What to do with an overtly exasperating colleague in computer science?

Question

I am working on my undergraduate research project in an HPCC. Whoever does programming knows this is a "teach-yourself" skill. Everytime a problem arises, you search the wealth of tutorials and guides on the Internet, you ask in an specialized forum and the last resource is to consult a colleague (everybody is very occupied with their own business).

A few weeks ago, a new MSc candidate entered our lab and was assigned a seat next to me. She seems unable to go beyond very basic commands like executing a for loop in a Bash script. We helped during the first weeks, but she is making no progress. I myself handed her down a couple of tutorials and scripts. What exasperates us is that the more frustrated she gets the more she starts mumbling expecting that someone will approach and solve the problem for her, not to mention that she interrupts every five minutes asking the same question we have answered several times. We spoke about it in the most kindly possible way, now she says we are bullying her!

She is completely unprepared and we are starting to think that she lied to get accepted by our supervisor. She is destroying the amicable relations in our lab and affecting performance of almost everybody. I decided to program the most I can in my laptop instead of using the HPCC, but a process that gets done in five minutes at home is solved in seconds in the lab.

We don't want to mess up her reputation with our supervisor, but we got to find solutions. Does anyone had a similar situation with a colleague? How would you solve it in the most diplomatic way?

UPDATE: Group meeting this morning. Our supervisor found out she has done nothing. She blame us. Our supervisor said it is not our business to solve her problems, she might have spoken them out a couple of weeks ago. She (our supervisor) asked her to separate an appointment to talk privately. Let's see what happens...

Answer 1

You have a legitimate complaint (you are not there to teach that stuff). However, that's surrounded by misconceptions that I'd like to correct. Philip Guo's has written excellently on the topic and I recommend his post; I'll apply his idea to what you describe (which is a part of your actual case).

There are two explanations for the situation you described:

• she's indeed incompetent on stuff she should know (as you're assuming);
• she just happens to not know bash and run into some obstacle learning it, but appropriate training would make her able to contribute. In which case, the (wrong) assumption "how could one possibly not get that by oneself" acts as a form of bullying. In fact, since women and minorities are less likely to know Unix culture, this attitude can easily be a source of discrimination (see Philip Guo's post).

And my HPCC experience was full of bash-like tools with similar problems. (You only mention bash once, and bash is a bad offender, but lots of this answer is more generally valid).

Whoever does programming knows this is a "teach-yourself" skill. Everytime a problem arises, you search the wealth of tutorials and guides in Internet, you ask in an specialized forum and the last resource is to consult a colleague (everybody is very occupied with its own business).

I've learned programming that way (up to Linux-kernel-hacking level, in case you're tempted to dismiss me), but that's not the only one. For all skills, teachers exist to help overcome "teach yourself" obstacle.

[...] She seems unable to go beyond very basic commands like executing a for loop in a BASH script.

Does she ignore languages she was taught in her courses, or does she have trouble getting bash? In the second case, I have news for you: it's extremely easy to be a genius and not get bash. I'm a PhD student who does get bash, surrounded by many smart students who get it less, in a programming languages lab.

One key problem is that bash violates so many unconscious assumptions that are valid in "sensible" programming languages—for instance, by lacking a parser. So that forgetting any space in if [ -f "$i" ]; then will result in unhelpful errors — and trying to get them will result in more frustration.

As Philip Guo explains:

There is a huge disconnect between the elegant high-level ideas discussed on the whiteboard (while presumably sipping cappuccinos) and the grimy, grungy, terrible command-line bullshittery required to set up a computing environment suitable for implementing those ideas in code. This gulf of execution is tremendously frustrating for highly-capable and motivated students who just didn't happen to spend 10,000 hours of their youth wrestling with nasty command-line interfaces.

Now, you're not supposed to teach her this stuff — the professor should check one has the skills on entry or setup training. However, I don't see how one can avoid unconsciously bullying somebody else with those misconceptions.

Now, how could it be that she has an issue with the whole lab? Guo answers again:

OK here's what gets me super pissed. Many commenters presumed that “real programmers” should be command-line experts... POSIX-flavored command-line experts, to be precise.

[...]

More generally, this notion that the only “real programmers” are those who have already mastered POSIX command-line-fu before they leave the university is a dangerous one, and contributes to the continued monoculture in software-based industries.

I've had the same idea—then being a PhD student gave overwhelming contrary evidence.

EDIT: Another insightful analysis of what's hard in learning POSIX (specifically, of the "long tail" problem) appears in an analysis of a completely different problem—learning to cook for geeks. Here it is. Enjoy.

Answer 2

I will try to present yet another approach than the ones in the answers above. I don't consider it a definite solution, rather a way to go that may work for someone and may not for someone else. (And also, I would much more prefer to speak to someone higher-up, if you can; that's what the other answers discuss.)

---

The thing is: Be honest, but without judging the other person's qualities at all. I had been in such a situation before: person coming to me with simple problems that fall in the "googlable" or "in every book" or "in every course" or "in every tutorial" category, again and again.

The honest answer was, for me at least: Well, we have discussed a similar thing before. I'm sorry but I have got other things to do as well. I'm willing to help you with the burdens, but please, try to [point out some good resources here] before approaching me, and if you approach me, please make it specific enough so that we don't lose too much time solving the problems.

It is true, honest but not rude. After all, it is your time, your studies and you yourself who'll go on the market to sell your results, and you need to be productive. If this is a serious annoyance that makes your productivity drop significantly, they need to know it.

Answer 3

This is a obviously difficult situation, perhaps in part due to being an undergraduate researcher, with an MSc candidate the subject of your question. As you said that this is starting to affect the productivity of the lab though, this is obviously a matter which may affect the PI or supervisor.

I've been faced with a similar situation before, although the dynamic was somewhat different, in that while officially of the same level as the person in question, I effectively was their mentor or "secondary unofficial supervisor". In that case, I had a duty to raise the concern with their formal supervisor (in an informal manner), to alert them to the concern. They were grateful at being alerted early to the situation, rather than being left out of the loop, although this will vary in different cases, depending on their (desired) level of involvement in the goings-on of the lab.

If someone is struggling though, I do feel that you have an academic responsibility to a colleague to offer assistance if appropriate, but that has obviously been done here, to the point of detriment of others' work.

If you have a good working relationship with the supervisor (I am presuming it is also your supervisor), you could probably bring this to their attention casually/informally, although this obviously may not be the case - I had a completely colleague-like relationship with mine, and would act as the main contact supervisor to his project students), and any problems or concerns (including with other graduate students, as I was the most experienced) would be raised. In that case however, this information was always understood to be simply to help them as much as possible.

Perhaps you could discuss the concerns with a (no doubt already aware) experienced colleague in the lab. I am presuming that there's someone who is generally looked up to for advice by those working there. I'm sure they would be happy to discuss the matter in private and advise (or perhaps talk to the supervisor if they have a good relationship with them).

Answer 4

You are a student, not a mentor, advisor, teacher, or TA. And therefore it is not your responsibility to teach other people to do their job (you can help, but only if you want so). When the time will come and your advisor will ask you to present your results, I highly doubt that "I have not finished my task, but I helped A and B to do their task" will help you.

It looks like you have done more than expected. Taking into consideration that some of your attempts were considered as bulling I would stop giving any help.

If you want to make it in a more diplomatic way and do not want to speak with your supervisor I think about two possible ways:

1. Start answering the questions with an increasing delay: answer the first question, for the second - tell that you need to finish some stuff and will be able to think about it in 15 minutes, the third in 1 hour and so on.

2. Tell that you will be able to do this tomorrow, so ask her to write her problem in an email so that you will be prepared on the next day. This way the third person (your supervisor) would be able to see whether the issues she is failing with are really trivial (who knows, maybe only you think that it is easy) and you would have a list of all problems she faced and asked your help.

If the person tells "I want it now, stop all you things and go help me", you have to explain that the word does not work this way.

P.S: But please do not assume things like this: "she lied to get accepted by our supervisor". There are a lot of explanations apart of she lied, so this is a really strong accusation.

Answer 5

Let's say you have taken the final "asked a colleague" step, what's the next source of help if that still isn't useful? It sounds like that is the action she should take.

If you are not sure yourself, then that is probably the problem. There should be another source of information past just your colleagues.

If the answer by that final source of information is, "Look it up" then support that answer yourself. Ask "Have you looked it up?" She may say she didn't find it. Ask "Have you asked on the [such and such] forum? She may say she didn't get an answer. If she has done this, and you don't know the answer yourself without looking stuff up - you can choose to either show her how you do your process for looking something up - by having her looking it up but you just tell her verbally what you would do - or refer her back to the final source of information saying, I tried looking it up but still can't figure it out.

• Most people like to "take control of the keyboard" to teach, but that is probably a poor teaching method for her. Guide her if you choose too, but don't do it for her.

• Is she asking the same questions over a particular process? Have her write it down somewhere safe and then if she asks again, refer her to the notes. She may not realize everybody doesn't memorize everything.

---

I myself was a student who would be afraid to ask above the colleague level because after the first time of "Look it up" it felt like that means, "I will not help you", which is not true - they are just trying to make the best use of their and everybody's time.

Answer 6

Be careful about the bullying part. Even if it's completely unfounded, if she complains about bullying to a supervisor or even the department, you will have to explain yourself. And if people believe her but not you, you are in trouble. There is no guarantee that people make sensible judgments.

If you raise the issue with a supervisor in a way that leaves a paper trail, you are not only doing something to address the immediate situation, you are also protecting yourself from a situation that is much worse than the current one.

Also keep in mind that communication is tricky. If you say "she starts mumbling expecting that someone will approach and solve the problem for her", then that is almost certainly not the way she experiences this. Involving a third party (the supervisor) can resolve communication issues, or mismatching expectations.

Answer 7

This person sounds like a real-world help vampire. The problem is, you don't have the buffer of an SE site to prevent her vampirism, and she seems to have learned very well how to leech help off of those around her - recognizing that her muttering will get people to pay attention to her, if for no other reason than to stop it, and knowing that repeatedly asking the same question will get people to eventually just do the work for her.

This is an entirely unhealthy habit on her part, and you should do everything you can to not feed into it. Don't help her unless it is absolutely necessary. Don't offer to write any script for her. Don't direct her to anything but the most basic of search engines. You may even want to consider playing dumb if she insists on you giving her help, and if that doesn't work, then consider asking for or sitting in a different seat (I can't imagine you have assigned seating at the Graduate level, but if you do you should be able to ask the professor for a change of seating).

Above all else though, do not feel guilty about cutting yourself off from this person. They may try, very effectively, to guilt trip you into offering help, but do not feed into it. As you said, this is a class where it is expected that the student will do a lot of self-teaching (and I hope that's true, or else she should be asking the professor, not you, for help) and she's going to have to learn, the hard way, that self-teaching means exactly that - finding a way to teach yourself.

Answer 8

There are several good answers here already. For example, be honest with her. I really like this idea but it might not work for a simple reason:

No matter how hard you try, if there is no will to change on her side, you will not succeed. There is one person who can force her to change her attitude and that is of course your supervisor. I think safest solution here is to let the supervisor know about the situation and let him/her handle the situation.

The point is, you are in a tough situation and I encourage you to act very wisely. You need to consider few aspects carefully.

For instance, be mindful of sexist behavior/bullying/humiliating. I am not saying that you are being patronizing or any other accusations. However, experience is telling me these types of situations have good potentials for misinterpretations. For this reason, I believe it is wise to handle this situation through the senior staff sooner or later.

The rather difficult part here, is how bring this situation to your supervisor's attention. Perhaps you can directly tell them about the situation and ask them to handle it. This highly depend on the relationship you have with you supervisor and your position in the lab. Be sure that he/she knows that you have done more than enough to solve the situation in the lab.

Another solution would be involve the supervisors indirectly. For instance next time she comes with a new question, ask her to email it to you and CC the supervisor. If the answer to the question is very trivial your supervisor will automatically be alerted about the quality of new msc candidate. A better scenario would be that she will try to solve the problem herself, when she face the case of supervisor's involvement. You should carefully and delicately pass her to the supervisor.

Couple of notes:

1. She might be in a hard situation as well. New programming paradigm, new person in the town, just coming from bachelor, feeling a bit vulnerable, etc. Maybe, just maybe, she is actually trying.
2. Do NOT assume she is lying. You are not in a position to judge her and this is a very strong accusation and quite unprofessional.

Answer 9

I think this calls for some serious talking with all the parties involved, including your supervisor.

One possible approach to keep things civilised and without pointing fingers, is to set up a private chat for the lab, and everybody asks through that. This way, you don't interrupt anyone asking a question, and potentially anybody can answer you. This also gives some broad exposure to the questions (maybe you didn't know that one of the new people to join the group is an expert in the technology you have a question about, and you wouldn't have asked him).

Applied to your particular case, it creates a paper trail of her questions, spreads the load among more people, and forces her to think about the question (I managed to answer a couple of questions I had while writing them on Stack Overflow). When she asks in person, politely point her to the chat, where she will get the answers. It is important that you make this a firm rule, with no exceptions.

If she still refuses to use this chat, you change your subjective complaint (she asks too many too basic questions) to a very objective one (she is not asking through the chat).

Answer 10

Suppose she asks you how to open a file in Python. Rather than telling her about the "open" command, go to her computer and google "how to open a file in Python." She'll quickly learn that she should be asking Google instead of you.

If her question requires more than 30 seconds of thought for you to answer, you can also say "I don't know, why don't you check the Internet? Someone there will probably know the answer."

Answer 11

I don't know much about the interpersonal climate in your lab, but if you feel confident to solve it with humor, I can help you.

Hand her 20 question vouchers. Sounds ridiculous but worked perfectly for my daughters some years ago. 20 vouchers per day are plenty, but it forces her to think about every question. It also gives her the possibility to proof, that she isn't asking to much questions, which is smoothing her way to improve her "teach-yourself" skill.

Answer 12

Is she pretty? This may sound like the most sexist question possible, but in some STEM courses/disciplines particularly with a low female quota this can be a large contributing factor to not getting used early to organizing your workloads and learning yourself. Because those male coeds who get along well with their stuff have little better to do than fall over themselves providing help and support in study groups.

Been there, done that, got no T-shirt (800 coeds starting with me, about 16 female students among them). Now depending on the college and courses you are in, you still need the skills and brains to pass. There are hard limits to how far you'll get just by being someone to look at. Which is not restricted to females, actually: it is well-known that taller and/or more self-confident men convey much more competency than their less visually gifted counterparts. But they are less likely to be the victim of unabating "let me do as much for you as I can" attack waves.

Whatever the reasons, it is becoming apparent that she is not well-prepared for the kind of workflow she is supposed to be part of. That does not have to mean she is dead baggage: one thing that comes to mind to me is oboists: one of the most time-consuming and exasperating part of playing that instrument is making your reeds. As an amateur, you easily spend an hour on a reed, and then it will work any amont of time from 5 minutes to several months. Several professional oboists have a side business of making reeds for amateurs. You can perfectly well perform on your instrument while never making a reed yourself. But it makes you dependent on the skills and availability of someone else. Of course, you'd not expect another colleague in your orchestra to make your reeds without compensation.

So at any rate, it is clear that your colleague here is not able to do her work without the involvement of others. That's a department and managing problem. You need to get together, figure out what is missing here, figure out what is required for jump-starting her into become productive in the manner expected in your department, figure out who can help with that in what manner and what effort, and figure out how to free the resources necessary to do that.

Basically there are three possible outcomes: a) she'll be more work than she gets done. This will not work out, separate. b) she'll require work on a continuing basis, but deliver more than one puts in. This requires proper work arrangements and understandings. One needs to evaluate whether the net benefit justifies the pay and the work disruption. c) she can develop into a productive member of your team once she gets over the culture shock and develops the necessary skills.

It is clear that this will not work without someone originally carrying some of the weight, and the self-organizing of that does not work well. Fingerpointing will not get this situation under control. Actively trying to fix this on a department level will keep the cost under control, will make it clear who needs to get credit and responsibility for carrying that initial cost, and will make it clear how to assess where one is getting.

This is an unproductive and aggravating situation for all of the involved. Conflicting plans, targets, and expectations are not going to help.