Hi all,

Well my viva is 2pm on Friday. I was all ready pretty terrified and now upon discovering a huge error on my behalf I have reached new levels of petrified.

I have spotted a major mistake - I mean MAJOR mistake in my stats. I have reported something as significant in one of my stats which is clearly, oh very clearly not significant (evidently in my sleep deprived write up state I misread the table). I have then proceeded to bang on about this significant statistic and its possible implications for a page and a half.

I genuinely do not know what to do or how to handle this in the viva. I was already preparing myself for a major corrections verdict / revise and resubmit but I now think full on failure is on the cards.

My PhD supervisor is pretty much useless, and attempts to contact him have failed since February. I know this is utterly my fault and the responsibility is at my door but any advice on how to handle this in the viva would be welcome.

Thank you in advance

It is one of many. I did two studies, one large qualitative and one quantitative. The quantitative analysis is structured around 5 propositions. The fifth one is where the error lies.

I genuinely feel physically sick. My PhD experience has been pretty terrible, with little supervision and what make me so angry is that this chapter was handed in to be reviewed by my supervisor and he didn't pick this considerable error.

I feel like the PhD viva will just be a humiliation and failure a sure thing. I have really tried to be positive but this is the straw that broke the camel's back in terms of morale.

Take a step back and breathe.

Yes, this is not an great situation, and ideally it is an error that should have been picked up by whoever read your thesis. It probably reflects worse on your supervisor than it does on you.

Sense of perspective - you've misinterpreted one result out of several, and written a couple of pages of error out of 200+. It is a mistake, but it doesn't have to define your entire thesis.

Those with experience of vivas here can advise on how best to handle the situation (you are not the first person to have made (and spotted) and error prior to via and it is not an automatic reason to fail).

There is plenty to be angry about, but if you can try to defer dealing with that till afterwards and keep a cool head, that would be best. It is done now, and what will determine the outcome of your viva is how well you handle the problem. Calm consideration is likely to help more than defeatism or blame, though those are perfectly natural responses.

They have already read your thesis now so all you can do is reply to their comments on the day. As it is one of many findings go through and be doubly sure it's the only error. Then make a note of all the pages where the error is mentioned, check your intro and conclusion. Make some notes (you can't get too detailed now) on how you would revise the thesis in light of this (it might be deletion of a section -some re-wording) and take it with you. If it is noticed and they bring it up just talk them through how you are confident it doesn't change the structure or integrity of the thesis and the other more significant findings, assuring them it's easily and quickly solvable. I'm not sure on if you bring it up if they don't - as it's a wrong statistic, obviously you wouldn't publish that as a result in an article but I don't know how you would personally feel it being left in your printed (un-revised thesis)? Anyone got any experience with this? Good luck.

Thank you for your swift replies. It is so nice to have some support and advice, as you can imagine I've had a pretty intense reaction to it today.

The statistic in question shows that there were differences between the two groups but these failed to reach significance at the conventional level. Therefore I think that there is a possibility that I can face the error without undermining the whole of the associated argument, which luckily is confined to one section in my chapter five and a paragraph in my conclusion chapter.

I am pretty terrified, as you can imagine. Just not sure how to manage it if they pick it up. To claim it was an oversight/typo shows lack of attention and consideration. Also I am worried they will think I did not have it checked by my supervisor before submission. Any suggestions on how to face this in the viva would be much appreciated!

Luckily I am exiting academia, so worse case scenario if a fail verdict is given I am not losing a career path, just a a considerable amount of self worth.

I think if they pick it up you just have to be humble and accepting but not too apologetic in that the rest of the work stands up. Rather than say it in a way that looks like lack of attention could you put it in a way that it was more to do with being a relative novice to the method used and you have taken this on board as a training development issue that you are confident won't be repeated (i.e the more you have put the method to practice the more you've learned how to interpret without error evidenced by your picking it up on your re-looking at the data and you are able to correct before publication)?

In the two months between submission and viva, as part of my viva preparations I went right through my thesis and as with any thesis there were mistakes. Eighteen were typographical errors and I mopped these up without fuss in preparation for any hardbound copy should I have been awarded minor corrections (which in the end I was).

However, I also found two moderately significant errors that I believed could have triggered a major corrections verdict. These I corrected in the script prior to viva, intending to say I was aware of the errors and they would not be present in the final printout. These turned out to be fairly straightforward and were just a matter of me mixing up a couple of tables of data in the discussion (the correct versions were in the results). The total error count from memory was around about twenty, including the two larger errors.

My supervisor told me to keep a log of any errors and bring that with me to viva. At viva, he then told me to put the error log away and not mention it the the examiners, who themselves found three minor typographical errors. I get the impression he'd talked to them beforehand and they'd already said they'd found very little.

My advice really depends upon what happens to you in viva. The distinct impression I got was not to draw attention to any errors and let the examiners find them unless there are major structural problems with your thesis that cannot be avoided. If raised, say how you intend to rectify the problem and discuss what changes you'll make to the thesis to correct it and any consequent changes to the conclusions.

It appears you error has been to misread a statistic, with a consequence that a few pages are incorrect. If 'minor corrections' are say three months at your Uni., then your error sounds rectifiable inside the three months.

You're pre-viva and problems will seem much greater in your head that they really are at this stage, so relax!!!

Ian

PART 2:

If they don't pick up on your error, what you do next is a matter of your own conscience. I personally would not want a major error in my thesis, so whether I could sneak it through with any requested minor corrections or not only you would be able to judge. However, bear in mind if you're caught (i.e. examiner goes right back through the thesis rather than just reviews the minor corrections - actually not too likely unless you have major corrections) you could end up in trouble.

However, others at and after viva just want rid of the thing and my own predecessor admitted there were errors the examiner didn't pick up on. That makes me question whether or not some examiners actually read right through people's theses.

Therefore, my advice is as follows. If it is a genuine, minor issue you can clear up quietly (and examiners don't want to be bothered with every comma that is in the wrong place), then do so without drawing attention to it and bringing on major corrections when the minor correction period will be more than enough to sort things out. Make a log as I did and deal with and admit to errors as and when they are raised by the examiners. If you feel the error is a major problem then it's probably better to admit to it and how you're going to deal with it as I've already discussed.

But whatever you do, don't place yourself in the position where you attempt to deliberately deceive the examiners over something that will significantly change your overall conclusions. That is a clear deception I myself would not tolerate (I have examined M.Sc. student's projects).

Ian

Thank you so much Ian. At least I have a plan of action now. It really only makes a difference to a couple of pages but I just think it undermines my ability to interpret statistical outputs correctly. I feel utterly terrible and hope it doesn't make them question the rest of the thesis.

I shall do what you advise. I have already got a list of typos and I have added these two corrections to the list also. I think I have a major corrections verdict at the best on my hands. I will have to wait and see.

I just want to thank you all for responding, it shows real kindness and I am so appreciative (my family do not understand at all and as I have already alluded my supervisory team, who are a husband and wife, are neglectful by nature).

I think I'm likening your mistake to reading the tails of a bell-shaped distribution (normal, student-t) as being the significant result rather than the body (or vice versa depending upon your data). If so, you've made the easiest mistake (and most basic) you can in statistics. I would guess most statisticians have been there at some time or other. :-)

You make your supervisors sound like people who stayed on to continue the student jolly long after their time as students was finished. :-)

Ian

Let's just put it this way, I wouldn't wish my supervisors on my worst enemy. I've always questioned how ethical it is to have a married couple supervising a doctoral student, as they inevitably back each other to the end. I am the only individual I know of with this particular problem.

My supervisors have had problems with getting their students over the finish line in the past, so I doubt I will be the last to suffer from their lack of interest and attention. I have already received an email informing me that neither will be in attendance at my viva on Friday.

Ah well, worse things have happened at sea - must push on and just hope for the best.

Hi.
Take along the corrected version, a colleague I know had something similar a few years back when a table printed incorrectly which changed results. In my ViVA some of my visuals were pulled apart, but I talked them through it. Hang in there, be strong. I have quite a few typos and a tough examination team but they didn't at any point say it was sloppy. I am now looking at the detail of amendments. Sending you positive vibes.

I passed with minor corrections :) Thrilled!

Thank you so much to all your suggestions - it really helped me today. Feel like I could sleep for a week now! :)

Well congratulations!!

And do tell us more details about the viva! Did they pick up on the stat error?