Created
May 20, 2012 13:44
-
-
Save doryokujin/2758196 to your computer and use it in GitHub Desktop.
prop.test
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# hand calculation での注意:二項分布の同等性検定は2×2の分割表によるχ^2検定に等しいのでそちらの計算方法を採用。 | |
# Q1.はサンプル数が少ないのでイェツの補正を行っている。 | |
###################################### | |
# Q1. | |
# Head Tail S | |
# A 4 6 10 | |
# B 6 4 10 | |
# S 10 10 20 | |
# | |
# hand calculation | |
# T = (20* (|4*4-6*6| - 20/2)^2 ) / (10*10*10*10) = 0.2 < χ^2(1, 0.05) = 3.84 | |
# ∴ 「差があるとは言えない」 | |
> prop.test(c(4,6),c(10,10)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(4, 6) out of c(10, 10) | |
X-squared = 0.2, df = 1, p-value = 0.6547 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.7294066 0.3294066 | |
sample estimates: | |
prop 1 prop 2 | |
0.4 0.6 | |
###################################### | |
# Q2. | |
# Head Tail S | |
# A 16 24 40 | |
# B 24 16 40 | |
# S 40 40 80 | |
# | |
# hand calculation | |
# T = (80*(16*16-24*24)^2 ) / (40*40*40*40) = 3.2 < χ^2(1, 0.05) = 3.84 | |
# ∴ 「差があるとは言えない」 | |
> prop.test(c(16,24),c(40,40)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(16, 24) out of c(40, 40) | |
X-squared = 2.45, df = 1, p-value = 0.1175 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.4397033 0.0397033 | |
sample estimates: | |
prop 1 prop 2 | |
0.4 0.6 | |
###################################### | |
# Q3. | |
# Head Tail S | |
# A 40 60 100 | |
# B 60 40 100 | |
# S 100 100 200 | |
# | |
# hand calculation | |
# T = (200*(40*40-60*60)^2 ) / (100*100*100*100) = 8 > χ^2(1, 0.05) = 3.84 | |
# ∴ 「有意な差があると言える」 | |
> prop.test(c(40,60),c(100,100)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(40, 60) out of c(100, 100) | |
X-squared = 7.22, df = 1, p-value = 0.00721 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.34579029 -0.05420971 | |
sample estimates: | |
prop 1 prop 2 | |
0.4 0.6 | |
###################################### | |
# Q4. | |
# Head Tail S | |
# A 400 600 1000 | |
# B 600 400 1000 | |
# S 1000 1000 2000 | |
# | |
# hand calculation | |
# T = (2000*(400*400-600*600)^2 ) / (1000*1000*1000*1000) = 80 > χ^2(1, 0.05) = 3.84 | |
# ∴ 「有意な差があると言える」 | |
> prop.test(c(400,600),c(1000,1000)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(400, 600) out of c(1000, 1000) | |
X-squared = 79.202, df = 1, p-value < 2.2e-16 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.2439407 -0.1560593 | |
sample estimates: | |
prop 1 prop 2 | |
0.4 0.6 | |
###################################### | |
# Q5. | |
# Head Tail S | |
# A 480 520 1000 | |
# B 520 480 1000 | |
# S 1000 1000 2000 | |
# | |
# hand calculation | |
# T = (2000*(480*480-520*520)^2 ) / (1000*1000*1000*1000) = 3.2 < χ^2(1, 0.05) = 3.84 | |
# ∴ 「差があるとは言えない」 | |
> prop.test(c(480,520),c(1000,1000)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(480, 520) out of c(1000, 1000) | |
X-squared = 3.042, df = 1, p-value = 0.08114 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.084791052 0.004791052 | |
sample estimates: | |
prop 1 prop 2 | |
0.48 0.52 | |
###################################### | |
# Q6. | |
# Head Tail S | |
# A 470 530 1000 | |
# B 570 430 1000 | |
# S 1000 1000 2000 | |
# | |
# hand calculation | |
# T = (2000*(470*470-530*530)^2 ) / (1000*1000*1000*1000) = 7.2 > χ^2(1, 0.05) = 3.84 | |
# ∴ 「有意な差があると言える」 | |
> prop.test(c(470,530),c(1000,1000)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(470, 530) out of c(1000, 1000) | |
X-squared = 6.962, df = 1, p-value = 0.008326 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.10474717 -0.01525283 | |
sample estimates: | |
prop 1 prop 2 | |
0.47 0.53 | |
###################################### | |
# Q7. | |
# Head Tail S | |
# A 8 12 20 | |
# B 48 32 80 | |
# S 56 44 100 | |
# | |
# hand calculation | |
# T = (100*(8*32-12*48)^2 ) / (20*80*56*44) = 2.60 < χ^2(1, 0.05) = 3.84 | |
# ∴ 「差があるとは言えない」 | |
> prop.test(c(8,48),c(20,80)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(8, 48) out of c(20, 80) | |
X-squared = 1.8491, df = 1, p-value = 0.1739 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.47129558 0.07129558 | |
sample estimates: | |
prop 1 prop 2 | |
0.4 0.6 | |
###################################### | |
# Q8. | |
# job non-job S | |
# last 95,100 4,900 100,000 | |
# this 95,000 5,000 100,000 | |
# S 190,100 9,900 200,000 | |
# | |
# hand calculation | |
# T = (200000*(95100*5000-95000*4900)^2 ) / (100000*100000*191000*9900) = 1.06 < χ^2(1, 0.05) = 3.84 | |
# ∴ 「差があるとは言えない」 | |
> prop.test(c(4900,5000),c(100000,100000)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(4900, 5000) out of c(1e+05, 1e+05) | |
X-squared = 1.0416, df = 1, p-value = 0.3075 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.002911256 0.000911256 | |
sample estimates: | |
prop 1 prop 2 | |
0.049 0.050 | |
###################################### | |
# Q9. | |
# (say = "自分ではっきり言う") | |
# say not S | |
# 5M 53 247 300 | |
# 20M 48 152 200 | |
# S 101 399 500 | |
# | |
# hand calculation | |
# T = (500*(53*152-48*247)^2 ) / (300*200*101*399) = 2.99 < χ^2(1, 0.05) = 3.84 | |
# ∴ 「差があるとは言えない」 | |
> prop.test(c(53,48),c(300,200)) | |
2-sample test for equality of proportions with continuity correction | |
data: c(53, 48) out of c(300, 200) | |
X-squared = 2.606, df = 1, p-value = 0.1065 | |
alternative hypothesis: two.sided | |
95 percent confidence interval: | |
-0.14075264 0.01408597 | |
sample estimates: | |
prop 1 prop 2 | |
0.1766667 0.2400000 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment