p-value
Hypothesis test output
A p-value is a statistical measure indicating the probability of obtaining results at least as extreme as those observed, assuming no real effect exists.
Low p-values (traditionally below 0.05) suggest the results are unlikely due to chance alone, leading researchers to reject the null hypothesis: The study found a p-value of 0.03, indicating statistical significance.
P-values don't measure the size or importance of an effect—only whether it's likely to be real.
Hypothesis check for averages
- ANOVACompares means among groups
- chi-squareStatistic for categorical data
- F-testVariance ratio statistic
- regressionReturn to former state
correlation · hypothesis test · parametric · nonparametric · degrees of freedom · sample size · omnibus · pairwise
First Greek letter
- thresholdPoint of beginning
- null hypothesisAssumed no relationship exists
- rejectDismiss as unsuitable
- type I errorRejecting a true null
false positive · statistical significance · confidence level · cutoff · accept · significant · nonsignificant · borderline
Allocation across recipients
- randomOccurring without pattern
- chanceLikelihood of occurrence
- likelihoodChance something will occur
- cumulativeIncreasing by successive additions
density · normal distribution · t-distribution · quantile · percentile · expected value · deviation · stochastic
Act of duplicating precisely
- reproducibilityAbility to repeat findings
- publication biasPreference for significant results
- peer reviewExpert evaluation before publication
- methodologyFormal procedural framework
empirical · evidence · findings · positive result · study design · experimental · observational · meta-analysis
t-test
Hypothesis check for averages
- ANOVACompares means among groups
- chi-squareStatistic for categorical data
- F-testVariance ratio statistic
regression · correlation · hypothesis test · parametric · nonparametric · degrees of freedom · sample size · omnibus · pairwise
alpha
First Greek letter
- thresholdPoint of beginning
- null hypothesisAssumed no relationship exists
- rejectDismiss as unsuitable
type I error · false positive · statistical significance · confidence level · cutoff · accept · significant · nonsignificant · borderline · marginal
distribution
Allocation across recipients
- randomOccurring without pattern
- chanceLikelihood of occurrence
- likelihoodChance something will occur
cumulative · density · normal distribution · t-distribution · quantile · percentile · expected value · deviation · stochastic
replication
Act of duplicating precisely
- reproducibilityAbility to repeat findings
- publication biasPreference for significant results
- peer reviewExpert evaluation before publication
methodology · empirical · evidence · findings · positive result · study design · experimental · observational · meta-analysis · preregistration
- experimentsexperiment, experiments, test, tests, trials
- research methodsexperiment, hypothesis, deduction, experiments, induction
- scientific methodexperiment, theory, data, hypothesis, observation
- statisticsmean, median, mode, variance, standard deviation
- data analysisstatistics, analysis, analytics, conclusions, mean
- biostatisticsANOVA, biostatistics, t-tests, confidence interval, correlation