RESEARCH · METHODS · EVIDENCE
Critical values and test direction
Browse statistics lessons
- Statistics for research
- Center and spread
- Distributions and standard errors
- Statistical inference
- P-values and interpretation
- Critical values and test direction
- Compare means with t-tests
- Read t-test output
- Compare several means with ANOVA
- Read an ANOVA table
- Analyze categorical counts
- Read chi-square output
- Describe an association
- Read correlation results
- Interpret a regression model
- Choose a statistical method
- Statistics resources
A critical value is a cutoff from a reference distribution for a specified test, degrees of freedom, significance level, and direction. There is no universal rule that a larger raw test statistic is always significant.
For a two-sided t-test, compare |t| with the appropriate upper positive critical value; a lower-tailed test uses a negative cutoff. F and many chi-square tests use an upper tail. Prefer verified software output with the method and assumptions documented.
References: NIST two-sample t-test