@Kamazeustra  
In the actual experiment the people are usually photoshopped together, so it’s exactly the same picture. The only differences are whether there are other people around, how many others, and where the person is placed relative to the rest of the crowd.
 
Though this being Pinkie Pie, one does have to wonder what she’s actually done.
 
 
@Background Pony #32F6  
Although there are some individual differences for specific trait preferences (some think it’s a way to avoid incest; you want someone similar to, but not too similar to, your parents), standards of general attractiveness are largely universal. “She’s very pretty, but she’s not my type.” A Pygmy tribesman who’s never seen a white person before would rate the relative attractiveness between paired caucasian fraternal twins the same way as white Americans or Europeans. Love is largely a way to get around that; once bonded, you no longer see your mate as “a male/female” rated against other similar individuals, but as “your mate <name>” who is then much more attractive to you than an identical-looking person of that sex would be if you didn’t already know and love her. (Or if the CMCs hadn’t given you a love poison.)
 
The Cheerleader Effect is different. Basically your brain is wired to perceive the group first, then drill down into the details of it, the individual people. We like groups, and we see that first. So that favorable impression rubs off on all the individuals once we start taking their individual faces in. Similar effects happen even with non-human objects. A circle surrounded by smaller circles looks bigger than the same circle surrounded by larger circles (The Ebbinghaus Illusion). We see first that the middle circle is “bigger” than its companions in the group, so our perception that it is “bigger” continues to taint our perception even when we look at it as its own unit. We see the same circle as larger when it was surrounded by smaller circles than when it was surrounded by identical ones. Even if we objectively know it’s exactly the same size, and both groups of circles are next to each other, the one surrounded by small circles still looks bigger than the one surrounded by identical circles.
 
It’s a survival feature. If you had to perceive and evaluate several individual lions in detail before building a mental construct that there is a group of lions in your path which you should avoid, your traits would be rapidly wiped out of the genome.

It also depends on the group. If someone in the crowd is more attractive by the viewer standards than the subject, they may prefer him or her when alone.

We can’t actually see the pictures. She could legitimately look better in the one where she’s with the others.