Social Scientific Method: Importances, Characteristics, Steps

What’s it: The social scientific method is an investigative technique or procedure used to understand humans and their interactions. Broadly speaking, it includes collecting, analyzing, and interpreting information across various social sciences, such as economics, politics, cultural anthropology, sociology, and psychology. Steps in social scientific research include:

• Determining the research topic
• Literature review
• Formulating a hypothesis
• Designing and conducting research studies
• Draw conclusions and report results

Meanwhile, to collect data, we can use surveys, laboratory experiments, observations, and ethnographic research. And the data or information collected can be qualitative or quantitative. Then, we may need statistical analysis to model and test our hypotheses.

Why are social scientific methods important?

The social scientific method is important for building social scientific knowledge, including economics. For example, we ask how interest rates affect household consumption expenditures? How big will an increase in interest rates impact changes in their spending?

Answering the questions above is not just asking one or two households and then drawing conclusions. However, we must use standard techniques to draw valid and unbiased conclusions. And that technique is what we call the social scientific method.

The method guides us on how to build hypotheses and collect data to test the hypotheses we build. Then, we interpret the results and generalize the results.

Allowing us to test existing theories or discoveries is the main reason the social scientific method is important. It helps us build independent and impartial research and, therefore, obtain unbiased results. Moreover, our results can be improvements or material for open debate.

What are the four characteristics of the social scientific method?

The scientific method we use must have the following four characteristics:

1. Logical
2. Confirmable
3. Repeatable
4. Scrutinizable

Logical. Questions or hypotheses must meet logical principles. So, when we do research, we base these questions on existing theories and references. So, when finished, we get a logical conclusion. For example, we may get results that differ from one theory but correspond to another. It can be material for refinement or open debate.

Confirmable. The results we get must be scientifically proven. Therefore, we must draw conclusions according to the data or information we observe.

Repeatable. Others can replicate our research, usually with a different subject or situation. Thus, they can test and draw conclusions about whether our findings can be applied to other situations or subjects.

Scrutinizable. Our research must be of high quality and resistant to criticism by other researchers. This is important to maintain high scientific standards, where researchers examine each other’s work. It balances the bias due to our subjectivity and perhaps also catches any fraud we might commit.

What are the steps in the social scientific method?

The scientific method offers us an approach to conducting research. It tells us the systematic and organized steps we must take to explore social problems. Thus, we get objective and consistent results.

The five steps in the social scientific research method include:

1. Determining the research topic
2. Literature review
3. Formulating a hypothesis
4. Design and conduct studies
5. Draw conclusions and report results

Determining the research topic

Before collecting data, we must define what area we will research. Then, we choose the problem we will research and ask the question we want to answer.

We may be able to get a topic from existing theory and explore it further from previous studies. Reading the conclusion section in a journal is the quickest way. There, the researcher usually describes the problems and limitations of the research conducted. Or they provide suggestions for further research. And we can start developing the topic from there.

Literature review

At this stage, we describe what literature we use. That includes other similar research on which we ask the question. To do so, we may have to visit the library more often to find the right book or journal. Finally, we read the theory or research results to which our research refers.

This stage is crucial to help us gain a broad understanding of the topic we are researching. In addition, we may be able to identify gaps in understanding the topic. Or, we get potential answers to our questions, on which we build our hypotheses.

Formulating a hypothesis

A hypothesis is a proposed explanation we put forward based on previous references. That becomes the starting point for us to investigate further. This is to test whether the potential answers we found earlier answered the questions we posed.

For example, we examine the relationship between interest rates and household spending. From the topic, we might formulate a hypothesis:

• H0: Interest rates have no significant effect on consumption expenditure
• H1: Interest rates have a significant effect on consumption expenditure

Design and conduct studies

At this stage, we determine the variables we use in the study. Since it is impossible to examine all variables, we isolate the other variables and assume they are unchanged (ceteris paribus). This is to focus on what we are researching; so it fits the hypothesis we want to prove.

In addition, we also make research designs. This is the basis for us before carrying out scientific research. It defines a concise and logical plan for answering the hypotheses in our study, thereby maximizing reliability and validity. The research design can be descriptive, experimental, correlational, or explanatory.

The next step is to determine the research instrument. These are tools or techniques we use to collect related data or information. It can be through surveys, observations, or secondary data collection.

The last thing is to do research. We must focus on finding accurate data. Thus, our research results are correct and by what we formulated.

Draw conclusions and report results

After getting the data and information we need, we must process it first. For example, we might input data into excel and tabulate it historically or in cross-sections. Then, we convert it to a statistical program such as EViews, SPSS, Stata, SAS, or Minitab.

Through statistical programs, we create descriptive statistics before creating models to test hypotheses. For example, we create statistics such as sample size, mean, and standard deviation. We may also need correlation statistics to determine the relationship between each variable.

If the correlation is a number, then preferably, we also make a scatter plot. The correlation only answers whether a variable is related to another variable. Meanwhile, the scatter plot helps us know the pattern, whether linear, quadratic, or exponential.

Building a model is the next step. It is to test the hypothesis. A regression model is commonly used to see the causal relationship between two or more variables. It has various variations such as:

• Linear Regression
• Nonlinear regression
• Panel data regression

Linear regression. This is to test the effect of the independent variable (e.g., interest rates) on the dependent variable (e.g., consumption expenditure), and the relationship between the two forms a linear pattern. It is simple linear if only one independent variable is used. And it is multiple linear if it uses more than one independent variable.

Nonlinear regression. It is used if the dependent and independent variables are related and form a nonlinear pattern, whether quadratic or exponential. In this case, the scatter plot helps us understand the pattern.

Panel data regression. It is used if we combine cross-section data and time series data (called panel data). Panel data means we observe the same sample over a certain period. Panel data regression is important to know whether the time effect has an impact or not.

After completing the model, we write and report the results in an academic journal. Earlier in the results section, we discussed descriptive statistics. Next, we describe the model we got and discuss our proposed hypothesis.