The methods section provides important information about how the study was done and who the participants are.
Research Design: 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 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.
Participants: 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 didnt choose to participate. The higher the response rate, the more likely it is the findings are representative of the actual population.
Measures. 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).
Approach to Statistical Analysis. 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.