I have 2 sets of summary of mean, standard diviation and n. I had to see if any

Question

I have 2 sets of summary of mean, standard diviation and n. I had to see if any difference. Hypothesis H0: u1-u2 = 0, Ha: u1-u2 'not equal' 0. I used MEGASTAT for tesitng. Not by hand. p-value (two tailed) came out. I though I had to divide the two tailed p-value, but somebody (he has a doctor degree) said "you don't need to, because you used MEGASTAT. MEGASTAT automatically calculated for you." But when I look at textbook solution at Chegg. It seems like I have to divide p-value (two tailed) to test levels of significance. Otherwise it does not match to the answer. Is Chegg right or Dr. Statistic right? WHY?

Explanation / Answer

Ans:

First of all,When we perform 2 tailed test,we always compare 2-tailed p-value with the significance level,you can not divide it by 2 to compare,that will give wrong interpretation.

You can check using t-statistic of the results as follows:

Say,t-statistic:

t=1.82

df=18

Use Excel formula for getting 2 tailed p-value as below:

p-value(2-tailed)=TDIST(1.82,18,2)=0.0854

Now,you have to use this 2-tailed p-value to compare with the given significance level say 0.05

As,p-value=0.0854>0.05,we fail to reject null hypothesis.

Now,in context of Megastat,it will give you 2-tailed p-value,and you have to compare that particular p-value with significance level(i.e. no need to divide that value,otherwise it will give you the conclusion for one tail test)

Related Questions

Navigate

• Browse (All)
• Browse I
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.