# state the null and alternative hypothsis

• August 5th 2011, 06:39 PM
wopashui
state the null and alternative hypothsis
Online ads targeting children were said to have decreased at least 56% between May and August 2000. A sample of 100 ads selected during this period showed a 45% decrease in online ads targeting children. State the null and alternative hypotheses and carry out the test at the 1% level of significance.

Here, which one shud be the null hypothesis $H_0$, and which one should be $H_a$? I'm confused.
• August 5th 2011, 10:24 PM
CaptainBlack
Re: state the null and alternative hypothsis
Quote:

Originally Posted by wopashui
Online ads targeting children were said to have decreased at least 56% between May and August 2000. A sample of 100 ads selected during this period showed a 45% decrease in online ads targeting children. State the null and alternative hypotheses and carry out the test at the 1% level of significance.

Here, which one shud be the null hypothesis $H_0$, and which one should be $H_a$? I'm confused.

Take a null hypothesis that the number of adds targeting children has decreased by 56%, and the alternative as the decrease is less than 56%.

However I don't think the data you have been given is adequate to construct such a test for a number of reasons.

CB
• August 6th 2011, 01:54 AM
HallsofIvy
Re: state the null and alternative hypothsis
Quote:

Originally Posted by wopashui
Online ads targeting children were said to have decreased at least 56% between May and August 2000.

The claim being tested is always the "null hypothesis".

Quote:

A sample of 100 ads selected during this period showed a 45% decrease in online ads targeting children. State the null and alternative hypotheses and carry out the test at the 1% level of significance.

Here, which one shud be the null hypothesis $H_0$, and which one should be $H_a$? I'm confused.
• August 9th 2011, 07:38 PM
wopashui
Re: state the null and alternative hypothsis
so shud I make $H_0: P= 56%$, and $H_a: P<56%$, however,it's at least 56%, does it mean p>=56%