Scribd
Upload
430 views5 pages

HW#4 For Montgomery Design and Analysis of Experiments

The document reports the results of statistical tests comparing test scores from two different birth orders (BO1 and BO2) and two different measurement methods (km and lm). For the birth order data: 1) The mean difference was -0.051 with a p-value greater than 0.05, indicating no significant difference. 2) Confidence intervals for the mean difference included zero, supporting no difference. For the measurement methods: 1) The mean scores were significantly different with a p-value less than 0.05. 2) The km data was normally distributed but lm was not. 3) The differences between the paired km and lm scores were normally distributed, allowing use of the paired t-test.

Uploaded by

studycam
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
430 views5 pages

HW#4 For Montgomery Design and Analysis of Experiments

The document reports the results of statistical tests comparing test scores from two different birth orders (BO1 and BO2) and two different measurement methods (km and lm). For the birth order data: 1) The mean difference was -0.051 with a p-value greater than 0.05, indicating no significant difference. 2) Confidence intervals for the mean difference included zero, supporting no difference. For the measurement methods: 1) The mean scores were significantly different with a p-value less than 0.05. 2) The km data was normally distributed but lm was not. 3) The differences between the paired km and lm scores were normally distributed, allowing use of the paired t-test.

Uploaded by

studycam
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 5

Individual Homework #4 Design of Experiment and Analysis 2.33. Manually BO1 BO2 6.08 5.73 6.22 5.8 7.

99 8.42 7.44 6.84 6.48 6.43 7.99 8.76 6.32 6.32 7.6 7.62 6.03 6.59 7.52 7.67 mean sd

diff 0.35 0.42 -0.43 0.6 0.05 -0.77 0 -0.02 -0.56 -0.15 -0.051 0.441

b) Ho = d=0 vs Ha: d0 to = d/Sd/sqrt(n) = -0.051/ 0.441/sqrt(10) = -0.3657 t0.025,9= 2.262 to < t0.025,9 No rejection of Null hypothesis. CI = to*Sd/10 d = -0.051 0.3159 The interval includes zero (0) so no rejection of null hypothesis. c) Ho = d=0 vs Ha: d>0 mean score of Date of birth 1 is greater than Date of birth 2 to = 0.3657< t0.05,9 =1.833 No rejection of Null hyp. Ho = d=0 vs Ha: d<0 mean score of Date of birth 1 is lower than Date of birth 2 to = -0.3657 < t0.05,9 =1.833 No rejection of Null hyp. R Solution a) Based on normal qq plot normal distribution of difference is reasonable.

Mean = -0.051 and s= 0.441

b) t.test(bo1, bo2, alternative ="two.sided", mu = 0, paired =TRUE, var.equal = FALSE, conf.level = 0.95) Paired t-test 95 percent confidence interval: -0.3664148 0.2644148 c). No evidence that birth order affect mean score (mean of differences) Ho: d=0 Ho: d>0 t.test(bo1, bo2, alternative ="greater", paired =TRUE) Paired t-test data: bo1 and bo2 t = -0.3658, df = 9, p-value = 0.6385 alternative hypothesis: true difference in means is greater than 0 95 percent confidence interval: -0.3065927 Inf sample estimates: mean of the differences

-0.051 Ho: d=0 Ho: d<0 t.test(bo1, bo2, alternative ="less", paired =TRUE) Paired t-test data: bo1 and bo2 t = -0.3658, df = 9, p-value = 0.3615 alternative hypothesis: true difference in means is less than 0 95 percent confidence interval: -Inf 0.2045927 sample estimates: mean of the differences -0.051 2.34. a) Yes, they are different. p-value = 6.767e-05 km=c(1.186, 1.151, 1.322, 1.339, 1.200, 1.402, 1.365, 1.537, 1.559) lm=c(1.061, 0.992, 1.063, 1.062, 1.065, 1.178, 1.037, 1.086, 1.052) t.test(km, lm, alternative ="two.sided", mu = 0, paired =FALSE, var.equal = TRUE, conf.level = 0.95) Two Sample t-test data: km and lm t = 5.3302, df = 16, p-value = 6.767e-05 alternative hypothesis: true difference in means is not equal to 0 95 percent confidence interval: 0.1649580 0.3828197 sample estimates: mean of x mean of y 1.340111 1.066222 b) p-value = 6.767e-05

c) t.test(km, lm, alternative ="two.sided", paired =TRUE) Paired t-test 95 percent confidence interval: 0.1700423 0.3777355 d) Data for Karlsruhe method appears to be normally distributed but data for Lehigh method is not normally distributed

e) Difference appears to be normally distributed based on normal qq plot

d) In the same way as the normality is a key assumption for one or two sample t test it is for a paired t test. In the case of paired t-test the samples are not drawn from independent populations. By taking the difference the bias (non independence) is relieved and the data is more normally distributed.

You might also like