botlhale klassen's Key Ideas from How to Lie with Statistics
by Darrell Huff

• Researchers can intentionally choose a group of people (or things) that have similar traits and thus 'prove' their point.
• This is deceiving because a report based on sampling must use a representative sample - every name or thing in the whole group must have an equal chance.
• Another famous method is choosing a very small sample.
• Depending on the population size, samples sizes must be at least 30% of the population.

There are 3 common averages: mean, median and mode. In some cases, the result of all 3 is the same number but in most cases the numbers are very different!

The Median (middle number) tells you more about the situation. On the left of this number is 50% of the sample on the right is the other 50% of the sample. Thus gives a good overview of what's happening as a whole.

But marketers normally report using the mean (the sum of all numbers divided by the number of units in the sample).

1. Not including the number of trials the experiment / research took - given a small enough number of cases, random 'by chance' occurrences can happen.
2. Leaving out the Degree of Significance - the demanded level is 1% which means there are 99 chances out of 100 that have an apparent difference.
3. The Range - this figure tells you the deviation of the sample from the average that is given.

Place little faith in an average or a graph or a trend when those important figures are missing

1. The Probable Error and The Standard Error - these tell you how accurately the sample presents the whole field (or population)
2. The Test of Significance (often calculated with the degree of significance)

To make things seem bigger than what they are, marketers can use these methods / tools

• The Gee Whiz Graph - basically adding more numbers on the graph to create an upward or downward trend that otherwise looked like a straight line.
• Bar Graphs - not using the same size for the bars or objects used

" If you can't prove what you want to prove, demonstrate something else and pretend that they are the same thing "

The general method is to pick two things that sound the same, but are not.

1. Look for conscious bias - ask "Who says so?"
2. Ask yourself "How does HE know that?" What methods were used (if any) to acquire that information?
3. Does it make sense - use your own intellect to see through the cracks!