I think most of us is familiar with this feeling. You can tell someone (who stands behind your back) is staring at you without any physical evidence.
Is it possible or is it just a matter of coincidence? How is this phenomena called? Was it ever proven or disproven?
There are some (in)famous experiments done by Rupert Sheldrake who claims that a so-called "morphogenetic" field is responsible for this sort of thing. Alas, his experiments had quite sloppy methodology.
The feeling itself is real, as most here will testify. But it has nothing to do with being actually stared at/observed.
http://www.scientificamerican.com/article.cfm?id=ruperts-resonance
Apart from some technical problems with Sheldrake's experiment, here is a partial explanation for why some people really believe they can feel when they are stared at:
Second, psychologists dismiss anecdotal accounts of this sense to a reverse self-fulfilling effect: a person suspects being stared at and turns to check; such head movement catches the eyes of would-be starers, who then turn to look at the staree, who thereby confirms the feeling of being stared at. [...] When University of Hertfordshire psychologist Richard Wiseman also attempted to replicate Sheldrake's research, he found that subjects detected stares at rates no better than chance.
So the conclusion is no, people cannot really feel somebody is watching them.
Cornell psychologist Edward Titchener (who studied under Wundt) tested the ability to detect unseen staring and published in Science (1898). His conclusion: people cannot detect when they are being stared at, though a great many believe they can.
Though others have tested this `ability' over the years (e.g., Rupert Sheldrake), the results are mixed, and when statistically significant results are got, this is usually because of sloppy procedure.
Titchener would be rolling in his grave to know this sh*t is still being discussed.
Here's a short summary (source):
[Titchener] reported that over the years he had conducted a large number of informal tests and found no evidence for this particular claim. As far as Titchener was concerned people were not able to demonstrate their widely held belief.
He went on to provide a very good normal explanation for why people have this belief. First he noted that humans have forward facing vision, which leaves us exposed to the rear and he suggested that when in a situation where you are forced to present your back to a group of people, that there may be some psychological discomfort in that. He went as far to say that our ancestors must surely have devoted constant care to the defence of their backs. Titchener claimed that this back vigilance is the first element of staring detection: that people protect their backs by being aware of the environment behind them.
Once the feeling to turn around has formed, it is followed by an executing of the behavioural component: turning the head around and examining the back environment. Attention moves across the back of the room, scanning it to update their information as to what is actually going on.
Titchener next turned his attention to what might be happening behind the individual. He noted that his students could be engaged in a range of different behaviours (playing with their hair, eating food etc) and that it should be expected that some of the people sitting behind may be staring in the general direction of the individual. When they turn around, they disturb the visual field for those people who happen to have been looking in their general direction. This movement is a strong stimulus for people sitting behind, which they are required to attend to. There may in fact be any number of people who have the person in their visual field and who suddenly respond to the movement of the person looking around.
It is this coincidence which Titchener argued is the basis for the belief. Why do people feel the special tingling in their neck though? Titchener remarked that this no different from the feeling experienced in the bottom after sitting down for a long time!
Here's the short description of the methodology adopted in several up-to-date researches, as it's described in Science Set Free: 10 Paths to New Discovery by Rupert Sheldrake:
"Since the 1980s the sense of being stared at has been investigated experimentally both through direct looking and also through closed circuit television (CCTV). In the scientific literature it is variously referred to as “unseen gaze detection” or “remote attention” or “scopaesthesia” (from Greek skopein, to view, and aisthetikos, sensitive).
In direct-looking experiments, people work in pairs, with a subject and a looker. In a randomized series of trials, blindfolded subjects sit with their backs to the lookers, who either stare at the back of their necks, or look away and think of something else. A mechanical signal—a click or a beep—marks the beginning of each trial. Within a few seconds the subjects guess whether they are being looked at or not. Their guesses are either right or wrong, and are recorded immediately. A test usually consists of twenty trials. These tests are so simple that a child can do them, and thousands of children already have.
In the 1990s, this research was popularized through New Scientist magazine, BBC TV and Discovery Channel TV, and many tests were conducted in schools and as student projects at universities. Altogether, tens of thousands of trials were carried out. The results were remarkably consistent.
Typically, about 55 percent of the guesses were right, as opposed to 50 percent expected by chance. Although the effect was small, because it was so widely replicated it was highly significant statistically. In more rigorous experiments subjects and starers were separated by windows or one-way mirrors, eliminating the possibility of subtle cues by sound or even smell. They were still able to tell when they were being watched."
Dean Radin published a meta-analysis of 60 different super-visioned experiments conducted over 33,357 trial tests, titled The Sense of Being Stared At: A Preliminary Meta-Analysis. The result is a 54% chance of hit.
It is clear that the mean hit rate stabilizes to just over 54%, where 50% is expected by chance. The FEM weighted mean effect size was significantly above chance, e = 0.089 ± 0.003 (mean ± standard error), z = 32.5, p = 10–232, and the distribution of effects was significantly heterogeneous, Q = 763.3 (58 df), p = 10–232.
