Scuba Diving Statistics

Folks, I see another misuse of diving statistics here I'll freak...

"Most scuba fatalities are found with their weight belts still on"

"Obesity is a leading cause of scuba fatalities"

"Bowling is more dangerous than scuba diving"

Come on, folks, let's get real here.

First of all, RAW NUMBERS tell you nothing. DAN statistics are RAW NUMBERS. You can't make useful conclusions from RAW NUMBERS.

For example: If there were a total of 80 fatalities in a year, and 70 of those were male, what does that tell you? If 70 of them were overweight, what does that tell you? If 70 of them were wearing weight belts when they were found on the bottom what does that tell you? If 70 of them were found with their fins on, what does that tell you? If 70 of them were wearing pink underwear, what does that tell you?

The answer to all of the above is NOTHING !!! You can infer NOTHING from those numbers. Well, maybe the pink underwear....

Only if you have additional data about the general population of divers (# of divers, # of dives, etc), and root cause of the incidents will these numbers even begin to have any significance whatsoever.

If 70 of the 80 fatalities were obese, but 95% of the diving population is overweight, then that 70 number might be expected and tell you nothing. However, if on the other hand only 5% of the diving population is overweight, then you may start to infer that obesity might be a factor. But until you know the population statistics and root cause information, you can infer nothing. If it turns out that the root cause of the fatalities were ascending too fast, then the obesity numbers could be coincidental and misleading.

My favorite is the ridiculous "safer than" statistics. "Safer than bowling" ranks up there with the most inane. If there are 80 diving fatalities, and 90 bowling fatalities, what does that tell you? NOTHING !!! You can't make even the slightest inference from that data. I'll leave it to you to figure out all the reasons why (if you can't come up with at least 5, then you're not thinking).

If you want to prove a point, don't use statistics unless you have all the associated data.

And if anyone cites bogus statistics again, I'll have no choice but to call the police

mccabejc:

Folks, I see another misuse of diving statistics here I'll freak...

"Most scuba fatalities are found with their weight belts still on"

"Obesity is a leading cause of scuba fatalities"

"Bowling is more dangerous than scuba diving"

Come on, folks, let's get real here.

First of all, RAW NUMBERS tell you nothing. DAN statistics are RAW NUMBERS. You can't make useful conclusions from RAW NUMBERS.

For example: If there were a total of 80 fatalities in a year, and 70 of those were male, what does that tell you? If 70 of them were overweight, what does that tell you? If 70 of them were wearing weight belts when they were found on the bottom what does that tell you? If 70 of them were found with their fins on, what does that tell you? If 70 of them were wearing pink underwear, what does that tell you?

Click to expand...

Jim, you're absolutely right that the raw numbers don't really tell you much that is meaningful without a knowledge of the entire population.

Much of my inference is from personal observation without benefit of actual numbers. I am at the dive park about 150 days a year. I get a reasonable picture of the representation of various divers in the park. I see that the vast majority of the deaths are of obese people. Does this statistically mean obesity causes death? No. However when I consider the entire population that I've observed, there is a distinct correlation between body type and death rate.

By the way... I DO think that bowling is much more dangerous than diving. Heck, last time I bowled a few lines, I hit at least four people and one was four alleys away! Those balls are heavy.

Bowling is definately more dangerous than diving.

I have never been beat up by a dive buddy.
The last time I went bowling my team beat me up for my fine score of 105.
My all time high game was 113, so no one invites me to go bowl.

As to numbers and statistics, we will NEVER have all the supporting data to produce the most useful indications but that doesn't mean that the information is worthless.

It has been said that there are lies, d**n lies, and statistics. The problem with statistics is that you have to be very careful how you ask the question and how well does the answer fit the question.

Then again, 87.352% of all statistics are made up on the spot. :dazzler1:

drbill:

Much of my inference is from personal observation without benefit of actual numbers. I am at the dive park about 150 days a year. I get a reasonable picture of the representation of various divers in the park.

Click to expand...

NOW we're talking !! What you cite is FAR more useful data. It includes not only population information, but also (with your local connections) root cause information. Now if only we could get that kind of information with the DAN statistics.

Hmmm, I'll put this through a multiple regression analysis and get back to you!
Actually if you had accurate raw data you could start to drill down to meaningful
data, causes, "x's" if you will. Problem is collecting meaningful data. I'm sure there
is no one source that you could reasonably compare bowling to scuba, "Apples
to Oranges." Your point is well taken about mis-use of statistics, however, there
are many factors that contribute to scuba accidents such as obesity, smoking,
which I can say, I'm a smoker "shame on me." Or how about a tri-athelete that
suddenly keels over from a stroke or heart attack, one too many Big Macs?
I'm sure there is a bell curve out there somewhere on scuba accidents, lets not
worry about it and go get some much needed bottom time!

mccabejc

To fill out the population data, I guess we could inspect every diver on every dive -- I sure don't want the job of looking for pink underwear though.

Seriously, I agree with you 100%. The "bowling" comparison is surely the most ridiculous. How do you get killed while bowling?

Carbon

pipedope:

Then again, 87.352% of all statistics are made up on the spot. :dazzler1:

Click to expand...

I thought it was closer to 83%. Oh well. Anyway, a more interesting and useful metric would be incident rates per X hours of activity or X number of participants. This gives you a much better basis of comparison. It's similar to people that freak out about shark attacks...the incidence rates for shark attacks is very low when you have the whole picture.

rtkane:

It's similar to people that freak out about shark attacks...the incidence rates for shark attacks is very low when you have the whole picture.

Click to expand...

Unless, of course, YOU are one of the statistics!

Have you heard that only 25 percent among the published articles containing a statistical analysis is a valid analysis.... I am not surprised at the news.....

mccabejc:

If you want to prove a point, don't use statistics unless you have all the associated data.

Click to expand...

87% of all statistics are made up.