Related Experiment Video
Updated: Feb 9, 2026

Evaluation of the Curing of Adhesive Systems by Rheological and Thermal Testing
Published on: July 3, 2020
Manipulating the Alpha Level Cannot Cure Significance Testing
David Trafimow1, Valentin Amrhein2,3, Corson N Areshenkoff4
1Department of Psychology, New Mexico State University, Las Cruces, NM, United States.
Discarding statistical significance testing and fixed alpha levels (p-values) is crucial for scientific progress. Relying on cumulative evidence from multiple studies, rather than binary decisions, fosters genuine discovery and robust findings.
Area of Science:
- Statistical methodology
- Philosophy of science
Background:
- Traditional hypothesis testing relies on arbitrary alpha levels (e.g., p=0.05).
- Recent proposals suggest lowering the alpha level to p=0.005.
- Both fixed and variable alpha levels present significant drawbacks for scientific discovery.
Purpose of the Study:
- To argue against the use of significance testing for accept/reject decisions in scientific hypotheses.
- To advocate for the abandonment of traditional p-value thresholds.
- To promote a shift towards cumulative evidence and robust study design.
Main Methods:
- Critical analysis of the limitations of significance testing and p-value thresholds.
- Discussion of alternative approaches focusing on study design and sample size.
- Emphasis on the importance of cumulative evidence from multiple independent studies.
Main Results:
- Significance testing, including proposed changes to alpha levels, is detrimental to scientific discovery.
- Blanket and variable alpha levels are equally problematic for advancing science.
- Sole reliance on p-values oversimplifies complex scientific inference.
Conclusions:
- Dispensing with significance testing altogether is a sensible approach.
- Inference should be based on cumulative evidence, considering factors beyond p-values.
- Rigorous evaluation of evidence requires assessing auxiliary assumptions, study design, and application implications, not just binary p-value cutoffs.
Related Concept Videos
Significance Testing: Overview
Critical Region, Critical Values and Significance Level
In hypothesis testing, a sample statistic is converted to a test statistic using z, t, or chi-square distribution. A critical region is an area under the curve in probability distributions demarcated by the critical value. When the test statistic falls in this region, it suggests that the null hypothesis must be rejected. As this region contains all those values of the...
Statistical Significance
Curing of Concrete
Curing Methods
Accelerated Curing of Concrete

