Correlation strength is the magnitude of a correlation — how strong the association between a factor and ranking position is — measured on its own, apart from whether that relationship runs positive or negative.
| Term | Correlation Strength |
|---|---|
| Category | Statistics and Evidence |
| Also known as | Signal Strength |
| Where it appears | Analysis results and legends |
What it means in RankGear
When RankGear correlates a factor against the ranking positions on a SERP, each factor earns a correlation coefficient that carries two separate pieces of information: a direction (positive or negative) and a magnitude. Correlation strength is that magnitude — how far the coefficient sits from zero, with the sign set aside. A factor scoring +0.42 and one scoring −0.42 have identical strength; they pull in opposite directions but with equal force. In analysis results and legends, RankGear uses this magnitude to order and shade factors so the strongest associations surface first, whichever way they point.
How to interpret it
Strength lives on a 0-to-1 scale. A value near zero means little linear association between the factor and rank; a value near one means a tight one. But magnitude alone is not the whole story — read it together with the sign (which direction the factor pulls), the sample size (how many results the SERP supplied), and whether the value is genuinely measured, missing, or a true zero. A strong-looking number built on only a handful of URLs is fragile, and two SERPs can yield the same average strength while differing in how spread out the underlying values are.
| Strength (magnitude) | What it tells you |
|---|---|
| Near 0 | Little to no linear association — the factor barely tracks position on this SERP. |
| Moderate | A visible association worth noting, but sample size and spread decide how much to trust it. |
| Near 1 | A tight association — the factor moves closely with position, though still comparative, not causal. |
Example
You run two analyses for the same keyword a week apart. Word count returns a correlation strength of about 0.30 in both runs, so at a glance the signal looks stable. Yet in the first run the top pages cluster tightly around a similar length, while in the second they scatter widely. The magnitude is the same, but the spread is not — and that difference is what tells you how confidently you can commit to a practical word-count target rather than treating 0.30 as a fixed rule.
Important considerations
- Correlation strength measures association, not cause. A strong magnitude never establishes that the factor makes a page rank.
- It is a comparative indicator drawn from the pages on one SERP, not a score Google assigns — read it relative to the other factors in the same run.
- Small SERP samples make magnitudes unstable; a high value on few results can vanish on the next run.
- Because strength ignores direction, always pair it with the sign so you know which way the factor actually pulls.
Related terms
Part of the RankGear glossary · how RankGear measures · the 870 factors.