How to read, select, process and conclude from a published study
Anyone can read a research paper but understanding the content requires entirely different skills which require some knowledge of science, the language in which the paper is written and a reasonable knowledge of the subject matter. Over the years, having conducted true research, written papers on literature reviews and conducted peer reviews, I have developed a system which provides a useful way of reading and assimilating information from some of the heaviest papers. I hope this offers some assistance for students and clinicians who may not find the world of academia easy.
Here are some of the questions worth addressing –
Ask yourself what you need from a paper. Do you need data – numbers, do you need methods, do you need outcomes showing the effectiveness of a method? How up-to-date should the paper be? Is there a limited number of papers on the subject you want to read, or are there dozens of papers? What does the study tell us? What is the outcome? What have I learned and is the paper a good paper? Is there a better paper and can I use this reliably?
Gaining access to a source
Studies or academic papers are published online or in regular journals. They may be open access for anyone to read or purchase from a source. The third outlet is through an academic institution, usually universities. All undergraduates, graduates and high levels of academia have access to university libraries. If you are a lay person or have no access to a NHS medical library or university library, you can purchase a paper directly from a journal source, which is often very expensive.
Searching and problems
Modern computer technology now allows us to search for papers. Access is easier if we know the author(s) or title. Where this is not known and we want to find a paper related to a subject, we must use a search engine. Google is one of the best-known search engines but there are other search methods.
Once you find a paper, you will want to know if it is relevant to your requirements. There may be other sources if you want different perspectives on a topic. These are easier to read and often have single authors; they are not necessarily objective and may even carry adverts. The reader must be aware of bias; the author takes one viewpoint and does not critique the information in any way within the paper. The other weakness is that there is no peer review. A panel or independent professional who can fact check and critique the suitability to cut out bias or copy from other sources to make it sound like the author was the source.
There may be articles that stand alone. These have no references and their design may not always follow the traditional anatomy of a study paper. The first thing we must do is separate a formal paper from an article. By all means add an article to your collection, using the information as a guide. The article describes a topic that should provide an overview and be easier to read. Of course, books and other articles are written about the subject so you can take your pick. I endorse reading articles as much as formal study papers but expect a fair amount of duplication or misleading information.
Do remember – not all papers are good, or are helpful or necessarily reliable
Why do you need to use a research paper?
If you are entering an academic field and need to present information for your formal educational contribution to a course or qualification, then you will need to argue your case. If you are giving a talk on a subject, your opinion alone will appear weak if you cannot quote (cite) a paper of merit. How do you find a paper of merit – well you need to ensure the source comes from a reliable publisher. The British Medical Journal is one example and has a wide readership. The Lancet carries a much higher academic rating and the complexity of the contributions are written at a level which is less likely to be easily comprehensive without some good foundation knowledge. For the most part, audiences require material at a simpler level of understanding, so clinicians benefit from simplifying research to support their discussion.
The title of the subject
The topic appears within the title and you need to find the common word or words that aid your search. Newspaper articles use word and sentence arrangements to drive your curiosity. News content may bear little resemblance to the topic. A research topic will use a similar approach to force your eye but will favour the subject and try to show what the paper is trying to achieve. Here are two papers covering limb length discrepancy.
Tollafield D R 1981 Limb Length Discrepancy. West Midlands Study Group Journal
Tollafield D R 1984a A Podiatric perspective in evaluating limb length discrepancy. Journal of Podiatry Association
The two titles above contain similar information. The 1981 paper is very clear about the subject of limb length, but the second article tells us the article looks at how we evaluate or measure limb length. The content might be very different but if we want a broad understanding of limb length, both papers might be worthwhile.
Clinical Photographic Observations of Plantar Corns and Callus associated with a nominal scale Classification and Inter-Observer Reliability Study in A Student Population. Tollafield DR. Journal of Foot and Ankle Research 2017.10,45:2-7
In the next example, corns and callus are the subjects, but this paper won’t inform the reader about corns and callus only whether one can classify corns and callus. The level is scientific and contains some modest statistics. The words ‘inter-observer reliability study’ tell us the paper is limited to the concept of measuring for a reliable or repeatable measurement.
Where do you look first?
Now that you have a title download your paper or read it directly off the screen. You may have several papers to read and find the information quickly may be important if this is the case. You need to recruit all the papers of relevance and discard those of less value. Once you have decided then print the papers off.
Anatomy of a published research paper
There is an anatomical scheme used for the publication of studies and research.
- Title and authors
- Summary or abstract
- Introduction or background
- Method
- Results
- Discussion
- Conclusion or summary
- References
- Author Details
Within this anatomical view of a research paper are subheadings, figures, tables and graphic representations. You can read original papers and read what is known as a literature search paper often called a systematic review or the more popular meta-analysis paper.
Limited searches and the value of the reference section
If the subject has limited resources, read as many as possible. You may well be considering an original topic, although remember there are many papers in languages other than English. South Korea has a good publishing record as other countries; many of these countries produce work in English.
If there are 50 papers, use the title first to guide you to the specific question because you will know how many are of little value. Next consider using papers between two dates: for example those from the last 5 years. Selecting dates will reduce your search but might miss an important paper. We will discuss avoiding missing the motherload by concentrating on references.
If you require to read more papers look at the references. First, you will find specific papers that the author(s) has/have used. There is often the mother paper, which is the first paper on the subject. I like to have the primary paper as it allows me to check back to avoid Chinese whispers. Many authors cite other authors who in turn, cite others. This forms secondary referencing and fails to acknowledge the source.
Loss of information and interpretation can arise. If J. Williams wrote a paper in 1914 and in 1970 A.R Jamieson wrote another paper on the same subject, then in 1991, R B Fletcher may cite Jamieson and leave out Williams as he did not bother to go back far enough. Good papers will show up with the same original authors so let us say B.J Turner writes an up-to-date paper in 2010 and cites all three previous authors; you know they have been more thorough. You can tell a good deal from references – their sources, the dates and whether they have undertaken a good review.
The meta-analysis is popular because it does 2 things effectively in the sense of helping selection. It shows how many papers exist on a subject in a specific category. It is not unusual to find hundreds to several thousand papers. Instead of quoting each paper the author(s) triage the papers down until they find the best papers using strict selection criteria.
Matthew et al (2019) show how the authors have reduced 3066 papers down to 170 then finally 22-25 papers
Barry G. Matthews, Sheree E. Hurn, Michael P. Harding, Rachel A. Henry and Robert S. Ware The effectiveness of non-surgical interventions for common plantar digital compressive neuropathy (Morton’s neuroma): a systematic review and meta-analysis in the J. Foot & Ankle Research
At this point, you can start to read the papers left over depending on the subject you wish to study. You want to read those papers with the strongest research validation. This includes papers with a controlled study, and a prospective study design rather than a retrospective; ideally, each paper will have a similar measurement method. If this were true, it would be wonderful. Most papers use a visual analogue scale (VAS). Others use different styles of questionnaires or measurements. Too many different methods make the comparison difficult. Poor methods of assessment design are relegated early as in the case of Figure 1.
The Abstract or summary
Many readers who desire to use material from papers just read the first section, the abstract or the summary. This section does break down the paper into a summary showing the broad findings. Use this section and look at the conclusion to see of this paper is what you need. I would say at this point never use a paper where you rely on the abstract alone. This happens with foreign papers. The abstract is in English and the remaining paper may be unavailable or written in a different language.
The abstract will show some data – the number of people in the study, the gender, age, the brief method and results. You have within your grasp a reasonable handle on the paper and can compare this to other papers so you reject those that are less relevant. You are in effect carrying out your meta-analysis. If two papers have a study of 98 and 105 people (the cohort) and a third paper have 30, you may prefer to concentrate on the larger numbered studies.
If I am writing something significant, I prefer to include all papers with a similar study design so I can make critical comments. Recently I wrote an article based on three papers. One from an orthotist, one from chiropractic sources, and the last from an orthopaedic source. The design study for the orthopaedic paper was more recent and better designed but it helped to contrast the other sources which had value for readers.
Make notes along the way
By writing notes from the abstract, you can read the rest of the paper, pay attention to the method and the results, and then refer back as needed. The conclusion should allow a clear explanation of the weak areas but you should be able to spot these. Make more notes so you can summarise the paper in your own words. Store paper summaries like this, as it is easy to forget or lose your source. I use an exercise book or blank postcards.
In some journals statistical analysis is fairly slim; in others statistical tests are packed. You should understand probability – p but it is not always necessary to be sufficiently adept with all the tests because you should be able to see from other results the value of the data collected. The subject of statistics is important and allows us to judge the merits of a paper. The smaller the p-value, the better the probability associated with the chance of the observation being useful.
In the final analysis, remember most single studies are limited to a narrow field of enquiry. Meta-analysis studies are helpful because they winkle out the strengths of different papers. If one paper gives you one useful message, that is perfectly reasonable.
As clinicians, all we can derive from papers is information that may explain one part of the puzzle. Most papers change little and the exciting part is when you have read 20 papers and finally found a message that stands out from the mother paper to the most up-to-date paper. You may have completed your systematic review, but you have the personal satisfaction of finding something hidden away. The more we read, the better we understand and realise that most papers change little. The value increases when combined with other papers and those of similar design.
Thanks for reading ‘research papers made simple’ by David R Tollafield
Published by Busypencilcase Communications Est. 2015
David was a former lecturer and consultant in podiatric surgery and has written and published many papers throughout his career. You can explore some of his published materials on this site books are available through Amazon.co.uk or Amazon.com
Recent Comments