I'm currently finishing up some work started by a postdoc who left a couple of months ago, so that we can get a publication, but I'm really worried that my experiment, which is running at the moment, won't work. My experiments of late have only been working about 75% of the time, which normally wouldn't be too bad, but the fact that someone else has managed to do them, and that I'm following their protocol, means that if I tell my supervisor that, for whatever reason, the results aren't how they should be, and that something has clearly gone wrong, then I feel incompetent. I have no idea why things aren't working, and obviously if I did, I could rectify it, but it feels like I'm pouring time and money down the drain and I have only myself to blame.

There, rant over :(

A normal response to failure I believe. Perhaps if you elaborated on the experimental design we could aid you ?

Is it possible that the protocol does not include all of the little quirky bits that the postdoc did for their experiments? It may be that they used a slightly different method that acheived reliable and consistent results that has not been divulged to anyone else. Don't feel that you have failed as it can be a number of things that can affect outcomes. Do you have any contact details for the postdoc to check the methodology is correct. Have you changed suppliers for any of the components used?

I think that you need to discuss this with your supervisor as sooner or later they will be expecting to see your results. It would feel alot worse telling them that after all this time nothing had actually worked . I think it is better to speak up early so that changes can be made to get this work done.
Good luck :-)

Thanks for your replies.

I think the quirky bits are included, but the problem is that I work with cells which can be unpredictable, and even my supervisor says they often "misbehave". Added to that the fact that I'm exposing them to something else of biological origin, and which itself is a little bit temperamental, and perhaps it becomes a bit more understandable.

My supervisors knows about the problems we're having, as I see her every week, and I'd much rather be open rather than spend 10 weeks pretending something is working when it sometimes isn't.

If you have followed the method to the letter - and it might be worthwhile re-reading it all and checking this is so, then you have found what anyone else following the method will find, which will not go down well with any journal as they would expect that the given method would work, so the fact that you have picked this up is important, not a failure. If you can set up a second run now, I would do so, just to check you are doing things correctly and to have a second set of results sooner than if you wait for completion. Write out what you have to do yourself, then you can see if you have missed anything, and if it still doesn't work then you have perhaps discovered an error in the script.

If you're working with cells, it could be anything really! I take it you've checked if you're growing them in the same stuff, not using too late a passage etc? Are the incubators working ok? Do you use any buffers that may be out of date? Any other chemicals etc. that may be out of date? Pipettes that may not be calibrated? Any substances that you use that are slightly different (i.e. from different suppliers) to the ones the postdoc used?

I have no idea what kind of experiments you are doing btw, just when you mentioned cells it just made me think of how many variables there are ;-)

Claudia, we're using a plant extract, and it's of a different lot number to that used by the postdoc. I've tested it and it seems to have a similar activity, although that's only one parameter and, being a biological extract, it could be quite different. I've email the results to the postdoc anyway, and she's going to let me know what she thinks...

I've been running some in vitro studies over the past few months, replicating the process a few times and even though I'm following the same protocols I still am getting some variability in the results. One prob I had early on was caused by a slight mis-measurment when I was diluting some chemicals I was using, it was only out by a  few micro litres but obviously it made a difference.