1. At this point it might be good to reemphasize that in this course, the primary focus is the major conceptual issues regarding research design: how to frame a study, convert theories into hypotheses, reliably and validly operationalize concepts as variables, develop measurement instruments, design questionnaires, draw a sample, collect data, and contrast the value of different data collection approaches such as surveys, experiments, archival research and ethnographic studies.
2. One major way of collecting data that relies on questionnaires is survey research. Survey research can be defined as the administration of questionnaires to a sample of respondents selected from some population. Turn of the century social work gained prominence through the survey technique using it to highlight social conditions across city neighborhoods. Journals were published to report the findings of various surveys.
3. Surveys can be used for descriptive, explanatory or exploratory purposes. Surveys based on probabilistic samples (such as random, systematic or multistage cluster sample techniques that ensure everyone an equal chance of being selected) are best for creating descriptive statistics of populations, such as a neighborhood’s residents or a welfare program’s client population. Nonprobabilistic samples (such as quota or snowball samples) are acceptable for explanatory and exploratory research when probabilistic samples are not feasible as when there is no list of program clients.
4. Surveys can be one-point-in-time, cross-sectional studies that provide a snapshot of the population being studied. They also can be over-time, longitudinal studies that provide analysis of what happens to the population being studied at several points in time, from a few months to several decades.
5. Survey research is probably the best method for collecting original data for describing a population too large to observe directly. Surveys almost always involve studying one individual at a time, with each respondent completing a questionnaire. Corresponding each individual is most often interviewed to report information about him or herself In such cases, the unit of analysis is the individual. Yet, a sample of individuals can be studied to have each one reports on their respective partnership, family, household, group, etc.
6. There are three main methods of administering survey questionnaires to a sample of respondents: self-administered, in-person and telephone. Self- administered questionnaires can be done in a variety ways such as arranging to have the respondents complete the questionnaires in a room or via the mail.
7. Each method has advantages and disadvantages. Mail and other self-administered questionnaires tend to be cheaper and quicker but have lower response rates, requiring extra follow-up. Mail and other self-administered questionnaires tend to be more effective in dealing with sensitive issues as in questioning about personal matters, but are less effective in dealing with complicated issues that might require clarification from an interviewer.
8. Interviewers need to be consistent and neutral when conducting interviews. They also need to be sensitive to differences between them and respondents.
9. Secondary data analysis is done when we analyze data already collected by others. Many good surveys, cross-sectional and longitudinal, already are available surveying national samples of key populations, such as families, youth, etc. While cheap and quick, use of such studies can have the problem of only including information that is not quite related to what the researcher actually wants to study, such as depression scores rather than anxiety levels for each individual.
10. Survey research in general has advantages and disadvantages. Advantages: (1) make the statistical description of large populations feasible, (2) offer findings that are more generalizable than those from experiments or other methods of data collection; and (3) flexible in the sense of allowing researchers to study many variables at once. Disadvantages: (1) emphasize gaining information on each individual at the expense of accounting for the broader social context, such as the culture or social structure; (2) emphasize reliability at the expense of validity of measurement of variables, such as when providing consistent but superficial measures of attitudes like levels of depression, self-esteem or many other attitudes; and (3) lack flexibility compared to more qualitative ethnographic methods of observation which can shift focus as the researcher learns more about the people being observed, for instance as when living with a group of homeless individuals over time as opposed to surveying them with a formal questionnaire. The advantages and disadvantages of the survey method highlights the need to triangulate by using more than one method to collect data on any topic.