In: Statistics and Probability
Does anyone have experience with during a lab on conducting a 2-sample t-test at a 1% level of significance to find out if the U.S. stock returns significantly outperform the U.S. bond returns using the monthly data covering the sample period of 1980-2017? Draw conclusions by both the P Value and Critical Value Method.
Arnavbha - I have an excel spreadsheet with data, but how do I get it to you?
You have to conduct right-tailed t-test for 2-sample
this can be done in Excel easily
go to data -> data analysis -> t-Test: Two-Sample Assuming Equal Variances
{If you are not seeing data analysis , you have to add plug-in}
select data both variables
choose alpha
the output will be something like this
t-Test: Two-Sample Assuming Equal Variances | ||
stock | bond | |
Mean | 36.135625 | 36.27625 |
Variance | 1902.812989 | 1923.616141 |
Observations | 8 | 8 |
Pooled Variance | 1913.214565 | |
Hypothesized Mean Difference | 0 | |
df | 14 | |
t Stat | -0.006429996 | |
P(T<=t) one-tail | 0.497480184 | |
t Critical one-tail | 1.761310136 | |
P(T<=t) two-tail | 0.994960368 | |
t Critical two-tail | 2.144786688 |
Note that this is just dummy result
Now look at p-value for one-tail as right-tailed is 1-tailed
if p-value < alpha, we reject the null hypothesis
using critical value method
if TS > critical value , then we reject the null hypothesis
take critical value of one tail
Please rate
Look for comment for help