Correlation tests the strength and direction of a relationship between two variables.
Regression goes a step further: it lets you model the relationship between a dependent variable and one or more independent variables, often using a line of best fit that lets you make predictions about your data.
Continue reading: What is regression?
Pearson’s r (the “Pearson product–moment correlation coefficient,” or simply “r”), is the most common way to compute a correlation between two variables. It tells you how two variables are related. Most statistical tools (like R or Excel) have a built-in correlation function.
The value of r ranges from -1 to +1. The sign of r (+ or –) indicates the direction of a relationship (whether a correlation is positive or negative), and the magnitude of r indicates the strength of the relationship (sometimes called the effect size).
What is considered a strong, moderate, or weak correlation varies by field. Many researchers use Cohen’s size criteria as a guideline:
| r value | Direction | Strength |
|---|---|---|
| Between –1 and –0.5 | Negative | Strong |
| Between –0.5 and –0.3 | Negative | Moderate |
| Between –0.3 and –0.2 | Negative | Weak |
| Between –0.2 and +0.2 | N/A | No correlation |
| Between +0.2 and +0.3 | Positive | Weak |
| Between +0.3 and +0.5 | Positive | Moderate |
| Between +0.5 and +1 | Positive | Strong |
Continue reading: What is Pearson’s r?
A correlation is a relationship between two variables: as one changes, the other tends to change as well. For example, coffee consumption is correlated with productivity: people who drink more coffee often report getting more done.
Causation, on the other hand, means that a change in one variable directly causes changes in another. To test whether coffee actually increases productivity, you could conduct an experiment: assign some people to drink coffee and others to drink water, and compare their task performance.
It’s important to remember that correlation does not imply causation. Even if coffee consumption and productivity are correlated, it doesn’t mean one causes the other. It’s possible that people who are working more are tired, so they drink more coffee.
Continue reading: What are examples of correlation vs causation?
Causation and correlation are two ways variables can be related.
Causation means changes in one variable directly lead to changes in another (i.e., there is a cause-and-effect relationship). For example, eating food (the cause) satisfies hunger (the effect).
Correlation means there is a statistical relationship between two variables—as one changes, so does the other. However, this relationship is not necessarily causal. For example, although a child’s shoe size and their reading ability are correlated, one does not cause the other (instead, they’re both influenced by a third variable, age).
Continue reading: What is the difference between correlation and causation?
The goal of explanatory research is to understand why something happens. This is often done by exploring a cause-and-effect relationship between two variables.
Examples of explanatory research questions include the following:
Continue reading: What are some examples of explanatory research questions?
The aim of explanatory research is to determine why a phenomenon occurs. This may be done using correlational or experimental research.
On the other hand, descriptive research captures the characteristics of something as is, without intervention.
Though these approaches may share data collection techniques (e.g., they both might use questionnaires), their overarching purpose is distinct.
Continue reading: What’s the difference between explanatory and descriptive research?
Explanatory research examines why something happens. This is usually done by assessing the statistical relationship between two variables.
Because statistical methods like correlation require quantitative (numeric) values, explanatory research generally involves quantitative data.
However, these quantitative explanations may be supplemented by data from, for example, interviews. This mixed methods approach may offer a more comprehensive explanation of a phenomenon.
Continue reading: Can explanatory research be qualitative?
Thematic content analysis is often defined in different ways, and the term is sometimes used interchangeably with thematic analysis or qualitative content analysis.
Qualitative content analysis focuses on systematically summarizing the prevalence of specific codes in a dataset, whereas thematic analysis not only codes data but also identifies themes and underlying narratives.
Thematic content analysis is sometimes defined as a hybrid between these two methods: much like qualitative content analysis, its purpose is to describe a body of qualitative data, but data are broken down into themes rather than more simple codes.
Because “thematic content analysis” can be interpreted in different ways, it’s important to provide a detailed description of your methodology if you choose to use this term.
Continue reading: What is thematic content analysis?
Reflexive thematic analysis is a type of thematic analysis that centers the researcher’s interpretation of the data.
Reflexive thematic analysis acknowledges the subjective nature of data interpretation—rather than focusing on uncovering some “ground truth” in the data, researchers are encouraged to engage with their data and use their own knowledge and experiences for interpretation and analysis.
Continue reading: What is reflexive thematic analysis?
In their 2006 paper, researchers Virginia Braun and Victoria Clarke outlined the following 6 steps for conducting thematic analysis:
Continue reading: What are Braun and Clarke’s 6 steps to thematic analysis?
