Reading a Research Article

This note is for those new to reading scientific research. Research designers and statisticians may find some of it more concrete than their conceptual frameworks. It further develops the information found in other files in the Notebook series.

One of the important skills college students must begin to master early is the reading, understanding and interpreting of research publications – whether print or electronic books, journals, reports, papers, etc. Doing so will prepare students for doing assignments and literature review papers (sometimes called library research). The information and knowledge read and collected for assignments and papers is often written in styles new to students and may involve the writers’ use of evidence and empirical research and a specialized academic vocabulary. The classic form of inquiry – the scientific method – is used in many disciplines.

The papers written for most undergraduate classes do not often involve original research of the types found in scholarly and scientific journals. Student papers are similar to the ‘review of literature’ sections of these papers. Sometimes undergraduates do write about their own research. This too is done more professionally when the student has good grounding in how to read the research and reports and literature reviews.

Before reading scientific research, one should think about the Basic Principles of Science*.

  • Uncertainty – Whatever is reported is always with this caveat: There is the strong probability that this is accurate, but we expect to learn from new research studies.
  • Probability (p) – Uncertainty is dealt with by measuring probability. This is statistical significance. Most scientists accept a minimum statistical significance of p < .05 which means there are fewer than 5 chances out of 100 that the results are due to chance. Researchers can only report what appears to be the case. Underlying all science is the concept that findings and conclusions never prove. Also, science cannot verify a negative – that something cannot or will not occur.
  • Power – The larger the study (large sample with many subjects), and the larger the number of studies with similar findings, the more likely the findings are not due to chance. In statistics, power is the likelihood of finding real differences if they are there.
  • Bias – It is hard to keep bias out of research. Some findings are contaminated by factors that confound the results or have undue influence on what was observed. These can be characteristics of the subjects, environmental conditions, faulty measurement tools, inappropriate sampling, one of the variables in the study, etc. Statistical methods can be used to control some kinds of bias but cannot correct for poor research design.
  • Variability – Measurement is one of the major difficulties of science. Any trait measured will vary from one subject to another or one study to another, albeit slightly. There is always the element of chance that the measurement was faulty. Statistics help determine the odds and control for it. The greater the odds against the observation being due to chance, the stronger the finding represents real differences.

*Adapted from the writings of Victor Cohn, a science journalist, writing for The Foundation for American Communications - www.facsnet.org.

Scientific research papers have a basic structure that is sometimes varied because of field of study, place of publication and author’s preferences and interests, but there are academic traditions that are important. What follows are some definitions and descriptions to help read and understand several types of academic papers.

Researchers more often say that they are studying something rather than 'doing research.' When looking for research information on a topic, use ‘study’ as one of your keywords, rather than ‘research’.

As indicated above, findings or conclusions never prove; they only support or do not support a particular thesis, relationship, activity, observation, hypothesis, theory, etc. Science is about observation and controlling the sources of uncertainty as much as possible with the methods available at that time. Scholarly writers in other fields gather their information in a variety of ways and they write their findings in literary essays and monographs. Their findings also may be altered by what is learned later.

Conclusions in all disciplines can change as more is learned from additional research. Knowledge in newer fields of research, with few completed studies, is more likely to change as more is learned.

Scientific and other scholarly papers may be lengthy and use specialized language. It pays to have a dictionary handy. It is difficult to skim a scientific article or scholarly treatise in the sciences, arts, humanities, social sciences or business.

The structure of a scientific article is likely to include many of these parts (sometimes with synonymous names):

  • citation –author(s), date, title of article, name of journal, volume, page numbers.
  • abstract – a brief summary of the work
  • introduction and review of literature – describes the purpose of the study, empirical questions, hypotheses, issues, events, phenomena in the study and its relation to previously published literature.
  • methods and materials – describes the participants, ideas or things studied, the methods and procedures for obtaining data (what was done with what or whom), and the statistical or other analysis to be used
  • findings – reports on results of the analysis of the data; what was found or not found. The statistical significance will be reported.
  • discussion and conclusion – puts the findings in the context of other related literature; describes what the author(s) believes was learned and what it means
  • limitations – authors may report problems with their procedures
  • recommendations – what the authors suggest should be done next by themselves or other researchers to further study this question. Sometimes suggestions are made for the application of what was learned in appropriate places in the wider social, business and technological community.
  • references – the literature mentioned in the review and discussion that was used while the authors performed their studies.
  • charts, tables, graphs, illustrations, photographs – these are used to clearly and succinctly report the data and findings. They often appear in the findings and methods and materials. Each requires its own approach to being read, just as the text does.

 

Answering the following questions and taking notes as you read can help focus and clarify a research study. Writing your responses in a notebook or on notecards (print or electronic) are good ways to absorb what is being described.

 

  • Does the title clearly identify the topic of the research?
  • What is the empirical question? Convert the title into a question.
  • Why was the study conducted (what is the purpose)?
  • What ideas are found in the literature review?
  • What was the hypothesis or hypotheses to be tested?
  • Who or what were the subjects?
  • What type of research was it? (i.e. case study, experiment, relationship study, predictive study)
  • What were the independent variables that were manipulated or classified?
  • What were the dependent variables that were measured and how were they measured?
  • What were the procedures of data collection (what was done to or by the subjects)?
  • If it is an experiment, how were extraneous variables controlled (held constant; randomized; etc)?
  • What were the results based on the hypothesis(es)?
  • What information is provided in the charts, graphs and tables?
  • What conclusions did the researcher(s) draw from the study?
  • What questions do you (the reader) have about why the study was done, the procedures used, the ethical practices of the researchers, or the conclusions drawn?
  •  

    **Adapted from a handout by Dr. Ron Feldt, Professor of Psychology, MMU