Visualisation Techniques

Learn about bar charts, scatterplots and more.

Scatterplot what

at least two quantitative attributes

Scatterplot how

express values as point marks using horizontal and vertical position

Scatterplot why

finding correlations between two attributes, providing overviews and characterising distributions

Scatterplot scale

hundreds

Scatterplot matrix why

find pairwise correlations, trends and outliers

Scatterplot matrix scale

less than 12 attributes, data points in the hundreds

bar charts what

one categorical attribute and one quantitative attribute

bar charts how

value attribute expressed through length and aligned in vertical position, nominal dimension separated through horizontal position

bar charts why

for lookup, comparing values and presenting trends

bar charts scale

typically between several to a dozen bars, limited to hundreds

bar charts consideration

if accuracy is important and averages are being used, add error bars and make it clear what metric is being shown

stacked bar charts what

two categorical attributes and one quantitative attribute

stacked bar charts how

bar glyphs where lengths represent the value of each category and are stacked on top of each other

stacked bar charts why

part of a whole relationship using absolute values, used to lookup values and find trends

stacked bar charts limitations

comparing the higher sub-bars is hard as there is no shared baseline

stacked bar charts scale

limited to hundreds of primary categories and a dozen secondary categories

dot chart/ dot plot what

one quantitative and one ordinal attribute

line chart what

same as dot chart with a line connecting the points

line chart why

good for showing trends

line chart scale

hundreds