Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Locked thread
Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

Tossing out any data that doesn't fit your agenda is at least a 200 level class

Have you even taken a stats class

Adbot
ADBOT LOVES YOU

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

Have you even taken a stats class

For the last two months I've been doing statistical analysis of asbestos air monitoring samples ten hours a day, you tell me

Nerses IV
May 4, 2007

by FactsAreUseless
Literally this week an extremely high sample count was discarded because the sampling cassette was dropped in to a glove bag, so it was thrown out as a meaningless outlier

Then they hit a sample that was just as high the next day

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

For the last two months I've been doing statistical analysis of asbestos air monitoring samples ten hours a day, you tell me

So why the struggle to grasp that using the mean when you have a huge outlier will give you heavily skewed data, especially in a small dataset?

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

So why the struggle to grasp that using the mean when you have a huge outlier will give you heavily skewed data, especially in a small dataset?

I'm not the one who asked for the mean, dumbass

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:


Then they hit a sample that was just as high the next day

Norway hasn't had another event like breivik and likely won't have one again.

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

Norway hasn't had another event like breivik and likely won't have one again.

Before Breivik they'd never had an event like Breivik and likely never would

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

Before Breivik they'd never had an event like Breivik and likely never would

Yes that's why it's called an outlier lol

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

Yes that why it's called an outlier lol

Yes and that's why outliers are important, and shouldn't be discarded out of hand just because they're not what you expected

dwarf74
Sep 2, 2012



Buglord

Nerses IV posted:

Tossing out any data that doesn't fit your agenda is at least a 200 level class
Well, here, if you're reluctant to properly note outliers in your data set, you can at least improve the honesty of the results with a time series - call it gun deaths per capita by year. Oh my that's a huge spike in 2011 for Norway, wow. Where's it sit for the remaining years of your data set? And if we are using it to make predictions of future gun deaths, what should the forecast look like for Norway? Should we predict 1.30 per 100,000 for 2018 and 2019?

Nerses IV
May 4, 2007

by FactsAreUseless

dwarf74 posted:

Well, here, if you're reluctant to properly note outliers in your data set, you can at least improve the honesty of the results with a time series - call it gun deaths per capita by year. Oh my that's a huge spike in 2011 for Norway, wow. Where's it sit for the remaining years of your data set? And if we are using it to make predictions of future gun deaths, what should the forecast look like for Norway? Should we predict 1.30 per 100,000 for 2018 and 2019?

Already the brilliant statistician has abandoned mass shootings entirely and is now focusing on gun deaths as they will provide a more convenient result

Nerses IV
May 4, 2007

by FactsAreUseless
Which stats class do they teach you about units mattering

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

Yes and that's why outliers are important, and shouldn't be discarded out of hand just because they're not what you expected

By keeping the outlier in you're obfuscating the underlying trend. Clinging to a one off event as a gotcha when nothing even close to Breivik's spree has occurred in seven years in Norway reeks of dishonestly.

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

By keeping the outlier in you're obfuscating the underlying trend. Clinging to a one off event as a gotcha when nothing even close to Breivik's spree has occurred in seven years in Norway reeks of dishonestly.

In the same thread as people citing the Gun Violence Archive, I'm dishonest for posting a chart that included all relevant data

dwarf74
Sep 2, 2012



Buglord

Nerses IV posted:

Already the brilliant statistician has abandoned mass shootings entirely and is now focusing on gun deaths as they will provide a more convenient result
Brother is that literally where we are, now, where you're going to pretend I'm suddenly talking about something different than the mass shootings we've been talking about all along?

Nerses IV
May 4, 2007

by FactsAreUseless

dwarf74 posted:

Brother is that literally where we are, now, where you're going to pretend I'm suddenly talking about something different than the mass shootings we've been talking about all along?

Ah okay when you kept saying "gun deaths" you were actually talking about something else and I just should have assumed that

Genius-level statisticians shouldn't feel the need to remain consistent when discussing their data, people can just fill in the blanks themselves

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

In the same thread as people citing the Gun Violence Archive, I'm dishonest for posting a chart that included all relevant data

Bad stats is bad stats. If you want to lay out a case against the gun violence archive thing then go ahead.

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

Bad stats is bad stats. If you want to lay out a case against the gun violence archive thing then go ahead.

I love the idea of posting a chart based on researching mass shootings by country from 2000-2014 and just leaving the big fucker out

e: To include it would be dishonest, you see

dwarf74
Sep 2, 2012



Buglord

Nerses IV posted:

Ah okay when you kept saying "gun deaths" you were actually talking about something else and I just should have assumed that

Genius-level statisticians shouldn't feel the need to remain consistent when discussing their data, people can just fill in the blanks themselves
This is seriously all you got left in the tank?

dwarf74
Sep 2, 2012



Buglord

Nerses IV posted:

I love the idea of posting a chart based on researching mass shootings by country from 2000-2014 and just leaving the big fucker out

e: To include it would be dishonest, you see
How's that time series coming along?

Nerses IV
May 4, 2007

by FactsAreUseless

dwarf74 posted:

This is seriously all you got left in the tank?

dwarf74 posted:

This is seriously all you got left in the tank?

Lol what the gently caress is this

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

I love the idea of posting a chart based on researching mass shootings by country from 2000-2014 and just leaving the big fucker out

e: To include it would be dishonest, you see

If you have an issue with basic statistical best practices take it up with math? Not understanding how and why such a large outlier actually gives you a less accurate picture of mass shootings worldwide is your problem.

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

If you have an issue with basic statistical best practices take it up with math? Not understanding how and why such a large outlier actually gives you a less accurate picture of mass shootings worldwide is your problem.

Imagine being this loving dumb

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

Imagine being this loving dumb

Must be rough living it every day champ

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

Must be rough living it every day champ

I'm sorry, but I can only say "I'm not the one who asked for per capita stats" so many times before I start thinking you just can't learn

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

I'm sorry, but I can only say "I'm not the one who asked for per capita stats" so many times before I start thinking you just can't learn

You sure seem hellbent on defending those stats tho

Nerses IV
May 4, 2007

by FactsAreUseless
Like, if "bad stats is bad stats" were actually true, you'd be tearing apart the Gun Violence Archive yourself instead of waiting for me to do it

But you agree with the Gun Violence Archive's numbers, so you don't

Per capita stats were just fine before I revealed the trick

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

Like, if "bad stats is bad stats" were actually true, you'd be tearing apart the Gun Violence Archive yourself instead of waiting for me to do it

But you agree with the Gun Violence Archive's numbers, so you don't

Per capita stats were just fine before I revealed the trick

Please do tear the gva stats apart. I'm sure for a mind as keen as yours it'll take no time at all

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

Please do tear the gva stats apart. I'm sure for a mind as keen as yours it'll take no time at all

How many mass shootings can you name where nobody died

VictorianQueerLit
Aug 25, 2017

Nerses IV posted:

Lol what the gently caress is this

The D&D special. War through attrition.

The way to do it is to never give up and win by default when people quit arguing with you. It doesn't matter if we are arguing about the specifics of the technical semantics of a tangent of a derail and they are trying to pretend that gun deaths don't matter to a gun death statistic when it doesn't even matter to this pointless argument as it would make America #3 instead of #4 on the list which still disproves their initial assertion.

Nerses IV
May 4, 2007

by FactsAreUseless

VictorianQueerLit posted:

The D&D special. War through attrition.

The way to do it is to never give up and win by default when people quit arguing with you. It doesn't matter if we are arguing about the specifics of the technical semantics of a tangent of a derail and they are trying to pretend that gun deaths don't matter to a gun death statistic when it doesn't even matter to this pointless argument as it would make America #3 instead of #4 on the list which still disproves their initial assertion.

God drat it dude you just ruined the surprise reveal

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

How many mass shootings can you name where nobody died

A mass shooting only implies that people were y'know, shot. They don't neccessarily have to die.

VictorianQueerLit
Aug 25, 2017

Nerses IV posted:

God drat it dude you just ruined the surprise reveal

:smug: I challenge you to prove that I was trying to "surprise" anyone by the webster's definition of the word. You want to go thirty posts about the intent of my wording? :smug:

As long as I'm still arguing I haven't lost. When discussing the amount of deaths by gun the amount of deaths by gun don't matter you see.

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

A mass shooting only implies that people were y'know, shot. They don't neccessarily have to die.

This is certainly true if you want to exploit massively inflated numbers from America's drug violence and make the statistics incomparable worldwide as nobody else tracks them that way

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

This is certainly true if you want to exploit massively inflated numbers from America's drug violence and make the statistics incomparable worldwide as nobody else tracks them that way

So the drug violence is in no way exacerbated by easy access to cheap, legal arms.

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

So the drug violence is in no way exacerbated by easy access to cheap, legal arms.

We should ban them, it worked for drugs

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

We should ban them, it worked for drugs

And guns are equivalent to drugs how exactly?

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

And guns are equivalent to drugs how exactly?

And drug dealers drive-bying eachother are equivalent to a school shooting how exactly?

Have Blue
Mar 27, 2013


Panther Like a Panther

Nerses IV posted:

And drug dealers drive-bying eachother are equivalent to a school shooting how exactly?

They're symptoms of an underlying flaw in our culture and laws that encourage and enable violence. If any pised off idiot can get their hands on a gun it's going to end badly regardless of underlying motive.

Adbot
ADBOT LOVES YOU

Nerses IV
May 4, 2007

by FactsAreUseless

Have Blue posted:

They're symptoms of an underlying flaw in our culture and laws that encourage and enable violence. If any pised off idiot can get their hands on a gun it's going to end badly regardless of underlying motive.

I often determine fixes to problems by ignoring the underlying causes

  • Locked thread