I guess that depends on the kind of selection. Why should something be changed if it is advantageous ? So there are different kinds of selection, the one that is promoting change and the other which is stabilizing / conserving the current state.

Cheers Thomas

As Thomas says, I think you mean the stabilizing selection, which favours the selection of the average phenotype (for example, if the selection is made for length, the shortest and longest individuals will be selected against). However, that doesn't mean it does not lead to evolution, I think. It leads to a more homogenous population, indeed, but since natural selection is an evolutive mechanism, if there is selection, evolution is happening. I don't think we should use "change" and "evolution" as synonyms. Of course evolution wouldn't happen if there was no variability at all in the population, or if this variability didn't lead to a better chance to survive, but, how probable is that this happens in a real population? I guess in pluricellular organisms, the probability is very low.

thanks Ingrid, now it makes sense.

No. Natural selection is mostly a conserving force in biology - quite unlike how its supposed"discoverer" in Charles Darwin - described it to be.  Basically, the great majority of variations are deleterious (harmful) to the organism and so those affected by them will survive and reproduce less than those without them.

This is a common source of confusion.  Selection is a process of sorting among available phenotypes.  For example, in one generation, fatter individuals may be more likely to survive.  Evolution is a separate process that relies on inherited change.  If the fatter individuals that survive became fat only because of the environment in which they occur (for example, better access to resources), then their offspring will not necessarily be fatter, on average, than the original population.  Stated another way, selection is an ecological process that occurs within a single generation, whereas evolution involves genetic changes that occur between generations. 

Selection and evolution are two different processes.  Selection is in principle regulated by environmental situations while evolution is in principle regulated by genome. 

Selection is a weeding process. The majority of mutations are deleterious, with the remainder predominantly neutral.  By deleterious, individual fitness is impacted, affecting survival or reproduction.

Regarding evolution - Examine HW conditions. It's a spatiotemporal state. Is genetic diversity within the population stable, that is, are alleles present in the population at exact proportions every generation? Are all individuals capable of freely mating - meaning - are alleles evenly distributed throughout the population? Any condition affecting the frequency of alleles is evolution. At it's simplest level, it is just change.  Are alleles mutating? Chromosomal rearrangements? Epigenetic / environmental factors? If at any time the conditions change, and they affect individual fitness, then the population is undergoing evolution.

Natural selection can occur without producing evolution if differences among individual phenotypes have not a genetic base, in this case selection will not have evolutionary impact. See above Christopher's example of "environmental produced phenotype".