英国论文代写网

当前位置: 首页 > 精品范例 > 美国论文代写essayassignment
美国论文代写essayassignment

美国代写-美国华盛顿大学经济金融硕士论文代写94分作业paper范例

美国代写-美国华盛顿大学经济金融硕士论文代写94分作业paper范例
  • 国家 : 美国代写
  • 级别 : 硕士
  • 专业 :
QQ768093293八年数万高分反馈成绩截图+写手资料证书水印照片曝光!史上最强联盟tutor+外籍+PHD上千精英联盟!邦典留学生论文代写网专业致力留学生英语论文作业的代写服务,无论英国论文代写还是澳洲论文代写,加拿大论文代写还是新西兰香港墨尔本悉尼温哥华等国家地区的论文都可以快速保质完成。专业留学英语论文代写.铁证如山数万截图100%好评!不满意退款!

详细描述

The Fisher’s Exact Test

Introduction

The Fisher’s Exact Test (R.A.Fisher, 1925) was proposed by a British statisticianR.A. Fisher in a form of an anecdote in 1925. During an afternoon tea, a colleague said that she was able to tell if the tea was pulled into milk or the milk was pulled into the tea because the order would influence the flavor. In order to test her words, Fisher got 8 cup of tea for her to taste and got the result. Yet, how to figure out if she had spoken true or just got good luck today? Fisher proposed a statistical method based on conditional probabilities theories to identify: the Fisher’s Exact Test. This assignment will show the method of the Fisher’s Exact Test, discuss the situations that this method can be apply to, and provide examples.

What is the Fisher’s Exact Test?

The Fisher’s Exact Test is a statistical method to analyze the contingency tables(University College, 1904-1922), which represent data of categorization in a 2*2 contingency table. And the Fisher’s Exact Test is able to examine if the association of these two kinds of classification is significant. An example that contains the full process of the Fisher’s Exact Test is presented as below:

Take the original situation that Fisher faced with. Say that the result of the colleague’stastes was listed as below:

 

Colleague’s Guess

The beverage pulled first

Milk

Tea

Total

Milk

A=3

B=1

A+B=4

Tea

C=1

D=3

C+D=4

total

A+C=4

B+D=4

N=8

Fisher calculated the probability of the event that his colleague got the result by chance:

Plug the data in the table into the formula above and get the result:

0.229 is the probability of the event, also known as p value (Stigler, 1986), that the colleague got the result only by chance, and compared it to the probability of type one error, usually set as 0.05 or 0.01. In this case, since 0.229>0.05, the conclusion that the colleague got the result by guessing is convincing and the assumption is accepted.

In light of this example, the Fisher’s Exact Test is a significance test to analyze the assumption based on the 2*2 contingency table. The test is able to calculate the exact probability of the assumption, and compare it to the fixed p value (probability of type 1 error, usually set as 0.05 or 0.01) to draw the conclusion that if the assumption should be accepted.

Why use the Fisher’s Exact Test?

Through the process above, we found that the Fisher’s Exact Test is an exact test, which means the p value can be exactly calculated. Moreover the distribution of the random variable is not requested. The most important character of the Fisher’s Exact Test can be applied on the situations with small sample size, in the situation above for example, there are only 8 samples. This is what other test cannot compared to, because other tests are usually take use of the approximation of the random variable ‘s distribution when the number of the samples is very large and take it as limit to infinite, chi-square test for instance. However, when the sample size is small, the approximation condition is insufficient. And the data are unequally distributed in the table, which will cause the p value based on the null hypothesis being low. In fact, theactual and approximated p values can be very different and even will lead an opposite conclusion, for data of small size, discretization, or unbalanced distribution.

When not to use it

It is noticed that the Fisher’s Exact Test can be and can only be applied when the columns and rows of the table are totally fixed, or the exact value of p cannot be calculated. However, situations of uncertainty seem very often in the real life. This is the first limitation of the Fisher’s Exact Test.

Moreover, when the sample group is very large, say 1000 samples for example, the calculate process of the Fisher’s Exact Test can be very complicated and time wasting. Even though the Fisher’s Exact Test can be applied in theory, in practical, people seldom use this method.

What to use instead of it and when

When the sample size is quite large, though the theory of the Fisher’s Exact Test is valid, the complexity of the calculation makes people give it up. In this situation, the Chi-square Test, which is invalid when the sample size is small, now can be applied. Since the distribution of samples (the null hypothesis is accepted) is approximate to a chi-squared distribution if the sample size is large enough. Actually the sample distribution will approach a chi-squared distribution as the size of the sample limiting to infinity. Thus, when the size of the sample is large enough, the Chi-square Test can be applied instead of the Fisher’s Exact Test. And in a 2*2 contingency table, it is called McNemar's Test.

Example

A hospital listed the data of the number of survivors suffering a certain kind of disease of each gender.

 

male

female

total

survive

A=9

B=4

A+B=13

dead

C=1

D=10

C+D=11

total

A+C=10

B+D=14

N=24

According to this table, the female death rate seems much higher than the male’s. Thus, we assume the null hypothesis is that the death rate of male and female are equal. The opposite hypothesis is that the death rate of female is higher than male. Use the Fisher’s Exact Test:

 

 

 

0.0042<0.01, thus we refuse the null hypothesis, and regard that the death rate of female is higher than male.

 

 

 

 

 

Reference

R.A.Fisher. (1925). Statistical Methods for Research Workers. Edinburgh: Oliver and Boyd.

Stigler, S. M. (1986). The history of statistics : the measurement of uncertainty before 1900. Cambridge, Mass: Belknap Press of Harvard University Press.

University CollegeDept. of Applied StatisticsL. (1904-1922). Drapers' Company research memoirs. London: Dulau and Co.

Agresti, A. (2002). Categorical Data Analysis. John Wiley, New Jersey

Ralph B. D'Agostino, Warren Chase and Albert Belanger. The Appropriateness of Some Common Procedures for Testing the Equality of Two Independent Binomial Populations. The American Statistician , Vol. 42, No. 3 (Aug., 1988), pp. 198-202.

Kenney, J. F. and Keeping, E. S. Mathematics of Statistics, Pt. 2, 2nd ed. Princeton, NJ: Van Nostrand, 1951.

R code for the Fisher’s Exact Test

> x=c(9,4,1,10)

> A <-matrix(x, nrow=2)

> fisher.test(A, alternative ="two.sided")渔船精确测试

景区简介
Fisher精确检验(r.a.fisher,1925)是由英国statisticianr提出。A. Fisher在形式的轶事1925。在下午茶,一位同事说,她能够告诉如果茶被拉入牛奶或牛奶被拉入茶,因为该命令将影响风味。为了测试她的话,他得到了8杯茶,让她品尝,并得到了结果。然而,如果她说真的,还是今天的运气好,怎么弄明白呢?提出了一种基于条件概率理论的统计方法,以确定其确切的检验方法。这项任务将显示的方法,该方法可以适用于该方法,该方法可以适用于,并提供实例。
什么是鱼的准确测试?
Fisher的精确检验分析列联表的统计方法(大学,1904),它代表一个2×2列联表的分类数据。而这些2种分类的关联性很有意义,而对其进行精确检验的检验则是可以检验的。一个例子,包含了全过程的渔民的精确测试如下:
以老渔民面对的情况。说的colleague'stastes结果如下:
同事猜测
先把饮料拉了
鲜奶
茶叶
鲜奶
= 3
乙= 1
一个+ = 4
茶叶
= 1
3 =
4 + +
一个+ = 4
2 + 4
= 8
他计算出他的同事碰巧得到的结果的概率:
将数据插入到表中,并得到结果:
0.229是大概率事件,也被称为P值(Stigler,1986),这个同事只是偶然得到的实验结果,并比较它的第一类错误的概率,通常设置为0.05或0.01。在这种情况下,由于0.229>0.05,得出的结论是,同事得出的结论是,猜测是有说服力的,而假设被接受。
在这个例子中,小鱼的精确测试是一个有意义的测试,分析的假设基础上的2 * 2应变表。测试是能够计算的假设的确切概率,并比较它的固定值(1或0.05)的个错误的概率,通常设置为0.01或)得出的结论,如果假设应接受。
为什么要用精确的测试?
通过上述过程,我们发现,该方法的精确检验是一种精确的检验,这意味着该值可以精确计算。此外,随机变量的分布是不要求。在小样本量的情况下,可应用于该试验的最重要的性质,例如,在上述情况下,只有8个样本。这是其他的测试不能比较的,因为其他的测试通常是利用随机变量的分布近似的,当样本的数量是非常大的,并把它作为无限的限制,例如,卡方检验。然而,当样本量小,近似条件是不够的。和数据的分布是不均匀的表中,将基于假设低导致P值。事实上,实际的和近似的P值可以是非常不同的,甚至会导致一个相反的结论,因为体积小,数据离散化,或分布不平衡。
当不使用它
它被注意到,渔船的精确测试可以和只能应用于当表的列和行是完全固定的,或无法计算的精确值。然而,在现实生活中的不确定性的情况下似乎非常经常。这是第一次限制了渔船的精确检验。
此外,当样本组非常大时,比如说1000个样本,计算过程中的精确检验可能是非常复杂和耗时的。即使在理论上也可以应用于实际,人们很少使用这种方法。
用什么代替它
当样本量相当大的时候,虽然理论上的精确检验是有效的,计算的复杂性使得人们放弃它。在这种情况下,卡方检验,这是无效的,当样本量小,现在可以应用。由于样本的分布(零假设被接受)是近似的卡方分布,如果样本量足够大。实际上,样本分布将接近卡方分布的样本限制到无限的大小。因此,当样本的大小是足够大,卡方检验可以应用,而不是鱼的精确检验。在一个2×2列联表,它叫做McNemar检验。
例子
一家医院列出了每一种性别的某种疾病的幸存者的数据。
男性
女性
生存
= 9
乙= 4
一个+ = 13
死了
= 1
10 =
11 + +
一个+ = 10
2 + 14
= 24
根据这张表,雌性大鼠死亡

邦典论文essaylunwen.com
QQ:768093293
FACEBOOK:
www.facebook.com/papercustom
微博weibo.com/essaylunwen
微信公众号

点击次数:  更新时间:2015-10-25  【打印此页】  【关闭