| Term | Practical definition |
|---|
| Alpha (α) | A chosen long-run Type I error rate for a statistical testing procedure, often .05. |
| Bias | Systematic deviation from the target quantity or truth. |
| Confidence interval | An interval produced by a procedure designed to cover the target parameter at a stated long-run rate. |
| Confounder | A variable that, under a causal model, influences both predictor/exposure and outcome. |
| Construct | A theoretical attribute not always directly observable. |
| Effect size | A measure of the magnitude of a difference or relationship. |
| Estimator | A rule or procedure used to estimate a population quantity. |
| Hypothesis | A testable proposition about a population quantity or relationship. |
| Independent observations | Observations whose dependence structure is adequately absent or accounted for in the model. |
| Operational definition | The procedure used to measure or classify a construct. |
| p-value | Under a specified null model and assumptions, the probability of results at least as incompatible with the null as those observed. |
| Power | The probability that a specified test rejects the null under a specified alternative, given its assumptions. |
| Reliability | Consistency or precision of a measurement procedure. |
| Sampling frame | The list or mechanism from which sample units are selected. |
| Standard deviation | A measure of spread among observations. |
| Standard error | An estimate of sampling variability of a statistic or estimator. |
| Validity | The degree to which evidence supports an intended interpretation or use. |
| Variable | A recorded characteristic that can take different values. |