PSYC 3000 Lecture Notes - Lecture 15: Type I And Type Ii Errors, Confidence Interval, Meta-Analysis

• School: Carleton University
• Department: Psychology
• Course: PSYC 3000
• Professor: Bruce Hutcheon

Document Summary

Confidence intervals and t test give exactly the same information. If the critical point is big then the ci will be big. Width of the ci is dependent on the sample size. The smaller the sample size, the bigger the ci. If the confidence interval fails to include the true population mean it would be a type i error. This will be determined by your alpha level. So 95% ci levels will have an alpha of 5%. If you lower your type i error, your ci levels become less precise. You will be more certain that your true population mean is within your interval, however the intervals will become larger. If a confidence interval embraces 0 than you have failed to reject the null hypothesis that treatment is useless. It means there is no statistical difference between the two samples. If the interval does not include 0 then you would reject the null hypothesis.