Evolution is such a complex dynamic. That it is the intersection, the outcome of the interplay of two separate factors, random mutation and natural [environmental] selection, is sufficient to defeat understanding on the lay level much of the time.
Misrepresentation, such as the depiction of an undifferentiated parade of creatures from fish through to monkey to human, of itself breeds much of the misconception. (two misleading aspects of that parade: its portrayal in terms of modern creatures only, and the lack of bifurcations.)
And without exposure to some of the underlying narratives, it's easy to make assumptions that are unwarranted. I've discussed complexity before (from Gould): how evolutionary increase in complexity of organisms is not a universal trend (some decrease in complexity; some remain remarkably stable over thousands of millenia). Gould discusses this at even greater length in his book Life's Grandeur; there is a more succinct discussion of this at New Scientist.
New Scientist has a whole set of evolution myth-busting articles, worth a scan through. Some of the useful points it makes:
- natural [environmental] selection isn't the only change mechanism - simple genetic drift can result in major changes. This particularly happens with small populations (bottlenecks). New Scientist claims this as the cause of most differences between humans and other apes*.
- adaptions don't have to be perfect. Good enough to survive is enough. Plenty of examples (such as the human eye) are less efficient than they could be (the octopus eye, a case of convergent evolution, developed more efficiently).
- not everything is explicable simply as an evolutionary adaption. The NS article on this discusses male nipples, smell, and behaviours as examples, giving reasons such as unpressured chance, side-effects of adaption, and vistigiality.
- evolution doesn't validate dog-eat-dog selfishness (as some arch-capitalists may posit). 'Survival of the fittest' - not a term from Darwin - is more about fitness for purpose than dominance; co-operative actions can often better favour survival.
- evolution doesn't axiomatically favour survival. The article discusses, inter alia, 'evolutionary suicide'.
That set of articles usefully refines understanding of many other points at risk of debasement by misunderstanding.
* I should note that recently reported claims of a major bottleneck in the human population, reduced to only 2,000 people at 70,000 years ago, have been disputed in a substantial article here. However, this point was not one of evolutionary change in humans, who have been stable as homo sapiens sapiens for a few million years.