As a rule, professional pollsters, opinion testers, and survey takers solicit thousands of individuals when exploring the answers to a question. They are chosen to represent either a certain segment of society or a broad range of the populace diversified in geography, income, ethnic background, and education.
The purpose of a survey may be to inform a manufacturer when test-marketing a new product or identifying a new market. Politicians use them to plan their campaigns and judge the mood of their constituents. Regardless, they are widely used because they deliver large stores of useful information quickly and efficiently.
Few student research-writers conduct such extensive surveys as the time, money and effort required is prohibitive. For smaller, less prohibitive surveys then, it is best to report the results of your survey in non-statistical terms.
It's one thing to say that "many of the students" who filled out a questionnaire on reading habits hadn't read a newspaper in the past month; it's another to claim such is true of "seventy-two percent" of the student population when you were only able to give questionnaires to the twenty-two percent who were in the dining hall the day you were there and half of those threw them in the trash on the way out.
A far more useful and reliable way to use a questionnaire is to think of it as a group interview. Use it when you want to collect the same information from a large number of people or when you're more interested in what a group thinks as a whole than what a particular individual thinks. Treat the information you collect as representative and use your findings to build an overall knowledge of the subject or to cull them for interesting or persuasive details and quotations.