All research proposals begin with a literature review. You need to read prior research about your topic in order to learn how to structure your study. Use library databases to find related journal articles, conference papers and books. Start with a broad topic, and then keep narrowing your search into more specific concepts as you go. For example, start with advertising, then narrow your search to advertising to children and finally narrow it to the psychological effects of advertising on children under 10. Find out how other researchers have studied this topic, and let that guide your research design.
Some topics require specific research designs. For example, you wouldn't conduct chemistry research the same way you would conduct history research. The hard sciences, including chemistry and math, always use quantitative research methods, which means they use numbers to describe a phenomenon. The social sciences, including history and psychology, use both quantitative and qualitative research methods, which means they can use statistics to describe something or they can use observation and interview techniques. Previous studies will help you determine if one specific method applies to your topic.
All research proposals are designed using a theoretical framework. When conducting your literature review, you should learn what existing theories were used and how they were previously applied. You can design a proposal that tests a theory or seeks to add knowledge to a theory. For example, if you want to study the psychological effects of advertising on children younger than 10, social cognitive theory can help you come up with appropriate research questions and testable hypotheses. The theory you use should be relevant to your topic.
Research design is largely established by the limitations of the researcher. You need to consider how much time you have to complete the study, your budget constraints, and your physical proximity to what you want to research. One researcher with a small budget can't conduct a national survey of children under 10, but one researcher can study a small sample of children under 10 at a nearby elementary school. A doctoral student in a three year program shouldn't propose a five-year study. Be realistic about what you can accomplish in the time you have.
Colleges and universities have Institutional Review Boards (IRB) made up of faculty and administrators to review research proposals that involve human subjects testing. This isn't limited to medical testing. If you plan to observe people, interview them or even give them a survey, you have to get permission from the IRB before you begin conducting your research. IRB requirements differ by institution, so visit your school's IRB website or talk to a faculty member to find out what you need in your proposal for IRB approval.