CO3353 Software engineering project management Additional material coursework assignment 1 2013-14

Page 1 of 4 International Programmes Computing and related subjects

Publisher interview (the client) Can you tell me what SciComms is and what it does?

Our business is publishing academic papers in highly specialist areas such as animal husbandry, particularly about sheep, and techniques for improving their well-being. We are looking at moving from a print-based business to an open access business model based on charging the author or their sponsors like publishers such as Springer are doing, rather than the reader, as, for example ACM does. We rely on the reviewers we use to establish and maintain our integrity and a large part of what we are asking you to do for us is about making the ‘user experience’ of the reviewer as enticing as possible using the new system.

Who should we talk to about the way the current manual system works?

Well, our editor, Dolina, of course, and perhaps one of our regular reviewers such as Dr. Collier. Dolina can give you names of some of our authors.

And can you tell me what you would like to see in the way of new or improved functionality?

The key to this is bringing all the different threads of information together into one simple system. We are calling the project MANTRA (MANuscript TRAcking) database. Status reports for each manuscript, each contributing author and each reviewer are vital. We also want to see all document-handling done via a secure online interface to MANTRA. Each type of user has to have a different set of permissions, apparently. Since we are monitoring authors and reviewers, the editor would like to be able to access their up-to-date contact information and keep a record of notes of meetings, times when they have let us down and so on. Some of this is highly sensitive information which we would not want the people we work with to see. We rate our reviewers on a scale of 1-7, for example. But you know academics, everything is scattered around the building in the most non-obvious places. It has to be better-integrated. And she would like to have a calendar showing people’s availability.

What would be the key factors that help you make a strong business case for the project?

What I said at the beginning is important, everything in one system, but the real benefit of that is the improvement of our document-handling processes. Did you realise that authors have to send us five copies of their manuscript, not one? And that it can take six weeks just to get the reviewer to agree to work with us on a paper? And only one in three of those papers get published?

The other thing that will be worth investing in is the ability to select and handle the reviewers more efficiently – matching them to a paper’s main topic using keywords, for example. I know I said earlier that everything goes through MANTRA but we need your advice as to whether that should be a separate system.

Editor interview Can you start by telling me how you work at the moment? You may find it easiest to think in terms of the manuscripts you receive and what you do with them.

Well, apart from dealing with phone calls and email about our forward-planning, such as special editions, and queries about formatting, the process really starts when I receive a manuscript from one of our contributing authors (a paper may have several authors but I only want to have to deal with one). I then select suitable reviewers based on what the author sends in the abstract and keywords. I then send them each a copy of the manuscript (we ask authors for five copies, one for me, three reviewers and one to go into archive), receive feedback and, as long as it is not too bad, negotiate changes with the author so that we can get to a final version that goes into production. That is assuming of course that the paper is in scope for the journal. It is pretty obvious normally when something is not right for our readers, and I just send them a polite letter explaining that it has been rejected.

CO3353 Software engineering project management Additional material coursework assignment 1 2013-14

Page 2 of 4 International Programmes Computing and related subjects

And what format are these manuscripts in?

Paper. There is the body text of the paper, which at this stage is anonymous, no author names, for example. Then each diagram, figure and table is submitted separately as artwork and we ask for a cover sheet with title, abstract and keywords. I nearly forgot – because of the need for anonymity we refer to each paper by a submission number which I

assign as part of the cover sheet. The cover sheet goes initially to the potential reviewers to see who wants to read the full paper. The production people have been asking for electronic documents for years now, but that’s one of the things they want you to sort out.

What excatly, please?

Being able to handle PDF files with everything in the right place from the beginning.

You started talking about selecting your reviewers. How does that work?

I’ve been the journal editor for over seven years now and I know the reviewers very well. I can normally get the ones I need fairly quickly, though it can get tricky if they are taking an extended break, travelling or whatever. And people move without telling us and their contact info changes. I tend to use people’s personal gmail or whatever, which is like a special favour to me. I keep things like that on a spreadsheet as it is not something I think everyone here should have access to. The only problem is that it all takes so long. I send abstracts to three people, two decline, I ask two more. It takes weeks. I would try more than two on the basis that at least one of them is going to say “No”, but I can’t go back to someone who is basically doing you a favour and say we don’t need her any more.

It must be very stressful.

Especially when I’m looking for perhaps ninety reviewers before we can start work on the next edition. Not ninety people, but three readers for each of, say, thirty submissions. The day is too short.

But it must be very satisfying when everything is in place?

Yes, it is really.

And what are the possible outcomes for the papers you receive?

Of the ones that get reviewed or the ones we reject as out of scope?

All of them.

Apart from those that are out of scope, the other type that we can get rid of quickly are those that get reviewed but are just not worth trying to repair. They are just flawed. The majority of manuscripts need more work before they can be published. That breaks into two groups – the ones with authors who are prepared to do the extra work and those who will not. If they do not agree the revisions, they join the reject pile. One or two actually object and complain that the three reviewers are all wrong! Then, very occasionally, you get one that goes straight into production. Oh, I should have said – the reviewers submit their reports to me and I take the final decision. There is no consensus process.

Going into production involves what?

Asking the author to prepare and submit camera-ready artwork before the production team start screaming.

CO3353 Software engineering project management Additional material coursework assignment 1 2013-14

Page 3 of 4 International Programmes Computing and related subjects

Contributing author interview I have just been talking to Dolina about her submission process and would like to clarify a couple of things with you.

OK.

Could you talk me through that from your perspective, especially in terms of areas where we could make things better for you?

Well, it starts with the rule that we have to send them printed manuscripts, in quintruplicate if that’s the right word. I produce documents for publication in LaTEX although some people seem to be happy enough with MS-Word. So why they can’t accept those I do not know.

Go on

Well, I’m also a reviewer and that is an area where SciComms has a lot to do. Why I can’t receive files by email, for example.

What about communication with the editor? How well does that work?

Well, Dolina is often out of the office, conferences and things, and it would be good to be able to see her diary. She’s actually pretty good at getting comments back to authors. I think the hold-up is normally the reviewers. Of course, not every paper gets accepted straight away. Sometimes we withdraw a submission if the feeling is that the reviewers are not very supportive. I’ve only complained about reviewer comments once, and didn’t get very far. I have a feeling that I do know who it was but obviously, Dolina is not going to tell me who it was.

Is there anything else we should know about?

Nothing major, but going back to the use of paper, it does make it more difficult when it comes to getting the finished version into production. I get the feeling that we are both making the same corrections in parallel sometimes.

Reviewer interview Can you explain the review process in more detail for me? The editor sends you one of the five manuscript copies and you decide whether you are the best person to review the paper, is that right?

Not really. We get the abstract first and only get the full paper once we have agreed to do the review in Dolina’s timescale which is normally four to six weeks. I prefer to send her a written report with comments for the author, but I know some people who comment on the paper copy itself, in a green pen.

But you are right in saying that it is my choice as to whether I can take on the review. Sometimes I am too busy, sometimes the paper is not really in my area. In fact, sometimes I wonder why Dolina chose me. The other problem is that if the paper gets a lot of comments, it can take time for the author to respond. Last year I took on a paper in the Spring and by the end of the summer I’d reviewed two sets of changes and ran out of time before leaving to go to New Zealand for three months. Dolina ended up faxing the last draft to Auckland for me to collect and I had to phone her to say I approved it.

Apart from trips to the other side of the world, can you identify any other sources of delays?

There are some bottlenecks in the sense that things have to be choreographed and it all works at the pace of the slowest person. The author doesn’t get any feedback from Dolina until all three of us have reviewed the paper, which is fair enough. But then we have to wait till the author has responded to all of our comments which can be frustrating. Especially when the author eventually decides to withdraw the paper.

Finally, what improvements in the review process would you like to see? As you know, SciComms are looking to implement an online manuscript management system.

The first thing I would expect would to be able to work with an electronic document, such as a PDF. Possibly even the three of us being able to comment on the same document so we can see what each other have said and simply comment “agree”. The other obvious thing is to be able to see what information the system holds about me. If Dolina has a way of

CO3353 Software engineering project management Additional material coursework assignment 1 2013-14

Page 4 of 4 International Programmes Computing and related subjects

matching reviewers to papers, it needs updating. The other thing that might help, although I doubt if Dolina would agree to it, would be a shared online calendar so I can see when one of the other reviewers anticipates missing the deadline so that I can review my own priorities. It is very frustrating giving up a Saturday afternoon in order to meet the deadline and then find that the comments do not go back to the author for another two weeks.

Notes