Hi. I need to complete hypothesis tests and i am unsure if i used the correct tests for each of the following:
1. Do people actually prefer Pepsi or Coke in a taste test?
2. Can people correctly identify Pepsi versus Coke?
3. Are people consistent with their prior preference? In other words, do they prefer the
same cola in their taste test that they had previously indicated was their preferred cola?
At first i was going to use two-tailed tests for all three. but then i started thinking maybe number 1 was a one-tailed test. now i'm really confused.
Since data will not be normally distributed, I would suggest using a X^2 (Chi square) test.
its a worksheet where i have to do the hypothesis tests though.
Hi there! I can help clarify the appropriate hypothesis tests for each of the scenarios you mentioned.
1. "Do people actually prefer Pepsi or Coke in a taste test?"
For this question, you can use a two-tailed test. Here's why: The null hypothesis (H0) would state that there is no preference between Pepsi and Coke, while the alternative hypothesis (H1) would state that there is a preference. Since you are not specifying the direction of the preference (i.e., whether people prefer Pepsi over Coke or vice versa), it is appropriate to use a two-tailed test.
2. "Can people correctly identify Pepsi versus Coke?"
In this case, you can also use a two-tailed test. The null hypothesis (H0) would state that people cannot correctly identify Pepsi versus Coke, while the alternative hypothesis (H1) would state that they can. Similar to the previous scenario, you are not specifying the direction of the ability to correctly identify, so a two-tailed test is appropriate.
3. "Are people consistent with their prior preference?"
For this question, a one-tailed test would be suitable. The null hypothesis (H0) would state that people are not consistent with their prior preference, while the alternative hypothesis (H1) would state that they are consistent. Here, you are specifically interested in whether people's current preference aligns with their previously indicated preferred cola. Therefore, a one-tailed test would be appropriate to investigate if people prefer the same cola as before.
It's great that you thought about the differences between one-tailed and two-tailed tests. Just remember that the decision should be based on the specific question you want to answer and the directionality you expect.