How to write the discussion chapter (and what NOT to write)

A well-written discussion chapter does more than restate results—it persuades readers that the findings matter, connect to the research questions, and fit (or challenge) what is already known. Without that work, the chapter risks functioning like a catalog of outcomes: readers may understand what was found but not why it is valuable, how it supports or contradicts prior studies, or what comes next for the field.

The discussion chapter’s job is to “situate” findings—linking them back to the literature presented earlier and clarifying how they answer the study’s research questions. A useful way to think about the difference is the ingredients-versus-cooking analogy: the results chapter lays out ingredients, while the discussion chapter explains what to make from them. That means the chapter should help readers evaluate the significance of the study, not just observe its outputs.

Across different university expectations and dissertation formats, three elements consistently belong in the discussion. First is explicit reference to the literature from the literature review—showing whether the findings support existing claims, contradict them, or extend them. Second is the researcher’s “voice,” which is more interpretive than the results chapter: it can include judgments like whether something was surprising, and it should also capture the reasoning behind interpretations. Third is the “so what,” which centers on answering the research questions clearly. Even if the results chapter already hints at how each question is addressed, the discussion is the final checkpoint that makes the mapping unmistakable.

Interpretation is where the discussion earns its place. Results sections typically stay objective and descriptive, but the discussion can interpret selectively—offering emerging explanations, suggesting why patterns occurred, and tying those ideas back to prior literature. The elements overlap in practice: an interpretation often relies on literature, and both ultimately serve the “so what” by clarifying how the study advances understanding.

Structure is flexible and depends on what the study highlights. There is no universal template, so the chapter’s backbone should reflect the study’s priorities. Common approaches include organizing by research question (each subsection tackles one question and discusses relevant findings, interpretations, literature links, and implications), by themes (major thematic categories become subsections), or by a guiding theory (the discussion is centered on how findings support, refine, or challenge that theory). In theory-generating approaches such as grounded theory, the discussion may go further—using the chapter to develop new conceptual claims, which increases the importance of careful interpretation and clear justification.

Two pitfalls stand out. The discussion should not introduce new literature or new findings. It is normal to discover new concepts while writing, but those discoveries must be incorporated back into the literature review so the discussion doesn’t spring surprises. Similarly, any findings or data excerpts used in new interpretations must already appear in the findings chapter; the only genuinely new material in the discussion should be the researcher’s voice and interpretation—not additional evidence or newly introduced studies.