There are several depositories that can help you find relevant research on firefighter health. Two of the most common are: PubMed: www.pubmed.com and Google Scholar: http://scholar.google.com/. For either of these, you can search by keywords. The Firefighter Life Safety Initiative of the National Fallen Firefighters Foundation also has compiled a database of references from the academic literature by topic areas that are relevant to the initiatives: http://www.lifesafetyinitiatives.com/. A good way to expand your literature search once you have found a relevant article is to look at what articles the authors site in their reference section. Some databases also provide similar references to your search. Some journals offer free online access. Other journals can be accessed through a university setting or the National Fire Academy’s library. You also can email the contact author directly to request a copy of the publication. Copyright agreements often limit the ability of authors to post full articles online for public use.

Journals can be in print and/or be published online. Some have public assess (the articles online) while others are by subscription only (typically subscriptions are held through libraries and universities). Authors submit the manuscript which is reviewed by 2-3 other scientists. The peer scientists typically review and critique the manuscript in detail and return comments to the editors. Editors either reject, accept or request revisions to the initial manuscript. If the authors receive a “revise and resubmit”, they have the opportunity to revise their manuscript in accordance with the reviewer’s comments and/or respond to the reviewer's concerns. The editors can then either accept or reject the revised work. Most journals publish only peer-reviewed articles. The importance of peer review is that scientists outside the research team have read, evaluated, and approved the publication before it is ever published. This ensures that the resulting manuscript is scientifically sound, quality work that is usually free of major defects.

The date of the publication is important because of how quickly science changes. To be up to date on current data, it is useful to refer to and use more recent publications (e.g. approximately the last 10 years)

Most publications have a corresponding author who provides his/her contact information. Feel free to contact the authors with questions – most are very open to hearing from those who are consuming their research. Author affiliations also are included. If you find publications you find useful, it is sometimes useful to look up the author as most authors publish within a topic area and have more than one relevant publication on a topic. The order of the offers matters as the first author is typically the primary or lead author. Depending on the tradition of the field, the author in the last position is often the senior author and/or the principal investigator of the study.

The abstract (which is usually available online) has a summary of the research and is similar to an Executive Summary of fire service reports. Typically the abstract includes a brief summary of what is already known about the topic and what is missing from the literature. Next, there is a brief description of the methods used and information about the sample participating. Results highlight the most important or key findings of the study. Finally, the end of the abstract usually summarizes the conclusions, provides future direction, or provides information about why the results are important.

In the introduction, the authors should summarize what is already available about the topic in the peer-reviewed literature. Typically, the introduction ends with the research question being studied and/or hypotheses about what the authors expect to find. Introductions are a great place to find additional resources. Data presented should be referenced with the original research where it was published. Check the reference section to find the original study results.

The methods section provides important information about how the study was done and who the participants are.

There are several different types of research designs and some are stronger designs than others. For instance, a cross-sectional design asks firefighters questions at one point in time and statistically examines the relationship between variables. Cohort studies follow people across time and measure domains at multiple time points. Randomized controlled trials randomly assign participants to groups and provide different treatments to each group. Comparisons are made between the participants in each group to see if groups respond differently. Some studies are reviews of the literature. A meta-analysis is a study that statistically combines the findings of several different studies and weights the findings based on how large the sample size of the study was – so, a small study would receive less weight than a large study – outcomes are standardized and combined so conclusions can be made about the literature in general on a topic.

It is important to know who the participants of the study were so you can identify how widely the findings can be generalized. For instance, if the study participants are all males, you would need to be careful assuming the results would be the same for women. If the firefighters were just from the South, it could be that the findings would be different in another area of the country. Sample size (typically noted as “N” or “n” depending on whether the sample is the whole group or just a portion of the whole group) also is an important thing to look at. The larger the study is, the more confident you can be in the findings. The response rate also is important because it tells you how many people of the ones solicited chose to participate. If the response rate is low, you need to be more concerned about bias in the findings. It could be that only those interested in the topic being studied participated and are different than those who didn’t choose to participate. The higher the response rate, the more likely it is the findings are representative of the actual population.

This section provides information about how the domains being assessed were measured – what instruments were used, whether the questions were created by the authors or from existing surveys, etc. It is typically preferred that scientists used existing measures that have been tested for reliability (the results are the same if you ask and re-ask the questions) and validity (the questions are measuring what they are trying to measure).

Usually, once an article has made it to the published literature, you can be confident that the statistics conducted are appropriate. Many editorial boards include statistical experts who review any questionable statistical analysis. More important than understanding the ins and outs of the statistical tests is knowing how to interpret them. Most commonly, statistics are reported with p-values. The p-value is the probability that a result would occur by chance. For instance, if p=.04, there is only a 4% chance that the results found would have occurred at random. P-values less than .05 are typically considered statistically significant by scientists.

In the results section, the authors outline the findings of each of the tests they have done. Results often are presented in both paragraph form and in table form. Table titles typically explain what data is being presented and usually statistically significant findings are noted.

The beginning of the discussion section typically presents a summary of the main findings of the study in plain language. This section usually also includes strengths and limitations to the study (every study has limitations so they don’t necessarily make the study less useful, but explain what should be kept in mind while interpreting results) and future directions (perceptions of the authors about what steps should be taken next and what other research questions remain).