The most obvious is to be selective in the data that is used to make a graph. Suppose I want to show how much more money I am making, I might chose to show my income. However, if I really want to show that I was doing well, I should also show my expenses. I can also be selective in the time period or in the way I average the data. In the previous example, perhaps I show my income as dollars per month, and don't show how many months I worked. The graph might look like I'm making more, but the opposite could be true if you don't know how many months I have been working. If you want to show that students arrive late to school, you might pick a period where there is a bad storm. You can change the scales on your graph to distort the correct interpretation. See related link for more examples.
Chat with our AI personalities
label the axises
Don't make graphs misleading!As for the answer: different scales, leaving out points, drawing extra lines with no meaning, confusing labels, ...Most graphs you see online are misleading, few are really good.
They usually contain a "break" in the graph, which would be on the left side of the graph.
In the case where graphs are of 3d objects or curves, sometimes it is hard to see what shape an object is, or how the curve is moving based on the perspective. You can fix this (at least somewhat) by drawing the graph from different angles or by presenting the graph within a cube
1. when its too detailed 2. does not start at zero 3. does not have equal intervals there are several more, but that's all i remember