Sorry for another quitting post. I really need some suggestions. I am at the end of my third year on a Computer Sicence PhD. I had a quite nice started, past my upgrade without problems, had my first paper accepted in a small conference on the 1.5 years. But due to funding problems, my supervisor and I decided to withdraw the paper. After that, nothing went right. I need to write a software to process data. I stacked at a technical problem for like several months and seriously depressed. Then I took a leave of absence for 8 months. I finally manage to solve the coding problem after returning study and starting to gather data. After I got the data, I found many other problems in my software. It looks like this route only can get some bad results.

I had no issues with my supervisor. He is a nice person. The only problem of him is he persuades me to do the route I don't like. But this is ok.

Now the situation is really bad for me. At the end of the third year, some raw data, a crap software, no results and no publication. Based on all this, I may be able to put together one and a half chapters that's all I've got now. I have one writing up year left. I am very doubt about whether I can finish it or not? Or I simply walk away from the PhD?.

I recommend you have a meeting with your supervisor and discuss options. Before that, you should take stock of what you have in terms of data and output, and determine whether you could actually shape it into a PhD thesis. You may be surprised at what you have, especially given the time you have been working on it. If you can find enough to put into 3 data chapters, then you have a PhD.

Don't worry about publications at this stage, you can always publish after you have your PhD. Conference papers are are a 'nice to have', but not essential, especially if you aren't on top of your thesis.

It would be a waste of 3 years if you quit, and you would always have regrets.

thx for the quick reply Barramack. The 1.5 chapters refer to data chapters. However, I am pretty sure this is what I've got now. I have another idea on pilot study stage. However, I recently found that method has been used in other ppl's publication.

I can explain it on a more clear way. I have: raw data, 80%-90% software for method 1, 50% software for method 2 (has been published by somebody else), no result. I need: two set of results from software 1-2. 2 may need huge amount of modification. evaluation + user evaluation.

Now what I am worrying is whether I should throw another year in and still cannot finish or just save this year for something else. My university have very restrict rules about submitting within 4 years. Also I already determined to do not still in academia since long time ago.

Does your supervisor think you are going to have enough data for a PhD?

We had a panel meeting in last Nove. After that, my supervisor and I had a decent chat about the rest of the phd. The assumption was, if everything goes really well, I would have the PhD. The data capture was ok, however, the processing software never reach the required level, which means I definitely need to spend much more time on it. Therefore, it is difficult to say now....may or may not.

Well check again. 6 months is a long time.

My PhD is not in computer science but either way I know that quitting the PhD at this stage is a BIG mistake and you'll regret it. Keep a good rapport with your supervisor and be honest about what you think and what you need to finish the PhD.

Is it possible that the problems you've encountered with your software and data are worth reporting? Of course, it always feels better to get the results we're hoping for, but when things go wrong, it's often the case that something worthwhile can be reported from that too, such as the reasons why the results weren't as expected, and what should be done in future studies.

Had a long chat with my supervisor. Things get worth. He said if I cannot get something novel from now to the end of this year. I can only write as a Mphil....I have to produce software or results like every two weeks....

Sorry to hear that. Do you think you are able to achieve this? What do you need from your supervisor/other people to help you?

I dont think there would be any real help. I have to do everything by myself, I am the only one who is doing my direction in the lab. I am in a very small research group, no RAs/Postdocs. My supervisor will not involve in any kind of coding. He may help with data collection, but just test equipments etc.

Well I think you need to identify what help you need and then find it from elsewhere if possible because I don't see how else you will be able to complete.