Feeds:
Posts
Comments

Archive for the ‘Scientific progress’ Category


Universities are always cash strapped, and over the last few decades there has been a drop  in the number of PhDs funded by the UK government. In essence stipends have been increased (absolutely necessary and credit to The Wellcome Trust for pushing this), but funding has not increased commensurately. So numbers have dropped.

The response of institutions has been to push for more overseas students and to develop, from internal monies various funding  schemes. However, generally this has only served to replace losses in numbers funded by government. Gone are the days when every PI expected a new PhD student every year. This has a detrimental effect on research culture, as it leads to centralisation, a reduction in diversity (in the broadest sense of the word) and fewer trained scientists. The UK relies extensively on scientists trained elsewhere for its R&D and we need to do more in relation to training. Granted science is a mobile and international profession, but without contributing proportionally to the global talent pool, our R&D may wither in the long run.

The problem is money: institutions have limited funds they can use for PhD studentships. Our European model, where PhD students are fully funded for a set number of years (usually 3-4, sometimes a few more) is in my view preferable to the US one, where students work to support themselves. This is for the simple reason that the latter model can lead to feudalism and abuse of power, which is well documented.

There is money available for institutions willing to take leadership on the Open Access Agenda.

An aggressive pursuit of an Open Access agenda, as has been done in Germany, The Netherlands and Sweden, and most recently UC, means cancelling subscriptions to the journals of the Big Four. This frees up a substantial budget, with no ill effect on research and scholarship. A portion of the funds would, of course, need to be used to hire librarians supporting document sourcing by UG and researchers and the balance to fund PhD students. I note that teaching UGs these skills is important, since most STEM workplaces (industry, where many STEM graduates and postgraduates work) do not have large libraries. One advantage of using internal funds is that one can select and so only take the very best candidates, rather than restricting enrolment to those with access to funding. While such proposals will likely be met with horror by a good many academic staff, once in place it appears that no one notices much change and continues to work productively. This is the experience of countries and institutions that have cancelled subscriptions to one or more of the Big Four.

Sadly, with my University running workshops on “How to get published in Nature”, it appears that we are 20 years behind the times and this simple and effective means to improve teaching and research (PhD students being an engine room for research) is unlikely to see the light of day here.

Advertisements

Read Full Post »


Following a request to comment on a future university-wide workload management scheme, I thought it useful to put my comments in a public place, as these schemes are proliferating.

Simple questions that should be answered before any effort is put into devising and implementing such a scheme:

1. The motivation question

What is it for and who benefits?

The answer to the first is ‘fairness’, and, therefore, the beneficiaries are the employees.  That is, such schemes will ensure that activity and productivity are quantified, so Bloggs, a well-known skiver, will be found out, and made to contribute to teaching and research. People buy into this, without critically examining the evidence: is Bloggs actually skiving? If Bloggs is not putting in a decent shift, perhaps there are good reasons for this, such burn out. In any event, do we need a detailed workload scheme to manage Bloggs? The answer is no, it is perfectly possible to manage Bloggs without such a scheme, as long as a Head of Department has some simple information: is Bloggs producing papers and is Bloggs teaching? If productivity on the first count is low, then Bloggs can pick up more teaching.

What the workload system does create is a hierarchy, a value system, which equates to a class system. So some colleagues are upper class, others middle class and yet others lower class. It not surprising that class ridden Anglo cultures that have never had a revolution in the modern (1789) sense love workload systems. This does nothing for fairness and poisons the workplace, reducing collegiality. The importance of a class system is evident from the tone and content of presentations by senior management. It also provides senior management with an ‘achievement’, so justifying what Piketty termed ‘super manager salaries’ that have spread from the private sector to the upper echelons of universities. As Piketty points out, the success of this new breed of managers is down to luck, not skill.

2. The dimensional analysis question

Is this measurable and what are the units?

Another simple question, which should be the first posed by a scientist, yet is rarely is. The answers have always been poor, unlikely to gain a pass mark.

Any workload management scheme needs quantitative measurements: teaching, administration and research activity are given values. These must have some sort of unit. One can argue that this is possible for the first two, in that qualitatively they can be measured in a unit of time. This can only be qualitative, particularly in the instance of small group teaching, since individual students are extremely variable in their needs, a simple consequence of humans being an outbred animal. For research, there are different units, and likely more than one. Reading could be given a time unit, writing not so. I challenge anyone to come up for a singular unit for supervision and training of research staff, since this is dependent on the individual and the project. Acquisition of resources is often simply measured in cash terms, nice and simple. But who gets the credit? PI? Co-I?. And how about Bloggs, who isn’t on the grant, but asked a key question that sparked the entire idea? Again we come back to the importance of collegiality. Damage that and you damage the entire teaching and research enterprise.

A workload scheme represents a metrification of employee activity. As such it is the cornerstone of lazy management, where rather than putting in the effort to make a professional judgement, an individual’s activities are boiled down to a score. This will fail to increase productivity or to make workloads fairer. I say that with confidence, because our own School/Institute scheme has failed on both counts. Reading a colleague’s papers on the other hand provides a rather different view, but that is something that senior management could not contemplate, because it takes them out of the decision -making loop.

I would recommend a read of this article on the failure of metrification to achieve its aims.

Finally, can anyone provide evidence that a workload management scheme increased teaching quality and the research productivity of an institutions? Any such evidence has to have the effects of new hires stripped out.

Read Full Post »


A discussion today with a student asking about the use of the Royal “we” in a report about his work. I agree, this is wrong. My suggestions were the first person singular and the passive. The passive gets a bad press in places, but it does work; the repetition of “we” or “I” grates, the latter particularly so because it can convey a strong sense of ego. Though as I pointed out, this depends how it is used. It was common in single author papers for the author to use “I”. The practice has disappeared due to multiauthorship and the urge to make scientific observations look objective. We finished by joking about the feudalism implicit in the use of ‘my laboratory’, as if this was some sort of sentient being, and then I wondered out loud whether one might not, in a multiauthor paper state:

“In experiment X (Fig. X), blogs demonstrated that….” And then later “In experiment Y (fig. Y) Doe indicated….”

Tonight a tweet from @UtopianCynic

UtopianCynic tweet

reminded me of my earlier conversation. Indeed, why bother with all the rubbish associated with authorship position? Why not have a list of authors and in the paper report who did what and who thought what?

It would then be clear (i) who pulled together the original hypothesis; (ii) who did the experiments; (iii) who thought up the interpretations of the data.

I think I might try this out.

This also solves the long-standing problem of blaming whoever is at the bottom of the pile when a paper is found to have manipulated data. Someone will be explicitly on watch and someone else will have done a particular measurement under that person’s watch.

It will be obvious who should walk the plank, and reaching for lawyers will only result in keel hauling, because it will be all written down and signed off.

 

Read Full Post »


A recent article on bioarchiv “Amending published articles: time to rethink retractions and corrections?” puts forwards ideas on how we might change the way we deal with retractions and corrections. (more…)

Read Full Post »


This is a question raised at the end of the excellent article by @Amy_Harmon regarding Open Access and preprints is can biomedical scientists evaluate each other without journals?

The short answer is a resounding yes.  Physical scientists and mathematicians have been posting much of their research as preprints on arXiv for a few decades, with no prejudice to their ability to evaluate the quality of work or of individuals.

The counter argument raised by many in biomedical sciences, from scientists to some journal editors can be boiled down quite simply: We are special and cannot possibly do this.

Various arguments are put forward, from competition (=fear of scooping) to intellectual property. These arguments are heard in many biomedical/biology departments, sometimes leading to quite heated discussions. It is also interesting to note that the defenders of the status quo are not necessarily the older members of the community.

There is a simple answer. Yes you are special, but not in the good sense of the word. (more…)

Read Full Post »


This post assembles various comments I have posted and other thoughts on sci-hub and access to the scientific literature. It finishes with some ideas about what we should consider keeping and some of my better experiences, as a consumer and producer of the scientific literature.

Some time between clay tablet and the PDF

Once upon a time manuscripts were hand written, double spaced (fountain pen as ever outperforming all other tools), graphs transferred to tracing paper using a rotoring pen and Letraset (also alive and well) used for symbols. (more…)

Read Full Post »


During a quick scan this morning of the “recent” comments on Pubpeer, an activity that I pursue regularly, as part of my reading, there seemed to be a lot more author responses.  So I counted.

70 articles featured with comments.

10 of these had an author response.

This is progress. I have no data, but my impression is that a year ago author comments were far rarer, maybe 1% or thereabouts. Now we are at 14%. Let’s hope this is not an anomaly, but a trend, and maybe in a few years papers without author responses will be in the minority.

Regardless of arguments about anonymity, etc., post publication peer review is growing, which is a sign of health in the scientific enterprise.

Read Full Post »

Older Posts »