Skip to content Skip to sidebar Skip to footer

Approaches, Methods, and Types of Techniques in Research

The research approach includes qualitative and quantitative approaches. The two approaches greatly affect the type and method of research. The qualitative approach can use descriptive and exploratory research types. The quantitative approach can use the types of explanatory, descriptive, predictive, and evaluative research.

While the methods used include content analysis, surveys, case studies, network analysis, communication/public relations audit, discourse/semiotic analysis, experiments, framing analysis, and other methods that appear as current trends such as phenomenology, ethnography, etc. However, this method for the S1 stage is rarely used, because the framework must be comprehensive.

Approaches, Methods, and Types of Techniques in Research
image source:

Research methodology

1. Unit of Analysis

Research that uses students who are general as the unit of analysis can take many research subjects because they can simply take one or two classes of students from one school without the hassle of visiting many places. In contrast to researchers who use special schools, they may only take a few research subjects because the subjects are classified as 'rare'.

2. Research approach or model

Survey research will use a large number of research subjects, while experimental research may use only a few research subjects.

3. The number of special characteristics that exist in the population

In determining the size of the research sample, researchers inevitably have to try to identify variations in the characteristics that exist in the population. To clarify this statement, it may be possible to present examples of various population conditions with variations.

Group A, level I management consists of 40 people, all women and all children of civil servants. Group B, level I production consists of 40 women and men, all children of civil servants. Group C, level I marketing consisted of 40 men and women, from families of farmers, traders, members of the Armed Forces, and civil servants.

From the examples of the three groups, the most homogeneous condition of the subject in the population is group A, followed by group B, then group C. In sampling, for group A the researcher may take only a few people. In a small sample, it can be seen that it represents all employees of that one level. For group B, samples must be taken from representatives of each type of group section, namely women and men. Furthermore, to determine who will be sampled from group C, the researcher must consider the family share of men from various families and women, as well as from various families. At least there must be a representative from:
  • A male employee of the farmer
  • The male employee of the merchant
  • Male employees of ABRI members
  • Male employees of civil servants
  • A male employee of the farmer
  • The male employee of the merchant
  • Female employees of ABRI members
  • Female employees of civil servants

4. Research limitations

Due to the availability of limited time, funds and personnel, researchers may be forced to limit the number of research subjects taken, namely carrying out sample research, namely using part of the population as research subjects.

What size sample should be taken in the study? In simple terms, it can be said that the larger the research sample, the better the results obtained because in a large sample the more realistic picture of the results will be reflected. In general, people think that thirty research subjects are the boundary between a small sample and a large sample. Thirty or less can be said to be a small sample while greater than thirty is a large sample.

Approaches, Methods, and Techniques Types or Types in Research - In determining the sample, researchers should always keep in mind the limitations of understanding the research subjects, research respondents, and data sources that have been applied in the previous section. If researchers think about data collection techniques, then what must be considered is the understanding of the respondents. If they will use a questionnaire, quite a lot of respondents can be taken. On the other hand, if the research uses interviews, of course, there cannot be many respondents because interviews are data that can be said to be 'elite'. After all, it requires a lot of time and effort. Likewise, if the researcher will use security techniques. To use this technique, special skills are needed for its implementation.

Several formulas can be used by researchers to determine the number of sample members. As a rule, if researchers have several hundred subjects in the population, they can determine approximately 25-30% of the total number of subjects. If the number of subject members in the population only includes between 100 to 150 people, and in collecting data the researcher uses a questionnaire, it is better if this number of subjects is taken entirely. However, if the researcher uses an interview (interview) or observation (observation) technique, this number can be reduced according to the sampling technique according to the ability of the researcher.

Research subjects, research respondents, and data sources are three things that have different meanings. Research subjects are objects, things, or people to which the research variables are attached. Respondents are people who can provide answers or information about variables. Data sources are places, people, or objects where researchers can observe, ask questions, or read about matters relating to the variables studied. Data sources can be broadly divided into people (person), place (place), and paper or documents (paper).

To determine the sample size, researchers must understand the meaning of "unit of analysis." What is meant by a unit of analysis is a unit that refers to the research subject. The unit of analysis is a factor considered by the researcher in determining the sample size in addition to the approach, the specific characteristics of the population, and the limitations of the researcher. The existence of characteristics or types of groups in the population determines the possible sampling techniques: random sampling, cluster sampling, stratified sampling, area sampling, proportional sampling, and twin sampling. (double). Besides that, there is a less reliable sampling technique, known as chance sampling. With this technique, the researcher only takes subjects that can be found by chance.


Not all broad publics can be sampled. A census is a collection of information from all members of the population. To distribute the census, we need to identify all the people or articles that created the information. The units that make up the population are defined as sample elements. The list of all sample elements is defined as the sampling frame.

Sampling frames

The sampling frame is usually incomplete. So we must assess the completeness and accuracy of each sampling frame used to collect information about the public, media placement, and so on. Technically, the sampling frame is defined as an actual public study. The difference between the ideal general study and the actual population defined by the sampling frame may not be that important. This is a matter of value judgment, which is based on research objectives. If this distinction is important, then we look for an actual and broad sample.

Advantages of the census

The advantage of using the census in targeting the public is that it is easy. We get definite information about the entire population, rather than having to make estimates about the population based on a sample. Of course, this can happen if the census is complete, the articles are also complete and all employees in the population are studied so the percentages are accurately described – no estimation is needed. But rarely is a census 100 percent successful. Because of the possibility of errors in the process of submitting articles. The disadvantages of using a census affect the sampling that underlies the research.

Source of bias

When a census collects information from less than 100 percent of the population or the general public, misinformation is a common form of bias. We cannot be sure that the error element is the same as the element entered in the census. If the element of error is small, then the potential for bias is small, and vice versa. To study the elements of error in a census or research sample, there are several questions that we must accept, namely: are people who did not fill out the questionnaire not interested in the newsletter? Do people who don't participate in research on newsletters look better than reality on the assembly line? The disadvantages of census research are time and cost. As the population or general population increases, the cost of gathering information increases proportionally. Research takes a long time to complete and usually time is not available. As the number of target publics or media placements increases, samples provide an efficient substitute for census research. Sampling allows us to make estimates or knowledge claims about the entire population, based on information gathered from the population.

Sampling and knowledge claim

We may decide not to collect information from each member of the target public, or analyze every newspaper article in a file set. This means that we must decide how to study some members of the target public (or several newspaper articles in a collection of files) to make knowledge claims about the entire population. A knowledge claim is a statement that describes the entire population but is based on information obtained only from a sample or subset of the population. Reliable knowledge claims about the population are highly dependent on the type of sampling strategy used. There are two categories of sampling strategies:
  1. Non-probability sampling strategies. This sampling technique allows us to limit and allow biases in understanding the populations from which they are derived. This strategy is useful for conducting exploratory research.
  2. Probability sampling techniques. This technique allows us to use probability theory to make precise knowledge claims about the entire population, based on a sample drawn from the population. Data sources can be broadly divided into people (person), place (place), and paper or documents (paper).
That's an article about Approaches, Methods, and Types or Types of Techniques in Research. Hopefully useful for all of us.

1. Griffin, Em. 1991. A First Look at Communication Theory. New York: McGraw-Hill
2. Sendjaja, Sasa Djuarsa, 1993. Teori Komunikasi, Jakarta: Univ. Terbuka
3. Mulyana, Deddy. 2001. Ilmu Komunikasi Suatu Pengantar Littlejohn, Stephen, 1996, Theories of Human Communication. Wadsworth Publishing Company Inc Belmont