2008 First Semester STATS课程官方问答贴

dk7

原帖由 水鱼皇后 于 2008-4-22 01:24 发表
18*10.5%...

就是用死亡率per thousand那个column 乘percentage那8个columns...得出来再做PCA？

Spot on! ; )

z-score

原帖由 阴雨天 于 2008-4-21 23:17 发表
给你10个数 算standard division 怎么算来着? 高中学过,忘的一干二净~
谢谢了 阿童木~

如果你用Casio fx-82计算器的话(新西兰高中经典黑计算器)，下面是计算standard deviation的步骤：

1. Press MODE button and select SD.
2. Press SHIFT MODE button and select STAT CLEAR then pressen = button to clear the stats memory.
3. If this is your data: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. Then presse 1 then M+ (directly above the ON button), then press 2 then M+ and so on and so forth.
4. Press SHIFT 2 (or 1 depending on your model) then 1 to get the mean, press SHIFT 2 then 3 to get the sample standard deviation.

阴雨天

原帖由 z-score 于 2008-4-22 08:21 发表


如果你用Casio fx-82计算器的话(新西兰高中经典黑计算器)，下面是计算standard deviation的步骤：

1. Press MODE button and select SD.
2. Press SHIFT MODE button and select STAT CLEAR then press ...

谢谢!! 另一个问题 怎样在学校的word 2007里输入那些希腊字母,比如说mu,代表standard division的那符号.
加减号又怎么输? X bar呢?

z-score

原帖由 阴雨天 于 2008-4-22 18:23 发表

谢谢!! 另一个问题 怎样在学校的word 2007里输入那些希腊字母,比如说mu,代表standard division的那符号.
加减号又怎么输? X bar呢?

这个在Word 2005就比较麻烦，我还真没用过2007写这些符号。

如果你找不到一个叫equation editor的plugin, 其实最好用手写。

阴雨天

原帖由 ~Kurosagi~ 于 2008-4-19 15:13 发表
pf  - the probability of female
pm - the probability of male
(1) pf - pm = the difference between proportion of female planners whose organisation contributes to costs for professional subscrip ...

got  exactly same answer wiz u! lol

阴雨天

原帖由 t-multiplier 于 2008-4-19 15:25 发表


(1) 最好加上the difference between the underlying proportion........
(2) 最好加上the difference between the sample proportion............

我觉得应该是situation a吧，因为问题是compare male ...

不是吧.........

z-score

male and female, New Zealanders and Australians都是经典的situation A

....oΟ药葫芦

希腊字母的话 用你输入法的键盘输入就可以    X BAR的记得在 输入法带的数学符号里应该有.. 加减号小键盘就有啊~~~

kirstens

原帖由 z-score 于 2008-4-22 20:02 发表


这个在Word 2005就比较麻烦，我还真没用过2007写这些符号。

如果你找不到一个叫equation editor的plugin, 其实最好用手写。

I'm far behind... I only know in 2003, there's insert / object / microsoft equation 3.0
I know many people have given up word and using latex to edit reports. Got to learn it some time.

z-score

原帖由 qian.qian 于 2008-4-23 20:13 发表

STATS 210 ASSIGNMENT 3 Question 4
(d)的R command应该是什么...
是2*ppois(9,5.5)?还是2*(1-ppois(9,5.5))?

> 2*ppois(9,5.5)
[1] 1.892445
> 2*(1-ppois(9,5.5))
[1] 0.1075549

应该是第二个，因为P-value小于1. ppois算得是lower tail probability, 我们要的是upper tail probability.

z-score

原帖由 kirstens 于 2008-4-24 10:02 发表

I'm far behind... I only know in 2003, there's insert / object / microsoft equation 3.0
I know many people have given up word and using latex to edit reports. Got to learn it some time.

嗯，其实我也是，我现在基本不用word, 一般写正式的文件我都用latex, 数学符号很漂亮．

kirstens

is it one tailed or 2 tailed test?

z-score

原帖由 kirstens 于 2008-4-24 14:47 发表
is it one tailed or 2 tailed test?

I think it is a two-tailed test, because it said to test if it is consistent with lambda = 5.5. What do you reckon?

kirstens

The question itself actually said H0=5.5
"The null hypothesis is H0:  = 5.5. Write down the
null and alternative hypotheses in terms of X and its parameters. Remember to specify the full distribution of X under the null hypothesis, and to consider whether the test should be one-sided
or two-sided"

and from the words "is consistent with", it does sounds like a 2 sided test.

but in the story, it used the discription such as “Minister urges raising driver age after high Easter toll” and "the situation was worse", and the number 9 is way larger than the mean 5.5, that's what make me think, is the question tend to a 1-sided test?

阴雨天

06年second semester  mid term test 上的一道题:
为什么说student(df=8) is more variable than student (df=3)这个statement 是false的?
我看course book上的图,明明是df 越大,图象的peak越高.难道peak高不能说明more variable么?
谢谢

阴雨天

原帖由 z-score 于 2008-4-24 13:38 发表


嗯，其实我也是，我现在基本不用word, 一般写正式的文件我都用latex, 数学符号很漂亮．

请问你说的这个latex怎么弄? 或者学校的哪个computer lab 有?

阴雨天

做个past paper屡屡碰壁
In 2001,41350 NZ students (60% female and 40% male) enrolled in a Bachelor degree programme. After 4 years, 66% of all these students had not completed the degree and were no longer enrolled for that degree programme, 70% of the males enrolled in 2001 were in this category.
After 4 years, the percentage of the females who enrolled in 2001 who had completed the degree or were still enrolled in the degree programme is approximately:
(1)37%
(2)22%
(3)30%
(4)65%
(5)49%
做了N多遍,死活得不出来正确答案...靠你了!! 正确答案是37% 可我一直得出来的是22%..能把你的2*2的那个table列出来给我看看么? 3Q!!!

阴雨天

2006 semester 2 mid term test的第18题麻烦给看下.
关于第3个和第4个选项:难道在dot plots diagram里还看出mean 和standard division? 还有这两个选项里to the nearest half kilometer这句话在题中怎么讲? 谢谢!

阴雨天

最后一个问题.
还是06年第二学期的midterm test最后一题:
一点头绪都没有. 请ATM指点下

z-score

原帖由 kirstens 于 2008-4-24 15:43 发表
The question itself actually said H0=5.5
"The null hypothesis is H0:  = 5.5. Write down the
null and alternative hypotheses in terms of X and its parameters. Remember to specify the full distr ...

Yep it should be two-sided then. What the ministers and others say should not be taken into consideration when forming the hypothesis.

z-score

原帖由 阴雨天 于 2008-4-24 17:17 发表
06年second semester  mid term test 上的一道题:
为什么说student(df=8) is more variable than student (df=3)这个statement 是false的?
我看course book上的图,明明是df 越大,图象的peak越高.难道peak高不 ...

Peak越大就证明更多的observations will be closer to the peak (mean), meaning less observations will vary about the mean, so generally the higher the df, the less variable the distribution.

z-score

原帖由 阴雨天 于 2008-4-24 17:32 发表

请问你说的这个latex怎么弄? 或者学校的哪个computer lab 有?

这个比较麻烦，我不建议为了一个assignment而学习latex. 要用latex写这份assignment估计要先读上两个小时的latex manual. 一般写比较长的thesis用latex比较划算，assignment这么短的就不用为了它来学习了。

z-score

原帖由 阴雨天 于 2008-4-24 19:01 发表
2006 semester 2 mid term test的第18题麻烦给看下.
关于第3个和第4个选项:难道在dot plots diagram里还看出mean 和standard division? 还有这两个选项里to the nearest half kilometer这句话在题中怎么讲? 谢谢!

这道题你要认为statement 3, the mean to the nearest kilometer is 1, then a two standard deviation interval would be the values would range from 0 to 2 if the standard deviation is really 1. But according to the 68%, 95%, 99.7% rule, two standard deviation will cover 95% of all observations, 从0到2这个interval太大了。Standard deviation 因该是差不多0.5，这样的interval能跟好的fit the data.希望你能明白我在说什么。

z-score

原帖由 阴雨天 于 2008-4-24 18:04 发表
做个past paper屡屡碰壁
In 2001,41350 NZ students (60% female and 40% male) enrolled in a Bachelor degree programme. After 4 years, 66% of all these students had not completed the deg ...

                                                         Male            Female

dropped out                 11578          15713              27291

Done, or still studying   4962             9097               14059

                                    16540           24810            41350

9097 / 24810 = 0.37

z-score

原帖由 阴雨天 于 2008-4-24 19:41 发表
最后一个问题.
还是06年第二学期的midterm test最后一题:
一点头绪都没有. 请ATM指点下

哦，这道题，当年晕倒无数人，我记得David Smith说两千人中就4个人答对了。

其实最简单的方法是看第一个statement和另外四个statement叙述方式不一样。它用了double negative, are different is not consistent, 这个错因为the two means can be different, 就像statement 3和5一样，the first mean could be bigger or small then the second mean, so they can be different and the first statement saying it is inconsistent must be false.

z-score

原帖由 qian.qian 于 2008-4-24 20:05 发表
所以 > 2*(1-ppois(9,5.5))
[1] 0.1075549 这个才是对的??

是的

加热的咖啡

阿童木, 能帮忙看下stats108的第6题 (a) 吗? 这个不太明白 谢谢了

ivy113

for Q4 what is distribution. and why we can calculate c under the seme condition. does c include
distribution. Please help me .thank you very much