Page 56 - Quantitative Data Analysis
P. 56
Quantitative Data Analysis
Simply Explained Using SPSS
Type I and Type II Error Example - 1
Suppose a researcher is interested in examining the safety of a drug.
– H 0 : The drug is unsafe.
– H 1 : The drug is safe.
• What is the type I error?
– Reject H 0 when true; more specifically, conclude drug is safe when in-fact
it is unsafe.
• What is the type II error?
– Fail to reject H0 when it is actually false; more specifically, conclude drug
is unsafe when it is actually safe.
• In this example, which error is more harmful?
– Would like want to make the likelihood of a Type I Error to be very small.
Type I and Type II Error Example - 2
• Suppose a medical researcher is testing blood to determine if it is
appropriate for use (contaminated or not contaminated)
– H 0 : The blood is not contaminated.
– H 1 : The blood is contaminated.
• What is the type I error?
– Conclude blood is contaminated when in fact it is not.
• What is the type II error?
– Conclude blood is not contaminated when it is contaminated.
• In this example, which error is more harmful?
– Type II Error is probably worse here.
• What if there was a blood shortage?
– If there was a serious shortage, you may take the increased type II error
risk as opposed to not getting the blood transfusion.
The Theory and Applications of Statistical Inferences 40