Why is the answer c when t-test>p-value
Coefficients** Standard Error t-Statistic p-Value JPY/USD change 0.2864262 0.291700144 0.981919 **0.330289
Q. Based on the results of the regression model shown in Exhibit 2, the best conclusion Garfield can make about a hypothesis that the coefficient JPY/USD change is zero is to:
Can you highlight which are the p values as the exhibit is unclear?
Ok, If i am correct this is Jordan Garfield case, for JPY/USD change p value is 0.33 which is above 5%, so we fail to reject the null hypothesis and this test in not significant.
You can look at it from a different point using t test which is given as 0.99 comparing this to the t critical it will be below t critical so you will fail to reject the null hypothesis as well.
You do not compare the t-test with the p-value.
You compare the p-value to the significance that would lead to a reject of the null hypo. A p-value of 0.330289 means that you would have to test at a significance of 33% to reject the null hypth… Since you’re testing at 5%, then you fail to reject.
(i may have some of the terminology above wrong, but the general idea is there… so dont shot me tickersu)
All the P value should be looked at to determine is whatever number in percentage times, ie .33=33 percent compared to whatever significance level which I can tell you that with a .33 p value means insignificance in about 99.9 percent of instances, hence we assume it is not different from zero. Correct.
You don’t compare the T-stat to the p-value. You either 1) compare the T-stat to a critical value or 2) you compare the p-value to a reasonable limit, usually 0.05. Since 0.33 > 0.05, you fail to reject the null, which is C.
I’m going to say this is generally incorrect and a misinterpretation of p-value.
Okay please tell me when you are testing at a 34 percent level of significance then? Because that’s what would be needed to be considered different from zero…
If p value is 0.33 which is above 5%, I fail to reject the null hypothesis and this test in not significant. is not p-value the lowest value to reject the null hypothesis, hence the 5% value should be lower not higher?
P-value is the lowest significance level at which the null hypothesis H0 can be rejected
So you can happily reject the null, so you reject the null at values of p or higher, but not at values lower than p
Hence the p-value is the lowest significance level you can reject the null hypothesis at , so you reject the null at the p value or higher, but not at values lower than p value
This statement (all of it, including the bold) is difficult to read because the sentence structure is not proper. Particularly unclear is your part “…determine whatever number in percentage times…” Maybe it is the grammatical structure, but it sounds like there is some misunderstanding of what a p-value actually represents.
By the way, in the real world, people who run 7 tests each at 5% significance level have roughly an overall significance level of ~35% for the set of tests. This is common, though not always good practice.
The bold is a true statement but it is not an actual definition of p-value (maybe an operational definition if you want to stretch it). This kind of definition leads to your confusion that alpha (significance level) should be changed from 5% based on the p-value.
Okay. Maybe the following is more clear: You will never, and I emphasize the “never” part, see a question on the CFA Level II exam where a p value of .33 is significant.
I agree with that statement. But, please elaborate on the explanation you provided for the concept of p-value (something about it tells you the percent…).
Respect … you are illustrating it in more details which is something i always like
I am engraving the basics deep in my mind till the big day.
Is your hypothesis running? Because you better go catch it.
I prefer the good ol’, “Is your toilet running? Well, you better go jiggle the handle.”