Research Methods | Definition and Types
Research methods are the procedures followed to collect and analyze data. You can think of them as the series of steps you would provide someone who wanted to recreate your study.
Research methods refer to the specific processes taken when conducting research—they’re a series of instructions on how to collect and analyze your data.
Research methodology, on the other hand, is broader in scope. It concerns the research methods used but focuses on the theory and reasoning behind why these methods were chosen to address a research question.
In brief, research methods are what you’re doing, and research methodology is why you’ve chosen these methods.
Academic journal articles will generally include a Methods section; more involved documents, like a dissertation, may instead have a Methodology chapter.
What are research methods?
Research methods are the systematic steps taken when conducting research. When addressing your research question, there are some key decisions you must make:
- What type of research will I conduct?
- How will I sample my data?
- What techniques will I use to collect my data?
- How will I analyze my data?
The best methods choices will be specific to your own research question and the time and resources you have available. Carefully considering these aspects of your research methods will help ensure that your results are valid and reliable.
Types of research methods
A key decision that will inform both your data collection and analysis is the type of research you choose to conduct. Most research can be categorized as qualitative or quantitative.
Quantitative research
Quantitative research uses numerical data to address a research question. These data might include measurements, survey responses, and experiment results. Quantitative data are helpful when you want to systematically test a hypothesis. However, not all phenomena can be described numerically, so quantitative research may not be possible in every situation.
Qualitative research
Qualitative research involves narrative, non-numerical data, which may include text, photos, videos, interviews, and observations. Qualitative approaches can provide deep and nuanced insight into a topic. They are helpful when studying topics that are difficult to measure or create hypotheses for, such as human experiences or new research areas.
Mixed-methods research
Mixed-methods research blends both qualitative and quantitative approaches. Combining the two can provide a deeper understanding of a research question, and the advantages of one may compensate for the disadvantages of the other.
Descriptive vs experimental research
Aside from whether to conduct a qualitative, quantitative, or mixed-methods study, you should also consider whether you will conduct descriptive or experimental research.
Descriptive research captures the characteristics of the group you are studying, whereas experimental research captures how changes in independent variables impact dependent variables.
Experimental research is tightly controlled, which helps eliminate the influence of confounds that may influence a study’s outcome and its ability to identify cause-and-effect relationships. However, in some cases experimental research may not be necessary, feasible, or ethical. Some of these limits can be overcome through quasi-experimental research (similar to experimental research but without manipulation of an independent variable).
Sample design
Because it’s usually not possible to collect data from every single person in the group you’re studying (the population), you must decide how you will select a subset of people (the sample) to study.
Sampling techniques can generally be categorized as probability sampling (everyone in a population has an equal chance of being included in a study) and non-probability sampling (some people are more likely to be included than others).
Some common probability sampling techniques include the following:
- Simple random sampling: The sample is totally random, so every member of the population has an equal chance of being included.
- Systematic sampling: Every nth member of the population is sampled, where n is the sampling interval.
- Stratified sampling: The population is separated into groups (strata) based on certain characteristics (e.g., age, gender), then these strata are sampled using another probability sampling technique.
- Cluster sampling: A population is divided into small groups (clusters), and clusters are randomly selected to be part of the sample
Some common nonprobability sampling techniques are as follows:
- Convenience sampling: The sample is selected based off of convenience (e.g., people the researcher knows, people passing by on a street).
- Purposive sampling: Individuals are selected because they possess specific characteristics.
- Snowball sampling: Existing subjects recruit new subjects.
- Quota sampling: Similar to stratified sampling (the population is first separated into strata based on certain characteristics), but individuals are chosen from these strata using a nonprobability technique.
Data collection
To properly address your research question, you must decide what type of data to collect. Primary research uses data you collect yourself (primary sources), whereas secondary research uses data previously collected by someone else (secondary sources). Secondary sources could include existing datasets, images, videos, written text, or other records.
Primary data afford a greater degree of control, but secondary data can be more convenient and less expensive. If secondary data are available that can be used to explore your research question, this approach is often a more efficient alternative to collecting your own data.
If you are conducting primary research, you need to decide how you’ll collect data. Some data collection methods are specific to quantitative or qualitative research, and some can be used for both.
| Data collection technique | Qualitative or quantitative? | Purpose |
|---|---|---|
| Direct measurement | Quantitative | To obtain information about a physical or biological property (e.g., temperature, heart rate, brain activity via fMRI). |
| Questionnaire | Either | To gather information about a topic of interest, which could range from political views to anxiety symptoms. Scales are questionnaires that provide a numerical score. |
| Interview or focus group | Qualitative | To obtain detailed information about an individual or group’s experiences or perspectives by asking questions |
| Observation | Either | To record behaviour during an activity. This could be done by a human observer (e.g., watching and noting how children interact) or by a computer (e.g., recording reaction time as people respond to stimuli on a screen). |
How you operationalize the constructs relevant to your work will determine how you collect data.
Data analysis
Before you start collecting data, you should have an idea of how you’ll analyze it. Formalizing your analysis plan helps ensure reproducibility and prevents “p-hacking” (manipulating your data or exploring different analysis methods until you find a favored result). Your analysis will depend on what type of data you’ve collected.
Qualitative data analysis
Most qualitative data analysis approaches involve organizing and reviewing data to extract recurring themes and ideas. Common analyses include content analysis, thematic analysis, and discourse analysis.
Quantitative data analysis
Quantitative data are analyzed using statistics. Descriptive statistics describe the data themselves, and may include information such as central tendency (mean, median, or mode) and variability (the range of values or standard deviation). Inferential statistics are used to test hypotheses and make predictions about your data. They involve a degree of uncertainty, because you’re attempting to generalize the results from a sample to a population.
How to write a methods section
When you’re writing a research paper, the purpose of the methods section is to describe the steps you followed with enough detail that someone else could redo your study and (hopefully!) get the same results. Your methods section should also demonstrate the validity of your study—were you properly measuring what you intended to?
The specific details you include in your methods section will depend on your field and research approach. The methods section may also be broken into subsections. It can be helpful to refer to relevant examples when formatting your methods section.
Consider the following questions when writing your methods section (note that they won’t all be relevant to every study!).
- Who (or what) did I collect data from? How did I select this group? Are there any relevant descriptive statistics I should provide (e.g., age, sex, gender)?
- Were any participants or data points excluded from analysis? Why?
- How did I incorporate research ethics into my work?
- What was the time course of my study? Were data collected during one session or repeatedly over a period of time?
- How did I operationalize (measure) the variables of interest in my research?
- What data collection techniques did I use?
- What procedures were followed when collecting data?
- What equipment, stimuli, or materials were used? Are there any technical specifications the reader should know (e.g., equipment manufacturer)?
- How did I analyze my data? What software programs or libraries (including version number) did I use?
Frequently asked questions about research methods
- In research, what is the difference between methods vs methodology?
-
Research methods are the steps you follow when conducting research. A methods section should describe the type of research you’re conducting, sampling techniques, data collection methods, and data analysis.
Research methodology instead focuses on the theory behind your research methods and why you chose them to address your research question.
Though people sometimes use the terms method and methodology interchangeably, they are not the same. Methods describe how you conduct your research, and methodology describes why you chose these methods.
- What are common research methods in psychology?
-
Most research methods in psychology are quantitative: numerical data are used to address a research question. Quantitative approaches include the following:
- Experimental research
- Quasi-experimental research
- Survey research
Some psychology research is qualitative (i.e., uses nonnumerical data to explore a research question). Qualitative research approaches include the following:
- Interviews and focus groups
- Case studies
- Survey research
- Observational research
Mixed-methods research combines qualitative and quantitative approaches.
- Why are research ethics important?
-
Research ethics are principles that guide scientists, helping them distinguish right from wrong when conducting research. Research ethics help protect the people involved in scientific studies and ensure the integrity of scientific research.
- What are descriptive research methods?
-
Descriptive research is a research method that aims to uncover the characteristics of a population or phenomenon. Research questions can be addressed using techniques like surveys and observation.
Examples of descriptive research questions include the following:
- What percentage of people in a community experience food insecurity?
- What is the most popular social media platform for people under the age of 25?
- How many cars does the average American family own?
Descriptive research can answer what, where, when, and how questions but not why questions. Both quantitative and qualitative methods can be used for descriptive research.
- What is operationalization?
-
Operationalization is when you define how a variable will be measured. Operationalization is especially important in fields like psychology that involve the study of more abstract ideas (e.g., “fear”).
Because fear is a construct that cannot be directly measured, a researcher must define how they will represent it. For example, in studies involving mice, fear is often operationalized as “”how often a mouse freezes (i.e., stops moving) during an experiment.”
Operationalization can be used to turn an abstract concept into a numerical form for use in quantitative research.
Some operationalizations are better than others. It is important to consider both reliability and validity (how consistent and accurate a measurement is, respectively) when operationalizing a construct.
