A good visualization takes a complicated issue and uses clever visual aids to make it more clear. The beauty of the visualization below is that with a few dots and lines, the basic process of science is illuminated from the first gathering of evidence to our closest approximations of truth.
Using this framework, let’s consider an example: evolution.
In the beginning, let’s say that we find fossil evidence of some organism. As we continue to diligently dig through the strata, we find more and more fossils that look slightly different, appearing to come from the same organism. But if these specimens are separated by geologic time, where are the differences coming from and why? We attempt to draw lines of theory between what we have found, positing that some overarching process is causing a change in morphology (namely, natural selection). Other researchers have similar ideas, and they too are offered up as an explanation for the fossils we found. For example, the idea that individual organisms can pass on traits that they acquire during their lifetimes (“Lamarckism”). But as we gather even more evidence (perhaps from genetics or biology), these other theories are ruled out as inconsistent with our observations.
We eventually whittle down our theory of evolution by natural selection to one that both agrees with our observations and experiments and makes the fewest number of assumptions (the famed principle of Occam’s Razor). We then apply our theory to other observations that we make, to see if our theory is consistent and can make predictions about the world. For instance, evolution by natural selection predicts that we will not find a morphologically modern fossil organism in the same geologic strata as an ancient one (think dogs and dinosaurs). Current evolutionary theory has survived such falsification, and gains explanatory strength.
In the end what we are left with is a picture of the world, informed by observation and experiment, that tries to approximate the truth as closely as possible. Most of our scientific theories are not complete, but because they work so well, make testable predictions about the world, and agree with past and present observations, we can assign temporary agreement between them and how the world actually works.