We're in a strange situation right now.
Introductions and discussions still need citations, methods sections still need to use validated scales, and you cannot reinvent the wheel during a PhD. We've got to build on the existing non-replicable literature, but simultaneously build more reliable and replicable findings.

How do you know what you don't need to read unless you read it?
It is hard to tell whether a paper is problematic without actually reading the paper.

What to skip when reading a paper

When reading review papers or theory papers, you might skip sections based on interest or expertise. For example, if a review paper discusses implications for an area of research that isn't relevant to you, you might skip that section. For example, I am a non-clinician so, when reading reviews, I often skip any "implications for clinical practice" sections.

When reading review papers or theory papers, remember no single review is authoritative. Reviews and meta-analyses are some of the most valuable papers to read, but remember that authors still write from their own perspective. Review papers also involve some storytelling about the field.

When reading original research, if you are already familiar with the literature, you might skip paragraphs in the introduction. Once you're familiar with an area of research, you can skim or skip entire sections because you have already been "introduced" by reading other papers. Personally, I tend to read the first sentence of each paragraph, then skip any paragraph describing research I already know.

When reading original research, fully read the methods and results. Methods and results are the sections where you learn the most important details about the study. These are also where flaws are most likely to readily appear. The methods and results sections tend to be dry and procedural, which makes them less obfuscated by storytelling. They provide the clearest descriptions of what the study actually did.

When reading original research, the discussion section is often the most dubious part of a paper. Skipping to and reading only the discussion section is insufficient. Unfortunately, summaries presented in discussion sections tend to lack detail and precision that can only be found in methods and results sections. Additionally, during the publication process, editors and reviewers often fail to catch overgeneralizations in discussion sections. If you only read the discussion section, you may miss methodological flaws and important limitations that the authors themselves didn't notice. I have read many papers where the discussion section treats a result as statistically and clinically significant, but a close reading of the results section reveals that the result was non-significant or clinically irrelevant. Some authors wrongly treat non-significant p-values (generally 0.06 – 0.10) as significant by calling them "trending" or "marginal". Other authors fail to consider the effect-size of their findings, claiming too much out of an impractically tiny effect.

Index

Return to Graduate Psychology

Jump to Effect-Sizes

See also How to write a paper for publication