graa statistics!

I assume this is for a class. What course is it for? It probably wouldn''t help to give you an intro to bio stats book recommendation if you''re taking advanced industrial statistics. (Though, I find that industrial statistics books tend to give explanations in words more often than other stats books.)

Yes, it's for a class Statistics for Science and Engineering.

This damn book keep using symbols all the time. Check this out, exactly from the book, Chapter 8.1, right in the beginning of Chapter 8:

A statistical hypothesis, or just hypothesis , is a claim or assertion either about the value of a single parameter (population characteristic or characteristic of a probability distribution), about the values of several parameters, or about the form of an entire probability distribution. One example of a hypothesis is the claim u = .75, where u is the true average inside diameter of a certain type of PVC pipe. Another example is the statemet p < .10, where p is the proportion of defective circuit boards among all circuit boards produced by a certain manufacturer. If u1 and u2 denote the true average breaking strengths of two different types of twine, one hypothesis is the assertion that u1 - u2 = 0, and another is the statement u1 - u2 > 5.

Just tell me...tell me.. why did you choose .75, why p < .10, why u1 - u2 = 0, and why u1 - u2 > 5, where in the hell did you get 5 from, and what are you talking about!! You examples don't make any sense!

[edited by - alnite on April 16, 2004 7:32:53 PM]

Oh geez...I could probably at least attempt that if my eyes didn''t feel like they were gushing blood.

That''s not a very good book, me says. Poorly written, it looks like.

I wonder if Mathworld has a stats section...why, yes it does.

---

capn_midnight | Captain Midnight | deviantArt
ACM | SIGGRAPH | Generation5

quote:

---

Original post by alnite
A statistical hypothesis, or just hypothesis , is a claim or assertion either about the value of a single parameter (population characteristic or characteristic of a probability distribution), about the values of several parameters, or about the form of an entire probability distribution. One example of a hypothesis is the claim u = .75, where u is the true average inside diameter of a certain type of PVC pipe. Another example is the statemet p < .10, where p is the proportion of defective circuit boards among all circuit boards produced by a certain manufacturer. If u1 and u2 denote the true average breaking strengths of two different types of twine, one hypothesis is the assertion that u1 - u2 = 0, and another is the statement u1 - u2 > 5.

Just tell me...tell me.. why did you choose .75, why p < .10, why u1 - u2 = 0, and why u1 - u2 > 5, where in the hell did you get 5 from, and what are you talking about!! You examples don''t make any sense!

---

Unless there''s extra context that you haven''t included in your excerpt, it looks like all of those are just examples, meant to show you what "hypothesis" means when used in a technical, mathematical context. Assuming they are just examples, the numbers don''t mean anything, they''ve just been made up as part of the example.

John B

quote:

---

Original post by JohnBSmall

quote:

---

Original post by alnite
A statistical hypothesis, or just hypothesis , is a claim or assertion either about the value of a single parameter (population characteristic or characteristic of a probability distribution), about the values of several parameters, or about the form of an entire probability distribution. One example of a hypothesis is the claim u = .75, where u is the true average inside diameter of a certain type of PVC pipe. Another example is the statemet p < .10, where p is the proportion of defective circuit boards among all circuit boards produced by a certain manufacturer. If u1 and u2 denote the true average breaking strengths of two different types of twine, one hypothesis is the assertion that u1 - u2 = 0, and another is the statement u1 - u2 > 5.

Just tell me...tell me.. why did you choose .75, why p < .10, why u1 - u2 = 0, and why u1 - u2 > 5, where in the hell did you get 5 from, and what are you talking about!! You examples don''t make any sense!

---

Unless there''s extra context that you haven''t included in your excerpt, it looks like all of those are just examples, meant to show you what "hypothesis" means when used in a technical, mathematical context. Assuming they are just examples, the numbers don''t mean anything, they''ve just been made up as part of the example.

---

Right, and I figured it out about a couple of minutes after posting this. But can you still see how the author uses the symbols? Can''t he just say "10% of all circuit boards are malfunctioning", instead of "Let p be the number of defective boards, p < .10"

quote:

---

Original post by alnite
]Right, and I figured it out about a couple of minutes after posting this. But can you still see how the author uses the symbols? Can''t he just say "10% of all circuit boards are malfunctioning", instead of "Let p be the number of defective boards, p < .10"

---

keeping consistent terminology is important. once you get used to it, it gets much faster.

---

capn_midnight | Captain Midnight | deviantArt
ACM | SIGGRAPH | Generation5

Here''s the bad news: That is about as clear as it gets when dealing with advanced statistics. You can get clearer texts for lower level stats books, but they probably won''t be able to help you out with your current course.

And yes, I''ve seen worse in terms of pure math symbol overload.

quote:

---

Original post by alnite
Can''t he just say "10% of all circuit boards are malfunctioning", instead of "Let p be the number of defective boards, p < .10"

---

Taking this example and running with it, p < .10 is not the same as saying "10% of all circut boards are malfunctioning". The hypothesis is that the total proportion of circuit boards that are defective is less than .10, Which is distinct from the proportion of defective circuit boards being equal to .10. By using mathematical notation it''s easier to differentiate the two situations. And the tests of significance for equal to and a less than are fairly different animals.

In the end, questions you are going to get will be like: "S = { .... } what is the significance level for p < .10." You won''t care about if p is the proportion of defective circuit boards or number of people who prefer Pepsi to Coke. Basically by having < and = instead of saying the hypothesis in pure english, you make it clearer what you''re going after.

Put another way, the hypothesis is "p < .10" and the rest of the text is just window dressing.

You want symbols? Try discrete math sometime. Blagh. Often, the board didn''t have a single word on it - just a set of symbols, numbers, and letters, that was occaisionally, maybe almost comprehensible.

Once you learn the symbols, however, it makes it much more straightforward and clear - it''s just the initial learning curve that hurts. The textbook was pretty good, too.

I wrote my final this morning, and I am glad to be done with it.

discrete structures/math is absolutely the most mind-numbingly fucking stupid pointless bullshit class you will ever take in college, ever.

i seriously dragged my ass through that class kicking and screaming and in the end, still have no clue what the hell that class was all about, or how the hell i have a better understanding of computer science now.