The discussion chapter interprets your findings given your literature review. It doesn't repeat findings. It explains what they mean. Here's a step-by-step structure that works.

Step 1: Restate Your Main Finding and Situate It Relative to Comparable Research

Open the discussion with one sentence naming your main finding. Not a vague statement. A specific claim.

Example: "The primary finding of this study is that peer-led mental health support groups reduce staff burnout more effectively than standard occupational health services, but only when institutional leadership actively endorses the intervention."

Then, immediately situate this finding relative to the most considerable comparable study in the literature. Do you align with most previous research or not?

Example: "This finding aligns with Jones (2020), who found that organisational endorsement moderated the effectiveness of peer support interventions. However, our effect size was larger than Jones reported (Cohen's d = 0.67 versus d = 0.42), suggesting that the particular context of healthcare may amplify the impact."

This opening tells the reader immediately what your finding is and where it sits in the existing knowledge.

Step 2: Interpret the Finding

What does this finding mean? What might explain it? This is where you move from description to interpretation.

Example: "Why might institutional leadership endorsement matter so much? Two mechanisms seem plausible. First, when leaders publicly support peer groups, attendance increases and stigma decreases. Staff perceive that their managers value peer support, making participation seem safe and productive. Second, when leaders endorse peer groups, they allocate time for them. Staff participation requires protected time away from clinical duties. Without explicit leader support, peer groups occur in staff breaks and seem like an optional extra rather than a core support service. The effect size in this study suggests the endorsement mechanism is substantial."

You're not just reporting what happened. You're reasoning about why it happened.

Step 3: Compare With the Literature

Does your finding agree with most previous research? If yes, why might that be? If not, why might you differ?

Example: "Three previous studies (Smith, 2019; Brown, 2020; Lee, 2021) found similar effects of peer support when institutional backing was strong. However, four studies (Davis, 2018; Wilson, 2019; Garcia, 2020; Martinez, 2020) reported no considerable effect of peer support regardless of context. The difference may reflect variation in intervention quality. Studies reporting no effect often lacked structured protocols for peer facilitator training. This study provided explicit training, which may explain stronger effects. Alternatively, differences in occupational groups studied (healthcare versus other sectors) might matter. Healthcare workers face specific stressors that peer support might address particularly effectively."

You're explaining what agrees, what disagrees, and plausible reasons for disagreement. This demonstrates sophisticated thinking.

Step 4: Address Unexpected Findings Separately

Every dissertation has at least one finding that surprised you. Don't bury it. Examine it with curiosity. Unexpected findings, handled well, demonstrate analytical maturity.

Example: "One unexpected finding was that burnout reduction persisted at six-month follow-up even for the 40% of participants who didn't continue regular peer group attendance after the first month. The literature suggests that benefits of support interventions typically decay once support ends. Several explanations are possible. First, the early peer group experience may have triggered internal changes (increased help-seeking, normalised discussion of stress) that persist even if attendance stops. Second, participants who stopped attendance may have found alternative support. Third, burnout may have reached a threshold and stabilised rather than continuing to decline, creating an illusion of persistence."

An unexpected finding examined thoroughly shows you're thinking critically about your data rather than just reporting what fits your hypothesis.

Step 5: Address Your Limitations Honestly, Not Defensively

Every dissertation has limitations. State what they're without apologising or over-explaining.

Example: "This study was conducted in five hospitals in the South of England. Healthcare settings vary considerably by region, healthcare system, and resource availability. Urban hospitals have different staffing pressures and demographic profiles than rural hospitals. NHS hospitals operate under different constraints than private healthcare settings. The extent to which these findings generalise beyond the South of England NHS context is unknown and would benefit from replication in other regions and healthcare systems."

Not: "Of course, like all studies, this study has limitations, and while these limitations must be acknowledged, they don't basic undermine the strong and important findings that emerged from this careful research."

The first version is honest and specific. The second is defensive and padding. Choose the first.

Step 6: State the Implications

What does your finding mean for theory, practise, policy, or future research?

Key Considerations and Best Practices

For theory: "These findings complicate the standard change management literature, which emphasises communication and planning as the primary drivers of successful change. This study suggests that psychological contract management is equally important. Change management frameworks should explicitly address the psychological contracts individuals hold with their organisations and attend to perceptions of breach."

For practise: "Healthcare organisations implementing peer support should ensure that institutional leaders explicitly endorse the initiative. This need not require extensive resource allocation. Public statements of support, protected time for participation, and inclusion of peer support in staff development frameworks create the endorsement signal without substantial additional cost."

For policy: "Health policy currently incentivises organisations to provide occupational health services. Policy makers might consider whether peer support should be mandated as complementary to standard occupational health, particularly in high-stress sectors."

For future research: "This study examined one specific peer support model. Future research should compare different peer support models (professional-led versus peer-led, group-based versus individual, formal versus informal) to identify which elements drive effectiveness."

Common Structural Errors to Avoid

One enormous discussion section with no subheadings. Use subheadings to signal structure. Each heading should name a claim you're making.

Subheadings that name themes rather than analytical claims. Wrong: "Peer Support." Right: "Institutional Endorsement as a Moderator of Peer Support Effectiveness."

Repeating findings verbatim rather than interpreting them. The findings chapter reported your results. The discussion explains what they mean. If you're copying sentences from the findings chapter into the discussion, you're not interpreting.

Interpreting findings without reference to the literature. Your discussion should constantly compare your findings to what's already known. "In contrast to Smith (2019), we found..." "This aligns with the meta-analysis by Jones (2020)..." Keep literature in mind.

Introducing new literature in the discussion. Everything you cite should have appeared in your literature review. The discussion interprets findings given the literature you've already introduced.

Expanding scope. If your dissertation examined hospitals in the South of England, don't discuss what your findings mean for primary care or international healthcare. Stick to your scope.

Frequently Asked Questions

Q: How long should my discussion chapter be? A: For a 10,000-word dissertation, the discussion is typically 1,500 to 2,000 words (15 to 20% of the total). For a 15,000-word dissertation, 2,000 to 2,500 words. For a 5,000-word dissertation, 800 to 1,000 words. Your discussion should be longer than your introduction but roughly similar in length to your literature review.

Q: What if my findings don't match my hypothesis? A: Address this directly in the discussion. You expected X, but found Y. Why? This is interesting. Non-considerable findings can be as meaningful as considerable findings if you interpret them thoughtfully. "We hypothesised that burnout would be lower in the intervention group. This hypothesis wasn't supported. We did find lower stress levels in the intervention group. The literature distinguishes between burnout and stress. It's possible that peer support reduces stress but not burnout, or that our sample sizes were insufficient to detect burnout differences. Future research should clarify this distinction."

Q: Should I make recommendations in my discussion? A: In some disciplines yes, in others no. Social science dissertations often include recommendations in the discussion or a separate recommendations section. Science dissertations typically don't. Law dissertations often do. Ask your supervisor what's expected in your discipline.

---