Size of research groups - is bigger better? Takeover bids and spontaneous combustion

Bigger isn’t better

From David Parker

Philip Ball’s article ‘Family values’ discussed some of the pitfalls of large research groups. Unfortunately, there seems to be a belief that to ‘succeed’ in many areas of chemistry, especially in synthesis, you must have a big group. This dogma needs to be challenged. 

A group’s total productivity may well increase with size, but this comes at the expense of efficiency, with respect to both human and financial resources. Rather than assessing a group’s total output, we should examine outputs and their quality against key inputs. For example, you might expect a group’s publication output to be one paper per person per year, but the reality of multi-authorship reduces this. We need a rigorous assessment of productivity and research quality. The 2008 research excellence framework exercise used the number of PhDs per person entered as a factor contributing to the research environment score – perhaps not a measure of quality at all?

There is a ‘critical mass’ for any group: a size limit within which knowledge, expertise and know-how can be passed on effectively. The key person in this information transfer process is the research leader – it is their duty to create a supportive working environment and to set out the key challenges, to paint the historical perspective and inform PhD students of prior work: what alternatives were considered and where the blind alleys are. It is their privilege to inspire their students, to define their intellectual space and to launch their publication record. But as a group’s size increases, the amount of time that each PhD student enjoys with the supervisor inevitably diminishes. 

A partial solution to this issue is to impose limits on the number of PhD students that an academic is allowed to supervise: three per year might be considered a reasonable maximum. And at any time, no more than 10 candidates should be registered with a given supervisor. 

Larger groups also select for a particular type of student and dissuade those who feel they will be given less attention in large groups, especially if there is a highly charged atmosphere with researchers seeking to ‘prove themselves’. Could it be that, for example, female graduate and PhD scientists are more discerning and less keen to be foot soldiers in a competitive, target-oriented research environment, thus contributing to the observed gender bias in academia? These aspects are worthy of further debate and scrutiny, along with acknowledging the innate differences between young males and females in self-confidence and ambition, and their ability and desire to change location at key times of personal development.

In choosing success factors, the following output data sets are better considered: the career paths of group members; the number and nature of publications per head per year; the percentage of PhD students completing within four years; the gender balance in the group; the quality of group PhD theses; the number of poster or oral prizes gained at conferences by group members; the number and nature of patents and licences granted or issued; the nature of competitive prizes awarded to the supervisor or group members; the frequency and type of invitations to lecture for the leader and senior group members, both at national and international conferences. In these days of innumerable audits, in might also be of interest to estimate the unit cost per article, in an effort to assess ‘fiscal efficiency’. 

D Parker CChem FRSC

Durham University, UK

Stripped of science

From Anthony Lipman

In the June editions of RSC News and Chemistry World, I was interested to read articles concerning the UK government’s funding commitment to supporting UK scientific research and Pfizer’s attempt to take over its rival AstraZeneca.

How can the UK government be said to be support UK science while being a cheerleader for an American company taking over a significant UK research and development company? It is well known that the best way of suppressing competition from a rival is to take it over, strip its assets and then close it down or move the operation overseas to cut costs.

Over the past 60 years or so, many UK chemical and pharmaceutical companies have been closed down, with the consequent loss of thousands of scientists.

The French government openly supported Sanofi Synthélabo’s hostile bid for its rival Aventis, a UK company, in 2004. In 2013, Sanofi in Dagenham Essex, formerly May & Baker, closed its operations. I worked there in the 1950s and I was so sorry to see its demise.

Unfortunately, successive UK governments have only paid lip service to supporting science, despite the actual cost benefits accruing from science and innovation. 

A Lipman MRSC

Chelmsford, UK

Gunpowder plot sunk

From Tom Smith

Mike Baldwin’s letter suggesting that ships carried the ingredients of gunpowder rather than gunpowder itself must surely be wrong.

Not only is the process of making gunpowder dangerous and time consuming: (‘Hang on chaps, we can’t engage the enemy yet, we need to make some gunpowder’), but the typical millstones used even at that time weighed 8–10 tons.

There is a wealth of literature on the on-land production of blackpowder at the time, and even then they recognised the inherent hazards that would have made on-board manufacture out of the question.

On the other hand, maybe the movement of two giant mill stones was the real cause of the Mary Rose capsizing!

T Smith

Via Facebook

Junk’s hidden gems?

From Clive Delmonte

It is always a pleasure to read Philip Ball’s contributions. ‘Is junk DNA all garbage’continues the series.

After several billion years of evolution on Earth, only the very brave would assert that ‘junk’ DNA has no useful function.

It seems to me that we might make progress if we were to attempt a preliminary register of the roles that need to be matched against hereditary functions. For example, it has long been known in cellular biology that alleles form pairs with very high selectivity in meiosis prior to separation. How this is done remains a mystery. I hypothesise that long lengths of ‘junk’ DNA in genomes are actually the identifiers that allow like alleles to recognise each other. A test of this suggestion would be to identify the ‘junk’ DNA base sequences to learn whether paired alleles have similar or very similar nucleotide sequences.

This might lay the foundation of much wider studies. For example, despite its title, On the origin of species is primarily about the origin of varieties and not the origin of species at all, which is quite a different matter. We still barely understand, at the molecular level, how new species form, or indeed, what constitutes a recessive or a dominant gene, chemically speaking.

C Delmonte FRSC

Norfolk, UK

Fire afoot

From Norman Groocock

David Jones’ article about spontaneous combustion reminds me of a couple of incidents that took place at my first place of employment. 

In 1956, I worked at the British United Shoe Machinery Company (BU) in the quality control laboratory of their accessories division. The company made toepuffs, which are the stiffeners at the front of the shoe that stop the leather from collapsing onto the toes. One of the materials used was cloth impregnated with nitrocellulose. When one of the BU’s competitors was transporting some rolls of their product through the Suez Canal, one of the rolls spontaneously combusted, with predictable results. This caused a flurry of activity at the BU to see whether this could happen to any of its products. 

We set up an experimental rig outside the laboratory window where steam coils were pushed into the centre of small rolls of material, with a temperature probe towards the outside of the roll. We could soon predict when spontaneous combustion would occur and a small group would gather for the trials to watch the rolls ‘take off’. 

Another type of toepuff was made by impregnating cloth with wood rosin. The rosin was quite expensive and after the toepuffs had been stamped out of the roll, the rosin was extracted from the offcuts with benzene – there was a huge tank of benzene next to the production plant for this purpose. The laboratory and the recovery plant were sited next to the canal and over time the hollows in the canal bank were in-filled with cloth waste from the extraction process and covered over with soil to give an even path. Eventually, weeds grew on the path and some bright spark thought it a good idea to kill these by scattering sodium chlorate on them, again with predictable results and predictable panic. 

N Groocock MRSC

Bakewell, UK

Sample selection

From Sam Logan

Your anonymous correspondent wrote as if the only necessary characteristic of a valid sample was its size. It is equally important that it should be a random sample. 

Decades ago, Alfred Kinsey filled tens of thousands of forms with his notes on the sexual practices of humans. His selection of subjects was based solely on those who volunteered, so his samples, while they were undoubtedly huge, were far from random. Thus, his findings are largely disregarded, even if they sold very well at the time!

S R Logan CChem FRSC

Portstewart, UK