 In many decisions, people do not only affect themselves, but also some other people. For example, doctors decide which medicine to prescribe to patients. These decisions do not only affect the doctor's profits, but the health of the patients. Another example would be in elections. So voters decide whether to vote for a policy or not, and this decision does not only affect her own tax payment, but also the social outcome of the election. In these decisions, if one knows that what is better for her or himself is also what's better for other people, then it's an easy decision. If the decision maker knows that what is better for herself is harmful to others, this will be a difficult decision for many people, because the decision here is whether you want to benefit yourself on the cost of others. So these are two extreme cases. In one, you know that being selfish harms others, and in the other one, you know that being selfish doesn't harm others. And there's a middle ground. That is when you don't know the consequences of your decisions on other people. And oftentimes it's the case in reality, and oftentimes we have information available for us to collect in order to learn about a situation. In this information acquisition, by collecting information, you can affect your beliefs. You can affect your beliefs about whether being selfish harms other people or does not harm other people. Our research question here is, in this situation, how are you going to acquire information? And in particular, how are you going to decide when you will stop collecting more information? So our first goal in this study is to review in the data how people acquire information in such decisions. What we use is a laboratory experiment. We invite people to the lab, and we first give them some money in dormants. Then we pair them two by two into pairs. So one of the pairs is always a decision maker. We call them the dictators. They decide between two options, and these two options have different consequences on both themselves and also this other person who we call the receiver. The receiver is complete passive. They don't make any decision. They just receive the consequences of the dictator's decisions. Let's say these two options the dictators can choose from are called X and Y. In our experiment, one of these two options will reduce the receiver's payment, but dictators don't know which one. What I do know is that option X is always better for themselves. It generates additional payments for the dictators themselves. Before making this dictator decision, the dictators can collect information about a harmful option to the receivers. This information we design is firstly, costless. They don't pay for it. Second, the information here is noisy. It contains some noise. This means that information can sometimes be correct, sometimes be incorrect, but most often it's correct. This means if they collect a lot of information, they will know the truth. But each piece of the information they collect can either tell them that X, the self-rewarding option, harms the other person, the receiver, or it can tell them that X, the self-rewarding action, doesn't harm the receiver. In the experiment, we record each piece of such information that is received by each dictator and a dictator's decision to continue or to stop acquiring information after each piece of information. This data allows us to analyze the information acquisition strategy of the dictators and in particular how these decisions to continue or to stop depend on the past information that the dictators have received. Besides the experiment, we also use a theoretical model to study why people do what they do in an experiment, why they acquire information in such a way. In the experimental data, we find that people base their decision to continue or to stop acquiring information on the previous information that they have already received. So if the previous information predominantly suggests that being selfish doesn't harm others, they tend to stop acquiring more information. But if the previous information suggests that being selfish would harm other people, they continue acquiring information. So this means that people don't always acquire all the information that is available to them. What they do is that they stop acquiring information when previous information generates some desirable beliefs for them, make them believe that they can behave selfishly without harming other people. But if previous information is the opposite, then they will seek information, they will continue acquiring information. In our theoretical model, we show that this information acquisition strategy would be optimal for a person who does not genuinely care about others but cares about her own belief that she doesn't harm other people. And at the same time, she wants to capture the actual profit for herself. For this person, if the previous information has made him believe that being selfish doesn't harm others, then she will stop acquiring information just to avoid future information that might be the opposite, that might destroy this desirable belief. But if she has previously received a lot of information suggesting that being selfish would harm other people, then she will continue hoping that future information will be different than the previous ones and will support her selfish decision. It's important to understand how people collect information in social decisions, and it is also very important to understand the consequences of this information acquisition strategies that people use. You might think that information acquisition strategies motivated by selfish interests must be bad for the society, must be bad for other people, but our data shows the opposite. In our data, if your selfish interest is not involved in a decision, you don't acquire that much information, so you often make a bad decision for the other person. But once your selfish interest is involved, remember we say that if you have received bad news, the news that you don't like before, you keep going. When this effect is large, then you can become more informed in your decision later, and consequently, you make a better decision for the other person. So the receiver in our experiment are home less often if you acquire information in this motivated way compared to if you acquire information as if your own interest is not involved in the decision. Today, people have access to rich information, and it's often at their discretion how to acquire this information, what information they acquire, determines what they believe, and what they believe determines what they do, what decisions they make. Therefore, information acquisition can be used by individuals as an instrument to influence their own behavior, and these behavior have outcomes on the society as well as the decision makers themselves. In our study, we investigate one aspect of information acquisition, that is how people decide when to stop acquiring information, but there are many other aspects of information acquisition. For example, from which information source would people like to acquire information? And we study one decision context where information plays a role, that is social decisions, decisions with outcome or with consequences on other people, but there are other contexts where information might matter. For example, whether people get themselves tested for deathly diseases and so on. So many researchers have studied different aspects of information acquisition in different contexts, but there is still room for more research to be done for us to understand how people utilize information and the ability to acquire information to influence their own decisions.