1. An advertiser is experimenting with a new color scheme and conducts a study to test its effectiveness. In which situation could they use a hypothesis test for matched pairs?

They track each customer's spending habits in the old platform and then they change the color scheme to see if spending habits go up or down for each consumer.

They randomly expose consumers to one website when they land on their page: the old one with the original color scheme or the new one with the updated color scheme. Then they measure to see how much people buy.

They show consumers both options: theoriginal color scheme or the updated color scheme. They let consumers decide which one they like better and then select the color scheme that most customers prefer.

The following is for questions 2, 3 and 4:

Not much is known about the use of cannabis while pregnant because for obvious reasons, expecting mothers cannot be randomly assigned to use or not use this drug while pregnant. In samples of mothers in the US who do use cannabis, the comparison is usually flawed because the samples have other important differences, namely that the mothers who use are often also poor. In a first study of it's kind, a researcher matched mothers who did and did not use cannabis so that each pair had a very similar economic status.

2.

Given the design of the study and the question of interest, which of the following 4 computer outputs is relevant to use? O Paired T-Test and CI Sample N Mean StDev SE Mean Marijuana Use 10 77.59 23.38 7.39 No Marijuana Use 10 78.02 17.36 5.49 Difference 10 -0.43 11.83 3.74 954 upper bound for mean difference: 6.43 I-Test of mean difference = 0 (vs U): T-Value - -0.11 P-value - 0.5\" From the output we learn that: O The data do not provide sufcient evidence to reject Hg. In other words, based on the data we cannot conclude that the mean health scores for babies born to moms who use cannabis is higher than the mean health scores for babies born to moms who do not use cannabis. O The data provide sufficient evidence to reject Hg and, thus, conclude that the mean health scores for babies born to moms who use cannabis is higher than the mean health scores for babies born to moms who do not use cannabis. O The data provide sufficient evidence to reject Hg. We therefore conclude the data do not provide evidence to conclude that the mean health scores for babies born to moms who use cannabis is lower than the mean health scores for babies born to moms who do not use cannabis. O The data do not provide sufficient evidence to reject H]. We therefore conclude that the mean health scores for babies born to moms who use cannabis is higher than the mean health scores for babies born to moms who do not use cannabis. The researchers analyzed the data and obtained the following output: Summary statistics: Sample n Mean Variance Std. Dev. Std. Err. Clean 50 35.3 41.438774 6.4372954 0.91037107 Dirt 50 38.7 41.030613 6.4055142 0.9058765 Hypothesis test results: Difference Sample Diff. Std. Err. DF T-Stat P-value Dirt-Clean -3.4 0.3736199 49 -9.100158