A research project always begins with a question, something the researcher wants to know more about. Often, that researcher has also developed a hypothesis, a general answer to the question that needs to be verified or disproved through research. Psychological researchers employ many different research methods, including case studies, questionnaires, observation, and laboratory experiments. We can look at each of these methods through the lens of a possible research question.
Let's say a researcher wants to figure out how the behavior of only children differs from the behavior of children with siblings. The researcher develops the hypothesis that only children are more responsible and show fewer behavior issues, but he needs to verify that hypothesis.
One of the researcher's chosen methods is the case study. He focuses in on a single family and observes them closely over several weeks, interviewing the mother, father, and child, spending time in their home, and talking with the child's teachers. The researcher records his impressions meticulously, knowing that they will prove to be useful data. The researcher also realizes that a single case study is not enough to prove or disprove his hypothesis. He will need several other case studies.
The researcher also requires a broader range of data that he decides to get from a questionnaire or written survey emailed to a large group of families, both those with only children and those with multiple children. The researcher constructs his questionnaire carefully, asking for information about children's behavior patterns, family structure, discipline methods, expected responsibilities within the family, and the fulfillment of responsibilities. The researcher realizes that the data provided by his questionnaire will be more general and less reliable than that provided by his case studies, for questionnaires require subjects to report their own behaviors and can, therefore, be less accurate.
Since the researcher would like more data from his own observations, he sets up a session in which he invites several children, some only children and some siblings, to play together for an afternoon. The researcher posts himself in a corner out of the way, and the children quickly forget about him. He watches closely, observing how each child behaves and taking copious notes. The data he receives is valuable, because it covers the unscripted actions of several children from different families.
Finally, the researcher decides he wants to observe the children in a more controlled situation, so he sets up a laboratory experiment in which he can control the variables and the number and status of the participants and observe their behavior in particular circumstances. In this case, he selects a group of only children to observe and a control group of children with siblings. He places each child in a room, one by one, and gives the child a series of simple tasks to accomplish. He then observes how responsibly (or not) each child performs the tasks. This gives him another set of data with a much more specific focus.
When he has finished collecting his data through these research methods, the researcher evaluates what he has discovered and judges whether his hypothesis is valid or not.