Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

After successfully completely the course HLTH 204, Pamela Lillian Isley (aka Poi

ID: 3315412 • Letter: A

Question

After successfully completely the course HLTH 204, Pamela Lillian Isley (aka Poison her own study to test the claim that the median human body temperature is less than 98. the following data. Use significance level of 5%. 3. lvy) conducted 4o °F. She collects 96.6 97.1 97.6 97.2 98.2 97.5 96.5 98.1 98.6 98.6 97 97.0 98.8 . What is the hypothesis of interest? . What is value of the test statistic and pvalue when using the Sign Test (via the Exact Binomial Test)? What is the conclusion from the Sign Test (via the Exact Binomial Test) for the researcher's claim? What is value of the test statistic and pvalue when using the Signed Rank Test? What is the conclusion from the Signed Rank Test for the researcher's claim? For the given context and data, name the parametric method that would have been used if normality was given? Since normality is not given or known, would a parametric or nonparametric method be more appropriate to test the researchers claim? Explain.

Explanation / Answer

Hypothesis of interest:

To test if the median human body temperature is less than 98.6

H0: median body temperature = 98.6

H1: median body temperature < 98.6

Test used: Sign test

Values calculated using R as follows:

require("BSDA")
data <- c(96.6, 97.1, 97.6, 97.2, 98.2, 97.5, 96.5, 98.1, 98.6, 98.6, 97, 97, 98.8)
SIGN.test(data, md = 98.6, alternative = "less" )

data: data
s = 1, p-value = 0.005859
alternative hypothesis: true median is less than 98.6
95 percent confidence interval:
-Inf 98.19558
sample estimates:
median of x
97.5

The value of the test statistic = sample median = 97.5

p-value = 0.005859

level of significance = 0.05

p-value is less than 0.05

so, we reject Ho

Conclusion: We conclude that the Median body temperature is less than 98.6

When using signed rank test (wilcoxon tset)

In R:

> wilcox.test(data, alternative = "less", mu = 98.6, paired = FALSE)

Wilcoxon signed rank test with continuity correction

data: data

V = 1, p-value = 0.002539

alternative hypothesis: true location is less than 98.6

The p-value = 0.002539

P-value is less than 0.05

So, we reject Ho

Conclusion: We conclude that the Median body temperature is less than 98.6

If normality would have been given,

we could use one sided t-test.

Since normality is not nown,

A nonparametric method would be more appropriate to test the researcher's claim.

Because if the assumption sof normality is violated in the data,

it would lead to incorrect and unreliable results if we use a parametric method.