As a researcher, I have been living in a bad-model situation for many years now, and the only real problem I see, is that it took all those many years to understand it. There was no counselling available when I was young (relatively) and full of hopes. Therefore, I have been passionately thinking about a way to let my personal experience be of some help to others and I realized that it is not that easy. Indeed, getting things right in science is so much based on a certain “luck factor” (the right lab, with the right project, in the right moment etc...) that no generalisation is allowed.
However, it might be useful to lay down few indications about how a “bad model”, also known as a bad project, looks like. First of all, a bad model is one which needs to be adjusted on a weekly basis for years before (if ever) producing results. This might be one of the first things to check when you are going to accept your first post-doctoral position, with eyes full of dreams. A very subtle hint may come from the fact that your project has been running for 12 years (before you), and by half a dozen people but it somehow remained “promising”. If this is the case, pack your bag, forget the whole thing, go to the next interview and do not look back.
A bad model might also consist in an experimental framework that, whatever the quantity of data produced, is never sufficient to get any collateral or small publication done, not even a microscopic one. A little publication every now and then, helps the mood, sustains the student, reinforces a thesis: incredibly, it does almost only good. But obviously, a small publication is useless, or at least you will be told that it is. You do not need that, forget about little papers, you are primed for big stuff: keep working on whatever is promising, big, giant enormous or hot, even if it looks like you entering the porn-industry. Reality is that is your boss who does not need small publications, not you. Reality is also that a good project, coming from a good intuition and backed up by knowledge, generates several both small and big publications: only bad project do not.
Another indication of ongoing “bad modelling”, is having the group leader worrying too much about consumables. If your boss is confident in his model, he will not care at all about any consumable, because he is so secure about the incoming big publication, with big impact, that so much more money is going to come from grants and so on. If you trust your model, what you want is having the data on your bench quickly, not cheaply. Particularly, considering that in the current world, slow science does not exist. The boss worries because he knows, maybe deeply or unconsciously, that you are spending his money shooting in the dark to a moving target. There is no other possible reason to worry. Therefore, when you start being questioned about how many blots you are doing without stripping/re-probing, think very well about it. When home, think again, and possibly start looking for another position somewhere else.
A bad model is also the one that regularly gives results which are slightly or widely different from what it predicts. A bad model does not follow the natural sequence of any scientific hypothesis: preliminary investigation>model re-shuffling>data consolidation. No. The bad model maintains a characteristic status of perennial preliminary investigation, whatever the age or the people involved. High-risk, preliminary investigations are by nature short events. Any project looking as an extended preliminary investigation is inevitably destined to fail: with only statistically-based exceptions.
I hope these few indications will help all the fellow young and hopeful minds when choosing a place for their Ph.D. or for a post-doc. When it was my turn I did not have any advice of the sort and my only feeling was of extreme flattering, because I was being offered a position, a job, and particularly, trust. Well, do not ever forget that when accepting a position in science, trust is not given to you but simply exchanged. So, take your time to think just a tiny bit more before giving your trust and time away, because unfortunately, if things do not go well, there is no way to get them back.
Marco
