04-16-2012, 10:32 AM

Originally Posted by CaldwellYSR

Yes they can be misused but it seems like it shouldn't be legal for people to misuse them like that. Not telling the whole truth is just as deceitful as telling a complete and total lie.

Ah, but therein lies the problem (this seems to be my favourite phrase in this thread). How do you determine which statistics are more "credible" if all of them have been presented with supporting statistical proof? Sure, the statistics they present might appear to be junk statistics (and I agree with you on that) but who has the right to say that "statistic A is more credible than statistic B." Like everything else in life, statistics, although arrived at through mathematical method, are subjective in their application.

04-16-2012, 10:52 AM

Originally Posted by vansmith

Therein lies the problem. Two different statisticians can use the same data set and come up with different results. Numbers, despite beliefs to the contrary, aren't neutral when used to make a point. One group making ostensibly false claims can back up their results with seemingly credible statistical tests.

I think my use of the words statistics wasn't the right one, I should have used bad science as the words since my beef is not the ambiguity of statistics but the blatant misuse of science to make claims not backed up by any real science

For instance, claims that a product can protect against radiation from Fukushima, where the actual science says that such radiation reaching the USA from Japan is barely adding a fraction of a fraction to background radiation

04-16-2012, 11:04 AM

I agree that it's easier to discern the difference between bad science and good science although it's certainly possible for that line to be blurred. The very fact that people still use science to claim that climate change doesn't exist is evidence of that.

However, I think it's easier to refute bad science over bad statistics.

04-16-2012, 11:12 AM

I prefer not to think of it so much as "Bad Science" but rather using "Alternative Logic Processes".

Having decided in advance what the desired research outcome is allows application of sophisticated data filtering and closed loop induction fields (aka circular reasoning) to support said thesis.

04-16-2012, 11:29 AM

Originally Posted by Razormac

Having decided in advance what the desired research outcome is allows application of sophisticated data filtering and closed loop induction fields (aka circular reasoning) to support said thesis.

This is why the methodological component (and, depending how you spin it, the intro) of proposals can be the most difficult since you have to be confident about what you're looking for without sounding as if you already have the answer.

04-16-2012, 01:46 PM

That's why some one said "the best thing a scientist can be is confused or wrong"

04-16-2012, 11:21 PM

Originally Posted by vansmith

Ah, but therein lies the problem (this seems to be my favourite phrase in this thread). How do you determine which statistics are more "credible" if all of them have been presented with supporting statistical proof? Sure, the statistics they present might appear to be junk statistics (and I agree with you on that) but who has the right to say that "statistic A is more credible than statistic B." Like everything else in life, statistics, although arrived at through mathematical method, are subjective in their application.

You can tell if statistic a is more credible than statistic b if statistic a is using the correct and relevant numbers to start with. Which in the example there's no way they were Although most of the time it's going to be pretty hard to tell the difference that's for sure.

04-17-2012, 01:08 AM

Originally Posted by CaldwellYSR

He says 7000 kids drop out of school every school day. Now I have a VERY hard time believing that is true, but for the sake of this argument I will let that slide. He continues to say "that is 26 students per second".

No he didn't. He said one drops out every 26 seconds.
LeBron James Addresses High School Dropout Crisis In New PSA (VIDEO)

Here's the actual commercial:
NBA All-Star LeBron James Featured in New Public Service Advertisement Encouraging Americans to Let Students Know We Believe In Them

I wonder what the legality is regarding people who go on a rant and mis-quote their source?

04-17-2012, 03:58 PM

Originally Posted by lifeisabeach

No he didn't. He said one drops out every 26 seconds . . . I wonder what the legality is regarding people who go on a rant and mis-quote their source?

They force them to become Politicians, for whom it is second nature . . .

04-17-2012, 05:59 PM

Came across this today and thought of this thread. It's George Carlin on advertising BS.

04-18-2012, 11:47 AM

Originally Posted by lifeisabeach

No he didn't. He said one drops out every 26 seconds.
LeBron James Addresses High School Dropout Crisis In New PSA (VIDEO)

Here's the actual commercial:
NBA All-Star LeBron James Featured in New Public Service Advertisement Encouraging Americans to Let Students Know We Believe In Them

I wonder what the legality is regarding people who go on a rant and mis-quote their source?

Ha!!! fail on my part. I guess that's what made it seem so ridiculously false.

So if 1 student drops out every 26 seconds that's 3324 students in 24 hours and the statistic is still pretty far off (86,400 seconds in a day divided by 26 seconds equals 3323.07... students). Thank you for clearing up my failure though.

04-18-2012, 12:06 PM

As of 2008, there were 16,375,000 high school students in the United States (source). There are roughly 200 school days a year. Using your value, that means that roughly 664,800 students will drop out in a year. That's 4.06% of the high school population. This government website pegs the figure at 8.1% (as of 2009). It would seem then that the value in the commercial is an underestimation. Using the values I've provided, it's closer to one student every 13 seconds (if I've done the math right).

04-18-2012, 12:34 PM

Originally Posted by CaldwellYSR

Ha!!! fail on my part. I guess that's what made it seem so ridiculously false.

So if 1 student drops out every 26 seconds that's 3324 students in 24 hours and the statistic is still pretty far off (86,400 seconds in a day divided by 26 seconds equals 3323.07... students). Thank you for clearing up my failure though.

No. First off, he said 7000 students drop out every SCHOOL day, of which there are an average of 180/year. According to the links provided, approximately 1.3 million students drop out per year, which comes to approximately 7,222 per school day. So he's right so far. So let's go with the 7K figure. 7,000/24=291.67/hour. Divide again by 60 minutes and you have 4.96/minute. Divide by 60 seconds and you get .08 dropouts/second. Multiply that by 26 and you get.. wait for it… 2.08 dropouts per 26 seconds!

But wait! He said 1 dropout per 26 seconds, when it comes to 2 dropouts per 26 seconds? Well the school year may be 180 days, but it may as well be 365 for all it matters. Extend that dropout rate to the full calendar year and you get… wait for it… 1 dropout every 26 seconds.

So basically, assuming the number of dropouts is accurate (and I'm quite certain you are in no position to present more reliable numbers), then the ad was on the money. The breakdown isn't inflated. If anything, they could have been somewhat more consistent by saying "1 student drops out every 5 school-seconds" (assuming 7 school-hours per school-day), but that would make no sense whatsoever. Not to the average person anyway.

Of course I can hear it now… "then they should say 3,500 students drop out per day". If the school year was indeed 365 days/year, then there would be a LOT more dropouts due to burnout. As it is now, you can't exactly drop out on a weekend or on break, so breaking down the numbers by school-year makes more sense, and is a figure people understand. But "school-seconds" isn't a figure that exists.

04-18-2012, 12:58 PM

Originally Posted by lifeisabeach

No. First off, he said 7000 students drop out every SCHOOL day, of which there are an average of 180/year. According to the links provided, approximately 1.3 million students drop out per year, which comes to approximately 7,222 per school day. So he's right so far. So let's go with the 7K figure. 7,000/24=291.67/hour. Divide again by 60 minutes and you have 4.96/minute. Divide by 60 seconds and you get .08 dropouts/second. Multiply that by 26 and you get.. wait for it… 2.08 dropouts per 26 seconds!

But wait! He said 1 dropout per 26 seconds, when it comes to 2 dropouts per 26 seconds? Well the school year may be 180 days, but it may as well be 365 for all it matters. Extend that dropout rate to the full calendar year and you get… wait for it… 1 dropout every 26 seconds.

So basically, assuming the number of dropouts is accurate (and I'm quite certain you are in no position to present more reliable numbers), then the ad was on the money. The breakdown isn't inflated. If anything, they could have been somewhat more consistent by saying "1 student drops out every 5 school-seconds" (assuming 7 school-hours per school-day), but that would make no sense whatsoever. Not to the average person anyway.

Of course I can hear it now… "then they should say 3,500 students drop out per day". If the school year was indeed 365 days/year, then there would be a LOT more dropouts due to burnout. As it is now, you can't exactly drop out on a weekend or on break, so breaking down the numbers by school-year makes more sense, and is a figure people understand. But "school-seconds" isn't a figure that exists.

Nope I have no arguments for that... looks legitimate to me and I'm glad it is. I'll gladly bite the bullet and admit failure here. Thanks for pointing out the validity.