbers of sodas consumed vs. days exercising week. A couple of outliers appear on this scatterplot (30 sodas, 7 days/exercising) and (24 sodas/0 days exercising) which will affect the correlation coefficient.
The correlation coefficient of 0.152 is displayed below:
<result 7>
Since the absolute value of r is (Table A5 in the textbook), we can conclude that there is a statistically significant correlation, but it does appear to be a very weak one as displayed by the scatterplot.
<data 1>




Summary statistics:


Correlation between days/exercise and softdrinks/wk is:
0.15173393 