Cross variables · Atlas Abierto

Level

Spearman correlation (ρ) for each pair of variables. n is the number of places with data on both.
	Mean net income per person2024	Rent · annual change (IPVA)2024	Tourist dwellings · % of all2026-05	Annual inflation · CPI2026-09
Mean net income per person	—	-0.18n = 48	-0.19n = 52	0.15n = 52
Rent · annual change (IPVA)	-0.18n = 48	—	0.48n = 48	0.11n = 48
Tourist dwellings · % of all	-0.19n = 52	0.48n = 48	—	-0.16n = 52
Annual inflation · CPI	0.15n = 52	0.11n = 48	-0.16n = 52	—

The correlation describes provinces, not people, and says nothing about cause. With about 50 provinces, values near zero are usually noise. If the two periods differ, the comparison is between different moments. Each pair uses only the places with data on both variables, so n can differ from one cell to the next.

The pair with the strongest correlation is Rent · annual change (IPVA) with Tourist dwellings · % of all.

Study report

This study measures whether rent · annual change (IPVA) is related to tourist dwellings · % of all across 48 provinces in Spain.

The data show a moderate positive relationship (ρ = 0.48) that is unlikely to be chance (p < 0.001), though with many exceptions.

Roughly, the two variables share 23% of their variation (ρ²). The rest depends on other factors.

The two variables come from different periods (2024 and 2026-05).

The result describes provinces, not people.

Two variables moving together doesn't show that one causes the other: a third factor may be behind both, and proving a cause needs a different kind of study.

What it means

Correlation

Measures whether two variables rise and fall together. It runs from −1 (one rises as the other falls) to +1 (they rise together). Near 0 there is no ordered relationship.

Spearman (ρ)

Correlation computed on the places' order rather than their values. It copes well with extreme values and curved relationships. It's the one used for the reading.

Pearson (r)

Correlation computed on the values. It measures whether points line up on a straight line. One very extreme place can move it a lot.

p-value

The probability of seeing a correlation this large if there were really no relationship. Below 0.05 it's usually called significant. It doesn't say whether the relationship matters.

n (places compared)

How many places have data on both variables. With few, the result is unstable; with many, almost any difference is significant.

Ecological fallacy

A correlation between provinces or municipalities says nothing about individuals. Richer municipalities having more older residents doesn't mean older people are richer.

Trend line

The straight line that best fits the points (least squares). It shows the overall direction, not a prediction for each place.

Rank chart

Replaces each value with its position (1st, 2nd…). It shows what Spearman measures, and extreme values don't dominate.

Correlation matrix