The verification of hypotheses: a hypothesis comparing average null hypothesis with a sample? Hello, I'm stuck on a homework problem, it is for political science and is a statistical work. The question is:
Imagine that we are interested in the effects of political advertising and run an experiment in which individuals are first given only to the part identification number and a few positions about a candidate and then asked to rate the candidate on a scale of 0-100. Then they presented a series of ads about the candidates are again asked to rate the candidate on a scale of 0-100. The average difference is 14.66 and the standard deviation is 5.21. The sample size is 15. Suppose 0.05 significance level.
A. What are the null and alternative hypotheses?
B. Calculate the value of the test statistic.
C. What is the value?
D. State your conclusion.
Thank you!
If we assume that this difference in the HTE score follows a normal distribution, we can perform the following analysis.
H0: μ = 0
Ha: μ> 0
stat test = (14.66 - 0) / (5.21/sqrt15)
10.90
It's already a great value and you can already reject H0. But just to go through the procedure.
From the tables T (df = 14), the critical value of Virginia
1.76
Since the path STAT test exceeds the critical value, we can reject H0. The campaign made the difference.
This is a test of paired data. Therefore, the null hypothesis is always average (after - before) = 0. I suppose you want to see if the Commericals make a positive difference. Therefore, the null hypothesis is: average (after - before)> 0.
Test T: T = (14.66 - 0) / (5.21/SQRT 15) = 10.898
T critical degree of freedom = 15-1 = 14 See T value in a table (mine is in the office, I can not do it for you.)
If T> T Test criticism (and I suppose it is), null is rejected. This means that there is a positive influence on political advertising.
Posted on April 21, 2010.
