What does a Data Story look like | My Assignment Tutor

What does a Data Story look like? There are many examples in the DI-igo site,  https://groups.diigo.com/group/uts-aei/content/tag/data-story and we can find many others in the media. A Data Story is a story where the main ideas come from the data and from asking questions about the data. To write a Data Story we need to begin with data, analyse it, look for patterns in charts and tables, and write about what we observe, being clear about the evidence we give for any arguments or statements. Then we compare our findings with published work to see how our conclusions compare with what others have written about on that topic. Here is the outline of how a data story could be written from data that we collected in class from students in a recent semester.  The data is in an Excel spreadsheet that you should look at first, to see how I analysed it and made lots of charts and tables. Note that I have not included in the written report all the tables and graphs that I generated from the data. I include the tables and graphs that best illustrate the main points of the story. For your AT2 we suggest that you do the same – have a spreadsheet where you start with raw data and do your analysis, then choose interesting parts from it for including in the written report. I have used subtitles to help the reader find their way around the story. “How normal is our class?” Introduction The class for Arguments, Evidence and Intuition consists of about two hundred students from many different faculties and degree courses in the University of Technology Sydney. As a data collection exercise, we measured and counted and recorded quantities and qualities about ourselves early in the semester. From these measurements, 113 were considered complete records after eliminating errors. This is the data story about these data. The question I am asking is, “how normal is our class?” What are our fitness levels? The class members were asked to rate their own level of fitness from five options, and the summary is illustrated here with a table and a pie chart: Table 1   Self-reported fitness levels for AEI Spring 2018 Couch potato – not at all fit76.19%Not fit2320.35%Average fitness6456.64%Very fit1513.27%Elite athlete43.54%Total113 Figure 1   Pie chart showing levels of fitness of students in AEI, Spring 2018 The shape of the distribution of these data, especially in Figure 2, shows that the largest group are those considering themselves to be of “average” fitness, with smaller numbers above and also below average. Figure 2   Distribution of fitness levels as reported by class members, AEI Spring 2018. What are the comparable results for the general population? The Australian Bureau of Statistics conducts many surveys of the Australian population and publish these on their website. This table shows comparable results from their 2011 research, (ABS, 2011) and indicates levels of exercise taken: Table 2: Extract from Table 11_1 of the Australian Health Survey: First Results, 2011–12 Persons 18-24Sedentary643600Low657600Moderate472700High415500Total(b)2,189,400 To make a comparison I have considered the “Sedentary” category to be equivalent to the “Couch Potato” category of the class data and the “Low” category to be equivalent to the “Not Fit” category of the class data. Then I considered “High” in this table to be equivalent to our “Very Fit” and “Elite Athlete” categories combined. Figure 3 then shows the comparison. It appears that our class has a higher proportion in the average fitness categories. However the picture could be different if we put the “Not Fit” into the “Sedentary” section. This illustrates the limitations of comparing data from different data sets. Figure 3   Comparisons of fitness levels. At the high end, the ABS data indicates that 19% of the age group 18-24 years do a high level of exercise, and for our class this compares well with the 17% who report that they are “Very Fit” or “Elite Athletes”. Does fitness influence other measures? A high level of fitness usually is accompanied by a lower pulse rate. According to Dennis Kravetz (2014), “The average person has a resting pulse rate of between 70 and 75 beats per minute. Fit people who get lots of aerobic exercise having resting pulse rates in the 50s and 60s. Some professional athletes have resting pulse rates as low as the upper 30s. On the other side, unfit people have resting pulse rates of 80, 90 or more beats per minute.” I compared the pulse rates of students in the top two fitness categories with those of students in the lowest two fitness categories and the results were in this direction but not so marked:   Figure 4   Box plots showing differences in pulse rates (heart rates) for very fit versus not fit sub groups of the class. For both groups, the pulse rate increased after a small amount of exercise, in this case ten “stand and sits”. How do the fit people stay fit? Do our fitter students visit the gym more often? Yes, as shown in Table 3 there is a big difference in the way students use the gym. The Elite Athletes and the Very Fit mostly attend a gym at least once per week with more than half of them attending a gym at least three times a week. By comparison, those in the least fit categories hardly ever attend a gym. (Barely 1 in 6 attend a gym.) Table 3, Gym visits per week. Gym visits per week by students who consider themselves Elite athletes and Very Fit     none 3 16% one 4 21% two 2 11% three 3 16% more than three 7 37% total 19 100%Gym visits per week by students who consider themselves “not fit” or “couch potatoes – not fit at all”.   none 25 83% one 4 13% two 0 0% three 1 3% more than three 0 0% total 30 100% Is there a relationship between height and fitness? Do fitter people tend to be taller? There does not seem to be a link for the data in this class. The average of the Not fit group (couch potatoes and not fit) is 169 cm and the average height of the very fit and elite athletes taken together is 172cm. The box plots overlap (see the spreadsheet). Is the class “normal” in terms of travel modes? The most popular mode of transport to uni for this class is the train 56% followed by bus (19%). According to the City of Sydney Community Profile (2017) these figures are 22% and 12% for Sydney commuters who travel to work. The difference is most likely due to the low numbers of uni students who use cars to commute, which is 22% in the survey for workers. Is there a relationship between height and armspan? According to the Sabine de Brabandere (2017) a person’s arm span is roughly equal to their height. To explore the relationship between these two continuous quantitative variables, a scatter plot was prepared after deleting one datapoint that did not seem to belong. The scatter plot does show a linear relationship between these two variables, so it makes sense to calculate a correlation coefficient, which is 0.83, indicating a strong correlation between those variables. It does seem however that many armspans are lower than heights, so this was investigated with side by side boxplots. There seem to be some armspans that are shorter than the heights. (Figure 6) Figure 5   Scatter plot showing relationship of armspan and height. Figure 6, comparisons of the distributions of armspans and heights. Looking at the quantitative variable of height The summary statistics for the quantitative variable “Height” is listed here, along with a histogram. The boxplot is in Figure 6 above. There are no outliers, meaning that there are no students who are much shorter nor much taller than the majority of the class. Half the class are above 171 cm and half are below. Minimum149Lower quartile165Median171Upper quartile178Maximum194  IQR Interquartile range13    Mean171.3717Median171Mode178  Standard Deviation9.447755 Figure 7. Histogram of heights of people in the class. Pet ownership Finally, we look at the number of pets in the households where our students live. The majority of these households have no pets. (Figure 7), and about three in five households have at least one pet. According to the RSPCA, “Australia has one of the highest rates of pet ownership in the world. About 62% of Australian households own pets.” On these measures, our class does not reflect pet ownership and this may be because student households are often shared households with tenants moving in and out, so less likely to have pets. In addition, students may not have the time nor the money to afford to care for pets. Figure 7 – number of pets in the household. Summary In summary, our class has a spread of fitness levels that broadly compares to the Australian population of similar age. The fittest students have lower heart rates than the least fit, both before and after exercise, and are far more likely to attend a gym and to go there more often. Like most Sydney residents, the train is the most popular way to commute. Arm spans are related to heights (as expected). Our students don’t have many pets compared to the Australian average. Number of words – 1336. References City of Sydney Community Profile (2017) Accessed on 20th August 2018 at https://profile.id.com.au/sydney, https://profile.id.com.au/sydney/travel-to-work Kravetz, D. (2014) How Heart Rate Is Related to Fitness and Longevity  THE BLOG 12/17/2013 06:09 pm ET Updated Feb 16, 2014, accessed on 20th August 2018 at https://www.huffingtonpost.com/dennis-kravetz/heart-rate_b_4428266.html de Brabandere, S. (2017) Scientific American (2018) https://www.scientificamerican.com/article/human-body-ratios/ ABS – Australian Bureau of Statistics – (2011) http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/4364.0.55.0012011-12?OpenDocument


Leave a Reply

Your email address will not be published. Required fields are marked *