How to summarize research

"You don't really understand something well enough if you cannot explain it cogently to your grandmother!" - Unknown

For a scientist or a researcher, the ability to summarize a work to a target audience effectively and succinctly is an important and useful skill. However there are different media (text, video, image) for conveying the research, each medium with different word limits (100 to 100 thousand words) and targeted towards a different audience (general public to experts). Here are the different media for conveying research that I discuss in this article:

• Abstract - max. 100 words - general audience
• Precis - max. 300 words - scientists but not from the field
• 1 slide - eqv. 1000 words - scientists from the field
• 3 minute video - eqv. 3000 words - general audience
• Poster - eqv. 10,000 words - scientists but not from the field
• Paper - max 30,000 words - scientists from the field
• Thesis - max. 100,000 words - general audience

I discuss the skill of summarizing research with an example project. I have summarized one of my PhD projects in all the above media listed.

Abstract - max. 100 words - general audience

This medium is typically used in conferences meant to engage public. The reason for this brevity of this medium is now readily understood as the attention span of general public is very narrow. When targeting a general audience we need to avoid all jargons as much as possible or we need to explain any jargons used. Sometimes we cannot avoid jargons as they convey certain aspects of the research which are used in common parlance. Given the brevity of the medium it is difficult to expand on all ideas and huge generalizations are very much required to stay within word limits yet convey the core message. Here is my research as an abstract in 100 words for general audience:

My work impacts half the world who develop age-related hearing loss with difficulty understanding speech in noise. To understand how a healthy brain solves this problem, I need to record from single cells in brain (neurons) suitable only in animals. Monkeys are best suited given our similar auditory brains. I use sounds without semantics & perform non-invasive brain scans (fMRI) to show that monkeys use similar brain regions as humans to separate overlapping sounds. I'm the first to develop an animal model of sound segregation. Now, I can record from neurons in monkeys and understand how human brain solves this problem!

Precis - max. 300 words - scientists but not from the field

This medium is typically used in conferences with wide appeal to huge section of scientists and researchers rather than a specific field. Since the audience is scientific it is OK to use common jargon with expansions which is helpful to meet the word limit. Generalisations of the result and sweeping statements are best avoided since the research community is interested to scientifically judge our work. However it is difficult given that the audience is still not experts in the topic and some assumptions about their knowledge is made to convey the results. Couple of references are commonly used. Here is my research as an Precis in 300 words for scientific but not from the field audience:

More than half the world’s population develop age-related hearing loss with difficulty understanding speech amidst background noise. However, we do not yet fully understand how the brain performs auditory figure-ground segregation (AFGS), a fundamental process of auditory scene analysis pertaining to transformation of an acoustic signal into object based representation. So we intend to record from neurons in a rhesus macaque model performing AFGS. Thus we need to establish that macaques utilize similar brain regions as humans to perform AFGS.

To compare macaque brain activations underlying AFGS with humans, we employ artificial sounds that are equally relevant to both species that simulate the challenges faced in real-world listening yet are devoid of semantic confounds. These synthetic stimuli have a ‘figure’ with simultaneous onset of numerous (defined coherence) tones repeating in time, which overlaps spectro-temporally with ‘ground’ made of randomly varying tones. Our behavioural experiment in two macaques showed increasing figure detection performance with increasing coherence as seen in humans. Thus macaques can segregate ‘figure’ from ‘ground’ in these stimuli.

We employed sparse temporal design to acquire fMRI in three visually fixating macaques presented with these artificial sounds. We used General Linear Model to reveal areas that systematically co-varied with coherence. We found that bilateral rostro-lateral belt and parabelt in macaque auditory cortex are employed in AFGS similar to non-primary auditory areas along bilateral superior temporal gyrus in humans. Thus our study [1] has developed rhesus macaque as an animal model of human AFGS and provided spatial priors for targeted neurophysiology.

Ref: [1] Schneider, Dheerendra, et al., Scientific Reports, 2018. PMID: 30560879

One slide - eqv. 1000 words - scientists from the field but non-experts

This medium tests our ability to convey our results visually in the form of a single slide or an image. A picture is equivalent to a summary in 1000 words. One image is so short but it can be targeted towards both scientific experts as well as non-experts. So my thousand word summary as one slide or image is shared below. Here is my research as 1 slide for scientific non-expert audience:

How does the brain solve the cocktail party problem? from PradeepD32

Three minute video - eqv. 3000 words - general audience

This medium is generally reserved to convey the results to an general non-scientific audience given the possibility of engaging them in 3D illustrations required to convey some kind of research. This medium typically tests our ability to present results in the form of a video with use of illustrations. A one minute video is equivalent to a summary in 1000 words. So my three thousand word summary as a 3 minute video with illustrations is shared below. The video editing was done using ShotCut software. Here is my research in a 3 minute video for general audience:

Poster - eqv. 10,000 words - scientists not from the field

A poster is equivalent to a summary in 10,000 words. The target audience could be researchers in one's broad field or subject. Sometimes the poster is also created towards a non-scientific or scientific but non-expert audience. A poster aimed towards scientific expert audience typically contains sections: abstract, introduction, aims, material and methods, results, discussion. A poster aimed towards non-expert audience should avoid this construct and present the research in the following sections: importance, questions, sections specific to research, conclusions, future directions. This revised construct is aimed to capture the imagination of the non-expert audience. Here is my research as a poster with 10 sections or 30 points for a scientific audience but who do not belong to my field:

A monkey model of auditory scene analysis from PradeepD32

Paper - max 30,000 words - scientists from the field but non-experts

A paper is generally targeted towards a specific audience: subject-matter experts. However this depends on the audience of the scientific journal in which it is published. Depending on the audience, use of jargon and assumptions about their knowledge can be made. However given that every idea or result is referenced, even a non-expert scientist if they wish to can read a paper written for a expert audience though it might take a lot of time and effort to understand the results when compared to a situation when the same paper is published in a journal meant for general scientific audience. The title of the good paper should convey the key message of the research if possible. There is a specific word length for each journal that differs across fields. Here is the link to my research in five thousand words as a published paper for scientific non-expert audience::

Reference: Schneider F*, Dheerendra P*, Balezeau F, Ortiz-Rios M, Kikuchi Y, Petkov CI, Thiele A, Griffiths TD. Auditory figure-ground analysis in rostral belt and parabelt of the macaque monkey. Scientific Reports 2018, 8, 17948. (* - equal authors)

Link: https://www.nature.com/articles/s41598-018-36903-1

Thesis - max. 100,000 words - general audience

A science-based PhD thesis is usually written in at most 100,000 words targeting a general audience assuming they have no knowledge on this topic. All requisite background to interpret the work is provided in the general introduction (chapter 1) and general material and methods (chapter 2) chapter of the thesis. The actual research will be presented in one of the results (chapter 3) chapter. Generalisations of the implication of the research are usually reserved for the general discussion (chapter 5) chapter of the thesis. But the target audience of a thesis is typically the examiners who are established researchers working on the same topic as the thesis. It is also written towards satisfying the requirements of a degree as set by the Institute / Graduate School / Faculty / University and aim to demonstrate to the examiners that the examinee is an expert in the topic of research. So my fifty thousand word PhD thesis is shared below. Here is my research work presented in the Chapters 1,2,3,5 of my PhD thesis targeted towards a general audience::

Title: A primate model of human auditory cortical analysis of auditory objects 

Link: https://www.researchgate.net/publication/332116741_A_primate_model_of_human_cortical_analysis_of_auditory_objects