Prof. Dr. Larry AdamsAcademic, Author & Researcher
How to Conduct Quantitative Research · Chapter 22 of 33

Part IV — WRITING THE DISSERTATION OR RESEARCH BOOK

Chapter 18. Discussion and Integration with Literature

18.1 What the discussion must do

The discussion explains what the findings mean, how they answer the research questions, how they compare with prior evidence, what theory they support or challenge, why unexpected results may have occurred, and what can reasonably be concluded. It should not merely repeat the results chapter or claim that every hypothesis was “proved.”

18.2 A robust discussion sequence

  1. Briefly restate the study aim and the main finding in relation to each objective.
  2. Interpret the magnitude, direction and uncertainty of the estimate.
  3. Compare the result with relevant empirical studies and explain agreement or disagreement.
  4. Explain theoretical implications and whether the conceptual framework needs refinement.
  5. Consider plausible contextual, methodological or measurement explanations.
  6. Discuss alternative explanations and possible confounding.
  7. Distinguish statistical from practical significance.
  8. Explain what the design permits and does not permit the reader to infer.
  9. Transition to limitations, implications and recommendations.

18.3 Interpreting unexpected or null findings

An unexpected result is not automatically a failure. Check data quality, coding, precision, measurement performance, assumptions and model specification. Consider whether the estimate is compatible with both small and meaningful effects. A non-significant result may be informative if the confidence interval rules out effects that would matter in practice; if the interval is wide, the study may simply be inconclusive. Avoid post hoc stories presented as if they were predicted in advance.

18.4 Link findings to theory

Explain whether the observed pattern is consistent with the mechanism proposed by theory, whether the evidence challenges a boundary condition, and whether alternative models remain plausible. A single observational study rarely confirms a theory in full. Strong discussion recognises uncertainty and makes the contribution proportionate to the design.

18.5 Avoid common discussion errors

  • Do not claim causality from a cross-sectional association.
  • Do not treat a p-value as the probability that a hypothesis is true.
  • Do not ignore non-significant or contradictory results.
  • Do not compare studies without noting differences in population, measures and design.
  • Do not introduce a new major analysis without explaining why it was needed.
  • Do not claim novelty solely because the study was conducted in a new location.
  • Do not recommend interventions that the evidence does not support.