Skip to content

Troubleshooting.

2009 August 18
by Hélène Martin

It’s somewhat ironic that computer scientists often get asked for computer help that they can’t provide.  The truth is that software-writing ability generally doesn’t depend on utilitarian knowledge of Windows 2000, optical mice, wireless access points or printer drivers.  A number of computer scientists in fact take offense when asked for general computing help.  I guess they see it like asking a biologist for help troubleshooting a microscope or something though I’m not sure the biologist would be quite as offended.  I suspect they may act offended because they honestly don’t know.

I’ve always had a place in my heart for IT and happen to take pleasure in banging my head against strange computer (or more often, user) glitches.  And as unglamorous as diagnosing computer slow-downs or malfunctioning peripherals may be, I’d go as far as saying that my troubleshooting abilities are some of the most valuable skills I have.  What does it take to teach those?  Why don’t more people have them?

I recently received an e-mail from someone I really like and respect.  Our conversation went something like:

  • Short e-mail asking if I could help her fix her backspace
  • I ask whether it’s a mechanical problem and if so recommend popping out the key and cleaning it
  • She answers with more details: the backspace key doesn’t delete a selected region of text but works otherwise
  • I ask what program it’s happening in.  By this point, I suspect a Word or Office setting so I ask her to try in notepad, Outlook and Word
  • It’s only in Word
  • After a little poking around, I find the “typing replaces selection” setting

This is representative of the kinds of problems I’m asked to address — the issue is that most users don’t realize what their problem actually is.  In this case, the backspace key itself was fine but a software setting had changed.  It turns out that searching for “overwrite word selection” in Google provides a solution from Microsoft as the first hit.  Does it take a sophisticated understanding of the way computers work to know what questions to ask or things to try?

I’d like to try to help my students gain troubleshooting abilities but I’m not sure how to go about it.  Some of the things I’d like them to think about:

  • Eliminate hardware first (“is it plugged in?”  “Is the cable ok?”  “Is the socket ok?” “what if you plug it in elsewhere?”)
  • Applications have their own settings (“does it happen in application X?”)
  • Some applications are related (IE-Windows, the Office suite)
  • Rebooting solves a lot of problems
  • The right search query solves a lot of problems.  Coming up with it requires asking lots of questions until you have ideas on what might be causing the problem (hardware, settings, interaction between programs, malware)

Having “book knowledge” of troubleshooting isn’t particularly helpful, though.  I guess I can reproduce the backspace issue, for example, pretty simply and have them walk through how they would address it.  I wonder what students would do with that…

6 Responses leave one →
  1. Craig permalink
    August 18, 2009

    I would assert that 90% of “knowing how to fix computers” is knowing how to ask the right question. Awareness of which search terms are likely to drill down to what you’re looking for is both a learned skill and a problem of vocabulary. I was much less effective at helping people in Japan because I just didn’t know what words to search for in Google.

    • Hélène Martin permalink*
      August 18, 2009

      Interesting anecdote. I think you’re absolutely right that asking the right question is most of the battle!

  2. August 18, 2009

    Your introduction reminded me of the first class session of my Linguistics 101 class a few years ago. The professor asked us to guess what the most common question he received when he told a stranger that he was a linguist. There were a couple odd guesses from my classmates, but I guessed it was “How many languages do you know?” and I was right. As a polyglot yourself, it might not be a terribly maladroit question, but my professor only knew English fluently (though he admitted that he knew a little Japanese), so it was a consistently awkward way of starting a conversation at a social event.

    As for troubleshooting, in particular, I think your example (along with any number of other examples you or I could produce) just go to show that there are significant differences between people in their willingness and ability to analyze problems (technical or otherwise), and even to be able explain whatever problems they have (i.e. constructing the right Google query to get the intended search results). It’s kinda fascinating how, I imagine, most people out there manage to get through life without being able to answer a large portion of the questions that may go through their mind every day, assuming those questions even occur to them in the first place.

    Hi! :D

  3. Sylvia permalink
    August 18, 2009

    This is an interesting topic you bring up.

    And I totally agree with Craig about the asking the right question. I also think that the reason computer scientists are better at doing that than other people is that we are around computers all the time and have experience using them and sometimes breaking them (whether intentionally or by accident). Over time we learn to not be afraid of the computer and learn how various things interact with each other (even if only on a surface level, such as key settings being different for various applications). When it comes to solving the problem we have ways to narrow things down and maybe a few guesses as to what’s wrong. And if we need to we know how to ask Google too. Other people always seem to think that the problem is something really complicated and even if they searched they either wouldn’t get the right query or wouldn’t understand the solutions that came up. So they don’t bother. They ask their tech-savvy friends instead.

    I don’t think there’s really a way to teach troubleshooting. People just have to know that often it’s not that complicated and they can do it themselves if they tried.

    Perhaps if you had a list of common problems and how to troubleshoot them, it would give students a starting point from where to go or perhaps ideas as to what could have went wrong. Show them that it’s not that complicated.

    (And here are some pictures for relevant amusement:
    L’problem
    Building experience!
    But really, computers are nice. )

  4. Hélène Martin permalink*
    August 31, 2009

    Hi, Jim! That’s a really good analogy you make… especially given my background. You are good, sir! Here’s the thing, I actually get a lot more annoyed when people ask me how many languages I speak when I say I have a linguistics degree than when I’m asked to fix a computer because I have a CS degree. In a way, I feel like the latter is ok — I should understand my field’s tools. The former is kind of like asking a marine biologist how many lions he’s dissected because, well, he’s a scientist who works with animals. Or something. Linguists generally study how either a particular language works or what abstract principles govern the process of language. As Jim knows.

    Part of the reason I’m ok with being asked to fix a computer is exactly what Sylvia brings up. I should be able to figure it out because I’m a problem solver. I don’t understand these other folks who find it beneath them to care or know: if they can’t figure it out, then they’re just not very good at solving problems, which is Not Good!

  5. Aliya Walji permalink
    September 8, 2009

    Ok this is clearly not as thoughtful a reply as the others but still, somewhat on topic. I’m thinking of printing this and giving it to my boss:

    link to xkcd.com

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS