The population of internet users are rapidly increasing and the interest of the internet survey is also increasing. Recent years has seen a transition from traditional modes of data collection into internet survey. Some surveys are administered with mixed modes of traditional data collection methods and internet survey, and some surveys are conducted through internet only instead of traditional modes, such as telephone survey, postal survey, face-to-face interviews and so on. However, one of most crucial parts of a survey is the reliability of the collected data and internet survey is no exemption. Changwon National University has been annually conducting a survey of new students and transfer students with almost same contents of questionnaire. The survey is a longitudinal survey and it had been administered by paper-pencil surveys until 2009. In 2010 the survey was administered through internet. Every students has to login with student ID number and the last 7-digit of national identity registration number, and complete the 2010 survey before registration their courses. If they leave any question without being answered, then could not move to the registration site for courses. This study explores the distortion of responses using the new students survey of Changwon National University, which could occur when the survey responses are not confidential. We find that the distortion of responses occurs from the questions with social desirability pressure, pressure of winning favor with the researcher, and pressure of explaining their situations. There are no distortion of responses from the questions which are describing simple opinions or simple facts, for example, the place they plan to live while in school.