0892: "Null Hypothesis"
Moderators: Moderators General, Prelates, Magistrates
0892: "Null Hypothesis"
Alt: Hell, my eighth grade science class managed to conclusively reject it just based on a classroom experiment. It's pretty sad to hear about milliondollar research teams who can't even manage that.
There he goes again, preaching Null Theory.
Look at me, still talking when there's Science to do.
 Time Kitten
 Posts: 28
 Joined: Mon Jul 19, 2010 4:38 am UTC
 Location: Behind you!
 Contact:
Re: 0892: "Null Hypothesis"
Eight grade science? isn't that the year they revisit the brontosaurs? Though of course, second grade is where all the real serious study of paleontology is done.
Re: 0892: "Null Hypothesis"
Is it just me or are there a lot of statistics based comics lately? I wonder why...
"I think nature's imagination is so much greater than man's, she's never gonna let us relax."
 RebeccaRGB
 Posts: 336
 Joined: Sat Mar 06, 2010 7:36 am UTC
 Location: Lesbians Love Bluetooth
 Contact:
Re: 0892: "Null Hypothesis"
Where are these kids going to school? I didn't learn about Null Hypotheses until college. And then I made a comic about it five years before Randall:
Stephen Hawking: Great. The entire universe was destroyed.
Fry: Destroyed? Then where are we now?
Al Gore: I don't know. But I can darn well tell you where we're not—the universe!
Fry: Destroyed? Then where are we now?
Al Gore: I don't know. But I can darn well tell you where we're not—the universe!
Re: 0892: "Null Hypothesis"
I once asked out a statistician.
She failed to reject me <3
She failed to reject me <3
President of the peoples republik of the internet.
screw your coffee, i download my java!
screw your coffee, i download my java!
 TappingTheLine
 Posts: 14
 Joined: Fri Nov 02, 2007 5:07 pm UTC
Re: 0892: "Null Hypothesis"
I don't get it. Maybe it's because I just came back from my tabletop RPG where my character is bad at think hard and I haven't gotten my brain started back up yet, but I don't get what the punchline is. Is this comic just punchlineless preachyness?
"My parties are clothing optional, if you are offended by casual nudity I'm not entirely sure how you manage to get along with me...." chaosreigns.com/quotes

 Posts: 87
 Joined: Sun Apr 10, 2011 10:17 pm UTC
 Contact:
Re: 0892: "Null Hypothesis"
I had never heard of the Null Hypothesis before today. >.< Time to go Wikiing!
Re: 0892: "Null Hypothesis"
I'm assuming the punchline here is that he heard in a study somewhere that the 'Null Hypothesis' (of that study) was disproved. It's a classic comic of ignorance and misunderstanding.
Re: 0892: "Null Hypothesis"
TappingTheLine wrote:I don't get it. Maybe it's because I just came back from my tabletop RPG where my character is bad at think hard and I haven't gotten my brain started back up yet, but I don't get what the punchline is. Is this comic just punchlineless preachyness?
Well, the punchline is debatable in whether it is good, but it is there. This one isn't preachy at all.
In statistical analysis of experiments, you'll see two phrases crop up quite a bit. "Reject the null hypothesis" and "fail to reject the null hypothesis." If you reject the null hypothesis, then you have enough evidence to think that your hypothesis is true (with some margin of error). If you fail to reject it, then you do not have enough evidence from this experiment to think that your hypothesis is true (with some margin of error).
In very early labs for science classes you'll often conduct basic experiments where you "reject the null hypothesis." So here, we see the guy talking about how they "rejected the null hypothesis years ago," treating it as a static hypothesis across all experiments that can simply be refuted.
 phlip
 Restorer of Worlds
 Posts: 7572
 Joined: Sat Sep 23, 2006 3:56 am UTC
 Location: Australia
 Contact:
Re: 0892: "Null Hypothesis"
TappingTheLine wrote:I don't get it. Maybe it's because I just came back from my tabletop RPG where my character is bad at think hard and I haven't gotten my brain started back up yet, but I don't get what the punchline is. Is this comic just punchlineless preachyness?
The null hypothesis is a statistical method... it comes from the fact that "X is likely" is difficult to prove directly, but "X is unlikely" is easier to formalise. So instead of proving "X is likely", you prove "notX is unlikely". The terminology is that "notX is likely" is the Null Hypothesis, and in order to prove what you're actually after, you work to disprove the null hypothesis. If you succeed, you've proven that X is likely... and if you fail, you haven't proven anything... you haven't proven that notX is likely, you just haven't been able to rule it out. Hence Rebecca's comic above  failing to reject the null hypothesis... failing to prove that "I am wrong" is wrong.
So the joke in the comic is a variation on an old joke... a child is doing maths homework and goes to their parent and says "I need some help with my homework, I need to find the lowest common denominator." And the parent responds "Really? They were looking for that back when I was in school, haven't they found it yet?" Just with the LCD replaced with a more advanced maths concept.
That is, the joke is in a character confusing one instance of a concept for the concept as a whole  confusing "the lowestcommondenominator in this puzzle" with "the lowestcommondenominator in general" or "the null hypothesis of this experiment" with "the null hypothesis in general".
Code: Select all
enum ಠ_ಠ {°□°╰=1, °Д°╰, ಠ益ಠ╰};
void ┻━┻︵╰(ಠ_ಠ ⚠) {exit((int)⚠);}

 Posts: 68
 Joined: Thu Apr 15, 2010 5:46 pm UTC
Re: 0892: "Null Hypothesis"
This one had quite a few chuckles from me. A part of me wonders if my adviser would think it was funny if I said this... but then I remember that I am supposed to be adept at this by now and he would probably just become terrified.
If only I didn't care what he and my committee thought.
If only I didn't care what he and my committee thought.
"God does not play dice with... Yahtzee!"  Little known quote from Einstein
Re: 0892: "Null Hypothesis"
I wouldn't have appreciated it much a year ago, but I've studied so much statistics lately that it's hilarious.
If nothing else, it should shut up the "remember when xkcd was funny?" crowd. Assuming they're part of the target demographic of this particular one.
If nothing else, it should shut up the "remember when xkcd was funny?" crowd. Assuming they're part of the target demographic of this particular one.

 Posts: 37
 Joined: Wed May 20, 2009 8:48 pm UTC
Re: 0892: "Null Hypothesis"
He's just trying to increase his counternull value.
Re: 0892: "Null Hypothesis"
the Null Hypoithesis basically means you assume there's 'nothing interesting here': There's no conclusive evidence that the thing you're testing for does anything / pushes anything in the direction you want it to. You generally reject it when your data suggests results this odd (assuming the null hypothesis) are less than 5% likely (called '95% confidence level', 95% and 99% are common). This does NOT suggest your alternative hypothesis is 95% likely to be correct. It could be a fluke, it could depend on something completely different that is just correlated to (tends to occur at the same time as) the thing you're testing for.
The comic works in several ways: Highlighting common ignorance about terms very important to how technology/science move on; this should arguably be taught to everyone at school. A particular null hypothesis is rejected in any study that has an interesting result.
It could also be criticism of a perceived trend that threatens sound science: Big industries (and some political / religious groups) are accused of throwing money into big studies (statistics works best with big numbers of test cases, preferably you also want to test against expected sources of error... that's a lot of work, so bigger is better) until they have the results they want. The comic takes it one step further: A 'big study' has disproved the concept of the Null hypothesis and killed sound science as we know it.
The comic works in several ways: Highlighting common ignorance about terms very important to how technology/science move on; this should arguably be taught to everyone at school. A particular null hypothesis is rejected in any study that has an interesting result.
It could also be criticism of a perceived trend that threatens sound science: Big industries (and some political / religious groups) are accused of throwing money into big studies (statistics works best with big numbers of test cases, preferably you also want to test against expected sources of error... that's a lot of work, so bigger is better) until they have the results they want. The comic takes it one step further: A 'big study' has disproved the concept of the Null hypothesis and killed sound science as we know it.
Last edited by Iranon on Fri Apr 29, 2011 6:14 am UTC, edited 1 time in total.
LEGO won't be ready for the average user until it comes preassembled, in a single unified theme, and glued together so it doesn't come apart.
Re: 0892: "Null Hypothesis"
It's rare that i dont get an xkcd at all, and this was one of those. it's common that the comics contain nuances i miss, so i still come here anyway to check. statistics was so dull i dropped it two weeks in. it is a handy set of tools that i regret not knowing.
once the joke was explained, i chortled. (my ewe key sticks so i'm having to switch for words like 'chortled'.)
http://www.explainxkcd.com/ is too slow, fora folks have it first.
the webcomic i ran into today and read front to back was http://www.sandraandwoo.com.
once the joke was explained, i chortled. (my ewe key sticks so i'm having to switch for words like 'chortled'.)
http://www.explainxkcd.com/ is too slow, fora folks have it first.
the webcomic i ran into today and read front to back was http://www.sandraandwoo.com.
Re: 0892: "Null Hypothesis"
On reading the first sentence I was expecting a serious objection to NullHypothesis testing as a method of discovering knowledge, there being some legitimate complaints with it.
But then it turned into a funny comic.
But then it turned into a funny comic.
I don't think you realize that for me this is real.
 Plasma Mongoose
 Posts: 213
 Joined: Tue Feb 01, 2011 1:09 am UTC
 Contact:
Re: 0892: "Null Hypothesis"
So I take it that Null Hypothesis is NOT related to the Double Blind testing then?
A virus walks into a bar, the bartender says "We don't serve viruses in here".
The virus replaces the bartender and says "Now we do!"
The virus replaces the bartender and says "Now we do!"
Re: 0892: "Null Hypothesis"
Mavrisa wrote:Is it just me or are there a lot of statistics based comics lately? I wonder why...
I'd say it's just coincidence, but apparently today that argument isn't valid
Re: 0892: "Null Hypothesis"
It can be, depending on the experiment. In fact, any sort of testing/experiment should have a null hypothesis, including double blind tests.Plasma Mongoose wrote:So I take it that Null Hypothesis is NOT related to the Double Blind testing then?
Re: 0892: "Null Hypothesis"
Aw yes, I remember when I rejected my first null hypothesis.
Those were the days.
Those were the days.
Check out my web interzones powered by Web 3.0 technology running on Mozzarella Foxfire:
Spoiler:
Re: 0892: "Null Hypothesis"
Ahh, null hypotheses. I always thought statistics and methodology should have been the two most interesting, and almost most important, classes in my social studies. Unfortunately our faculty disagreed and made them boring, dense, uninteresting and incomprehensible. They also dropped basic philosophy from the curriculum, giving us more time for marxism and postmodernism ...

 Posts: 4
 Joined: Wed Nov 26, 2008 5:06 am UTC
Re: 0892: "Null Hypothesis"
Heh.. I have another one to bring in to class and show my stats professor. Both this and the "Jelly beans cause acne!" strip were posted immediately before I took a stats test, too. I'd need a larger sample size of stats comics to show if the timing is statistically significant though.
 WolvenSpectre
 Posts: 2
 Joined: Fri Apr 29, 2011 11:09 am UTC
 Location: Saskatchewan, Canada
Re: 0892: "Null Hypothesis"
WOW... I made it to halfway through my first year of Engineering (had to leave because of money/health issues) and graduated high school with a couple of University Credits and so many High school credits that they had to redesign the size of the transcript forms and I had never heard of this. We usually are proud of our education here in Canada, but talk about a FAIL... The terms we used to use were Default Position, Subject Hypothesis, or just plain proposed stance.
Mind you statistics was not a huge part of the mathematical curriculum and we weren't even offered Calculus in my High School. University level Computer Science and Physical Education, but no Secondary Level Calculus.
Mind you statistics was not a huge part of the mathematical curriculum and we weren't even offered Calculus in my High School. University level Computer Science and Physical Education, but no Secondary Level Calculus.
Re: 0892: "Null Hypothesis"
Everything is connected at the spiritual level, dude. The null hypothesis is false.

 Posts: 201
 Joined: Tue Nov 03, 2009 6:16 am UTC
Re: 0892: "Null Hypothesis"
achan1058 wrote:It can be, depending on the experiment. In fact, any sort of testing/experiment should have a null hypothesis, including double blind tests.Plasma Mongoose wrote:So I take it that Null Hypothesis is NOT related to the Double Blind testing then?
They are correlated, in that both are part of a good experiment involving people, but they are not causally related. Whether or not the determination of "success" in an experiment uses the Null Hypothesis does not tell you if it was doubleblind, though it does increase the likelihood of it being doubleblind.
Doubleblind testing is testing where neither experimenter nor subject knows the nature of the thing being tested (whether it's a placebo or the real medicine, whether the washing powder is brand X or Y, etc.). This is mostly limited to testing of humans, and used to remove the possibility of the experimenter accidentally informing the subject. Only afterwards are the people involved informed. (Commonly, there is also a third party who does nothing but receive information about the experiment, and can stop it if the lack of information causes somebody's life to be unduly at risk).
The Null hypothesis is the standard hypothesis that the result of the experiment is null  that there is nothing out of the ordinary (that the medicine works no better than the placebo, that the currently existing mathematical models describe a phenomenon correctly, etc.). The experimenter needs to prove, with some level of certainty, that the Null hypothesis is wrong: that the experiment can't be explained by what was known before the experiment. Here, for example, the Null hypothesis is placed extraordinarily high  since no model exists where one may fall down the appeal chute, the observation of one falling down the appeal chute is rejected since it can be explained by previous models (weather balloon).
Re: 0892: "Null Hypothesis"
Reminds me of review graduate students research reports. The abstract of one red as follows:
This research study will explore reasons for truency within the afroamerican religious community using the case study research methodology... The study failed to reject the null hypothesis.
WTF?
This research study will explore reasons for truency within the afroamerican religious community using the case study research methodology... The study failed to reject the null hypothesis.
WTF?
I will not attack your math, just your epistemology.
You think you have it bad, I teach Intro to Project Management to Undergrads.
You think you have it bad, I teach Intro to Project Management to Undergrads.
Re: 0892: "Null Hypothesis"
I rejected the Null hypothesis in favour of the Aleph Null hypothesis.
There just seems to be more to it.
There just seems to be more to it.
Re: 0892: "Null Hypothesis"
I actually get this one and find it clever.
"The future is the only kind of property that the masters willingly concede to the slaves"  Albert Camus

 Posts: 105
 Joined: Mon Aug 20, 2007 11:30 am UTC
Re: 0892: "Null Hypothesis"
On a side note, is it normal that I've always loved and been interested in Probabilities but always hated Statistics?
Re: 0892: "Null Hypothesis"
TappingTheLine wrote:I don't get it. Maybe it's because I just came back from my tabletop RPG where my character is bad at think hard and I haven't gotten my brain started back up yet, but I don't get what the punchline is. Is this comic just punchlineless preachyness?
To simplify what others have said, the null hypothesis is something that's different every time.
A similar joke:
An algebra student's homework wrote:Solve for x.
1. x+2=3x+4
x3x=42
2x=2
x=1
2. 5x4=3x2
5(1)4=3(1)2
9=5? What gives?
Not named Dennis Miller.
Re: 0892: "Null Hypothesis"
SolkaTruesilver wrote:On a side note, is it normal that I've always loved and been interested in Probabilities but always hated Statistics?
No. It's a tdistribution with 46 degrees of freedom.
Re: 0892: "Null Hypothesis"
phlip wrote:TappingTheLine wrote:I don't get it. Maybe it's because I just came back from my tabletop RPG where my character is bad at think hard and I haven't gotten my brain started back up yet, but I don't get what the punchline is. Is this comic just punchlineless preachyness?
The null hypothesis is a statistical method... it comes from the fact that "X is likely" is difficult to prove directly, but "X is unlikely" is easier to formalise. So instead of proving "X is likely", you prove "notX is unlikely". The terminology is that "notX is likely" is the Null Hypothesis, and in order to prove what you're actually after, you work to disprove the null hypothesis. If you succeed, you've proven that X is likely... and if you fail, you haven't proven anything... you haven't proven that notX is likely, you just haven't been able to rule it out. Hence Rebecca's comic above  failing to reject the null hypothesis... failing to prove that "I am wrong" is wrong.
So the joke in the comic is a variation on an old joke... a child is doing maths homework and goes to their parent and says "I need some help with my homework, I need to find the lowest common denominator." And the parent responds "Really? They were looking for that back when I was in school, haven't they found it yet?" Just with the LCD replaced with a more advanced maths concept.
That is, the joke is in a character confusing one instance of a concept for the concept as a whole  confusing "the lowestcommondenominator in this puzzle" with "the lowestcommondenominator in general" or "the null hypothesis of this experiment" with "the null hypothesis in general".
Thanks
Title: It was given by the XKCD moderators to me because they didn't care what I thought (I made some rantings, etc). I care what YOU think, the joke is forums.xkcd doesn't care what I think.

 Posts: 3
 Joined: Fri May 15, 2009 7:35 pm UTC
Re: 0892: "Null Hypothesis"
I think people are missing the logical paradox portion of the punch line. That is, you have a big study trying to disprove the null hypothesis, so they have a null hypothesis along the lines that that rejecting the null hypothesis is meaningful and, according to the strip, that null hypothesis was rejected which means that rejecting the null hypothesis is not meaningful, which means that the method they used to disprove the null hypothesis was not meaningful, which makes the whole study meaningless.
I also agree with the other analysis that there's also the conflation of various null hypotheses, but I think that one is a bit more obvious because of the clarifying alt text.
I also agree with the other analysis that there's also the conflation of various null hypotheses, but I think that one is a bit more obvious because of the clarifying alt text.
Re: 0892: "Null Hypothesis"
lly wrote:So here, we see the guy talking about how they "rejected the null hypothesis years ago," treating it as a static hypothesis across all experiments that can simply be refuted.
Ah, that was it. Thank you! Yes, it is mildly funny. The alttext is basically the same.
 SpringLoaded12
 Posts: 350
 Joined: Wed Oct 08, 2008 1:58 am UTC
 Location: Guarding the Super Missile
 Contact:
Re: 0892: "Null Hypothesis"
The girl at the desk has interesting hair.
I didn't get this one until after looking up the Null Hypothesis.
Somewhat unrelatedly, does anyone else often find that if they don't get one of the jokes in the strips at first because they don't know the concepts that make the joke, that within a few weeks they are taught those concepts and thus can get the joke?
For example, a long time ago there was a chart showing the paths of various fictional hurricanes, each of which was named based on its path. One of them was called Hurricane cos(x), and its path was a cosine wave, but I did not understand the joke until a week or two later when I was taught how to graph trigonometric functions.
I didn't get this one until after looking up the Null Hypothesis.
Somewhat unrelatedly, does anyone else often find that if they don't get one of the jokes in the strips at first because they don't know the concepts that make the joke, that within a few weeks they are taught those concepts and thus can get the joke?
For example, a long time ago there was a chart showing the paths of various fictional hurricanes, each of which was named based on its path. One of them was called Hurricane cos(x), and its path was a cosine wave, but I did not understand the joke until a week or two later when I was taught how to graph trigonometric functions.
"It's easy to forget what a sin is in the middle of a battlefield." "Opposite over hypotenuse, dipshit."

 Posts: 404
 Joined: Sun Nov 22, 2009 10:23 pm UTC
Re: 0892: "Null Hypothesis"
Stellazira wrote:I had never heard of the Null Hypothesis before today. >.< Time to go Wikiing!
Seconded. Were you a victim of Catholic schools as well? As to why it took so long and cost so much to disprove it, Government Money.
I am Lord Titanius Englesmith, Fancyman of Cornwood.
See 1 Kings 7:23 for pi.
If you put a prune in a juicer, what would you get?
See 1 Kings 7:23 for pi.
If you put a prune in a juicer, what would you get?
Re: 0892: "Null Hypothesis"
Baron Sloth wrote:I think people are missing the logical paradox portion of the punch line. That is, you have a big study trying to disprove the null hypothesis, so they have a null hypothesis along the lines that that rejecting the null hypothesis is meaningful and, according to the strip, that null hypothesis was rejected which means that rejecting the null hypothesis is not meaningful, which means that the method they used to disprove the null hypothesis was not meaningful, which makes the whole study meaningless.
I don't think that's an intended reading. Even if he had read a big study which set out to disprove the validity of nullhypothesis testing methodology, nothing says that study itself used nullhypothesis testing methodology, and it wouldn't make any sense to assume it did unless something in the comic said so.
But I don't think the point is that he read a big study which set out to disprove the validity of nullhypothesis testing methodology at all; I think the point is that he read a big study which disproved it's null hypothesis, and misinterprets that to mean that there is one specific hypothesis called The Null Hypothesis which has been disproven and yet is still discussed in schools, rather than different null hypotheses being tested in each experiment.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy)  The Chronicles of Quelouva (my fiction)
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy)  The Chronicles of Quelouva (my fiction)
Re: 0892: "Null Hypothesis"
dcollins wrote:SolkaTruesilver wrote:On a side note, is it normal that I've always loved and been interested in Probabilities but always hated Statistics?
No. It's a tdistribution with 46 degrees of freedom.
>46 degrees of freedom
God help us all.
Check out my web interzones powered by Web 3.0 technology running on Mozzarella Foxfire:
Spoiler:

 Posts: 5
 Joined: Mon Mar 02, 2009 5:14 am UTC
 Location: New York
Re: 0892: "Null Hypothesis"
IIAOPSW wrote:I once asked out a statistician.
She failed to reject me <3
That is so clever and cute!
Sometimes the comments here are better than the comic itself, in my opinion.
 bmonk
 Posts: 662
 Joined: Thu Feb 18, 2010 10:14 pm UTC
 Location: Schitzoed in the OTT between the 2100s and the late 900s. Hoping for singularity.
Re: 0892: "Null Hypothesis"
Once in philosophy class we disproved the Null Hypothesis. So now I know the world means something.
The question now is what does it mean?
I'm just afraid that the Universal Correlation is an inverse correlation.
The question now is what does it mean?
I'm just afraid that the Universal Correlation is an inverse correlation.
Having become a Wizard on n.p. 2183, the Yellow Piggy retroactively appointed his honorable self a Temporal Wizardly Piggy on n.p.1488, not to be effective until n.p. 2183, thereby avoiding a partial temporal paradox. Since he couldn't afford two philosophical PhDs to rule on the title.
Return to “Individual XKCD Comic Threads”
Who is online
Users browsing this forum: Alsn, AndrewGPaul, DwayneSa and 48 guests