How to write a research paper. By Mark Bush

When to Publish … and Where?

Mark Bush

When to Publish? … As soon as possible

but …

• Your thesis will benefit from peer review • Examination often smoother • Reviewers often stimulate new ideas • Your research/academic career will be based upon peer review. Best to develop the culture early.

When to Publish?

Only when there is a story, conclusion or take-home message

Avoid the tendency to try to publish a lab report (i.e. I took these measurements and this is what I saw)… … what does your data tell you, or what does your model predict?

When to Publish?

Publishable work may be, for example:

•Development of a tool or technique, such as a theoretical model or experimental method.

•Design and testing of a device

•Development of a new material, and associated tests/measurements to assess its performance

Publication in science should add new knowledge that will be of value to the community (new or refined tools, devices, materials or techniques).

Example – Modelling of Longitudinal Cracks in Teeth Amir Barani (PhD student)

Task: Can we effectively model the evolution of cracks in enamel?

Requirements/Steps: Finite Element Method with Crack Growth – XFEM Materials Science – Properties of tooth components? Simplified structure? How simple is good enough? Mesh density? Crack initiation and advance strategies?

Compares well with experimental results. Predicts the sudden acceleration of crack at a ‘critical load’. Demonstrates the effect of tooth radius (R) and Enamel thickness (d).

Results:

A considerable amount of work and innovation to get to this point…. Could this be published now?

The work so far amounts to development of a technique (eg a model or an experimental technique). It will be of value to the community. Many journals and particularly conferences will welcome reports of these developments – it is certainly publishable

However, the work will have more impact if you put it into context – provide the big picture:

• How can the model be used? What does it tell us about the physics that we cannot easily see experimentally? What does it predict?

Use the model to study the system of interest and learn something new. or

• What do the experimental measurements tell you? A new phenomenon or relationship? What are its implications? How might this phenomenon be used/applied?

Back to the tooth fracture modelling example:

We can now calculate the load to produce a longitudinal crack in a tooth. So what? What anthropologists need is a means to predict the bite load an extinct animal was capable of producing (implications for diet and lifestyle) What we have is a technique that can calculate the load a tooth of given size and enamel thickness has undergone if a longitudinal crack can be seen in its fossilised remains. But it requires a full finite element analysis, skill and considerable computing resources. How do we bring the two together?

Fracture Mechanics: Load = C.T.R.d

1/2

Constant

Material Property

Tooth radius Enamel thickness

If we know C we have a simple means of estimating the load that a cracked tooth has experienced, which is a measure of the bite load capability of the animal. We cannot determine C from experiment – dealing with extinct species (no teeth to test), but we can use our FEM model to determine the relationship between Load, R and d, and calibrate the equation by determining C, giving a far more general result than would be possible experimentally.

Back to the tooth fracture modelling example:

Comparison with predictions from jaw mechanics

A. Barani et al, Mechanics of Longitudinal Cracks in Tooth Enamel, Acta Biomaterialia, 7(2011)2285–2292

Summary: Your personal situation will vary – fundamental, applied, theoretical, experimental, etc, but the same principles apply: • Attempt to publish only when you have something new or innovative to report – don’t publish to a schedule, let your publications be outcomes driven. • Find the key ‘take-home message’ resulting from your work. Make this your conclusion.

• Try to put your work into a broader context and think about its applications - the audience will be broader and the impact higher.

• Aim for high impact journals – reviewing will be more rigorous, your citations over time generally higher.

Where to Publish?

Journal • Generally most rigorous form of peer review • Don’t be shy, aim for high impact factor – impact of your work will be a factor in your later success (high IF implies high citation rate) - job selection panels and promotion committees pay attention to citations, reputation depends on impact and citations – H (Hirsch) Index measures productivity and impact. - you can always modify and submit to another journal if unsuccessful - support new and emerging journals only when it makes sense, and when your career can afford it. - IFs vary according to field, because citation rates vary with field – check IF lists for your field. • Select a journal appropriate to your paper -seek advice from supervisors, colleagues and journal websites.

Where to Publish?

Conference • Conference experience important – verbal presentation of your work, direct peer review, new ideas, chance to meet or hear the best in the field, networking. Try to get to at least one conference (hopefully international) during your candidature. • Quite variable in standard of review – can be almost none – you should check. • But, lead time is usually much less than journal. Often useful for presenting incomplete results and conclusions. • The fact is, even if so called ‘full peer review’, the number of papers for review at one time can lead to less stringent assessment and feedback. Conference papers are often not regarded as highly as journal papers for this reason. • Generally far more limited circulation than journals – tends to attract fewer citations. • Conference participation may vary from abstract to full paper, poster to on-stage presentation.