It is commonplace between scientists and lay people alike to treat a set of objects as a singular group. For example, if I buy a box of cheerios, I don't say I just purchased 2,000 cheerios. I say I bought a box of cheerios. Simple enough and justified. Similarly, when I refer to the inhabitants of say, India, we just call them Indians. Or a trunk, root system and leaves as a tree. Again, justified. It seems to be natural for humans to group things together and refer to them as one. However, this grouping soon breaks down when used beyond anything but a convenience device. For example, if you refer to Indians as having such and such a property that is discrimination (or racism). If you say all cheerios are round, you are clearly wrong. Some cheerios are malformed. Others are broken during shipping and by the time they reach you are nothing more than dust or arcs.
The problem is even though we group things together, they still retain individuality. An Indian is an Indian no matter what. But attaching properties to a group name is fallacious on the grounds that you cannot take empirical evidence and draw a priori conclusions from it. In other words, you must not draw general conclusions from experience. Kant (and many others I'm sure) points this out in the "Critique of Pure Reason". He also points out that the universe can be split into two basic halves: the half that exists, and the half that doesn't. In other words the half that exists is the stuff of the universe: matter. The half that doesn't exist is thought. Ideas. Things like justice, harmony, friendship, scientific theory, etc. They don't have a tangible representation in space. Likewise, matter has no component of thought. It is literal. It is the wonder of the human mind that brings the two together to form understanding. All contingent of course, because thought is just that: thought. Ideas are a way for us (humans) to make sense of the world and thought never represents what is real or the way things really are. And this is why grouping is dangerous. We do it. It's absolutely necessary. But we must be careful how we apply it and we must never forget the practical (real) aspect of the things we apply our thoughts to. Because when you forget that the real exists, you cease to have any understanding of it.
To illustrate I will give you two analogies. The first is about a wall of bricks. The second is a flue epidemic in England. A brick wall is made up of (no surprise) many individual bricks. But we still refer to it as the singular 'wall'. Fair enough. We even refer to the bricks as all being the same. And in many respects they seem to be. In an idealized fashion they are the same shape, same structural characteristics, same chemical properties of bonding, etc. They fit the same mold. Or in other words they embody the idea of what a brick is. But when bricks are made, the raw materials are slapped into a mold, cooked and cured. On average the mix of raw materials is the same, but in practice they do vary slightly. The distribution might vary. The exact placement of molecules and elements is different from brick to brick. Each brick is unique and has its own individuality. So the brick wall is actually composed of many different, unique bricks. When a car crashes into this wall, the entire wall doesn't collapse or cease to exist. It takes the brunt in the bricks. Some are smashed, some aren't. Some fall down, some don't. When we repair the wall we don't replace the whole thing. Some bricks will have to be replaced, and some will not. So we see that it is convenient to refer to the stack of bricks as just a 'wall'. But when it comes to practical application of anything to that wall (crashing into it or repairing it), we absolutely must refer to its individual components.
Pretend a flu epidemic strikes England and many, many people become ill. Does this mean that every single person in England has the flu? No, that's preposterous you say. Now pretend there is a particular social problem endemic to the black population of the United States. Let's say this problem is high teen pregnancy. Does this mean that all black teens get pregnant? Likewise, does this mean they are all crack-whores, or that their men are thieves? Of course not for the exact same reason that not all English people have the flu.
Making assumptions based on groups is false. There are no circumstances under which it is acceptable to treat people or things based upon their membership in a group (willing or unwilling). It is important that we remain cognizant in how we characterize groups because there is a fairly good chance we can be proven wrong. In addition we, as humans, should shift our focus away from grouping things together and towards treating things as individuals, especially other people.