How to write an honest but effective abstract for scientific papers

Due to lack of training in scientific writing and sometimes unethical practices, abstracts are often poorly written, lack critical information, and sometimes contain spin. An effective abstract provides brief but adequate information on the purpose, procedure, results and implications of a study. This paper discusses the key components of a good abstract, offers recommendations on reducing spin in abstracts, and analyzes three abstracts written by the author with the goal of helping young researchers write honest and effective abstracts. © 2019 The Author(s). Published by Elsevier B.V. on behalf of African Institute of Mathematical Sciences / Next Einstein Initiative. This is an open access article under the CC BY license. ( http://creativecommons.org/licenses/by/4.0/ )


Introduction
Scientific African is a multidisciplinary journal that provides a platform for the cutting-edge research from the African continent.The goal of Scientific African is to be a regional journal with a global impact showcasing excellent science through publication of high-quality research.However, the attainment of research excellence in Africa is often hindered by insufficient research support, lack of mentorship, and poor research infrastructure [4] .Hence, our editorial team is committed to promotion of research excellence by helping African researchers improve their scientific writing skills and experimental design.
The publication of research papers in peer-reviewed journals is important for timely and trustworthy dissemination of data.We engage in scientific research to advance knowledge or present solutions to problems.Publishing a research paper is simply a means to attaining this goal, it is not the goal.For that reason, researchers should ensure that their research papers are read by other researchers and policymakers.Most people read the abstract of a research paper before deciding on whether they should invest their time reading the whole paper.Hence, it is important for researchers to make sure the abstract is an accessible and honest snapshot of the whole research paper.A good abstract is like a traffic sign on the edge of a busy highway; easy to see even in the chaos of a rush hour, easy to understand, and accurate ( Fig. 1 ).

Anatomy of an effective abstract
In addition to being clear and brief (preferably less than 250 words), a good abstract has four critical components; (i) a brief background that presence the problem the research paper is addressing, the state of science, and the purpose of study, (ii) a brief experimental procedure that shows the exactly reader what the authors did to address the problem, (iii) the principal results that shows what the authors found in the study, and (iv) the significance of the study that shows the global implications of the results the authors obtained.
Writing a good abstract takes practice as can be noticed by the differences in the abstracts I wrote in 2014 and 2019 ( Fig. 2 ).Although Abstract A (see Sanganyado et al. [3] . was succinct at only 210 words, it had grammatical errors and some critical components were missing.However, Abstract B (see Zhuang et al. [5] .showed some improvement, the only problem was a single use of a superlative in describing the results.It is important to note that by the time I wrote the second abstract, I had prepared or contributed to the preparation of 15 abstracts.

Writing an honest abstract for your manuscript
A good scientific abstract should be objective and accurately present the results and their meaning or significance.The scientific endeavor is currently fixated on novelty and high impact, and this has resulted in an unhealthy obsession with the fantastic.A previous study found 107 out of 128 studies investigated contained at least one example of spin in the abstract [1] .The authors identified spin as 'misleading reporting, misleading interpretation and inadequate extrapolation of the results' through use of things like causal language, linguistic spin, insufficient statements on societal/practical implications, and absence of negative results ( Table 1 ) [1] .
Table 1 provides some recommendations for authors to ensure their abstracts are honest and true representative of the objectives, experimental procedure and results of the study.Briefly, authors should ensure that the principal results are accurate by avoiding misuse of statistical significance, overlooking negative results, overgeneralization, and unsubstantiated claims of causation.Authors should avoid using words and expressions such novel, for the first time, significantly higher/lower, and superlatives.A scientific study should arouse interest based on the merits of the research quality and not empty promises and falsehoods.

Search engine optimization
An effective abstract should be discoverable through search engines.After all, people use search engines (Google, Google Scholar, Bing) and scientific databases (SCOPUS, Web of Science) to search for research papers.These search engines use algorithms to rank the results for the search input.A result that is relevant is often ranked higher.My review on chiral pharmaceuticals (see Sanganyado et al. [2] . is ranked first on Google and second on Bing probably due to my organic use of the phrase chiral pharmaceuticals ( Fig. 3 ).I used the phrase five times and its variant twice and offered its brief definition.Relevancy is often determined by the presence and organic use of keywords.Therefore, researchers should identify the keywords of their research (normally the keyword is the research topic) and use them in the abstract naturally.The authors should be aware of the limitations of their instruments, research procedure, sample size, etc. when explaining the meaning of their results.

Inadequate interpretation
Claim an effect when there is no statistical significance observed The effect of some factors is overstated even though no statistical significance was observed in the study.
Authors should avoid writing that a statistically significant results were found but report the p-value or effect size instead.Claim a cause and effect link between a result and an event.
Principal results are stated as if the authors found a direct link between an event, procedure, factor with the outcomes.
Authors should only make a causal claim when they conducted a mechanistic study to validate the cause and effect link.Claim there was significant difference without conducting corresponding statistical test A set of treatments are compared by mere observation of trends on a graph or difference in values of the results obtained.
Authors should only claim differences between treatments following an appropriate statistical test.Furthermore, authors should avoid using superlatives such as treatment A was higher than treatment B when no statistical tests were conducted.Overlooking the relevance of a result and focusing on statistical significance Authors emphasize statistical significance rather than the effect size for a bigger picture.For example, a wastewater was treated to increase pH from 5.0 to 7.0 using an adsorbent or biological treatment.If a difference of 0.2 was observed with a p value of 0.003, it is disingenuous to report the p-value.
Authors should focus more on the relevance of the effect size and not the statistical significance.

Overgeneralization
Results are generalized to different populations, scenarios, applications or outcomes beyond those directly investigated in the study.
Authors should extend their findings to circumstances of which they are meant to be applied only.

Inadequate implication for practical application
Authors claim the developed techniques, product or procedure can be used for practical purposes with no experimental validation.
Authors should ensure if they claim any practical application, their experiments would have validated it.

Human reader optimization
A good abstract should be readable.They are several writing bad habits that often render abstracts unreadable, examples include; (1) Long sentences -authors should write short sentences and cut any clutter.
(2) Too many abbreviations -abbreviations should be fully described and limited to at most three.Too many abbreviations confuse the reader.If possible, the authors should avoid use of abbreviations in the abstract.(3) Undefined technical terms -Scientific African is a multidisciplinary journal.Therefore, authors should minimize jargon in the abstract and use it when clearly and succinctly defined.(4) Poor connectivity/structure/organization -the background, objective, experimental approach, and significance of study should be aligned.

Conclusion
Since an abstract is a window to a study, authors should ensure that the abstract is accurate, honest, readable, and discoverable.This paper has shown that a good abstract should quicken the interest of the reader without spinning for attention.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.A good abstract is like a road sign that clearly and accurately shows what lies ahead.

Fig. 3 .
Fig. 3. Organic use of the chiral pharmaceuticals in Sanganyado et al. [2] .and the resulting ranking on Google Search.

Table 1
[1]mon types of spin in abstracts and how to reduce them (adapted from Lazarus et al.[1]).