Tag Archives: data

Delhi Elections: Exit Polls vs Results

Leave a reply

The results are out and the winner is ‘Aam Aadmi Party’. A start-up that growth hacked itself to become the second highest seat winner. Here is how 2 exist polls compared to the results

exit_pollsvsresults

 

While Today’s Chanakya Exit Polls data were closest to the results, Times Now (except for their spot on BJP Exist Poll data) are way away from the results. Good Job ‘Chanakya’ !!

Now to take a macro view of the Chanakya exit polls and results

 

Fifteen Quotes on Data

Leave a reply

Here are some of my favorite quotes for anyone fascinated by data

  1. “A point of view can be a dangerous luxury when substituted for insight and understanding” – Marshall McLuhan
  2. “Anything that is measured and watched improves.” – Bob Parsons
  3. “Data beats emotions.” – Sean Rad, founder of Ad.ly
  4. “True genius resides in the capacity for evaluation of uncertain, hazardous, and conflicting information.” – Winston Churchill
  5. “Data is a precious thing and will last longer than the systems themselves.” – Tim Berners-Lee
  6. “Facts do not cease to exist because they are ignored.” – Aldous Huxley
  7. “For every two degrees the temperature goes up, check-ins at ice cream shops go up by 2%.” – Andrew Hogue, Foursquare
  8. “If you don not know how to ask the right question, you discover nothing” – W. Edward Deming
  9. “If the statistics are boring, you’ve got the wrong numbers” – Edward Tufte
  10. “In God we trust. All others must bring data.” – W. Edwards Deming
  11. “If you torture the data long enough, it will confess” – Ronald Coase
  12. “Without big data analytics, companies are blind and deaf, wandering out onto the web like deer on a freeway.” – Geoffrey Moore 
  13. “Let the data set change your mindset” – Hans Rosling
  14. “Data is the new Oil” – Clive Humby

    and my favorite

    If we have data, let’s look at data. If all we have are opinions, let’s go with mine. – Jim Barksdale

PS: Top 15 is the new top 10

Book Review: How To Lie With Statistics by Darrell Huff

Leave a reply

Stumbled upon this book while searching for a basic level book on ‘Statistics’ and the blurb and reviews caught my attention so I decided to pick it up.

How to lie with statistics

(How To Lie With Statistics – Darrel Huff)

It’s a petite book with less than 150 pages. It’s not really a book on ‘Statistics’ but on how Statistics are used and abused both consciously and unconsciously by advertisers, surveyors, sales people and others alike or ‘Statisculation’ (Statistical Manipulation)

Here’s a little summary/review of the book:

  1. Many a times the surveys/findings that we read/hear about are bound to be incorrect because of the bias while collecting samples. Either the sample chosen for doing survey isn’t well defined or doesn’t reflect well on the overall set.
  2. Averages – Mean/Median/Mode . Choosing an average to depict some statistics is very important and for the layman’s eye it is very easy to draw wrong inference in some cases by mean or average without knowing about the range. Average salary alone for example doesn’t reflect the information correctly so its important to know the range and also the median.
  3. Missing little figures: Using a very small group to draw a trend, not providing enough information about a particular % in question to even as much as not labeling an axis in a graph. Often one needs to question the completeness of the statistic provided.
  4. Keep an eye on Probable and Standard errors with measurements in question for ex: IQ of 101 and 98 could just be 101 and 98
  5. A difference is only a difference if it makes a difference
  6. Graphs can be manipulated in various ways namely cutting the graph from bottom(to reduce the visible area), changing the measures of axis to make the graph looks better or worse
  7. In charts with more than 1 D, using double the area to show double dimension is misleading. Varying size of objects to show increased number often gives an impression of the change in actual size of object in question over time.
  8. Semi-attached figures, post hoc and many more

How to talk back to a statistic ?

Ask the following questions

  1. Who says so ? (To understand biases)
  2. How does he know? (Source and reliability of information presented
  3. What’s missing? (Actual numbers in case of %s, base in case of index or other things)
  4. Did somebody change the subject? (Switch between raw figures and their conclusion)
  5. Does it make sense? (Apply commonsense on the data, and try extrapolating the trends to check sanity)

A quick read for the initiated which will change how you see statistics used around and hopefully save you from deriving false conclusions