"It turns out that we are very bad at judging who our friends are," said Erez Shmueli from Tel Aviv University in Israel.
"And our difficulty determining the reciprocity of friendship significantly limits our ability to engage in cooperative arrangements," said Shumeli, who conducted the study in collaboration with researchers from Massachusetts Institute of Technology in the US.
"We learned that we cannot rely on our instincts or intuition. There must be an objective way to measure these relationships and quantify their impact," he added.
They examined six friendship surveys from some 600 students in Israel, Europe and the US to assess friendship levels and expectations of reciprocity.
Also Read
They then developed an algorithm that examines several objective features of a perceived friendship (that is, the number of common friends or the total number of friends) and were able to distinguish between the two different kinds of friendship - unidirectional or reciprocal.
"We found that 95 per cent of participants thought that their relationships were reciprocal. If you think someone is your friend, you expect him to feel the same way," said Shumeli.
"Reciprocal relationships are important because of social influence. In this experiment that analyses different incentives for exercising, we found that friendship pressure far outweighed money in terms of motivation," said Shumeli.
"We found, not surprisingly, that those pressured by reciprocal friends exercised more and enjoyed greater progress than those with unilateral friendship ties," he said.
Researchers found that their "friendship algorithm" determined with an extremely high level of accuracy the reciprocal or unidirectional nature of a friendship.
The findings were published in the journal PLoS One.