Computer swallows research grant applications
By Robyn Marshall
New research grants in medicine for 1999 throughout Australia were announced on November 20. As usual, only 25% of grant applications to the National Health and Medical Research Council (NHMRC) were funded. Also as usual, a number of promising research careers in medicine and science were destroyed because some researchers, who are totally dependent on this money for their salaries, failed to get a grant.
The NHMRC is the federal governmentfunded body which dispenses most of the money for medical research. Grants are becoming more and more difficult to obtain; there has been an overall cut of 50% in research funds over the last 10 years.
This year the application process for NHMRC grants was a total shemozzle. Administrators in the Howard government decided to change the process, so that instead of all applicants sending 1215 copies of a paper application to Canberra, applications would be made through the internet.
A few simple tests were made. Every potential applicant had to log on at a certain time and date at the beginning of last year. Even though that failed and the connection crashed several times, they went ahead.
The result was disastrous.
The NHMRC GrantNet program was a fancy set-up, worthy of the most obsessive computer nerd. The first page had a pretentious logo of a rotating world with Australia in the centre. However, no-one had given a thought to the practical problems.
The grant had to be written up while connected to the net. This meant few people could get into the program because so many people around Australia were logged on at any one time for hours on end. People waited until 2am to log on because this was the only time available as the March 6 deadline approached. Even these work hours wouldn't bring about a successful logon.
Grant applications always include a section where a list of all publications of the major applicants have to be listed over the last six years. With three researchers who may have 150 publications each, the listing of these publications in a particular format on the GrantNet became a nightmare. The grant deadline had to be extended four weeks, till the end of March.
If you had a problem, you could ring a 1800 number. One researcher hung on the phone for five hours waiting to get through as the problems were becoming acute; at 5pm, he was told the office had closed. This useless facility cost the government $15,000 in phone bills over the three months in was in operation.
The mainframe couldn't cope with the volume of such large files: there were 150 applications from one university alone.
Just before the applications were opened at the beginning of 1998, the NHMRC office in Canberra was moved into a newer building, with much less space. A number of transfers and sackings of experienced staff took place because it was thought many fewer staff were necessary with this whizbang technology. This was another disastrous decision.
As the number of complaints piled up (hundreds of irate emails arrived every day), it was soon realised that Â鶹´«Ã½ of the grant applications could not be completed on the internet and would have to be e-mailed. There were so many mismatches of software programs that the GrantNet and the mainframe just couldn't cope.
This meant new staff had to be hired to put together emailed versions with half-completed internet applications. The new staff were totally inexperienced and untrained, unable to decipher whether applications on cardiology were in the same category as entomology. The new office workers had to do the work sitting on the floor and sifting through papers and grant applications because of the lack of space.
Research grants are supposed to be confidential because new ideas and experimental approaches are all important. Confidentiality was another casualty as research grant applications were sent for verification to the wrong applicants.
In the middle of this chaos, the head of the NHMRC administrative unit decided to take early retirement and bow out with a golden handshake. Another experienced public servant had a nervous breakdown and disappeared from the scene.
In May, research applicants were still being sent requests for versions of their grant applications.
Usually only 50% of applicants are interviewed, but because of the mess, all who submitted applications were given interviews to maintain some sort of impartiality. Interview sessions began in July, a bevy of senior scientists and clerical staff travelling around the country to interview applicants in every state capital. (It's not clear whether this personal contact is intended to exclude those who are not presentable, have a twitch or look obviously mad, or to determine if you're male or female.)
Several researchers arrived for their interviews to be confronted with the fact that even after all this time, the interviewing panel had not received the appropriate documentation of proposals. These interviews had to be rescheduled at a later date by forming a new interviewing panel and flying researchers around the country, all at the expense of the taxpayer.
In the end, Price Waterhouse was contracted to do a study of what went wrong, for a large fee of course. It found that GrantNet was not userfriendly (surprise!) and over-designed, and that no obligations had been written into the contract, so there is no possibility of suing the company for breach of contract or inefficiency or anything. The NHMRC committee had been sold a pup.
Secretaries of research administration bodies from all the universities were flown to Canberra for a one-day conference on the problems, again at considerable cost to the public.
Literally millions of dollars had been spent on this farce. No details reached the general public, no apology was given to those on whom this debacle was inflicted. And even less money gets spent where it is needed, on scientific research.
For the next grant applications, the whole GrantNet scheme had to be scrapped and a fresh start made.