1
$\begingroup$

good evening every body

I have a question about t-statistics in first place I need to use t-statistics to estimate the population mean from a sample set, as you see I have a data set the mean and standard deviation are known for sample but not for the population.

how can I use t-statistics to determine the population mean I have X bar and S and n values for sample n=45
how can I compute the degree of freedom and how can I determine the confidence level? thanks for your time huge regards

$\endgroup$
1
$\begingroup$

$df=n-1$

You need to set the confidence level yourself. $\alpha =.05$ is standard.

Confidence interval: $$Upper limit=Xbar+\frac{tquantile(1-\alpha/2,df)*S}{\sqrt{df}}$$ $$Lower limit=Xbar-\frac{tquantile(1-\alpha/2,df)*S}{\sqrt{df}}$$

Where tquantile is the quantile from a t-distribution chart or obtained from statistical software. It is 2.015368 so you have:

$$Upper limit=Xbar+\frac{2.015368*S}{\sqrt{df}}$$ $$Lower limit=Xbar-\frac{2.015368*S}{\sqrt{df}}$$

In R:

> qt(.975,44)
[1] 2.015368

Note: I'm not sure if you need to divide by $\sqrt df$, depends on whether S is the standard dev of the sample or the individual observations...

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.