Exercise #2: Related-Samples t-tests

Surprised by her results on the first study, the same researcher decides to conduct another study focused on assessing the effectiveness of prayer. This time, she wants to avoid any human factor that could emerge as a confounding variable (e.g., placebo effect, discouragement etc.) so she used a sample of 20 bush babies (primates) who were infected with rhodesiense, a parasite that causes fever headaches, confusion, affects muscle coordination and disturbs the sleep cycle.These animals were diagnosed and then prayed for by a group of intercessors for 1 week, after which their symptoms were re-assessed to see if there was any improvement.Consequently, the symptoms of the primates were measured beforeand after the week of prayer.

The researcher would like to determine whether being prayed for made a difference in the health of the primates. In variable terms, the researcher wants to know if there is a mean difference between the before and after the week of intercession.

File titled "SPSS Assignment #2 - Related Samples T-Test data".

Step 1: Setting up your data

1. Input the data into SPSS or import it.
2. Label and code your data

before Width: 8

Decimals: 2

Label: health scores before prayer

Values: None

Missing: None

Columns: 8

Align: Center

Measure: Scale

Role: None

after Width: 8

Decimals: 2

Label: health scores after prayer

Values: None

Missing: None

Columns: 9

Align: Center

Measure: Scale

Role: None

For a repeated or paired-samples t-test, the normality tests are conducted on difference scores, not the values of the paired groups themselves. To accomplish this, you first need to compute the health scores difference and store them in a new variable.

Once that's done you will then instruct SPSS Statistics to perform the tests of assumptions on this newly created variable, which contains the difference scores. In this example, the new variable you will create will be called difference.

To make this new variable, follow the instructions from the following tutorial: "SPSS - Compute difference variable "

Note: You must deduct after from before., [before - after].

Step 2: Answer the following questions

1. What are the independent and the dependent variables?
2. What are the scales of measurements of the independent and dependent variables?
3. What are the null and alternative hypotheses?
4. Are the hypotheses directional? What kind of test should we use (one-tailed or two-tailed)?
5. Name the assumptions that must be met for a dependent-samples t-test (refer to your textbook)
   1. Has the first assumption been met? Why or why not?
   2. Conduct tests of normality using the difference variable. Copy and paste the Shapiro-wilks table Has the assumption of normality been met? Why or why not?

Tips: to conduct the normality tests with the difference variable, when you get to the 'Explore' tab, simply transfer the difference variable into the dependent list.