Prform a paired-samples t-test for experiment p 01 or 99

Question

A group of 30 participants is divided in half based on their self-rating of the vividness of their visual imagery. Each participant is tested on how many colors of objects he or she can correctly recall from a briefly seen display. The results of this experiment are below. (Note: You will need one column in SPSS for the IV with two levels [vivid/less vivid] and one for the DV.)

Vivid	Less Vivid
18	14
17	12
16	13
15	10
14	8
13	8
9	7
8	6
7	5
6	4
5	3
11	2
13	5
18	11
18	14

Enter the data into SPSS and perform an independent samples t-test of the null hypothesis that vividness of visual imagery does not affect the recall of colors; use alpha = .01 (or 99% confidence), two-tailed. (Hint: The default alpha level is not .01; see options in the test menu.) What is your statistical conclusion?

SPSS output:

Results write-up:

What are some of the limitations that prevent you from concluding that the visual imagery causes improved color recall in this type of experiment?

An educator has invented a new way to teach geometry to high school students. To test this new teaching method, the educator matches 16 tenth-graders into 8 pairs based on their grades in previous math courses. Then the students in each pair are randomly assigned to either the new method or the traditional method. At the end of a full semester of geometry training, all students take same standard high school geometry test. The scores for each student in this hypothetical experiment are as follows:

Traditional	New
65	67
73	79
70	83
85	80
93	99
88	95
72	80
69	100

Perform a paired-samples t-test for this experiment (p = .01 or 99% confidence, two-tailed). Is there a significant difference between the two teaching methods?

SPSS output:

Results write-up:

Find the 99% confidence interval for the population difference of the two teaching methods. Explain how these confidence intervals confirm the results of the t-test conducted in SPSS.