11.2 Confidence Intervals

11.2.1 One-Sample t Test

For one-sample t tests, R Commander produces confidence intervals automatically. Unfortunately, there is no way to produce this confidence interval in SPSS.

Watch the following video on how to interpret confidence intervals produced by R Commander for a one-sample t test. If you would like to see a transcript of this video, you can do so here.

Note: This video uses the Wong dataset from the carData package. If you need to review how to load the dataset into R Commander, review the video on R Commander basics. The section on how to import data begins at 3:32.

To report a confidence interval for a one-sample t test in APA format, add “95% CI for true mean [?, ?]” to the end of your results. You can use either “Inf” or the infinity symbol for infinity (∞).

A one-sample t test found that coma survivors (M = 87.6, SD = 15.1) had performance IQ scores that were significantly lower than 100, t(330) = -14.96, p < .001, 95% CI for true mean [-Inf, 88.9]

11.2.2 Correlation

For correlation analyses, R Commander produces confidence intervals automatically.

Watch the following video on how to calculate confidence intervals for a correlation test in SPSS, as well as how to interpret these confidence intervals. If you would like to see a transcript of this video, you can do so here.

Note: This video uses the Freedman dataset from the carData package. If you need to review how to load the dataset into R Commander, review the video on R Commander basics. The section on how to import data begins at 3:32.

To report a confidence interval for a correlation test in APA format, add “95% CI [?, ?]” directly after the correlation value.

There was not a significant correlation between city density and crime rate, r(98) = .11, 95% CI [-.09, .30], p = .27.