Glossary of statistics terms
Plain-language definitions of the terms you meet in StatLab output.
- Case
- One row of data — usually one person or one observation.
- Variable
- One column of data, such as age or gender.
- Value label
- Text shown for a numeric code (1 = Male).
- System-missing
- A blank cell; no value.
- User-missing
- A real code you declared as missing (e.g. 9 = no answer).
- Scale / ordinal / nominal
- Measurement levels: numbers with equal intervals / ordered categories / unordered categories.
- Mean, median, mode
- Average; middle value; most frequent value.
- Standard deviation (SD)
- Typical distance of values from the mean.
- Standard error (SE)
- SD of the sample mean across repeated samples: SD/√n.
- Confidence interval (CI)
- A range that would contain the true value in 95% of repeated samples.
- p-value (Sig.)
- Probability of results at least this extreme if the null hypothesis were true.
- Degrees of freedom (df)
- Number of independent pieces of information behind a statistic.
- Effect size
- How large an effect is, independent of sample size (d, r, η², V).
- Null hypothesis
- The “no difference / no relationship” statement a test tries to reject.
- Skewness
- Asymmetry of a distribution; 0 = symmetric.
- Kurtosis
- Heaviness of the tails compared with a normal distribution (0 = normal).
- Outlier
- A value far from the rest; in a boxplot more than 1.5 box lengths outside the box.
- Listwise / pairwise deletion
- Dropping a case if any variable is missing / only for the pairs where it is missing.
- Post-hoc test
- Pairwise comparisons after a significant ANOVA, adjusted for multiple testing (Tukey HSD).
- Homogeneity of variance
- Groups have similar spread; checked with Levene’s test.
- Multicollinearity
- Predictors that are strongly correlated with each other; checked with VIF.
- Cronbach's alpha
- Internal-consistency reliability of a set of items.
- APA style
- The American Psychological Association’s reporting format for statistics.