Outline

• Review three kinds of t-tests

Below are five problems similar to ones you've seen in the last few labs.

Below are some problems that are similar to ones you've seen in the last few labs. But now you're going to have to figure out on your own which t-test to use for each problem and then conduct that test using SPSS to answer the questions (on your own time, you should also use these problems to practice doing the computations by hand). For each question make sure to label and answer:

(a) Type of test you ran
(b) Null and alternative (researcher's) hypothesis
(c) The results of your test (reject/fail to reject null hypothesis & if there is a difference). If you find a difference, indicate what the difference is (i.e., which one was higher).
(d) Copy and paste a graph illustrating the results of your test onto the word document.
a. One-sample t-test: create a simple bar chart showing values of individual cases.
b. Paired-sample: create a simple bar chart showing summaries of separate variables.
c. Independent-sample: create a clustered bar chart of the showing the scores broken down by the grouping variable.

(1) A marine biologist is comparing the size of Great White Sharks in the Pacific and Atlantic Oceans to determine which ocean has the larger sharks. He takes a sample of 20 sharks, 10 sharks from each ocean, and measures their lengths. The measurements for the 20 sharks are listed below:

(2) A behavioral psychologist wants to know if food acts as a good motivator for rats to learn a maze faster than normal. She places a food pellet at the end of a maze that the rat can smell while working through the maze. She puts 8 rats through the maze and records how long it takes them to find the food at the end. She already knows that without the food, rats as a population take an average of 28.9 seconds to run the maze (with a normal distribution). Using the timing data recorded below, determine if the rats learn the maze faster with the food pellet than without it.

Times in seconds (n=8): 25.6, 29.0, 23.1, 25.5, 28.7, 26.5, 25.4, 23.9

(3) Two groups of participants (n = 10 per group, total N = 20) were given a problem solving task. One group was told they had 5 min to complete the task. The other group was not told they had a time limit, but was also given 5 min to complete the task. For both groups, the number of puzzles solved in the 5 min period was measured. These data for the 20 participants are listed below. Conduct a hypothesis test to determine if an announced time limit affects the number of puzzles solved. Based on the outcome of the test, what can you conclude about the effect of an announced time limit?

(4) Does caffeine reduce depression? Participants in this study were 10 people who regularly consume something containing caffeine each day. During the study, however, each participant was barred from consuming caffeine not provided by the experimenter. They came to the lab two subsequent mornings and were given a pill. The pill either contained caffeine or was a placebo (i.e., each participant received both pills but on different days). The order of the pill received was counterbalanced (i.e., 1/2 received the caffeine pill first and the other 1/2 received it last). They completed a depression scale at the end of each day. Based on the depression scores below (higher scores mean more depression), does caffeine appear to reduce depression?

(5) Whether a winning team can be purchased is a debated topic in baseball. Many major league team owners spend a lot of money on talented players to put together a team that they hope will win. Using the data below on 20 teams, conduct a test to compare batting averages for samples of players from the 10 teams with the highest payroll and teams with the 10 teams with the lowest payrolls to see if this practice is justified.