WRITING EMPIRICAL PAPERS
your results relate to previous research? Remember your study was designed to address a gap in existing research
literature; comment on how your findings fill that gap. Relate the results to the theories you introduced in the
Introduction. Your findings are just one piece among many -- resist the tendency to make your results the final story
about the phenomenon or theory of interest. Integrate the results and try to make sense of the pattern of the findings.
In some ways, decisions about how to organize your Discussion section come down to an awareness of what
your paper is fundamentally about. For example, imagine you did a study looking at the correlation between
perfectionism and self-esteem. The original focus of the project and the core area of interest in the paper could be
perfectionism. If this were the case, you would write about self-esteem as one of many interesting things to understand
about perfectionism. Alternatively, you could do the same study but with self-esteem as the core interest. In this case,
you would write about perfectionism as one of many interesting things to understand about self-esteem. This decision –
what is my paper about – is especially critical when considering what areas of the research literature you want to
emphasize and what ideas for future research you want to include.
If your results did not support your hypothesis, a central goal in the Discussion section is to explain why not.
In addressing this issue, the required elements about limitations and future research are often interwoven with each
other, and with the explanation of what you found. Consider the narrative flow of the paper. For example, in an effort to
fill in the blanks for each required element, many novice writers produce a Discussion section that is repetitive and lacks
depth. For example: “We did not find a significant correlation. This is inconsistent with our prediction. We mostly likely
did not find a significant correlation because our sample size was too low. A limitation of our study is that our sample
size was too low. Future research ideas include obtaining a larger sample.” Do not force an artificial separation amongst
your ideas. Rather, generate a Discussion section that is coherent and flows naturally. Most of the time, writing with this
mindset does end up creating the form you need for the Discussion section as the bottom half of the hourglass (moving
from your specific findings to broader implications).
If your results did support your hypothesis, the sections on limitations and future research are typically more
encapsulated (separate). With significant results, a central goal of the Discussion section is to re-examine the research
literature in light of your findings. What is the importance of your new discovery? Every research study has limitations, if
only because of the methodology used (e.g., correlational research cannot establish causality). So, you will be examining
your work for limitations. Future research can take any direction you wish. Now that we know the answer to this
research question, what is the next gap in the literature we should address?
Be thorough when you think about the possible limitations of your research, and be judicious about what you
choose to include. For studies that do support their hypotheses, common areas to examine include possible "third
variable" explanations, unmeasured mediators, and/or issues with the generalizability of your results. For studies that do
not support their hypotheses, there are three major areas to examine: (1) the logic of the research hypothesis; (2) the
materials you used to conduct the study; (3) your sample.
1. Reconsider the logic of your hypothesis. You may not simply say, “we were wrong.” After all, you presented a
thorough logical argument in favor of your hypothesis in the Introduction. Rather, there may be previously
undiscussed nuance or detail that influenced your results. For example, could there be unmeasured moderators?
2. Consider the materials you used in your study. In particular, think through how you operationalized the
constructs, the specific procedures used, and the possibility of self-report biases.
3. Consider your sample. Students’ favorite limitation to report is an unrepresentative sample, Think more deeply
than that. Remember you are trying to explain why you did not support your hypotheses. Perhaps you have
restriction of range problems because you did not have participants scoring at the high (or low) end of a
particular scale. Put another way, why do you think your sample lacked variability? Ideas for future research can
address these limitations but watch out for the tendency toward repetition (refer back to the poorly written
Discussion section example above).
Be specific when discussing limitations. For example, if you claim that a third variable might affect your
correlation, tell the reader what that third variable is and how it affects the results. If you think that the use of a
convenience sample (and thus, a non-representative/random sample) is a limitation, you must explain what segment of
the population might respond differently than did the participants in your sample and why. Avoid listing every possible
limitation or qualification you can think of. Rather, what are the points other people might most likely notice? What are
the points that have the strongest implications for future research?