Measuring Reaction Time Researchers wanted to determine whether the reaction time (in seconds) of males differed from that of females to a go/no go stimulus. The researchers randomly selected 20 females and 15 males to participate in the study. The go/no go stimulus required the student to respond to a particular stimulus and not to respond to other stimuli. The results are as follows:
(a) Is it reasonable to use Welch’s t-test? Why? Note: Normal probability plots indicate that the data are approximately normal and boxplots indicate that there are no outliers.
(b) Test whether there is a difference in the reaction times of males and females at the = 0.05 level of significance.
(c) Draw boxplots of each data set using the same scale. Does this visual evidence support the results obtained in part (b)?
Step 1 of 5) Measuring Reaction Time Researchers wanted to determine whether the reaction time (in seconds) of males differed from that of females to a go/no go stimulus. The researchers randomly selected 20 females and 15 males to participate in the study. The go/no go stimulus required the student to respond to a particular stimulus and not to respond to other stimuli. The results are as follows: (a) Is it reasonable to use Welch’s t-test Why Note: Normal probability plots indicate that the data are approximately normal and boxplots indicate that there are no outliers. (b) Test whether there is a difference in the reaction times of males and females at the = 0.05 level of significance. (c) Draw boxplots of each data set using the same scale. Does this visual evidence support the results obtained in part (b) If we reject, H0: m1 = m2, we can look at the bounds of the confidence interval to determine which mean (m1 or m2) is greater. For example, if the interval contains only positive numbers, then x1 - x2 is positive, so we have evidence that m1 7 m2.
