Marks - Any graphical symbol visible on a display medium - it is the most primitive building block that can encode some useful information. i.e the points, bars, lines on a graph.
Channels - The attributes of a mark, such as its position, shape, size, or colour. It’s called channels because it’s serving as channels in which we can encode underlying data values.
(Adapted from Ignatius & Senay, Visual Concepts)
There has been a lot of research an visual techniques. Studies have shown that for quantitative data, showing them through scales helps people much better understand the information. For example the scales that we’ve seen in the bar graphs, and this is in comparison to colour or area which were a little less accurate.
But for categorical data, using colours like red or blue actually works really well. However as you vary how many shapes and colours and data points you are using, the cognitive load increases and breaks down so in general
You should have about 5 different types of shapes and sizes at most in your visualizations!
Quantitative Data:
Data that can be counted or measured, and given a numerical value. Examples: distance, rainfall in ml, weight etc.
Ordinal Data:
This type of data is classified into categories within a variable that have a natural rank order. However, the distances between the categories are uneven or unknown. For example, socio-economic status (high-income, middle-income, low-income), or rating satisfaction (“extremely dislike”, “dislike”, “neutral”, “like”, “extremely like").
Nominal Data:
This is data that you can't really order in a meaningful way but can be labelled or classified into mutually exclusive categories within a variable. Examples: hair colour, country, race etc.