DARRELL HUFF wrote the book “HOW TO LIE WITH STATISTICS” to make people able to ask questions about statistics is a proper way it makes people understand many hidden ploys of statistics. It is a manual to understand some cheating in statistics.
HUFF says in this book, the secret language of statistics is employed to sensationalize, inflate, confuse and oversimplify. If a writer do not use words honestly and readers do not understand that. The result is only nonsense.
here is a indication in this book is-“A well warped statistics is better than hitter’s “BIGLIE”
This book is to know how statistics is used to deceive.
Chapter-1: The Sample with the Built-in Bias:
Here is a description is about bias of sample. All kinds of statistics is based on some sample and every sample has some kind of bias. Though a person want not to make any bias, the build-in bias comes from the respondents who do not reply honestly. An example, according to writers, is a readership survey of two magazines. Respondent were asked which magazines they read most-‘true story’ or ‘Harper’s’ Most respondents replied that they read “Harper’s. Now there were publisher’s figures that showed very clearly that ‘true story’ had much more circulation than ‘Harper’s’ refuting the results from the sampling. the reason for this discrepancy is that people were not willing to respond due to their own bias. So summary of this chapter is that there is always a bias in the sample.
Chapter-2: The well-choose Average
The main theme of this chapter is there are three average-mean, median and mode. When a person says about average, there are many scopes for deceiving. So when someone see an average, he/she should ask what kind of average? average of what? who is included?
An example is that in a village 25 people have been as respondent to study their monthly income. Their total income’s sum become 4,0,0,000 Tk. So mean or arithmetic average of their income is (4,00,000¸25) 16,000 Tk. So now people think that every body’s income is about 16,000 Tk. but in reality among 25 people 5 people’s monthly income is 40,000 Tk. each and rest 20 people’s sum of income is 2,00,000 Tk. in this case the average is arithmetic mean.
And if we arrange them (the income) from high to low then middle of the serial is median. And if we see that some people have 5,000 Tk. income equally then it is mode.
Chapter-3: The little figures that are not there
Huff wanted to say in this chapter how to sample data is picked up in a way to prove the result and picking the sample data can mean picking a sample size that gives the kind of results we are looking for,.
He discuss this with a very important issue for parents is my child is normal or abnormal. HUFF talks about the ‘Gessel Norms’, where Dr. Arnold Gessell stated that most children sit erect by the age of two. And this produced pain among parents whether their children are normal or not. This happens when the average has been changed from average to an exact figure. It is mentioned that there is a range of age in which a child sits erect, the reader were assuaged and that is where the little figures disappear. So one should ask if there is a range.
Chapter-4: Much ado about practically nothing:
This chapter is about errors is measurement. There are two measures for measuring error that are probable error and standard error. The probable error measure the error in the measurement based on how much off is one’s measurement device. If one, for example, was using a measure scale that is 3 inches off a foot, then his measurement accurse trials is+/-3.
Chapter-5: The Gee-whiz Graph:
This chapter is about how to manipulate a graph so that it should and inflated or deflated picture. Some tricks include messing out the measure of the exist, do not lable the axis leaving only numbers and hence letting the reader make his/her own assumption.