Understanding Scientific Research [1/2]: How to Read a Study and Interpret Results



A skeptical mind is always aware that scientific research is progressive, and that new findings are constantly emerging and can be used to change the way things are. The ability to interpret and evaluate research in context is critical in the process and understanding the results.

All studies depend on their own context. This is a crucial point and the reason why the public can be confused by contradicting studies and in some extreme cases distrust or reject science altogether. For example, 2 studies that apparently contradict each other can both be valid, under different circumstances.



Research plays an important role in advancing any profession or any decision-making process. This process should be developed following scientific laws, principles, and concepts derived from scientific research. These principles, theories and concepts are all subject to change as new research emerges, scientific papers are constantly published. 

New research can and should produce changes in one’s approach to a problem, so it is important to keep and open eye and open mind and be free of any dogmatic and fundamentalist thinking. Keeping up to date allows us to stay on the cutting edge of knowledge and advance any practice, professional or otherwise. Ultimately, research increases our understanding of a giving subject.

Introduction

The first section of any scientific manuscript is the introduction in which the authors develop the hypothesis tested by their research design. With this purpose in mind, a concise and short review of the scientific literature is outlined as a basis for the development of their specific hypothesis.

This introduction sets up the study’s context and relevance. Potential criticisms addressing the problem, hypothesis, or methods should be outlined as well.

Methods

This section is greatly important to understand how a giving hypothesis was tested and the design conducted, and possibly provides valuable information for other scientists to replicate the study, or make improvements.  The methods section should explain the approach to the problem, and research design to test the hypothesis presented in the introduction.

Detailed and specific information is presented on the type of subjects, specific types of equipment used, explanations for the procedures, and how they gathered data and then applied statistical analysis to test their hypothesis.

Independent variables for the study design and the dependent variables measured are also explained in detail, as well as their rationale selection. All procedures must be explained in great detail so others can replicate the study.

This section usually ends with comments on the statistical procedures used to analyze the data generated. Finally statistical significance is defined (P ≤ 0.05.)

Results

The results section can be the most important if the design and methods are strong. Findings are often presented in terms of p-values and/or confidence intervals, along with means and standard deviations of effect sizes (it depends on the field and specific question of inquiry). However since subjects may not respond equally, individual data for individual responses of subjects, if possible, should be also presented.

Discussion

Here the authors outline the paper’s importance. They interpret the results, and relate it to the scientific literature, giving it context and meaning. In other words, the study's findings must be put in context of the literature as a whole, so as to avoid cherry-picking. Strengths and weakness of study design should be expressed, and suggestions for future research. New questions may arise from the study needing further research.



If you see that a scientific article does not follow these simple guidelines then it is likely to be a bogus or flawed article.

A well-substantiated report is as vital to science as is a well-designed study (1). A sound, well-discussed report might be as necessary as a proper research design, if not more important. A high-quality original study needs a good deal of interpretation and discussion.

The report must not be poorly written and sentences in the introduction and discussion must be backed by references. Stating a scientific claim without substantiating it using proper references renders the statement unreliable.

Unsubstantiated scientific allegations might blur the distinct lines between scientific facts / theories and unproven ideas / theorems.

Context and practical applications

Ultimately, the purpose is to extract relevant information to improve a decision-making process, methods and or techniques, with the cutting edge data available.

We must realize that results from a study relate to the context, specific situation or conditions studied. Putting research into proper context is crucial in the process of translating it into practical applications.

So, here is where it is important to understand the context of the study which is related to the independent and dependent variables.

A closer look at Independent and dependent variables

Independent variables are factors controlled or selected to be held constant. Independent variables can be manipulated to see whether they alter the dependent variables’ response and if possible to understand its influence.

A “confounding variable” is a variable that is not controlled but has an independent effect. This lack of control can obfuscate the interpretation of the results. Ideally, the independent variables (sex, age, training status, temperature, nutrition intakes…) of the study should match the situation and population as closely as possible, this way such results can be applied to a specific situation or group of people.

A dependent variable, or “a response variable,” is measured in response to the set of independent variables in the study design. These variables can not be controlled, however they act as the outcome variable for the study.

The validity and reliability of a variable is also important. A measure only has validity if it measures what it is supposed to measure. Reliability on the other hand, relates to how consistent a measurement is, the consistency in repeating measurements. The variance should be no greater than about 5% between the two values for the measurement to be accurate. Heart rate, core temperature, oxygen consumption, are a few examples of dependent variables.

Independent variables can be endless, therefore all studies depend on their own context, as noted before. This is one crucial point and the reason why the public can be confused by contradicting studies and in some extreme cases distrust or reject science altogether. For example, 2 studies that apparently contradict each other can both be valid, under different circumstances.


Would you like to know more? Subscribe.

Summary of 50 articles with 49.167 words and 1321 references on
Exercise and nutrition