FatCamera/E+/GettyImages
Pearson’s correlation coefficient, denoted as r, quantifies the linear association between two continuous variables. Its value ranges from –1 to +1, where –1 signals a perfect negative relationship, +1 a perfect positive relationship, and 0 indicates no linear correlation. Researchers typically compute r using statistical software such as SPSS or SAS to ensure precision, especially when reporting findings in peer‑reviewed publications.
Choose two variables that are measured independently to avoid bias. The first is usually the dependent variable, while the second is the predictor or exposure of interest.
For large datasets, manual calculation becomes impractical, so use software or a scientific calculator. The mathematical formula is available in the reference section below.
An r near zero suggests that the variables do not share a linear relationship. Such results can highlight variables that may not influence each other.
A positive r approaching +1 indicates a strong linear trend: as one variable rises, the other rises proportionally. Interpretation must consider the study’s context.
A negative r approaching –1 reflects an inverse linear trend: as one variable increases, the other decreases by a corresponding amount. Context is again essential.
Interpret r within the specific research question. For example, an r of 0.912 denotes a very strong positive association, which could suggest a causal link warranting further investigation. Conversely, the same r in a well‑established relationship might flag data errors or design flaws.
Determine statistical significance by comparing r to critical values from a correlation table. Degrees of freedom equal the number of paired observations minus two. Look up the critical value for α = 0.05 (95 % confidence) or α = 0.01 (99 % confidence). If |r| exceeds the critical value, the relationship is statistically significant.
Use confidence intervals for r to assess population correlations in addition to point estimates.
