Turning Research into Outputs: Thesis, Papers and Beyond

This article is part of a PhD introduction series for new students in the Department of Materials, as organised by Professor David Dye (Director of Post Graduate Studies).

Introduction & ‘outputs’

Let’s consider what is involved in research and how we share this with the world, or more particularly “how can you successfully demonstrate that you have contributed some ‘output’ from your work?”.

Firstly, as scientists what do we consider as outputs?

The list is vast and may include:

· PhD Thesis — typically a ~40,000 word (must be <100,000 @ IC) document that will be examined at the end of your programme.

· Papers — a series of disseminated documents that convey and test new ideas to the wider scientific community. These have the rigour of ‘peer review’ and are the ‘gold standard’ of research.

· Data — you are going to test some ideas, and in doing so you are going to be generating data. Recording this successfully and conducting analysis is the ‘day job’ of a scientist. How can you manage data capture, storage and archiving effectively?

· Books/book chapters — a digest of good ideas that can be more in-depth than one off papers and can be targeted towards different audiences.

· Talks/seminars — an ‘open’ forum of communication that you can use to air new ideas and discuss them amongst your peers/members of the scientific community/wider society.

· Outreach/public engagement/advocacy — conveying why your science is important, how it helps society and inspiring the next generation is very important for the health of the scientific community.

· Patents/widgets/know-how — these are the commercial/technological outputs of your work that aim towards a competitive advantage for you, your sponsors, and UK plc.

· Methodology/techniques/code — underpinning new scientific discoveries, especially in Materials Science and Engineering are a series of sophisticated techniques and approaches. Documenting these for the next generation of scientists is important to enable new advances.

· You — A PhD is a training programme and no matter your destination (academia, industry, finance, public sector) and so your training in ‘the scientific’ method and how we try to think is important. The faculty celebrate development of our graduate cohort and enjoy following them as they continue on their career paths.

This is likely a sub-list of the many outputs on offer, but hopefully it gives you a flavour of potential. There are many more to consider, especially as you will all have different objectives, focus and paths through the PhD programme. If there are others that you come across and feel should be included, please let me know and I’ll update the list!

To draw a focus back towards the objectives for discussion here, we are going to examine some of the ‘core business’ of the PhD. This will target how we can be demonstrate effective ‘day-to-day’ science and getting your PhD. This is the process of discussing our experiments, writing up our work and completing the PhD thesis.

What is a PhD thesis?

As a PhD student one of your outputs, required for successful completion of a PhD programme, is the generation of a coherent and interesting PhD thesis that will be examined (and hopefully passed) by your examination panel. To do this it is worth having a rough idea the shape of a thesis. A thesis should describe your personal contribution to science from your studies and represent three-four years of hard work.

To paraphrase the Imperial College PhD regulations:

The thesis shall:

(a) consist of a candidate’s own account of their investigations

(b) form a distinct contribution to the knowledge of a subject

(c) afford evidence of originality, by the discovery of new factors and/or the exercise of independent critical power

(d) be an integrated whole and present a coherent argument

(e) give critical assessment of the relevant literature; describe the method of research and findings; include discussion of findings — together with an assessment of how this advances the field; demonstrate a deep and synoptic understanding of the field of study; and demonstrate objectivity and capacity for judgement in complex situations

(f) (typically) be written in English

(g) demonstrate research skills

(h) be of a standard to merit publication in whole or in part, or in a revised form (e.g. as a monograph or as a number of articles in journals)

In practice, your thesis tells a scientific ‘story’ of your contribution to science. It is not historical and you may find that the experiments you perform in the lead up to submission of your thesis are the best to clarify your general hypothesis and focus the scope of the work and therefore merit being introduced rather early on in the thesis.

You might want to consider the following structure (but note this varies from group to group, so take your supervisors advice on board too!):

(1) Introduction + Literature Review

(2) Materials and Methods

(3) Results 1

(4) Results 2

(5) Results 3

(6) Discussion

(7) Summary and conclusions

The number of results chapters will reflect your project scope and nature of the work, so take this as indicative. Fewer results chapters will require more ‘meat’ within the project and a greater depth of interrogation. It is unlikely that you are going to require more results chapters, as PhD study requires significant depth of study and a joined up/coherent ‘story’. Some mixed results/discussion/methodology chapters may arise, especially if you are developing some new approaches. I have been quite specific here to include a separate discussion chapter as a PhD thesis, at least in the UK, is supposed to provide a coherent ‘story’ and one skill that you need to acquire is the ability to distil your results and put these into context with respect to your introduction. This is a difficult skill but is very important to develop if you wish to continue within science.

Writing your thesis will take time. The shortest time is likely to be ~2 months, and the longest time can be one year. The trick is to keep on writing. Redrafting, editing and modification can happen once things are written down on paper and keeping a good record of progress, ideas and current research as you progress through the programme. Not everything you write will end up in your thesis, but the process of writing something every day is important both to ensure that you submit on time, and that you feel that progress is being made.

Setting a research agenda

Now that we have discussed one significant output — the thesis — you could now consider, together with your supervisor, what is reasonable for your research agenda within the PhD?

A clear objective, or theme, to your research is essential. You need a coherent ‘story’ underpinning your thesis. You have to be excited about this project (at least at the start, and on average throughout your studies) as this is going to be your day job for the next three years!

Your agenda must be tractable within your PhD project (i.e. 2–4 years full time research in the UK system) and needs to be appropriately resourced (equipment, time, & collaborations) within your research group. The PhD thesis must consist of ‘novel’ research, and can involve collaboration with a wide variety of partners and employing state-of-the art facilities. If you need to plan hands-on experiments then you need to ensure that you have scoped the research within your means, especially if this requires access to expensive/difficult to access kit such as multi-user synchrotron facilities, top end electron microscopes, or very well used laboratory equipment.

To tell the story of your research, you need to take ownership of the project early on and be the individual driving progress in your particular vein of enquiry. This is not your supervisor’s job and as a point of pride it should never be! As part of this, you will work together with your supervisor, who should be your greatest ally and supporter in this endeavour, as well as your research group and wider collaborative network to deliver new insight and understanding.

During your scientific journey, the contributions of others in your work is likely to be essential in the current climate of ‘big science’ and working with others is a very important skill to develop early on, as regardless of your choice of career this is a significant transferable skill developed further during your PhD studies. This means that aligning your research agenda with the goals of your group, especially if resource intensive, is very important. If you are a successful PhD student and continue on a path in academia independence is a path earnt through many (more) years of hard work and dedication.

To keep on track and to understand how ownership of a project evolves, within my research group year 1 = delivery of initial idea of mine by the PhD candidate; 2–3 = development of independence and a good candidate takes the lead, often asking me for advice/support rather than direct day-to-day guidance. The goal of a PhD is to provide a thesis, understand some new science, and to develop yourself towards the next step in your career, which will often be far more independent than the last!

While the ‘grand vision’ of your research will be scoped towards a grand achievement and unlock new understanding, you need to break this down into individual tasks or mini-projects towards this aim. In many cases these projects will still be very ambitious and they will open new and exciting lines of research along the way and the grand vision is open to changing. However, the process of breaking a larger task into smaller reasonable projects is essential in ‘chipping away’ at the problem and leaving you with the satisfaction of completing mile-stones along the way. These mile-stones are often aligned with the output of key research outputs, such as papers, talks or reports, along the way and are likely to result in significant contributions, or individual, thesis chapters. Keeping track of your progress through the project is essential and this is often managed with a Gantt chart or similar.

Scoping projects towards individual papers will require discussion with your supervisor and research group to ascertain what is important and reasonable to report on to the wider scientific community. In particular you will need to gauge whether it is reasonable to aim for one high impact publication, which is very risky but exciting, or to tackle several more self-contained avenues of research that lead to a range of papers exploring the grand theme of your thesis. If you are looking to stay in academia, noting that only ~3.5% of PhD students will land a permanent academic research job, and 0.5% going on to professor [1], then having a strong academic track record in the form of completed papers is important. Having these peer reviewed and out in the literature puts your name on the map and provides you a significant ‘leg up’ on your competitors in the next job search.

As your thesis is a novel piece of research, it is reasonable to expect at least one peer-reviewed paper in a high quality journal (in addition to the potential of conference proceeding papers) as a result of your work, and for some groups this may be in collaboration with others members of the group. Many PhDs will contribute towards 2–3 from their studies, and a rare few will have several more.

A research paper

When you’ve conducted some exciting research, what next? How can you turn this into a peer-reviewed paper and share this with the scientific community. There is a strong case to be made that all PhD students should contribute to the publication of new research papers, as your work is necessarily novel, a paper is a clear scientific output, and that if the research is not shared with the wider community it has not effectively been conducted.

You will read several hundred papers throughout your PhD. Find a paper in your field and examine its contents. In particular pay attention to the following:

(1) What is the core ‘message’ of the paper? What idea of the authors explored, tested and argued?

(2) How have they presented their findings? What figures, tables and data have they presented to support their claims?

(3) How do they argue that these claims are important, in light of existing knowledge and their methodology?

(4) How have they ensured that their paper is scientifically rigorous, in terms of describing their methodology and assumptions?

(5) How have they bounded their results with a reasonable case about the scope of their findings, the error on their results and the limitations of their methodology?

In answering these five key points you can now consider your own findings and decide on how best to write up your work.

Have an idea of where you would like to publish and read their author instructions. These vary from journal to journal and are worth checking in advance. Next install a reference manager (such as Endnote, Mendeley, Zotero, BitTex) and families it with your favourite writing package (typically Word or Latex) and outline section headings common for your journal of choice.

Results, Figures and the Main Story

The first point of writing up work is to describe your findings accurately, clearly and unambiguously. This will require you to generate figures that tell your ‘story’ effectively, avoid ambiguity and make it easy for the reader to take home the key message. There is an art to effective figure generation and there is significant literature and books available on this subject (see further reading for some suggestions).

In generating figures and reporting on your findings, it is essential that you keep your reader in mind and remember that they do not know your experiment as well as you do. Therefore keeping things simple, clear and reproducible within the figures is essential. Once you have good figures, test them out on your close colleagues, typically within group meetings or in discussion with your supervisor and pay attention to their comments or queries and ask yourself whether you can improve the presentation of your data (or whether you need more data to convince them!).

Once the figures and figure captions are complete, the results section is essentially complete. Add text to the results to provide a narrative to the story of your hypothesis and outline the highlights of your findings. This is not the place to hypothesise about what the results will mean, which can be directed towards the discussion, but instead a chance to sign post other researchers of salient points.

Methodology

Now that the results section is largely complete (at least for now!) you should continue with the methodology section. This section will enable other researchers to repeat your findings, and hopefully come up with the same figures that you have generated. There is further scope for them to build upon/develop and employ your methods in further studies. This is typically a rather dry section of any paper but it is essential to ensure that your paper is honest to the science that you have undertaken. There is a balance between describing your methods in minute detail as compared to providing enough information for a reasonably skilled researcher to replicate your experiments and results.

Introduction, literature review and discussion

Now that you have the methods and results written up, I would focus on the literature review section and discussion.

The literature review/introduction should lead the reader through the key scientific arguments that lead to the open question that is your hypothesis and motivate the reader into understanding why you have undertaken this study. It should fairly represent key findings of other workers within your field, and join together their work towards the general narrative that outlines the scope of your work. It is important that you write the introduction with two readers in mind: (1) your scientific peers who have a detailed understanding of the nuances of all the literature within the field — these are likely to be the easiest to please; (2) individuals who are less specific readers and may be interested in new research within a theme and therefore may need to be guided on the importance (societal and scientific) of your work, as well as being less familiar with jargon and ‘common’ approaches within your specific discipline.

Writing the discussion section is an opportunity for you to highlight how your results answer your hypothesis, as well as how they build upon the previous research you highlighted within the introduction. In the discussion you are free to speculate, within reason, the scope and importance of your findings and to finesse arguments that direct future studies. There is a balance here, you want to outline potential directions in the field and illustrate that you understand the relative importance of your results, while avoiding overt speculation.

Finally, you can wrap up by writing the abstract and conclusions section. These are factual reports of what is in your paper, its importance within the literature and your major findings. These are surprisingly difficult to write and best left until last. Your supervisor will be able to advise you on how to write these (and you are likely to have some experience from writing abstracts for conferences and similar as you progress).

The paper is now largely complete. Don’t forget to include acknowledgements, include appropriate contributors and adhere to journal guidelines (e.g. highlights, keywords, contribution states etc. as appropriate).

The Author List

The author list is important for science, as it highlights who has contributed to the work (and is important for your future career). If you are included as an author for the paper you have a responsibility for the scientific credibility of the work, as well as asserting that you made an active contribution to the study [2–3]. Papers are added to scientific bibliographic databases (web of science, google scholar, SCOPUS, etc.) and attributed to authors within this list. The presence of your name and position in the list is important for future career prospects as well as highlighting how much of a contribution each author has made. Please note that the author order is often a complex juggle between personalities and is best discussed early on in writing, often in private with your supervisor to start.

For a materials science publication, the author order follows a general trend. If you are writing a paper on your work, you have likely to have contributed the most to the study and therefore are probably going to be first author. Your supervisor has provided guidance, resource, ideas, and advice throughout the research and their contribution merits ‘last author’ position. In general the first few authors are those who have contributed the most ‘hands-on’ to the work, including lab-work, preparing reagents, microscopy, code, and analysis. The remaining ‘middle’ authors are those who provided essential advice, discussion, or specific contributions (e.g. support at a beam-line). Those towards the back of the authorlist have acted more in a supervisory role for parts or the whole of the project.

In general it is better to be generous with authorship, as you will want to build good will and acknowledge appropriate contributions. This order of authors can vary by field, and some fields will prefer alphabetic author orders and an appropriate statement of contribution within the manuscript. Have a look at the papers you are reading to get a feel for the process.

The ‘corresponding author’ (CA) position is subject to some discussion and is used differently amongst groups. Often it is merely a signpost for who to contact by any interested party in the next ~5 years. Some groups will therefore opt to have the lab head as the CA. In other cases, a post doc will be rewarded with CA status to reflect their positive academic lead in the project. This is an item worth discussing with your supervisor before the paper is more widely shared.

Data

New guidelines for government funding of research are increasingly highlighting the importance of good stewardship of data and lab books. In particular there is now a general expectation of a ‘data statement’ that should be included within your paper. This will often be a signpost of where to download data or code associated with the work. The minimum contribution should contain raw data included in tables or graphs, so that these can be extracted avoiding tools like data-thief (http://www.datathief.org/) in extracting data for subsequent analysis. A more generous contribution could include all raw data captured from the experimental programme.

The most important point concerning data management is that data are useless without appropriate meta-data that describes how the data was collected and what each column contains. This drives researchers towards improved data management practices and plans which should be discussed with your supervisor early on in your research career. Keeping of an electronic lab book is a good idea and capturing snapshots of analysis code throughout your studies is extremely helpful.

When you are in the process of submitting your paper, you may find the repository service offered by CERN (http://zenodo.org/) useful for uploading data, attaching a digital object identifier (DOI) and providing access for future researchers.

Finally, your data is important. (1) Establish a good back-up strategy early on; (2) Establish a good storing strategy, with the idea that someone else could follow-up your project should you win the lottery.

Thesis vs papers

The audience and context of your writing is important to consider. This is especially important when considering the difference between a thesis and peer reviewed papers. In summary, the thesis is a document that will (hopefully) result in a PhD and is there to convince your examiners that you are worthy of your degree, and there are specific requirements as outlined above. An academic paper is a more direct contribution to the scientific literature and is subject to scrutiny through external peer review and comments in future publications

The thesis therefore may include a longer literature review, especially to introduce ideas and topics for discussion, as well as ensuring the examiners that you are worthy of a degree in materials science and engineering. It can include more details, such as experiments that did not go quite according to plan, and include more speculation as this reflects your viewpoint on your chosen topic as an individual investigator. It is a sole-authored document, with clear input and advice from your research group and in particular your supervisor.

A paper is a curt, timely, stand alone, and important contribution to the scientific literature. It will (likely) be widely read and considered over many years. It must pass through peer review. It must fit the aims and scope of the journal in which you wish to publish. It must reflect the combined inputs of the authorship. It will be subject to significant redrafting, red-pen and discussions on focus and style. It may be modified as it bounces from one journal to another (if you get through peer review the first time with little comment it is likely that you were not aiming high enough!). It will be reviewed and changed in response to particular comments from a reviewer.

A sloppy thesis may try to consist of several papers ‘strapped’ together to form a coherent document. This betrays the opportunity for you to demonstrate that you are a rounded individual who can grasp the grand picture of the project and can tell a coherent and interesting story that reflects the hard work and talent of your three years of study. That being said, having papers ‘in the bank’, written up and reviewed puts you in a very strong position for your PhD viva, as well as noting that all post doc job hunters have a PhD but not all of them have papers!

Summary and closing thoughts

Doing science requires hard work in the lab or at the computer, as well as writing it up for others to digest, interpret and understand. There is a moral and ethical obligation on a publicly funded scientist to communicate science back into the community and if you fail to do so you will run into trouble for your future career, especially if you stay within science but also to reflect the hard work and contributions of your collaborators. There are two major forms of output that will assist with this endeavour: (1) your PhD thesis; (2) peer reviewed papers. They have different goals, structures and readership. Both are exciting to construct and knowing that you have contributed to the scientific record is one of the greatest and simplest pleasures in being a scientist.

Dr Ben Britton, 04/10/2015, Department of Materials, Imperial College London

E: b.britton@imperial.ac.uk

T: @bmatb

W: http://expmicromech.com

References, resources & further reading

[1] The Scientific Century: securing our future prosperity report by the Royal Society

https://royalsociety.org/~/media/Royal_Society_Content/policy/publications/2010/4294970126.pdf

[2] http://hms.harvard.edu/about-hms/integrity-academic-medicine/hms-policy/faculty-policies-integrity-science/authorship-guidelines

[3] https://www.imperial.ac.uk/research-and-innovation/about-imperial-research/research-integrity/authorship/

Writing — Papers, Theses etc.

The Science of Scientific Writing by Gopen and Swan

https://www.americanscientist.org/issues/pub/the-science-of-scientific-writing/1

YouTube — In praise of technique: Judy Swan at TEDxCMU

https://www.youtube.com/watch?v=1pzjxYCwb08

Elsevier — Publishing Campus

https://www.publishingcampus.elsevier.com/

The Wellcome Trust — Science Writing Series

http://blog.wellcome.ac.uk/science-writing-tips/

American Journal Experts — 12 Tips for Accelerating Manuscript Publication

http://www.aje.com/en/author-resources/articles/12-tips-accelerating-manuscript-publication

‘Best’ Publication Strategies — Best Publication Strategy in Science

http://www.smartsciencecareer.com/best-publication-strategy-in-science/

Representing Data

Why Should Engineers and Scientists Be Worried About Color? by Rogowitz and Treinish

http://www.research.ibm.com/people/l/lloydt/color/color.HTM

Tips on effective use of tables and figures in research papers by Rodriguez

http://www.editage.com/insights/tips-on-effective-use-of-tables-and-figures-in-research-papers

Visual Display of Quantitative Information (pdf book) by Tufte

Data Management

Data Management Planning

http://datacentrum.3tu.nl/en/planning-research/data-management-plan/