Statistics

On 07/18/2011, in Career, Informal Education, by Jordan Wilson
image_pdfimage_print

Whenever I see statistics used in support of an argument I get a little suspicious.

And I am often reminded of a quote attributed to Benjamin Disraeli: “There are three kinds of lies: lies, damned lies, and statistics.”

Now I am not saying that statistics are necessarily false. But I am saying that data can often be manipulated to meet the needs of its user. 

When you are pursuing your career, managing your own business, or assessing investment opportunities, always maintain a healthy degree of skepticism towards any data presented to you by those you are dealing with.

The onus is on you to determine the validity of the data as relates to your own circumstances. If you do not, you will be at a disadvantage in your affairs.

An example of interesting statistics comes from a study I read this week.

How do you feel after reading the following statistics?

Each year for the past two decades, the U.S. Census Bureau has reported that over 30 million Americans were living in “poverty.” In recent years, the Census has reported that one in seven Americans are poor.

That is a lot of poor people.

A few lessons from this data.

Trust, But Verify

When assessing statistical data, I suggest you verify the actual numbers to the degree you think prudent.

Was the data prepared by a qualified party?

The more legitimate the source, the more you can rely on their hard numbers. In this case, I have some confidence that the U.S. Census Bureau numbers are probably correct. At least more confidence than if I asked my neighbor to calculate the number of poor in the U.S.

If the statistics were produced by a lobby group for the poor, I would have less confidence in the results. The reason is that the lobby group has a vested interest in showing a large number of poor people.

Was the methodology reasonable?

If sampling, can the sample size and composition be extended to the entire universe? What is the potential error?

Is the preparer or user biased in any way?

You must assess the motivations of those presenting you with data. Is it in their best interest to massage the statistics to best support their presentation?

Know the Underlying Data

Assumptions and Fine Print

I like to check the assumptions used in the calculations and read the fine print.

An advertisement I saw today on television said that if I bought their golf training system, I could add up to 10 more yards per drive. Hey, who could not use 10 more yards a drive?

But wait a minute. 5 yards is up to 10 yards, 10 being the maximum increase possible. And it only said that I could add yards, not that I would add distance. So if I bought the trainer, I could conceivably lose 3 yards a drive and still fall under their promise.

Now let us turn back to the U.S., where 1 in 7 people are poor.

Definitions

I always like to understand the definitions of terminology.

What a word or phrase means may differ between parties. Make certain everyone is on the same page.

What constitutes poverty in the U.S.? I know that I have my own definition. And yours might be similar or different from mine. The study’s authors have their own thoughts as well.

For most Americans, the word “poverty” suggests destitution: an inability to provide a family with nutritious food, clothing, and reasonable shelter. For example, the Poverty Pulse poll taken by the Catholic Campaign for Human Development asked the general public: “How would you describe being poor in the U.S.?” The overwhelming majority of responses focused on homelessness, hunger or not being able to eat properly, and not being able to meet basic needs.[1] That perception is bolstered by news stories about poverty that routinely feature homelessness and hunger.

So how does the U.S. Census Bureau define poor?

The Census Bureau defines an individual as poor if his or her family income falls below certain specified income thresholds, which vary by family size. In 2006, a family of four was deemed poor if its annual income fell below $20,615, and a family of three was deemed poor if annual income was below $16,079. In 2009, the thresholds were $21,954 for a family of four and $17,098 for a family of three. U.S. Census Bureau, Poverty Thresholds by Size of Family and Number of Children, at http://www.census.gov/hhes/www/poverty/data/threshld/index.html.

Now maybe this is reasonable and maybe it is not. My point of this post is not to debate an appropriate poverty level, only to show that statistics can change one’s perception of things. And that how one person views data may not be the same as another.

According to the U.S. Census Bureau, 1 out of 7 people is poor. But from the same statistical data, the study authors arrive at another conclusion.

The typical poor household, as defined by the government, has a car and air conditioning, two color televisions, cable or satellite TV, a DVD player, and a VCR. If there are children, especially boys, the family has a game system, such as an Xbox or PlayStation.

If you were presented with this summary, you may have a different view of the situation.

While the U.S. Census Bureau is a legitimate compiler of data, its conclusions will be considered off base by some users. Whether the results were published to achieve a certain desired effect (e.g., a substantial percentage of the U.S. population lives in poverty and therefore requires more government programs) or they were inadvertent in nature, is not something I can say.

I will say though, that even when I review data from legitimate and reputable sources, I look at the hard data myself and avoid succumbing to any pushed highlights.

And on that note, I want to mention that the linked study was published at The Heritage Foundation. They are a conservative think tank that promotes, in part, small government. So their studies may have a conservative leaning perspective.

Summary

Always ensure that the quality of data is good. That it was prepared by a legitimate and unbiased party, using sound methodology.

Always consider the motivations of those presenting data to you. How they see the world and what they are trying to accomplish will factor into their own analysis and conclusions. This may be a conscious manipulation or it may be inadvertent. Either means does you little good.

Always understand the terminology, assumptions and definitions that go into the statistical analysis. What you may associate with something may not be what the numbers really reflect. Or you may be swayed by the data presentation such that you agree with your counterparty.

Always perform your own statistical reviews. In business or investing, there is no substitute for conducting your own analysis.

1 Response » to “Statistics”

  1. Oh my goodness! an incredible article dude. Thank you Nevertheless I’m experiencing issue with ur rss . Don’t know why Unable to subscribe to it. Is there anyone getting similar rss problem? Anyone who knows kindly respond. Thnkx



© 2009-2017 Personal Wealth Management All Rights Reserved -- Copyright notice by Blog Copyright