What makes one data point stronger than another? Ask any researcher, social work domain or otherwise, and they’ll refer to the importance of avoiding the many forms of researcher bias, clearly expressing the problem and using properly structured methodology, among other important steps. Strategies for collecting reliable data may vary slightly depending on the nature of the study, but the underlying concept is this: nothing outside the natural environment can influence the data. If the researcher influences the data at all, then the correlation between the hypothesis and the results is weakened significantly.

So, how can researchers in the social work field strive for the highest standards of objectivity in their findings? The answer starts with the proper methodology. The structure of each study needs to conform to the environment in which the hypothesis was formed, and not the other way around. For example, if a researcher wants to test how participants respond to different management styles in the workplace, they would do best not to remove the person from their work environment – even if the researcher simulated one of their own.

In some scenarios, simply observing participants in their environments like the above example is appropriate. When it comes to perceptions and behaviors that aren’t dependent on the environment, maybe a simple survey would indeed work. In order to observe and collect the data as it exists in its natural domain, researchers in the social work field have to rely on several research methods. Let’s summarize some of the most common structures, starting with the already mentioned survey method.

Surveys

Especially when researchers have access to large participant groups, surveys are a simple, affordable and reliable research method. The structure is simple: participants answer a series of questions designed to test the researcher’s hypothesis. If the researcher wants to evaluate the effects of digital media on certain perceptions, for example, they could simply ask participants to express their thoughts on popular events and people. Researchers send out the surveys, aggregate the results and form their conclusions based on the trends within the data.

Like many of the following methods, it is not the implementation of the survey method itself that can be tricky, but knowing when and how to properly use it. If the topic(s) being covered by the survey can’t be addressed with simple questions and answers, researchers need to opt for more open-ended data collection techniques. If respondents feel embarrassed or incriminated by answering truthfully, they may skew the results – observational methods would prevent this issue in many cases.

Ethnographic Description

Like a probe sent deep into a planet’s surface to collect data unobtainable from the surface, ethnographic studies seek to immerse the researcher in another culture for greater insights into any number of behaviors and beliefs. Contrary to surveys, ethnographic methods are generally more time-consuming and costly. A researcher may travel across the world to live within a culture for weeks, months, or longer, adopting that culture’s practices to enrich their understanding. They bring all of their data back home, where they use it to help other groups coalesce with members of the researched group.

Ethnographic research models can overlap with others. As mentioned, the researcher will generally travel to an area and immerse themselves in the culture. This can include:

  • Informant interviews
  • Surveys and census data
  • Observation
  • Participation

Case Studies

Popular in the business world, case studies are also well-suited to research efforts in social work. Simply put, a case study is an example – a real-life scenario that provides a testing ground for a hypothesis. Researchers can examine data from an ethnographic study, for example, even if the study had a completely different objective, to demonstrate the social or individual impact (or lack thereof) of certain behaviors. Though everyday events can be justified as case studies, researchers are often hard-pressed to prove that there were no extraneous variables affecting the data, since real-life scenarios don’t occur in controlled environments.

Case studies are useful in many scenarios, but they tend to address a specific theory. They can be used throughout the literature review and research phases to accomplish the following objectives:

  • Practically demonstrate a theory
  • Call for more research
  • Debunk a hypothesis
  • Test research findings in the real world
  • Uncover new variables affecting the hypothesis

Single-System Design

Experiments in the medical field especially tend to follow a model that compares the results (of a drug, treatment, etc.) across two groups: the control group, which doesn’t receive treatment, and the experimental group, which does. In a single-system design, however, there is only one group. Oftentimes, this “group” is just one person. The person or group is generally studied over a long period of time to assess their response to different variables.

With no control group, though, how do researchers gauge the effects of any particular variable? By manipulating the variable itself, not the audience. Single-system designs test the effects of different independent variables (that which the experimenter applies) on the dependent variable (the result of these changes and the theory being tested).

Let’s say that a researcher in the social work domain wants to determine the effect of digital media consumption on antisocial behaviors. Instead of setting up a control group (no digital media consumption) and an experimental group (two hours of digital media consumption per day per participant) to test their hypothesis, the researcher will simply change the nature of the digital media consumption for a single participant, recording the results of each change.

Program Evaluation

This particular vein of research is highly relevant to social workers, who often ally with programs of all kinds as a way of increasing access to vital resources for their clientele. Whether a program is funded by the government or private investors, nobody wants to pour money into a project that isn’t positioned for success from the very beginning. A program evaluation gives everyone involved the opportunity to assess a program’s fitness across multiple dynamics.

This research method requires a comprehensive look at contemporary findings to prove the effectiveness of a particular program. Even after a program has launched, program evaluations can help to refine things for greater efficiency. The following list of questions will help to define the purpose and applications of a program evaluation:

  • Will this program work?
  • How much will the program cost per participant?
  • How can we expand the program?
  • Is there a better way to serve the program’s population?
  • Are there any disadvantages for program participants?

Needs Assessment

Needs assessments are also fundamental to the sociological perspective because they seek to identify deficiencies in certain populations. The population isn’t always defined by region, income level or ethnicity, but these three factors comprise the majority of cases.

This research method is integral to social work at all levels. A social worker in the field, for example, can use needs assessments to identify opportunities for improvement with an individual client. Conversely, researchers, program planners and executive-level social work professionals can apply needs assessments to entire communities in order to affect change on a larger scale. In either case, needs assessments are often used as part of the planning process when conducting research, creating resources or developing a plan of care for one person.

Randomized Trials

Finally, the randomized trial is one of the purest and most broadly applied experimental models. Randomization in this case refers to the manner in which participants are selected to become part of the control or experimental groups. The experiment is run in a very formulaic, easily reproducible and highly measurable fashion. The randomly assigned experimental group is subjected to the variable, whether it’s a treatment, a specific stimulus, etc., and the randomly assigned control group is not. The response of both groups are measured and compared, and when applicable, a new variable is tested in the same fashion.

In both social work and the medical field, this research model is most appropriate when responses are easily quantifiable. Comparing subjective responses between two groups yields less actionable and obvious information than, for example, a measurable change in blood pressure.

To reiterate, no research model is objectively “better” than the other; each has their own application. Properly selecting and applying a model (or a combination of models) requires researchers to comprehensively evaluate the subject’s environment, the nature of the data (subjective, objective or both?), the hypothesis, and so forth. When used correctly, the proper research method can introduce highly valuable findings that hold up against future inquiries.

Tim Kalantjankos

B.S. Sociology | University of Nebraska at Omaha

A.S. Physical Therapy | Clarkson College

September 2019

More Related Articles: