But first, what is a peer reviewed journal article? In a magazine, articles are submitted, selected and possibly modified by the editor, and published. Peer reviewed journals are like magazines for researchers and academics that publish reports of research. Researchers submit articles which are reviewed by other experts in the field before being accepted for publication.
Some databases refer to peer reviewed journals as academic or scholarly journals, but they are the same thing: credible journals in the academic field.
The typical peer-reviewed primary research article contains some or all of the following sections (and possibly other sections not mentioned here). Click each one to learn more:
The abstract provides a brief summary of the journal article contents.
Look at the abstract when deciding whether to select and use an article in your search results.
The introduction sets the background and context of the article, and may include the reason the research was undertaken.
Pay attention to this section on your first reading.
The methodology might not be included in some "discussion articles", but will be included in many research articles.
It gives explicit details of how the research was conducted, and should be sufficient to reproduce the research.
On your first reading you might skip over this section.
It's the research equivalent of a cooking recipe.
Results and analysis (sometimes merged, sometimes separate sections) give a detailed listing of the research results, and details on how the results were summarised and interpreted, and may include statistical details.
On a first reading you might skip over this section.
This section might not be present in some "discussion articles".
This is usually a discussion of the results or other sections of the paper, including potential flaws in the research and implications of the results.
Pay attention to this section on your first reading.
This usually sums up the outcome or recommendations from the research or discussion. It is the "bottom line" of the article.
Pay attention to this section on your first reading.
At this point, you might want to consider whether the results and the analysis justifies the conclusion, or whether a strong conclusion is being made from insufficient or uncertain results.
Good-quality journal articles will provide references to either support the ideas and arguments in the article, or to provide a source for the ideas and information used. In-text citations in the other sections will show where each reference has been used. On a first reading you might skip over this section.
If you are using a recent article, the references can be a good source for finding more articles on the same topic.
Secondary research appraises the quality of studies and often makes recommendations for practice. Evidence is not always available from secondary research, and searching primary research may be required.
| Secondary research | Primary research |
|---|---|
|
|
Click on the buttons below to watch brief videos on common study types.
Systematic reviews and meta-analyses are secondary studies. They find all the evidence on a topic, critically appraise it, assess it for a risk of bias and then summarise the findings. They are useful to guide practice.
Randomised controlled trials are experiments to test out a treatment or intervention. Participants are sorted into groups to receive either the treatment or a placebo. Ideally, nobody knows which group is getting the treatment (blinding).
Cohort studies investigate the links between a risk factor and a health outcome. Participants are chosen based on their exposure to the risk and compared with a group that was not exposed.
Case control studies compare people with a particular condition with those who don't have the condition. They try and find what factors might have led to the development of the condition.
A qualitative study looks at people's experiences and views. They look at words and what people say rather than numbers, and there are often much fewer participants required.
Click on the following key research terms to see their definition. You may need to understand and use some of these when assessing an article. You can also use the CASP glossary, Sage Research Methods or the Dictionary of Nursing Research for other terms not included here.
The treatment that is being tested out
The chance process where participants are assigned either to an intervention or control group. Each participant has an equal chance of being in either group.
Participants' outcomes and results are analysed according to the group they were originally randomly placed in, whether or not they completed the intervention or dropped out. This can help tell if there is an issue or problem with the intervention.
Where the people involved in a study don't know which participants are getting the intervention. Single-blinding is where the participants don't know whether they are getting the intervention or not, usually done by giving the non-intervention group a placebo. Ideally, all people involved should not know, including the researchers, the people delivering the treatment, the people assessing the outcome, and the people analysing the data. However, for some interventions this is not possible. This helps reduce bias, where people may believe an intervention will work and therefore not score fairly and objectively.
The characteristics of the participants at the beginning of the study. Ideally, the groups in the study should be as similar as possible, to reduce the chance that any effects of the intervention are due to something else. Different baseline characteristics are recorded depending on what is being studied, such as age, sex, weight, or other biomarkers.
The instructions and treatment given to the participant groups. Ideally, they should both be the same, apart from the intervention to allow a fair comparison.
The deviation of results from the truth because of the way the study was conducted. Many aspects of a study methodology, such as blinding and randomisation of participants, are designed to try and reduce bias as much as possible.
A range of values researchers are reasonably sure the true value is within, with a given assurance. If the confidence interval (or given assurance) is 95%, and the range of values given is 0.64 to 0.89, this means they are 95% sure the real result is between 0.64 and 0.89. In articles this often looks something like: 95% CI, 0.64 to 0.89. Generally, the smaller the range of values is, the more reliable the result will be.
The result of the experiment, or the effect of the intervention on the outcome the researchers were studying.
Factors that could influence the results of a study that are not because of the intervention. For example, if a study found that mothers with healthy diets had healthier babies, a confounding factor could be the mothers' income, as those on higher incomes may be able to afford healthier food. These can sometimes be adjusted for with statistics, but can also impact on the study's reliability.
A statistical measure of the probability of getting a result different to what has been reported. Usually, it is accepted that the results are real if the likelihood it is because of chance is less than 1 in 20, which is recorded as p<0.05. A p-value less than 0.05 is considered statistically significant, and a p-value over 0.05 is not.
Validity is the soundness or rigour of a study. This means the way the researchers have conducted the study is less likely to lead to bias, and the results are more trustworthy.
The National Collaborating Centre for Methods and Tools (NCCMT) have created videos on understanding some of these research terms as well as others you may come across.
Critical Skills Appraisal Programme. (2022). Glossary. https://casp-uk.net/glossary/
Powers, B. A., & Knapp, T. R. (2011). Dictionary of nursing theory and research. Springer.
The Study Skills website has some excellent resources to support you in understanding journal articles.