p-values
Probability under null
A p-value is a statistical measure indicating how likely your data would occur if there were no real effect.
Low p-values (typically below 0.05) suggest the result is unlikely due to chance alone: The study found a significant effect, with a p-value of 0.003.
P-values are widely used in scientific research but frequently misunderstood—a low p-value doesn't prove an effect exists, only that the data would be surprising if it didn't.
Testable proposed explanation
- null hypothesisAssumed no relationship exists
- alphaFirst Greek letter
- thresholdPoint of beginning
- significanceQuality of mattering
t-test · ANOVA · chi-square · regression · correlation · false positive · false negative · confidence interval
Quality of mattering
- thresholdPoint of beginning
- cutoffPoint of termination
- arbitraryBased on personal whim
- borderlineEdge of acceptability
marginal · nonsignificant · reject · evidence · conclusion · inference · misleading · publication bias
Deliberate learning effort
- experimentSystematic investigation procedure
- findingsConclusions from investigation
- resultsWhat follows from effort
- methodologyFormal procedural framework
replication · publication · empirical · validation · sample size · reproducibility · research design · data collection
Allocation across recipients
- likelihoodChance something will occur
- chanceLikelihood of occurrence
- randomOccurring without pattern
- samplingExtracting subset for study
expected · observed · cumulative · continuous · discrete · normal distribution · t-distribution
Improper application
- misinterpretationIncorrect understanding of meaning
- overrelianceExcessive trust in something
- confidence intervalsBounds for estimated parameters
- reformImprove by correcting faults
hypothesis
Testable proposed explanation
- null hypothesisAssumed no relationship exists
- alphaFirst Greek letter
- thresholdPoint of beginning
significance · t-test · ANOVA · chi-square · regression · correlation · false positive · false negative · confidence interval · degrees of freedom
significance
Quality of mattering
borderline · marginal · nonsignificant · reject · evidence · conclusion · inference · misleading · publication bias · selective reporting
study
Deliberate learning effort
- experimentSystematic investigation procedure
- findingsConclusions from investigation
- resultsWhat follows from effort
methodology · replication · publication · empirical · validation · sample size · reproducibility · research design · data collection · clinical trial
distribution
Allocation across recipients
- likelihoodChance something will occur
- chanceLikelihood of occurrence
- randomOccurring without pattern
sampling · expected · observed · cumulative · continuous · discrete · normal distribution · t-distribution
misuse
Improper application
- misinterpretationIncorrect understanding of meaning
- overrelianceExcessive trust in something
- confidence intervalsBounds for estimated parameters
- statisticsmean, median, mode, variance, standard deviation
- biostatisticsANOVA, biostatistics, t-tests, confidence interval, correlation
- Clinical Researchclinical guidelines, randomized trials, double-blind, placebo, control groups
- Statistical MethodsANOVA, confidence intervals, correlation, covariance, hypothesis testing
- Model Diagnosticsresiduals, diagnostics, leverage
- researchstudy, research papers, researching, archive, peer review
p-valuesmathematical proof
Logical demonstration of truth
certainty · definite · absolute truth · proof · certain
p-value, value, values, valued, valuing, valuable, valuables, invaluable, valueless