Wk 4 Discussion: The Importance of Relationships

Identify an interval-ratio variable from the data set that you will use for this assignment

Select Data from the top menu

Click on Select Cases

Click on Random Sample of Cases

Click on Sample

Click on Exactly

Enter 100 (or 400) cases from the first 1000

Click Continue

Click Okay

Note: your GSS set has now randomly selected 100 (or 400) cases for analysis

Select Analyze from the top menu

Click Compare Means from the dropdown menu

Select One-sample t-test

Drag your interval-ratio variable into the Test Variable(s) box

Enter 0 in the Test Value box

Click Options

Enter 95% in Confidence Interval Percentage (this is the SPSS default value)

Select Exclude Cases Listwise

Click Continue

Your Confidence Interval is in the One-Sample Test output.

Note 1: your sample size, N, may be less than 100. This is because the random sample may include cases that have missing values for your variable.

Note 2: repeat this process to select a random sample of 400.

Note 3: change your Confidence Level from .95 to .99 in the One Sample utility.

Note 4: Do not save changes to your GSS Dataset. If you do so your data set will be reduced to 100 or 400 cases.

Note 5: Make sure your variable is interval-ratio (or 'scale' in SPSS lingo; check the Measure column in the data set). 

Note 6: You should have 4 sets of SPSS Output in your discussion post:

Set 1: N = 100 and CL = .95

Set 2: N= 400 and CL = .99

Set 3: N = 100 and CL = .99

Set 4: N = 400 and CL = .95

Note 7: Discuss the relationship among the Confidence Interval, the Confidence Level and the sample size.

-What happens to the CI when the Confidence Level increases or decreases?

-What happens to the CI when the sample size increases or decreases?