Statistical Hypothesis Testing Decision Error
(Redirected from Statistical Decision Error)
Jump to navigation
Jump to search
A Statistical Hypothesis Testing Decision Error is a inference classification decision error that occurs when a statistical hypothesis test reaches an incorrect conclusion about the truth or falsity of a null hypothesis.
- AKA: Statistical Testing Error, Hypothesis Test Error, Statistical Decision Error, Inference Error.
- Context:
- It can typically manifest as Type I Error (rejecting true null) or Type II Error (failing to reject false null).
- It can typically arise from Random Sampling Variation even with correct statistical procedures.
- It can typically be controlled through Error Rate Selection and Sample Size Determination.
- It can often result from Test Assumption Violation in real-world data applications.
- It can often increase with Multiple Testing without appropriate error correction.
- It can often trade off between error types through significance threshold adjustment.
- ...
- It can range from being a Conservative Hypothesis Testing Decision Error to being a Liberal Hypothesis Testing Decision Error, depending on its threshold stringency.
- It can range from being a Single-Test Hypothesis Testing Decision Error to being a Multiple-Test Hypothesis Testing Decision Error, depending on its testing context.
- It can range from being a Parametric Hypothesis Testing Decision Error to being a Non-Parametric Hypothesis Testing Decision Error, depending on its test methodology.
- It can range from being a Frequentist Hypothesis Testing Decision Error to being a Bayesian Hypothesis Testing Decision Error, depending on its statistical framework.
- ...
- It can impact Research Validity through incorrect scientific conclusions.
- It can influence Clinical Decision Making through misidentified treatment effects.
- It can affect Quality Control Processes through incorrect defect detection.
- It can contribute to Replication Failure in scientific research.
- ...
- Example(s):
- Type I (False Positive) Error), such as:
- Ineffective Drug Approval Error concluding drug works when it doesn't.
- False Discrimination Finding Error detecting bias where none exists.
- Spurious Correlation Error finding relationship in random noise.
- Type II (False Negative) Error), such as:
- Missed Treatment Effect Error failing to detect real therapeutic benefit.
- Undetected Quality Problem Error missing actual manufacturing defect.
- Overlooked Association Error not finding true variable relationship.
- Multiple Testing Errors, such as:
- Family-Wise Error with at least one false positive among multiple tests.
- False Discovery Error with incorrect discoveries in large-scale testing.
- Data Dredging Error from extensive hypothesis searching.
- ...
- Type I (False Positive) Error), such as:
- Counter-Example(s):
- Correct Statistical Decision, where test conclusion matches true state.
- Computational Error, which involves calculation mistakes rather than decision errors.
- Measurement Error, which affects data quality rather than inference decision.
- Model Specification Error, which involves incorrect model rather than test decision.
- See: Statistical Hypothesis Testing Task, Type I Hypothesis Testing Error, Type II Error, Statistical Power, Error Rate, Multiple Testing Problem, Statistical Decision Theory.