Only adjacent-axis relations read directly

What it is

An adjacent-axis relation appears directly in the segment texture formed by paired records between two axes. With consistent directions, strong positive association often forms roughly parallel bundles, while strong negative association produces systematic crossings; nonlinear or grouped relations can yield more complex textures. Nonadjacent dimensions remain joined within each polyline but have no local region showing their paired distribution in aggregate. “Only” concerns direct ensemble reading: highlighting one record can still support tracing values across distant axes.

Why it happens

Each inter-axis region encodes one dimension pair, confining crossing density, angles, and bundles to adjacent pairs. Tracing many polylines across intervening axes requires sustained identity and invites mismatches after crossings. Texture also depends on direction: reversing one axis can turn a crossing pattern into a nearly parallel one without changing the data relation. Independent scaling, outliers, and axis spacing alter segment angles, so neither one slope nor a fixed crossing threshold directly estimates a correlation coefficient.

Studying it

Use data with known linear, nonlinear, and grouped relations, placing target dimensions adjacent or apart while varying inversion, scaling, spacing, record density, and highlighting. Test correlation direction, strength ordering, anomalous relations, and record tracing separately; measure accuracy, time, confidence, and mismatches. Numerical correlations or models provide ground truth only for the defined relation. Pearson correlation is not the sole truth for nonlinear patterns or mixed subgroups.

Where it stops holding

Adjacency does not guarantee readability: overplotting, missingness, outlier compression, arbitrary categorical order, and mixed groups can obscure a relation. Axis inversion can reduce crossings or expose a pattern but needs persistent direction labels, or readers may treat upward as universally higher. Scatterplot matrices or correlation matrices cover all pairs more systematically. Adjacent texture adds no evidence for causal interpretation.

Applying it

• List the dimension pairs required by the decision and place priority pairs adjacent. Use stable presets or linked views when one order cannot satisfy them all.
• Keep high–low direction, units, and inversion state visible at every axis; provide clear feedback and undo after a direction change.
• Offer an on-demand scatterplot or numerical summary for an axis pair to check whether a bundle comes from outliers, grouping, or transformation.
• Mark missing segments explicitly; do not interpret fewer crossings caused by broken lines as a stronger relation.
• Test adjacent versus distant placement and normal versus reversed directions, confirming that readers know the relationship does not change with layout.

Related

• Same group: U2.14.1 Parallel coordinates render each record as a polyline across axes · U2.14.3 Axis order decides which patterns can be found · U2.14.4 More records, more overplotting, less structure · U2.14.5 Each axis needs independent normalization
• Nearby: U2.05.1 Scatterplots suit the relation between two continuous variables · U2.10.2 Row and column ordering decides whether structure is visible
• Search terms: adjacent axes · axis-pair pattern · line crossings · axis inversion