Apparently when i tell people horrible things like the median income or the unemployment rate, they do not believe me. I need sources, and it occurs to me you know this shit, but these heathens will not take the word of a Holy Man. So where do you get your sources?
Quote from: Requia ☣ on January 19, 2010, 07:44:50 AM
Apparently when i tell people horrible things like the median income or the unemployment rate, they do not believe me. I need sources, and it occurs to me you know this shit, but these heathens will not take the word of a Holy Man. So where do you get your sources?
http://bls.gov is a good place to start...when the government itself starts jabbering shit like that, you know we're in trouble.
Also http://en.wikipedia.org/wiki/Household_income_in_the_United_States but this includes ALL forms of income, including government assistance.
Besides, mode income is what you want... http://en.wikipedia.org/wiki/Mode_(statistics) <---for those who have forgotten what mode is. Good luck finding an official source for that, though...for some reason, the BLS doesn't compile that number. You have to do it yourself with the available data.
sounds like a typical job for Wolfram Alpha, it was designed for queries like this:
http://www.wolframalpha.com/input/?i=median+wages+usa
I couldn't get it to calc the modal income though, nor am I sure whether the number it gives is the one you are looking for.
Quote from: Triple Zero on January 19, 2010, 12:35:44 PM
sounds like a typical job for Wolfram Alpha, it was designed for queries like this:
http://www.wolframalpha.com/input/?i=median+wages+usa
I couldn't get it to calc the modal income though, nor am I sure whether the number it gives is the one you are looking for.
Nope. That's the median wage, and it's depressing enough as it is.
The mode wage is the wage you are most likely to find yourself making, statistically speaking (ie, more people make that wage than any other wage.). This includes everyone from the CEO of Halliburton, all the way to the part time jizz mopper.
Last I calculated it (based on the admittedly bad data from the BLS), it was a whopping $19,500.
You can see why this number is never brought up. Neither party wants it.
yes I understand, I just wasn't sure if that number was the proper median wage either.
to calculate the mode wage, you would count the number of people in different income brackets, I suppose?
Quote from: Triple Zero on January 19, 2010, 02:26:52 PM
yes I understand, I just wasn't sure if that number was the proper median wage either.
to calculate the mode wage, you would count the number of people in different income brackets, I suppose?
Yes. The only people who do that are labor activists and the 5 remaining journalists in America, so of course nobody listens.
I decided to do some number crunching, but first I gotta make sure I got the right data.
Wolfram Alpha says the median annual income is $31410 (2007 estimate) but www.bls.gov says $42270 (2008 estimate). Now either it could be the different years, or someone gives the wrong data, however I tend to go with me looking at the wrong figures. Got an idea?
And yes, I'm looking for the mode, I'll get to that, but I was just comparing the numbers to see if I'm even looking at the right stats.
Also all these totals don't really add up and are screwy.
Quote from: Triple Zero on January 19, 2010, 04:57:45 PM
I decided to do some number crunching, but first I gotta make sure I got the right data.
Wolfram Alpha says the median annual income is $31410 (2007 estimate) but www.bls.gov says $42270 (2008 estimate). Now either it could be the different years, or someone gives the wrong data, however I tend to go with me looking at the wrong figures. Got an idea?
And yes, I'm looking for the mode, I'll get to that, but I was just comparing the numbers to see if I'm even looking at the right stats.
Also all these totals don't really add up and are screwy.
The BLS is in the business of telling people that everything is fine under the current administration (at any given time). They are useless propaganda-peddlers that would be put to sleep in a civilized society. Go with Wolfram Alpha.
never mind I was looking at the wrong column, that was the mean. the median would be roughly $32385.60 (by multiplying the median hourly wage with 2080 hours per full-time year).
anyway I need the BLS cause you need a lot of numbers to estimate the mode. Wolfram Alpha doesn't give me enough numbers.
I just did a few checksums and the totals seem to come out pretty sane-ish, 134.5M, 136.2M or 135.2M total employed people in the USA, depending respectively on whether I <sum by occupation-group by state>, <sum by state, all occupations> or the total number BLS gives on that link. funny how the official BLS number is almost exactly in the middle of those two sums :)
anyway, then I got a list of occupation groups per state, that's about 36k entries (!!!). if I'd make a histogram out of the amount of people in each state-occupation-group versus their mean (yes mean) annual income, the mode should roll right out.
at the very least it should give a pretty graph.
Quote from: Triple Zero on January 19, 2010, 05:18:26 PM
never mind I was looking at the wrong column, that was the mean. the median would be roughly $32385.60 (by multiplying the median hourly wage with 2080 hours per full-time year).
anyway I need the BLS cause you need a lot of numbers to estimate the mode. Wolfram Alpha doesn't give me enough numbers.
I just did a few checksums and the totals seem to come out pretty sane-ish, 134.5M, 136.2M or 135.2M total employed people in the USA, depending respectively on whether I <sum by occupation-group by state>, <sum by state, all occupations> or the total number BLS gives on that link. funny how the official BLS number is almost exactly in the middle of those two sums :)
anyway, then I got a list of occupation groups per state, that's about 36k entries (!!!). if I'd make a histogram out of the amount of people in each state-occupation-group versus their mean (yes mean) annual income, the mode should roll right out.
at the very least it should give a pretty graph.
I would be VERY interested in seeing your results.
i'm working on it, I gotta adjust the scale of the graph cause it's one of those exponential distributions (just like Taleb said it would).
wtf there's 26,000 Retail Salespersons in California that earn a mean annual income of $489,140!! what's a Retail Salesperson? Didn't RWHN used to be one?
that's the maximum mean annual income occupation group per state. wonder who they are.
well I got something like a graph but it doesn't look right. there's a bunch of state-occupation-group that got an income of $0, that can't be right, otherwise they'd be unemployed.
and I gotta check on the other really really low numbers. maybe that's second jobs. you can't get that from this data, it only lists the numbers per job, not per person. ah well it's informational still.
Would mode income include just people who are currently working or would it be for everyone in the United States including unemployed, minors, and retired people?
Quote from: Iason Ouabache on January 19, 2010, 06:17:21 PM
Would mode income include just people who are currently working or would it be for everyone in the United States including unemployed, minors, and retired people?
Everyone eligible to work should be included, if you want the answers to mean anything. Minors, full-time students, and possibly illegals should not be counted. Retired people should be counted at their yearly income (if any), and unemployed people should be counted at zero.
well fuck me, I gotta graph:
(http://img4.imageshack.us/img4/7413/2008usincomedistributio.png)
if I understood it right, the highest peak should be the mode, yes? that one's the $25,000-$26,000 bin. every bin (graph step block thingy) is $1000 apart. so those other two peaks to the left of it are $22k-$24k.
I might get a more clear view if I'd make the bins bigger?
also, to comment on Iason's question, as long as the mode does not include unemployed, minors or retired people, meaning that in this case, there are no minors or unemployed that make $25k a year, you can drop those off the histogram and get the same mode. because the peak of the histogram is still at the same point. same reason why you can ignore all the people that make more than $100k, unless the most likely amount to make in a year is more than $100k, it's not gonna affect the mode. in that sense, it's even more robust against outliers than the median.
Cannot see graph at work. I will look at it when I get home.
and one more, with a bit larger bin size ($2000):
(http://img19.imageshack.us/img19/7413/2008usincomedistributio.png)
Roger, is it Imageshack? Maybe you can see the images if I host em here:
http://000.blackironprison.com/2008-US-income-distribution.png
http://000.blackironprison.com/2008-US-income-distribution-2.png
I gotta get to the kitchen and make me and my gf some dinner now.
Looks good. What is the numerical value of the most common wage? The image is a little blurry.
in the second graph, there's two bars highest, they represent $22k-$24k with 9.82 million jobs and $24k-$26k with 9.85 million jobs.
Quote from: Triple Zero on January 19, 2010, 07:23:14 PM
in the second graph, there's two bars highest, they represent $22k-$24k with 9.82 million jobs and $24k-$26k with 9.85 million jobs.
That's within $3K of the numbers I have seen, and the numbers I came up with.
That basically means that the average Joe is pulling down a whopping $11.50/hour. Compare that with the
average income, and you get an idea of the sheer grotesquery of the wealth discrepancy in America.
The funniest part? 50% of the people in this bracket argue strenously FOR this discrepancy, because they believe all of Horatio Alger's lies.
Speaking of Horatio Alger, he had the BEST TITLE FOR A NOVEL. EVER.
http://en.wikipedia.org/wiki/Ragged_Dick
Quote from: The Good Reverend Roger on January 19, 2010, 07:26:11 PM
Quote from: Triple Zero on January 19, 2010, 07:23:14 PM
in the second graph, there's two bars highest, they represent $22k-$24k with 9.82 million jobs and $24k-$26k with 9.85 million jobs.
That's within $3K of the numbers I have seen, and the numbers I came up with.
well as I said on the previous page, that graph is probably a reasonable estimate of the US 2008 mode of annual income, but not entirely correct. it makes the assumption "one occupation, one person" but that's not always true. there's people with two jobs, and there's other jobs that may be counted as full-time in the data but aren't ... I dunno.
here's hourly wages:
(http://000.blackironprison.com/2008-US-hourly.png)
http://000.blackironprison.com/2008-US-hourly.png
figures it's the same shape, just different scales on the axes. [I cut off a bit of the far end to zoom in on the lower numbers, we're not really interested in the distributions of the highest incomes, afterall]
You are all awesome, thankyou.
Quote from: Triple Zero on January 19, 2010, 06:38:25 PM
also, to comment on Iason's question, as long as the mode does not include unemployed, minors or retired people, meaning that in this case, there are no minors or unemployed that make $25k a year, you can drop those off the histogram and get the same mode. because the peak of the histogram is still at the same point. same reason why you can ignore all the people that make more than $100k, unless the most likely amount to make in a year is more than $100k, it's not gonna affect the mode. in that sense, it's even more robust against outliers than the median.
I feel stupid for not realizing that. That's what I get for having never taken a statistics class. That's right, I did 3 semester of calculus plus 2 of differential equations but no stats. Don't ask me how that ended up happening.
I wouldn't have taken stats either, but apparently psychology majors need it. For like, facts, or something. :roll:
Quote from: Felix on January 20, 2010, 06:51:44 AM
I wouldn't have taken stats either, but apparently psychology majors need it. For like, facts, or something. :roll:
I was a chemical engineering major. You'd think that someone would have noticed at some point that I hadn't taken statistics.
Isn't mean/median/mode something like 3rd grade math?
I know it was elementary school when I learned it. Of course, its probably not on the Test so why would they cover it anymore.
Also, I don't know about the school Iason went to, but at mine statistics was 'the easy class people who can't do math take'. (Business, pre law, and economics majors) if you were actually expected to know it they taught it taught it alongside the regular course (Thermodynamics and the labs in physics) or a dedicated course for the major.
Quote from: Iason Ouabache on January 20, 2010, 06:10:11 AM
Quote from: Triple Zero on January 19, 2010, 06:38:25 PM
also, to comment on Iason's question, as long as the mode does not include unemployed, minors or retired people, meaning that in this case, there are no minors or unemployed that make $25k a year, you can drop those off the histogram and get the same mode. because the peak of the histogram is still at the same point. same reason why you can ignore all the people that make more than $100k, unless the most likely amount to make in a year is more than $100k, it's not gonna affect the mode. in that sense, it's even more robust against outliers than the median.
I feel stupid for not realizing that. That's what I get for having never taken a statistics class. That's right, I did 3 semester of calculus plus 2 of differential equations but no stats. Don't ask me how that ended up happening.
I'm glad that you understood what I mean, because that was a fucking trainwreck of an explanation from me :)
and Requia, I only learned about the mean and the median in school. I never came across the mode before. Maybe in college statistics class but I don't quite recall, cause if they did they glossed over it [possibly assuming people would have learned it in high school hehe].