Understanding how research is presented in published papers

Back

There are thousands of research papers out there and they can be a great resource to increase your understanding of a particular area, but they can also feel overwhelming if you are not familiar with the topic, the terminology or how to know if you can trust what is being said.  We have compiled a beginners’ guide so anyone new to this can spot key elements that are important to consider when reading a research paper. 

 

Let’s start at the beginning…

What is a research paper?

 

A research paper, or scientific publication, is a form of academic writing that is published in a journal to provide an in-depth analysis, evaluation, or interpretation of a research study or topic.

 

There are different types of research and these are organised in a hierarchy of evidence so that you can decipher the likely rigour and reliability of a research study that is being presented in a paper. The higher up the hierarchy, the more likely the study can minimise the effects of bias and ensure precision. If completed well, randomised controlled trials (RCTs) are seen as the gold standard trial for evaluating the effectiveness of interventions as they offer the most scientifically rigorous method of hypothesis testing (what you expect your research will find).

 

However, be aware that there are pros and cons to all types of research and the quality of a study can vary even if it is using a higher-quality methodology.

Key parts of a research paper

 

A research paper is divided into sections to break down the research that has been conducted into its separate elements.

Research title

The title should summarise the main aim and purpose of the research study and so should give you a good idea of whether this is a paper that will be of interest for you to read.

Abstract

The abstract is a summary of the research paper, outlining the main points of the study. This provides a ‘quick look’ option if you just want to get an overview of the study. This can be helpful for you to understand if this paper is worth reading more of, or if you are tasked with reading a lot of papers! However, despite the conclusions being stated in the abstract, the lack of detail and context setting means you cannot be sure of the quality of the evidence yet.

Introduction

The introduction is there to set the stage. This section looks to explain the current understanding of this area of research, what is already known or proven and why further research is needed. It provides the context for why this current study has been undertaken and is a useful section to increase your overall understanding of the topic in question.

 

One study is never going to give you all the answers so getting a broader understanding of the area will mean that you can see how the results of this study fit with current understanding. Do they offer a new perspective, or do they seem to conflict?

Methods

This is possibly one of the most important sections of the paper. The methods section provides information on the participants used and the study design. The study design should provide enough detail so other researchers can repeat the study if they want.

 

Understanding the data used, e.g., the participants’ demographic, such as their age, sex, ethnicity, health status etc., helps to decipher if the research is relevant to your community or population.

 

The demographic section can impact the reliability and applicability of the study results.

Figure based on ‘Anatomy of a Research Paper’ from wordvice.com

There can be lots of other information presented in the methods section, but some of the other key pieces of information that you should find are:

 

  • How long the study will last,
  • How participants were recruited,
  • What were the inclusion/exclusion criteria, and
  • What are the trial endpoints (the primary and secondary outcomes that are being measured). This is important to know as you want to be able to see that what was measured matches the objective of the study to ensure the research study has validity.

 

Validity: How well the research measures what the researcher intended to measure.

Reliability: Whether research methods can reproduce the same results multiple times

 

The reliability of a study is in part dependent on the sample size. The more participants a study has, the greater the reliability of the study. Due to demands such as time and funding a larger sample size may not always be feasible and therefore may require further research to check the reliability of the results.

 

The more heterogenous (diverse/dissimilar) a sample, the larger a sample you would potentially need to be able to analyse different subgroups of the population in the results.

 

A study should aim to have a large enough sample size to be able to say that the results are well enough powered, that they can detect a difference between the study groups if one is present.

 

Researchers need to determine the study sample required to avoid under- or over-resourcing their study to align with available finances and ethical considerations. Within the description of the sample, the researcher should also state the number of participant dropouts (also called attrition rate) as this may impact the validity of the results. For example, if you start with 50 men and 50 women and most of your dropouts are women the planned demographic will change and the study will be left with uneven research groups, which could unduly impact the results.

 

Compliance rates could also give clues to the potential real-world relevance of the intervention, a high dropout rate or low compliance would indicate that this intervention may not be something people could apply in their everyday lives as it was not possible to maintain for many, even within a research setting.

Results

The results section covers the key findings from the study. This section is data-driven, rather than subjective analysis. You will often see the results in different forms such as tables or graphs.

What to look out for in the results section

There are a lot of statistics in a research paper and understanding this is beyond the scope of this blog, but one of the main things you can look for is whether the results are significant.

Statistically significant

In research significant doesn’t always mean large it means was it statistically significant. So, you could have what appears to be a small difference that is still statistically significant.

 

Because of this you can also have statistically significant results that are not clinically meaningful. For example, if an endpoint was looking for reduced BMI (Body Mass Index) and they found statistically significant weight loss between the control and experiment groups, but that amount of weight loss was too small to make a meaningful difference to someone’s health then despite the significant result the real-world effect would be too small to justify implementing the intervention.

P-value

The p-value can help you to understand if the results are statistically significant as this number represents the threshold of significance. P stands for probability and measures how likely any observed difference between groups is due to chance. In simple terms, the p-value is the measure of consistency between the results of the study.

 

Most researchers set a p-value of 5% (0.05). This means your results have a 5% or lower chance of occurring under the null hypothesis to be considered statistically significant. The lower the p-value, the more statistically significant the result is. If the p-value is above 0.05 then this would not be considered a statistically significant result.

 

Null hypothesis: There is no effect on the population.

Alternative hypothesis: There is an effect on the population.

Confounders

A confounder is a variable that potentially impacts the relationship between the two variables being studied. For example, if you were studying the effect of eating processed food on cardiovascular disease events, smoking could be another variable that may impact the outcome. The researcher needs to have confidence that the results seen were due to only the processed food and not in fact ‘confounded’ or affected by the fact that the person smoked.

 

There are statistical adjustments that can be made to account for this potential effect, but it is important that this is considered, and these adjustments are made for the conclusions to be valid.

External validity

External validity also known as applicability or generalisability refers to the extent to which the findings of a study can be applied to a broader population under real-world conditions. A larger and more diverse sample is more likely to have external validity.

 

For example, if two studies showed a statistically significant result but one study was conducted with a large sample size and different ethnicities/global communities and one study had a small sample size using only women from a small community in Africa, then it would be fair to assume that the results of the first study are more likely to have external validity and be generalisable. Whereas, in the second study, we can only really conclude that this effect is seen in women in this specific African community and further research would be needed to see if we could replicate this in other groups.

Discussion/Conclusion

After presenting the results, the researcher will discuss them in the context of the research area and elucidate their interpretation of the results. This is where they can highlight the strengths and limitations of their study and suggest further areas of research to progress their findings or to make the research more generalisable.

 

The strengths and limitations section is key to understanding the quality and credibility of a study. It is expected that a study will have limitations, but these need to be explicit to support the correct interpretation and validity of the findings.

Conflicts of interest

If these exist, then they are usually disclosed at the end of the paper.

 

The most obvious conflict of interest (COI) is financial. Independent funding can be difficult to obtain in certain fields and therefore funding may be sponsored by a company with a vested interest in that area of research. This does not always mean that the research is invalid because often researchers are completing the research independently and the results are presented with transparency. You can use what you have learnt in this blog and some of the resources below to identify if the COI may be impacting the quality of the research.

Beyond the research paper

We can also rely on additional sources to validate the findings of research papers. These sources include:

 

  • Check if the trial was registered at clinicaltrials.gov in advance of the trial starting, so the plan for the trial is clearly stated and can’t be altered. This means when writing up the study the researcher will not be able to add in different measures or leave out findings that don’t suit their message.
  • Beware of who has written the press release. It can be that you don’t read the paper itself and only the press release, but sometimes the media report different conclusions to the researchers or leave out the nuance that is required to give context to the results. This is why learning the basics of reading a paper, means you can check this out for yourself, rather than relying on third-party messaging.
  • Check out if the journal that the paper is published in is credible. There are lots of ways of establishing this but key areas include:
    • The journal’s impact factor, which is used to measure the relative importance of a journal.
    • Scientific rigour being prioritised. For example, does the journal ask for research reporting guidelines such as STROBE, PRISMA or CONSORT to be followed?
    • Peer review process. This is where the paper has been under the scrutiny of independent experts in the same area of research to assess the validity and quality of the research.
  • Is there a lay-person summary available? Reliable research in the health and social care space should be made public to benefit all and a lay-person summary means that the research can be understood by anyone, rather than just those who are literate in research.

Summary

Reading and appraising research papers is a skill; it takes time, knowledge and practice, but by looking for these key areas and maintaining a critical mindset, you can begin to identify research that is credible and trustworthy and adds something of meaning to the area of research.

Helpful resources: 

CASP RCT Checklist 

Cochrane Methods risk of bias tool