An abstract must stand alone: it is often the only part a reader reads. A conclusion must synthesise: it is not a summary of what each section contained. Both are written last and both are commonly written badly.
Sign in to mark this article as read and track your progress.
An abstract is the compressed version of the entire report, written so that a reader who accesses only the abstract still has a complete picture of what was done, how, and what was found. It is not a teaser. It does not withhold findings to encourage the reader to read on. A reader who needs only the key result — a database indexer, a systematic review screener — should get it from the abstract.
A structured abstract uses explicit headings: Purpose, Method, Results, and Conclusion. Structured abstracts are required by many computing and security journals, and they reduce ambiguity about which sentence is a method and which is a result.
Purpose: one or two sentences stating the research problem and objective. Method: two to four sentences stating the design, sample, and analysis — enough for a reader to judge the study's rigour without reading the methodology chapter. Results: the key quantitative or qualitative finding, including relevant statistics where space allows. Conclusion: one sentence stating what the finding means in terms of the research question, and one sentence on the broadest implication or limitation.
An unstructured abstract covers the same content without the explicit headings, using prose instead. Many conference and thesis formats use unstructured abstracts, but the same content must still be present.
Too vague: "This study examines machine learning approaches for network intrusion detection." That's a topic, not an abstract.
No results: "The results are presented and discussed in the following sections." An abstract must include the result.
Overpromising the conclusion: "This study proves that federated learning is the best approach for privacy-preserving IDS." An abstract should hedge to match the actual scope of the study.
The most common conclusion error is restating what each section contained. "Chapter 1 introduced the background. Chapter 2 reviewed the literature. Chapter 3 described the methodology. Chapter 4 presented the results." This is a table of contents, not a conclusion.
A conclusion synthesises: it takes the findings from the results chapter and gives a direct answer to the research question stated in the introduction. The research question is the bridge between the two.
If the introduction asked: "Does the proposed personalised aggregation strategy improve F1 on non-IID federated data compared with FedAvg?" the conclusion answers it directly: "The personalised aggregation strategy achieved a mean F1 improvement of 4.7 percentage points over FedAvg on ten-client non-IID simulations, with a large effect size (d = 1.1), supporting the hypothesis that client-specific model components reduce the accuracy loss caused by data heterogeneity."
A conclusion may also:
An abstract reads: "Ransomware is a serious threat to organisations. This study investigates ransomware detection using machine learning. Various approaches were compared and results are shown. Future work will extend this study." Identify every structural problem with this abstract, and list the four pieces of information it is missing.
