What makes a good paper?

Again the Ergonomics and Human Factors conference has raised many great debates and questions.

I’ll try to put more thoughts down about other sessions soon, but wanted to quickly record our discussion this morning which was titled “Desert Island Papers: A retrospective view of classic E/HF papers”. We started off by considering four papers that the presenters (me, Pat Waterson from Loughborough, Steve Shorrock from Eurocontrol and Claire Williams from Human Applications) had noted had influenced them in their careers, and then as a group we identified some characteristics of ‘good papers’.

The papers we initially identified as influential were:

Bainbridge, L. “Ironies of automation.” Automatica 19.6 (1983): 775-779.

Wilson, John R. “Whose attitudes to which aspects of work? or Cool media for high participation.” Work & Stress 1.4 (1987): 385-395.

Wilson, John R. “Fundamentals of ergonomics in theory and practice.” Applied ergonomics 31.6 (2000): 557-567.

Dempsey, P.G. “Effectiveness of ergonomics interventions to prevent musculoskeletal disorders: Beware of what you ask”. International Journal of Industrial Ergonomics 37 (2007) 169–173

We then had a discussion around why these papers had influenced us so much, and what we could take as guidance in producing our own written work, whether for journal papers, blogs or external publications.

Points that made a paper ‘good’ included:

• Good writing – we didn’t quite manage to define what good writing was, but good and clear use of language and argument came through as themes
• Makes you think – you don’t necessarily have to agree with what is written in a good paper, but moving ‘beyond the obvious’ can be valuable
• Clear message – have a clear point to make, and probably not too many points. Probably the paper that had the best example of this was the ‘Ironies of Automation’ paper, which, in a nutshell, says, automation has some ironies, this is what they are, this is how they might be fixed. 
• Have confidence (but not arrogance) in the points you are making, and don’t feel the need to over-reference
• Have an opinion – we felt all four of the papers we discussed had a clear opinion, and made an argument around it – we didn’t have to agree with everything that was said, but that was OK
• Define your end point – papers can occasionally ramble and not actually lead to the final point(s) that is/are being made
• Think about where and how to use data to back up your point – we noted that none of the papers we chose were particularly data-heavy, but were also aware of the need to publish, and the expectation of data when publishing. 
• Set the research in a disciplinary context – the example of this was the Wilson 2007 paper which referred to other thinking around ‘cool media’, but also demonstrated how work in the discipline of E/HF referred to other areas
• Make papers accessible – both in their writing, but also practically using open access wherever possible
• Take advantage of impact/contribution statements – this can really help the reader, especially if they are not an E/HF specialist
• Make people do things/act – this could be specific design guidance, it could be a method that we then go and use
• Be part of a series of ‘stepping stones’ of evidence – understand how different bits of work fit together, and what the contribution of each is
• And finally, have a good title. Keep it clear and simple, and ensure the title states what is in the content of the paper – seems obvious but isn’t always the case!

This is by no means a definitive, or even an agreed list, but hopefully can provide some help or guidance to us as we reflect on our writing and our communications.