Take a look at "It's Who You Know." Find the part where the author writes "How could this have happened?" What effect does this question have on readers? (this is the question) (this is the article) PHILADELPHIA, Pennsylvania. People often say it's not what you know; it's who you know. Now, however, scientists are finding that in many cases, it's also who the people you know know. Your social network can influence the choices you make. So can the networks of people in your network. These choices include how you vote, and even what you like and dislike.

Michael Kearns is a computer scientist. He has studied the links between social networks and how people behave. In the past, most networking experiments had been done using computer models. Kearns, however, has developed methods for testing real people.

In one of his latest experiments, Kearns explored whether a small group of people can influence the votes of a large group. Kearns first gathered a group of 36 students. He then created several networks within the group. He did this by placing each student at a work station. Each work station was linked to several other students. Some networks contained 19 students. Other networks had as few as 3.

Kearns then told all 36 students that they would be voting for a color. They could vote for either red or blue. If everyone in the group could agree to vote for the same color, everyone would receive a reward. If the group failed to agree, however, no rewards would be given. Kearns held several rounds of voting. This was so that he could compare the results.

Kearns encouraged students to vote for a particular color. He did this by offering incentives. Kearns offered some students $1.50 for each round that red won. He offered these same students 50 cents for each round that blue won. The rewards were the opposite for other subjects.

There was a tension in the air, Kearns said. The students all wanted to agree. But they also wanted everyone to agree on their color.

The group had only one minute to decide. Still people came to some agreement in 55 out of 81 votes.

The larger group did not always win, however. In one round, for example, Kearns promised 6 students the $1.50 reward if red won. He promised 30 students $1.50 each if blue won. In this vote, blue should have won. More people wanted to vote for blue. However, red won the vote.

How could this have happened? The six people wanting red to win belonged to larger networks. They knew more people. This helped them.

People with large networks can influence others, Kearns said. These people can change how a vote turns out. This is the case even if the large group has a strong reason to go the other way.

Social networks have also been found to influence people's tastes. Duncan Watts is a networking expert at Yahoo! Inc. Not long ago, he ran a networking experiment. Watts first recruited 14,000 people and asked them to rank 48 songs. These songs were new. The people had never heard them before. The people saw one another's choices. The study showed that they changed their votes to match the group's choices. Watts then divided the people into eight groups. He did not allow the groups to intermingle. Then results ranged greatly between the groups. For example, one group's top choice might be ranked 42nd by another group.

"We [believe] things are popular because that's what people want," Watts said. However, he added, "This is showing that's wrong."

Other networking studies are also being done. These studies are finding that social networks don't only influence voting choices and tastes. They also shape people's buying choices. They shape their habits. They even shape their levels of health and happiness.