Numbers Don't Lie...Or Do They?
By Jeff Moffitt
@moffittjc (128824)
Gainesville, Florida
May 17, 2016 9:11pm CST
I heard an interesting story today on how you have to be careful when you are working with data and statistics. The story was about UNC...the University of North Carolina.
UNC is a pretty prestigious school, offering many great degree programs. If you took a look at all their degree programs, and look at the data showing how much all the graduates earn in each of the various degrees offered, could you take a stab at which degree offers the highest average salary of them all?
Would your guess include a law degree? A medical degree? Those are good ones, but they're not even close!
Based on the average salaries of all its graduates broken down by degrees, would you believe that a Geography degree from UNC has the highest average salary?
This is where you have to pay attention to the data set! Can anyone venture a guess as to why the data shows that Geography majors have the highest average salary? I'll give you a hint: think of one of the most famous figures to graduate from UNC, and think about what his major was!
If you guessed Michael Jordan, you are correct! Michael Jordan has a degree in Geography, and also has a net worth of over $1 billion! His reported income last year alone was over $100 million. As a result, just one person skews the data set and makes it look like Geography majors make the most money!
So anytime you're analyzing data, make sure you examine the details closely! You never know when a Michael Jordan is sitting in there skewing all your data!
8 people like this
5 responses
@ElizabethWallace (12069)
• United States
18 May 16
Figures don't lie, but liars figure. Someone somewhere want the data skewed, or they would set the parameters to include only those who made their money based upon the degrees.
3 people like this
@moffittjc (128824)
• Gainesville, Florida
20 May 16
I'm sure that if in truth those statistics were used, there would have been filters and parameters set up, or complex algorithms put in place to weed out those with no salaries and those will outrageous salaries. In this case, I think it was just used as a teaching example for data analysis and the need to make sure you are reporting accurate data that has not been skewed or tainted in any way.
2 people like this
@ElizabethWallace (12069)
• United States
21 May 16
@moffittjc Never assume that people do what is logical. I fight this urge. I usually assume that people are logical, but am rarely correct in this assumption.
2 people like this
@JESSY3236 (22199)
• United States
18 May 16
lol I think degrees don't really matter. Because if they really did then Michael Jordan wouldn't have been a basketball star.
1 person likes this
@moffittjc (128824)
• Gainesville, Florida
20 May 16
It's amazing how many people will go through 4 or more years of college to get a degree, and then end up going into a completely different field that has nothing to do with their degree. I see it happening all the time! Why waste all that money on tuition and room and board and books if you're not even going to use the degree you earned?
2 people like this
@JamesHxstatic (29410)
• Eugene, Oregon
18 May 16
It is important to know how skewed stats can be by some small fact like that.
1 person likes this
@moffittjc (128824)
• Gainesville, Florida
20 May 16
They say numbers don't lie, but the way the numbers are presented can lie. Almost any data can be skewed in a way that is still technically "true" or "accurate," yet show a completely different result than someone else presenting the same data in a different manner.
1 person likes this
@JudyEv (381759)
• Rockingham, Australia
18 May 16
This is very interesting. I know in some data we were looking at a long time ago the two highest and two lowest scores were ignored. It was thought this would give a truer average. No idea what we doing now. I'll probably remember at 3am what it was all about!
1 person likes this
@moffittjc (128824)
• Gainesville, Florida
20 May 16
In the world of statistics and analytical data, there are methodologies in place to "norm" the data. It all depends on how you set up your data field, and the algorithms you use to set the parameters of the data you are analyzing. Valid statistical analysis will always factor in that there is information that can skew the data. Although I don't remember exact details, I do remember talking about this phenomenon in my statistics class back in college.
1 person likes this
@Daljinder (23193)
• Bangalore, India
18 May 16
He sure messed it up. lol I mean the stats.
1 person likes this
@moffittjc (128824)
• Gainesville, Florida
20 May 16
Well he sure didn't mess it up for himself! lol I'd say he's doing pretty well!
1 person likes this
