Small Sample
by Reg on Jun 10, 2007 12:18 PM PDT
63 comments
Why do so many people say small sample size? Just say small sample. You don't need "size" in there. When you state that a sample is small, it's clear that you are referring to the size of the sample. It's an extra word that doesn't need to be there.
Go Kotsay!
Go Michael!!
Go Blanton!
Go Ellis!
Go A's!
0 recs |
63 comments
Email
Print
Comments
recommended!
by ohad on Jun 10, 2007 12:35 PM PDT reply actions 0 recs
The diary, or therapy?
by Nico on Jun 10, 2007 12:37 PM PDT up reply actions 0 recs
The funniest thing
about the therapy suggestion is that it was made by the guy who likes to pretend he's a teenage girl on this site.
by Reg on Jun 10, 2007 4:48 PM PDT up reply actions 0 recs
Are there still people on this site
by Nico on Jun 10, 2007 6:39 PM PDT up reply actions 0 recs
Did you call her on your cellular banana?
by Reg on Jun 10, 2007 8:06 PM PDT up reply actions 0 recs
you know...
If theoretically there were a person on this site who had split male/female personalities, I support that theoretical person.
I used to help run a transgender support group when I lived in SF, and a lot of those folks live very hard lives and get abandoned/disowned by their families/friends, even in SF...
by The Pilots Dared Me To Die on Jun 11, 2007 9:22 AM PDT up reply actions 0 recs
Yess! The world needs more grammar nazis ...
... because as it is, we have so few friends.
by The Dogfather on Jun 10, 2007 1:00 PM PDT reply actions 0 recs
Not grammer
that's grammar.
by green star oakland on Jun 10, 2007 3:13 PM PDT up reply actions 0 recs
would you belive
by MobiusKlein on Jun 10, 2007 7:41 PM PDT up reply actions 0 recs
Well, it's not spelling, in which you get an
by The Dogfather on Jun 10, 2007 3:13 PM PDT reply actions 0 recs
Spelling grades
by Nico on Jun 10, 2007 7:48 PM PDT up reply actions 0 recs
uh, actually, "size" usually *is* necessary
Do we say "Larry Davis's man-boobs are protruding because of small shirt"?
No, we say "Larry Davis's man-boobs are protruding because of small shirt size."
Or, alternately, "Larry Davis's man-boobs are protruding because of the smallness of his shirt," or "... because his shirt is small."
by monkeyball on Jun 10, 2007 3:48 PM PDT reply actions 0 recs
Add "his" before small shirt
and the first option is the best.
by Reg on Jun 10, 2007 4:58 PM PDT up reply actions 0 recs
Still not grammar.
It's simianantics.
by The Dogfather on Jun 10, 2007 8:30 PM PDT up reply actions 0 recs
This thought provoking diary...
has been brought to you by:
Anheuser-Busch
by smith1051 on Jun 10, 2007 3:57 PM PDT reply actions 0 recs
And by the creators of
by Nico on Jun 10, 2007 4:11 PM PDT reply actions 0 recs
So THAT's what SSL stands for!
by oblique on Jun 10, 2007 7:14 PM PDT up reply actions 0 recs
I know of many adult stores
in Northern California that can help men with their small sample length problems!
by The Pilots Dared Me To Die on Jun 11, 2007 9:23 AM PDT up reply actions 0 recs
I think this diary is an homage to
by Englishmajor on Jun 10, 2007 8:55 PM PDT reply actions 0 recs
Or weather balloons.
by mlleaimee on Jun 10, 2007 9:05 PM PDT up reply actions 0 recs
important note
you might not need the word size after "small sample", but you do need the extra exclamation point after "Michael!"
by Cutthemullet on Jun 10, 2007 9:15 PM PDT reply actions 0 recs
So to offer an answer to the question ...
A small sample could still be sufficient - "We'll just take a small sample and send it to the laboratory" - but a sample suffering from small size is specifically insufficient to make reliable deductions from.
by green star oakland on Jun 10, 2007 10:56 PM PDT reply actions 0 recs
No, not the same thing
I'm all for better grammar and editing out unnecessary words, but in this case I think you're wrong.
"Sample" and "sample size" are both statistical terms. The sample is the subset of a population that is actually measured. The sample size is the number of measurements.
When you complain of a small sample, it suggests that you're drawing a conclusion about a group but you haven't looked at enough members of the group -- for example, if you conclude "baseball players like video games" after observing five of them who do but not observing the other 1,000+ who may or may not.
When you complain of a small sample size, it suggests that you're drawing a conclusion about an individual quality but you haven't looked at that quality often enough -- for example, if you conclude "Joe Kennedy is a good-looking guy" after observing two pictures in which he looks hot but not looking at any other pictures in which he may or may not.
Most baseball statistics are trying to measure the probability of something happening based on a collection of instances in which it did or didn't happen in the past. One might reasonably argue that this isn't exactly an issue of sample size, but it's a similar idea -- whereas it's really not an issue of insufficient sampling at all.
by iglew on Jun 11, 2007 2:41 AM PDT reply actions 0 recs
Yes, sample and sample size are different
But that's not my argument. Sample size is appropriate to use when the size of the sample has not yet been identified. But in both of the instances you have considered (using sample data to draw conclusions about populations or using sample data to make predictions about individual characteristics/performance), "small sample" is sufficient and "small sample size" is excessive.
by Reg on Jun 11, 2007 7:11 AM PDT up reply actions 0 recs
mdl is right
"Sample size" is a subset, not a redundancy, of "sample".
by Nico on Jun 11, 2007 8:19 AM PDT up reply actions 0 recs
I'm right
I think you're both missing the idea of the original post and arguing a different point altogether.
When one states that a conclusion about a population can't be made because of "small sample size," the word "size" is unnecessary. Even if someone asks you to describe the sample size (which is correct), the correct response is "it's a small sample," not "it's a small sample size."
Small means of limited size. If you say small sample size, it's like saying "of limited size sample size."
by Reg on Jun 11, 2007 9:20 AM PDT up reply actions 0 recs
Rice paddies...
A paddy is a field that is flooded to grow rice.
So, saying "Rice Paddy" is redundant...
Yet, since many people would not know that a paddy is a rice paddy, so you might as well use the extra syllable to ensure efficient communication.
Does the same thing apply here?
by The Pilots Dared Me To Die on Jun 11, 2007 9:25 AM PDT up reply actions 0 recs
You're right ...
that one does not need to say "small sample size", but it's not actually less efficient, because, in essentially any circumstance, in order to make the phrase gramatically correct you'll need to add at least one word.
eg:
"Because of small sample size" becomes:
"Because the sample is small" and you've used just as many words to convey the same message in a more ackward manner ... congratulations.
by devo on Jun 11, 2007 6:11 PM PDT up reply actions 0 recs
Fine
Add a word. That's not a problem. But don't add a redundant word (like size). And I certainly wouldn't use "because the sample is small." I would say "because of a small sample."
Also, if you are going to critique someone's use of language, your argument would be more convincing if you didn't have numerous spelling and punctuation errors in your post.
Congratulations!
by Reg on Jun 11, 2007 6:42 PM PDT up reply actions 0 recs
I wouldn't talk--
<lapses into a comma>
by Nico on Jun 11, 2007 6:50 PM PDT up reply actions 0 recs
It's not redundant ...
David Eckstein is a small sample ...
Willy Mo Pena, on the other hand, is a much larger sample, but any judgements about his season to date will suffer from issues related to small sample size.
While it's true that one does not necessarily need to add size, it helps to clarify that we are dealing with the second case and not the first. It's also a much clearer, easier understood way of writing/speaking, so one would have to conclude any efforts to the contrary are simply an exercise in being contrary.
by devo on Jun 11, 2007 7:45 PM PDT up reply actions 0 recs
Using your logic
one would also have to conclude from your user name that you are a big fan of a certain band.
When you are discussing statistics and referring to a sample in that context as small, there is no need to clarify that the size of the sample is being discussed. If I said that you can't draw a conclusion about a player's hitting ability from a small sample, 12 ABs, anyone with a statistics background would know what I meant.
by Reg on Jun 11, 2007 8:04 PM PDT up reply actions 0 recs
Billy Sample was 5' 9"
by Nico on Jun 11, 2007 8:12 PM PDT up reply actions 0 recs
Of course,
because that sentance includes a number of other words that clarify the meaning of "small sample" and, taken together, do make "size" extraneous.
by devo on Jun 12, 2007 9:15 AM PDT up reply actions 0 recs
So you
were arguing about all of the other ways small sample is used on Athletics Nation? Let's be real here. Almost every time there is a statement about a small sample on AN, it's in a discussion about statistics and prediction.
by Reg on Jun 12, 2007 9:57 AM PDT up reply actions 0 recs
This is dumb ...
by devo on Jun 12, 2007 10:02 AM PDT up reply actions 0 recs
Disambiguate
somewhat redundant words are OK if they work to resolve any potential ambiguity in the sentence.
Many times I write text that seems obvious to me, but is really unclear to most readers. If folks use a few extra words and by so doing, become more clear, what is the fault?
(And yes, sometimes fewer words are more clear too. But work with me, please!)
by MobiusKlein on Jun 11, 2007 9:53 PM PDT up reply actions 0 recs
If your point is
that "small sample" is sufficient to get the idea across, well, OK. Bt 1 cn lv out svrl ltrs fm a sntc & stl gt t pt cros 2. I wouldn't say that makes it excessive to spell out every word properly.
It's kind of a question of what you're measuring. The way I see it, what we're measuring is the player's current ability to do something. Take OBP for an example. We're assuming that a given player has an inherent % chance of getting on base. The only way we have to measure that is to look at an at-bat, which will yield an answer of 1 or 0. If we take this measurement often enough, we can average them out and get a number with some predictive power.
Since the at-bats are the measurement, this is a matter of sample size, not sample.
But suppose there's some high school kid you're scouting and you want to know what his OBP was in high school. Maybe they didn't keep stats very well and you can't get all the information. If you calculate an OBP from the numbers you do have, I would say your small sample is why you haven't accurately measured his OBP, and then the small sample size is why the OBP doesn't necessarily measure how likely he is to get on base next time.
As I said before, neither term exactly matches what we're talking about with baseball stats, so there is no 100% right or wrong answer here. If you and others start saying "small sample", I'm not going to misunderstand nor complain that you're wrong. But you're questioning our use of "sample size", suggesting that it is redundant, and I don't think you've made your point. It's not redundant; it's indicative of a certain way of viewing the data.
by iglew on Jun 11, 2007 11:29 AM PDT up reply actions 0 recs
Once again
I'm not questioning the use of "sample size" unless it is preceded by the word "small." Once you identify a sample as small, it is no longer necessary to use the word "size."
It appears that you believe "small sample" is appropriate in some instances and "small sample size" is appropriate in others. I disagree. "Small sample" should be used in both instances.
by Reg on Jun 11, 2007 6:52 PM PDT up reply actions 0 recs
Well, we just disagree then
I don't see how you can acknowledge that "sample size" is a precisely defined term in and of itself, and then say that the "size" half of it becomes unnecessary just because a modifier is attached.
by iglew on Jun 12, 2007 12:41 AM PDT up reply actions 0 recs
No
In statistics, sample is a precisely defined term. Size is one aspect of a sample.
by Reg on Jun 12, 2007 6:11 AM PDT up reply actions 0 recs
is this still going on?
Stop arguing. The simple comment Reg made in the diary is correct. You all know it, but for some reason he fires all of you up into argue mode. Seeing the 'size' at the end (or maybe just that it became a cliche immutable phrase) can get annoying. Sometimes, it is necessary. Often it is not.
by dscel on Jun 12, 2007 8:18 AM PDT up reply actions 0 recs
I thought for sure this was a Rashid diary
by ArakSOT on Jun 11, 2007 9:13 AM PDT reply actions 0 recs
what about Sample size...i mean
you know it's small right?!
by griffisgurl on Jun 11, 2007 11:00 AM PDT reply actions 0 recs
Not on the internets, it's not.
They're all BIG on the web.
by The Dogfather on Jun 11, 2007 11:12 AM PDT up reply actions 0 recs
You're all wrong.
Joe Kennedy is just plain hot.
by blueconversechucks on Jun 11, 2007 11:14 PM PDT reply actions 0 recs
Hurray!
Random grammar arguments! Yessss!
by CTAsGirl on Jun 12, 2007 5:40 AM PDT reply actions 0 recs
Just put my personal identification number number
into the automated teller machine machine.
Just before heading to my class at the school of redundancy school.
by JediLeroy on Jun 12, 2007 8:49 AM PDT reply actions 0 recs
Ah, JediLeroy
The voice of reason. I'm glad somebody gets it. Great name, too.
by Reg on Jun 12, 2007 9:59 AM PDT up reply actions 0 recs
Thanks!
It's funny how I often get dismissed as a fanboy nerd.
by JediLeroy on Jun 12, 2007 12:45 PM PDT up reply actions 0 recs
Yes, but you are an alter-ego
pretending to be a streetwise tough guy. Now, if you can add some brilliant fiction masquerading as reality, you'll have it nailed. And (as I understand it) duped celebrities will follow in droves.
by FreeSeatUpgrade on Jun 12, 2007 12:48 PM PDT up reply actions 0 recs
Nobody even suspects
by JediLeroy on Jun 12, 2007 1:18 PM PDT up reply actions 0 recs
hey, baby...
there's a huge niche in the adult film market for a 600lb Mongolian woman.
I'd get into all the juicy details, but that would be going too far.
However, I've got a great slogan to promote the movie we shoot:
"One 600lb Mongolian woman... Nine midgets!"
by The Pilots Dared Me To Die on Jun 12, 2007 1:23 PM PDT up reply actions 0 recs
Here's
600 lbs of Mongolian woman. Click at your own risk.
by The Dogfather on Jun 12, 2007 2:50 PM PDT up reply actions 0 recs
Upon further review....
I happen to have a friend who is a professor of statistics at an Ivy League school (and lead author of a textbook on Bayesian data analysis). Out of curiosity, I posed the question to him to see what he thinks.
His response is that he doesn't understand my "attempted distinction" between small sample and small sample size since they're basically the same thing. When I pressed him on the exact question, he said that, yes, he would agree that "small sample size" is a redundant phrase.
So I guess my argument earlier in the thread was really just a case of me attempting to split hairs that even academic professional statisticians don't care to split.
I don't promise that I won't ever say "small sample size" again, but I'm ready to retract my argument and concede that on the question of redundancy, I was wrong and Reg was right.
by iglew on Jun 12, 2007 1:38 PM PDT reply actions 0 recs
Thanks for the update, mdl
I also teach statistics at a university, but I am certainly no match for your friend. If anyone wants to meet me for a beer and talk statistics at a game, let me know. Wait, that would be really lame. Let's just get a beer. Go A's!
by Reg on Jun 12, 2007 1:55 PM PDT up reply actions 0 recs
C'mon, you're really a ..
Supermodel , aren't you?
by The Dogfather on Jun 12, 2007 2:55 PM PDT up reply actions 0 recs
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2009 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
