Sleepless Americans

Sleepless Americans

June 22, 2019 DATAcated Challenge 0

Hello data family, Hope you guys are doing good.

Today I wanted to talk about a topic – “Sleepless Americans”. This data provides us information about the Average sleep hours of Americans from 2003 to 2017 by age and sex.

The dataset was first cleaned to remove ambiguities. The first change to the data was to merge or to say in Tableau language ‘group’ two age groups named – ’15 and above’ and ’15 to 24 years’ into one age group named – ’15-24′. I know it might lead to more questions, but this is what I felt was right.

Next, there was ambiguity with respect to the type of days as ‘All days’ constitutes of both ‘Non-holiday weekdays’ and ‘weekend days and holidays’. Same goes with ‘Both’ gender. So, in most of the analysis, they were redundant.


Some of the insights drawn from the dashboard shown above are the following (Please follow from the top left corner, clockwise) –

  1. Irrespective of the age group, women have higher average sleep hours than men by just 20-30 mins. (Men must be waking up early and cooking them nice breakfast, maybe! )
  2. This is interesting. I always love to find a relation between data and real-life events. If we hover over the projectile of the average sleep hours by years, we observe that till the age group of 55, the least value falls in 2004. You know why? I assume that it must be 2004 elections and George W. Bush had defeated  John Kerry in nail-biting competition. We also observe that leat value for elderly people falls in 2008 due to the financial crisis. (Page format by years)
  3. I made 2 new parameters named ‘positive error’ and ‘negative error’. They are just basic addition and subtractions of standard error column to the average sleep hour columns. This can be interpreted as the confidence intervals kind of errors (or maybe standard deviation of avg. sleep hours). (Page format by years)
  4. The fourth one says that people in general, tend to sleep at least an hour to an hour and a half more on ‘weekend days and holidays’ than the ‘nonholiday weekdays’.

Thank you. Keep analyzing, keep visualizing!

Sai Raghuram Kothapalli


Leave a Reply

Your email address will not be published.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