Correlational Studies

Correlational Study Psychology

Correlational analyses are widely used in psychological research. Correlation research is based on observations between variables; this means there is no experimental manipulation involved.

Correlational research aims to observe whether or not two variables are related and, if so, how strong the association is.

The steps that researchers take when designing a correlational study are the following:

1. Stating the research question.
2. Identifying the variables.
3. Writing of hypothesis statements.
4. Conducting the research and gathering data.
5. Analysing the data.

Types of Correlational Studies

Three types of correlation studies exist, and we will describe them in detail below, with examples. Further, the different study types will be evaluated, presenting the strengths and weaknesses of each one.

Correlational Studies: Naturalistic observation

In naturalistic observation correlation studies, researchers record observations of variables in a natural setting; this is a non-experimental method in which no variables are manipulated.

A strength of naturalistic observational research is that it allows researchers to observe participants in a natural setting. This makes it more likely that participants will show their real behaviour, increasing the results' validity. In laboratory settings, for example, participants may not behave as genuinely due to the setting itself.

However, some limitations should be considered, such as the difficulty in limiting confounding factors, which can affect and reduce the study's validity.

Correlational Studies: Survey method

The survey method uses surveys and questionnaires to measure the researchers' variables.

The advantages of this research method are that it is relatively inexpensive, does not take too much time, and can recruit many participants in a short time. The method usually uses random samples for recruitment, so the research results are more generalisable than other sampling methods.

However, respondents may answer in a socially desirable manner rather than honestly, which reduces the validity of the results.

Correlational Studies: Archival research

Archival research is a type of correlational research that uses secondary data, such as previous research, case studies, historical documents, and medical registries, to measure variables.

The advantage of correlational archival research is that it can be cheaper than alternative methods. Data is readily available, and researchers can obtain data that may no longer be collected, such as documents from historical periods.

Nevertheless, the disadvantages of archival research should be considered. While conducting archival research, the researcher has no control over data collection methods, making it difficult to determine if the data is reliable and valid. Another issue is some data may be missing that is needed for the research.

Correlational Studies: Interpretations

In the statistical analysis of correlation data, a correlation coefficient is calculated.

The correlation coefficient (r) values can range from +1 to -1.

A positive number indicates a positive relationship between the variables; if one variable increases, the other is also expected to increase.

A negative coefficient indicates a negative relationship between the variables. If one variable increases, the other is expected to decrease.

A coefficient of 0 indicates no relationship between the two variables.

The value of the correlation coefficient determines the strength of the correlation data:

• When r = 0, then there is no correlation.
• When r is between 0.1- 0.39, there is a weak correlation.
• When r is between 0.4 - 0.69, there is a moderate correlation.
• When r is between 0.7 and 0.99, there is a strong correlation.
• When r equals 1, then there is a perfect correlation.

Scatter plots are typically used to show the relationship between variables by plotting the data when reporting correlation data. Scatterplots allow us to visually see the strength of the correlation and the direction between the variables.

Fig. 1. The scatterplot shows a positive correlation between the two variables.

Correlational Study Cause and Effect

One of the main ideas researchers need to remember when conducting correlational research is that researchers can't infer causation in correlational studies.

It is possible that, in some cases, one variable does indeed cause the other one. Further experimental research needs to be conducted to support or disprove it in such cases.

Example of Correlational Research

Researching the relationship between variables has been in the spotlight of psychological research for decades.

A correlation study will start by defining the research question. For example, a study may examine the relationship between self-esteem and social anxiety. Based on previous findings, researchers may hypothesise that there is an existing negative correlation between the two.

Researchers then decide which inventories or questionnaires will be used to measure the two variables. After this, the correlational statistical test will be calculated.

The statistical analysis may provide a significant result in which the correlation coefficient is -0.78, allowing the researchers to conclude that there is indeed a negative association between self-esteem and social anxiety.

The researchers may plot their data on a scatterplot, so they and readers can visualise the results.

Regarding the causality effect, it is tempting to suggest that low self-esteem makes individuals experience social anxiety. And although this could be the case, it cannot be established with a correlational test.

Fig. 2. An Example of a scatterplot showing a negative correlation.

Correlational Study Advantages and Disadvantages Psychology

In this section, correlational studies' advantages and disadvantages are critically reviewed.

One of the main advantages of correlational research is that it is quick and easy to conduct. It does not require great statistical knowledge for researchers to be able to use it.

Furthermore, correlations can be tested for existing data, which can inspire future research and be helpful when the researcher may have limited access to the phenomenon, e.g. if it's based on past events.

One of the main disadvantages of correlational research is it can't establish whether variables are causally related.

Since correlational studies only measure the co-variables, other potential confounding factors are not considered. The confounding variables may be a better explanatory factor for the study's outcome, making it difficult to determine the validity of the results.