Was this thread useful?

Discussion in 'Comments and Suggestions' started by EStjarn, Mar 5, 2011.

  1. EStjarn

    EStjarn Senior Member

    Spanish
    Members are asked to look for previous threads before starting a new one. Depending on how general the topic is, the number of such threads may be anything from zero to several hundred. (I think the maximum number of results is approximately 1000 threads.)

    Ideally we're able to refine our search so that the number of results is not greater than what is practical in order to solve the query at hand. In practice - at least that's my experience - it happens not seldom that the number of threads to choose among is so great that the effect is discouraging. In those instances, I feel the problem is that we have no way of evaluating beforehand what's hidden behind each hyperlink; we have to open them, one by one, to find out about their quality, or applicability.

    While it is an advantage to have access to a vast amount of information, if we have insufficient means for discriminating among its pieces, and therefore choose not to make use of it because of the time investment required, a part of its value is lost.

    What I wonder is:

    1) Am I alone in experiencing this as a problem: to have no means of discriminating among previous threads when there are many such threads to choose from?

    2) If not, would it be possible to add to each thread a question asking the reader whether the information was useful, and present the frequency results of the responses along with the list of threads in order to make it easier to choose a relevant one?
     
    Last edited: Mar 6, 2011
  2. Paulfromitaly

    Paulfromitaly MODerator

    Brescia (Italy)
    Italian
    Something similar has been discussed in this thread

    SUGGESTION: "Tick" for solved thread?! ;)

    As a forero, I do share your annoyance towards the ridiculous number of threads about some very common topics.
    For some reason that still escapes me, WR rule #1
    is very often flouted and that leads to an abnormal number of duplicated threads.
     
  3. jann

    jann co-mod'

    English - USA
    On the French forums where I moderate, we routinely combine threads that are on the same topic. We try to improve thread titles so that it's easier to assess the focus of a given thread without having to open it and read it. And sometimes we remove old threads that are unlikely to be useful to anyone.

    You can help us with this process: when you find a thread with a bad title (too generic), when you discover an old thread with zero answers, when you see a thread that is really useless (not just for your particular case, but probably for everyone), when you find 6 threads that are all about the exact same expression/term.... please click the report triangle ([​IMG]) to let us know. :)
     
  4. EStjarn

    EStjarn Senior Member

    Spanish
    Although related, it's not quite what I had in mind. The suggestion there is that each thread would be tagged as either 'solved' or 'unsolved'. The suggestion here is that people looking for an answer - not just the original poster but anyone who opens a thread with the hope of finding an answer to a query - would be able to confirm that the thread in question has been useful.

    My experience is that when we look for a specific answer and try to find it among a list of threads, we often have to open quite a few that are disappointing to read, not necessarily because they don't answer the question posed by the OP or someone else, but because they lack the informative quality we're looking for.

    In the list of related threads, we have no possibility of distinguishing a useful and much-appreciated thread from one that perhaps doesn't have a proper answer. And furthermore, since they're listed according to alphabetical order, the threads with a title that starts with for example a "w" are bound to receive less attention than those that start with, say, an "a".

    Indeed, this suggestion is directed specifically to this problem. If we have too many old threads to look through in order to find out whether the question we have in mind have been answered previously, it's understandable (to me at least) that many will choose to start a new thread instead of going through all the old ones.

    Here's an example. Say we're a new member, and we have learnt we should be looking for old threads on the topic we're interested in before starting a new one. Say that the topic is the difference between 'which' and 'that' in relative clauses. We've learnt that in order to find old threads, we can submit both words in the search box at the top of the page. Here are the results: approximately 100 threads where the words 'which' and 'that' appear in the thread titles. Which ought to be a good thing because among those threads we should be able to find the answer we're looking for.

    However, it's going to take some time. The first task is to browse through the thread titles. (I'm doing that now.) To find the first relevant thread title takes about thirty seconds. It turns out to be Difference between "that" and "which" in relative pronoun. (I'm opening the thread.) As it happens, the thread is closed but contains links to other threads. One of those links seems perfect, it is even recommended. Unfortunately, the thread in question turns out to not exist, and the rest of the linked threads (there are three of them, and we open each one) aren't applicable.

    Back to our list of related threads again, looking for the next relevant thread title. At around search results #50 to #60 (which takes another minute or so to reach), it seems we should be able to find what we're looking for; the thread titles all start with the term "Relative pronouns". We open the first one, in which we are able to find one post (out of seven) that is informative. But we probably want more than that, so we try another one, this time choosing more carefully among the #50-#60 sublist, and we end up selecting Relative pronouns - that and which - word. It's found to not contain much information, yet it has one crucial feature: a link to yet another thread, from the French and English grammar forum, a thread which ends up being all we could ever hope for in regard to our particular query.

    How many of us have the patience to do this kind of research instead of starting up a new thread? If the answer is 'not very many', I feel it makes sense to try to help members as well as users in general to find the more informative threads.
     
  5. Paulfromitaly

    Paulfromitaly MODerator

    Brescia (Italy)
    Italian
    Once again I agree with you, EStjarn.
    However let me point out some other issues that may help understand why there are so many threads about the same topics as and why a good solution to this problem is still far from being found.
    In your example you assume that the first thing a forero does is search in the dictionary or in the previous threads.
    Unfortunately that's often not true.
    Some newbies as well as some (lazy) senior members don't bother to use the search function.
    They don't care if there may be a few dozen threads on the same topic.
    They don't care if the thread they are about to open will be useless or the 76th thread on a done-to-death topic.
    They don't care to choose a meaningful title for their thread.
    They just want an answer to their question or a quick translation of their sentence.
    Whether their query is going to be helpful to other people is none of their concern.
    Sometimes they even PM the mods and ask whether we can remove their threads from the forum once they've found what they were looking for.
    You can see why those people wouldn't even want to know what a "solved" or "unsolved" tag stands for.
     
  6. panjandrum

    panjandrum PongoMod

    Belfast, Ireland
    English-Ireland (top end)
    This is a difficult problem.
    It has become more difficult as the forums have grown.

    Possible solutions?

    Merge all the threads on a particular topic.
    That was great while the forums were young. But by now it would mean threads of many hundreds of posts about common topics (such as that/which).
    It still makes sense for more uncommon topics, but really does not help for those that are most popular.

    Use helpful thread titles.
    This is, of course, something we encourage. It always helps, though it is not the ultimate solution. For example, the titles of many of the threads about possessives have been changed to add more information. The same with threads about prepositions and some specific locations (hospital, school, etc).

    Invent threads that summarise the topic.

    Possible, but very labour intensive. A very small number of these have been created in English Only.

    Annotate thread titles - SOLVED.

    Possible, but difficult, labour intensive, and very subjective.

    This topic has rattled around the moderators' world since the dawn of time, without a satisfactory resolution :)
     
  7. EStjarn

    EStjarn Senior Member

    Spanish
    I don't want to picky here, but it seems I'm not getting through completely. A thread with a question that can be considered solved is not necessarily informative. Normally, however, an informative thread does solve the thread question. Just to make sure there's no misunderstanding of what I mean, I will try to be more explicit.

    The problem I'm addressing is:
    When we are searching for information and are confronted by an extensive list of threads, how do we distinguish the informative ones from the less informative without having to open every link to find out?
    My suggestion is to add a yes-no question such as 'Was this thread useful?' or 'Did you find this thread informative?' at the end of each thread to which all visitors are entitled and encouraged to reply by simply clicking a button. The aggregated results of those replies would be displayed along with the thread titles in the list of threads that shows up when we use the search function.

    The following is an example of how such a list could look like. It parts from the search results for the words 'which' and 'that', mentioned in post #4. So as not to overdo the example, only a few of the more than one hundred threads are included. They are arranged with respect to how much feedback they've received in the past, where a green-colored number represents the frequency of 'yes' (useful/informative) replies and a red-colored number represents the frequency of 'no' (not useful/informative) replies:
    that imaginative realism which admits that this is a new world... 1002 / 79
    That/which: the book <that/which> you showed me yesterday 300 / 30
    Back to the village - that where which - I was born? 280 / 3
    The way that/in which/how 99 / 190
    Relative pronouns - that and which - Word 98 / 47

    Use of Colon and “which” or a “that” 88 / 14
    in which = that ....in ? 50 / 15
    That/which: experiences <that/which> I know... 29 / 29

    How to use 'that which' 9 / 40
    That/which: ... a yellow tablecloth<,> <that/which> suits the rustic tableware 8 / 40
    no better university for my interests than one <whose/ that's/ which> name is attributed 0 / 3
    That/which: You may borrow a book <that/which> you think is useful
    0 / 2

    ...from which country is that... 0 / 0
    With or without "that/which + to be". 0 / 0

    Although a solution like this wouldn't tell exactly what to expect in terms of usefulness when we open a thread, it would give us a hint.
     
    Last edited: Mar 6, 2011
  8. Paulfromitaly

    Paulfromitaly MODerator

    Brescia (Italy)
    Italian
    I do understand what you're suggesting :)
    However your whole reasoning is based on the assumption that people use the search function or the dictionary before opening a new thread, but unfortunately that doesn't happen very often.
    If they did, we wouldn't have so many duplicated threads.
    If they did, we wouldn't need the rating system you're suggesting :)
     
  9. cubaMania Senior Member

    I'm not at all sure that you can know this to be true. Do the mods actually have a way to determine how many people did a search and found a thread that answered their question and therefore did not open a new thread?
    Suppose, for instance, that 1000 people were interested in the term "frzzzl". Maybe 995 looked it up in the dictionary or search, found their answer, and did not need to post a new question. Maybe 5 either did not look up the term, or did look it up but did not find their answer, and so each posted a new question with the title "frzzzl". The mods could know about the 5 people who posted a new thread, but how would the mods know about the 995 people who didn't? Those 995 would be the ones doing the rating of existing "frzzzl" threads.
    EStjarn's suggestion looks clever and promising to me, and is different from any I've seen discussed before.
    However, I don't know whether it would be practical to implement, nor whether it would actually work. But it is an ingenious idea.
     
  10. EStjarn

    EStjarn Senior Member

    Spanish
    Let's say it's building on the fact that we're supposed to use the search function before opening a new thread. Whatever the reason is for people not to do this, I don't think it would be to a disadvantage for anyone if its functionality was improved. If the improvement is substantial, and members in general find it easier to make such a search, it should show in the statistics, as it is likely that some of the members who don't heed the rule in question have experienced the same frustration as I do, except that they have chosen to confront the problem differently.
     
  11. Paulfromitaly

    Paulfromitaly MODerator

    Brescia (Italy)
    Italian
    Your suggestion is brilliant (and I can tell you that Mike K. has been working on a new functionality that should allow foreros to report threads that aren't useful), but I don't know whether it's feasible or not.
     
  12. susantash Senior Member

    Montevideo
    Español de Uruguay
    I Like ESTjarn's suggestion very much. I think it would be really useful for members who DO use the dictionary and the search function. I am one of those people and I´m sure there are many more members like me.
    I believe it's worth giving it a try. If some people don´t take the time to look for previous threads that's too bad, but there are many more who will, and will find this new function very useful.
     
  13. EStjarn

    EStjarn Senior Member

    Spanish
    With my limited experience of programming, I would say, 'yes, it is feasible'. However, since I have no idea what it's like to write code for a website of WR's proportions, I can't tell for sure. A comment on this from the administrator would be highly appreciated.
     
  14. MartyfromFrance New Member

    French - France
    Hello,
    Pardon my intruding in this thread and world of experts.
    After years of anonymous consumption, I thought I HAD to registrer on WRF (and actually did) to applause on this suggestion! A progressive priorisation by the people who actually need the information!

    Look at all the examples around, from Yahoo answers, Wikipedia discussions to Facebook Likes : time for WR to move more into "Web 2.0" ?

    I think this feature should even be available for visitors.
    I can understand there is less trust in a visitor's appreciation thant to a senior member's. Maybe, on this score, you will want to give more credit to seniority ? (give more "load/weight" to a clic from a member, event more to senior members, etc.)

    By the way, I haven't found yet how the WR Dictionnaries are currently updated (and by whom), but this "Useful" feature could also help a lot to update the Dictionnaries with the most relevant additionnal entries ?

    Furthermore, it could also help the mods find and deal with unuseful threads ?

    Hope this contributes, long live WR,
    Bye,
    Marty.
    "30.000 lemmings can't be wrong". (Or maybe they can, but in the long run, they nevertheless help perpetuate their specie ?)
     
  15. panjandrum

    panjandrum PongoMod

    Belfast, Ireland
    English-Ireland (top end)
    If you look at the dictionary entries you will see that these are sourced externally to WordReference. For example, the English Definition dictionary is the Concise Oxford English Dictionary. WordReference cannot update the COED.
     
  16. MartyfromFrance New Member

    French - France
    Oups. Thank you for the info, Panjandrum.

    (Maybe this is a pity, but I don't want to go off topic.)

    Some references are also tagged "WordReference English-French Dictionary © 2011", so I guess there is also internal editorial work : could the "Useful" feature described above by EStjarn help this work ?
     
  17. EStjarn

    EStjarn Senior Member

    Spanish
    Do you mean that these measures can substitute for the idea presented above?

    Is this request related to the French forums only, or is it applicable to all WR forums?
     
  18. jann

    jann co-mod'

    English - USA
    Yes and no. :) We don't have a working version of the idea you presented above. We do have working versions of the measures I mentioned. I do not mean to say that we should not think about ways to make the archives easier to search, and ways to make it easier to identify useful content in the archives. But I am not yet convinced that the "was this thread helpful? Y/N" method so widely employed on many sites with interactive content is the way to go here on WR. Fortunately, the methods I mentioned are by no means incompatible with an idea like yours.... whatever form it may take in the end.

    I cannot speak for the forums where I don't moderate... but I have trouble imagining that any of my colleagues on those forums would object if you reported threads with bad titles, or suggested merging two threads where the discussion was essentially identical. Judging "useless" is slightly more difficult, and I don't think we want a flood of reports about threads with zero answers (though the occasional report would be fine). :)
     
  19. EStjarn

    EStjarn Senior Member

    Spanish
    More specifically, what is it that you find unconvincing about the method? Also, could you give a few examples of the sites you're referring to that are employing it?

    I understand this as, 'I mean to say that, beyond what is already done, we should think about ways to make the archives easier to search, and ways to make it easier to identify useful content in the archives.'

    If this is a correct interpretation, apart from the methods in use today, what ways would you propose to make it easier to reveal the usefulness of individual threads?

    I can see this as an advantage of the existing methods over the proposed idea if they were to compete. However, as your post states, they aren't incompatible. I'd say they treat different aspects of a concept that might be termed 'searchability'.

    The methods you promote - combining threads, improving thread titles, and removing threads - should lead to a more relevant selection of threads as presented in the thread title search results. (Panjandrum comments on the methods in post #6.) Although members might contribute to this work by using the report triangle, the bulk of it should fall on the shoulders of the moderators. (The forums' guests, who make up some 95% of the users, would not be involved at all.)

    As I am in no position to ask of moderators to do more work than they are already doing to improve the searchability of the archives, I look at the other side of the equation, to which I myself belong - i.e. to the community of members and guests. And in it, I find an untapped potential: Each time a user reads a thread, it leaves them with a sense of satisfaction, disappointment, or indifference. To offer an outlet for that emotion in order to gather information represents a way of acquiring 'information about information', something which otherwise may not be easy to come by.

    This meta-information, as has been said, could be helpful to users in determining the most relevant threads for their specific inquiry, and in avoiding threads that are disappointing. Also, as Marty suggests in post #14, it may help identify threads to be removed. In addition, even a disappointing thread might be experienced less so if we sense that our opinion of it is of use to others. That is, although our efforts might have been wasted from our personal point of view, on a global level they weren't.
     
    Last edited: Mar 12, 2011
  20. jann

    jann co-mod'

    English - USA
    Just because member X found a thread to be useful doesn't mean that member Y will also find it useful. Perhaps they are working in different contexts. Perhaps one has a grammatical query about the structure of an expression, while another is looking for clarification about vocabulary, register, or usage. Perhaps one member a much better mastery of a foreign language than another. In my experience, rating answers often leads to rewarding oversimplification and dumbed-down explanations. I wouldn't want to encourage that dynamic here on WR. I am not saying that it would be impossible to implement a functional and useful system that made it easier to find high-quality threads. But I do not have the answer for how to do it, and I am not sold on the "Y/N" rating system.
    Most of the major online retailers that allow customers to review products and services include a "was this review/comment useful? Y/N" button so that readers can give their opinion on reviews. A number of major English online Q&A services do the same. No, I don't care to give examples, because I don't want my comments to be interpreted as an endorsement (or a criticism) of any particular site.

    I support any advances that improve these forums. I encourage anyone who has the time and inclination to think about such advances to communicate their ideas. Then it is up to our site administrator, Mike Kellogg, to vet these ideas. The transition from conception to implementation is not automatic, and many ideas that initially sound good are in the end imperfect or problematic on some level (which may or may not be technical) that will become apparent at some point in further conception, development, or implementation. Thinking about those things is our administrator's job. One of the reasons he created this C&S forum -- which he monitors regularly -- is so that he can glean ideas and suggestions about how to improve the site from members like yourself.

    I do not -- and do not intend to -- spend significant time thinking about how to implement a rating system. As a volunteer moderator, that is not one of my personal priorities here on WR.

    I respectfully disagree. Speaking from my own personal point of view as a moderator, I do not spend significant time searching the archives of the forums. I spend most of my time on threads that are currently active. Only when those threads lead me back to older discussions to I end up diving into the archives. This does not mean that I am not perfectly happy to clean up and improve the archives! But there is no project to go through the hundreds of thousands of old threads in a systematic manner. Instead, any clean-up I do is spot-wise. How else should I know that a certain topic would benefit from clean-up if not because a member who was searching the archives has reported it and requested that we take a look at some of the duplicate or poorly titled threads? You are certainly in a position to alert the moderators to work that requires our attention. Let me assure you that we will simply ignore your request for archive clean-up if we are swamped with more urgent problems in currently active threads. :p

    As I suggested before, removing threads is not a clear-cut matter. Remember that users who search our archives have many, many different purposes and objectives. WR is not just an interactive dictionary. It is also an incredible, searchable database of language in use. Even an unanswered thread -- perfectly useless to you if you were hoping to find the translation of that particular term -- may in fact be very useful to another member who was not actually interested in the translation question, but was instead searching for e.g., usage examples of a vocabulary word that the poster happened to include in his question, or syntax examples of a structure that appears as part of the poster's message.

    I don't mean to sound negative. I want to thank you for the time you've spent thinking about this, and for your obvious dedication to making these forums a better place! :) I simply mean to say that there are no easy answers. The reason this is still an issue is because it is not an easy one to solve, not because we haven't thought about it (as Panj said). There is a tendency to assume that things which exist in electronic format can be tweaked easily -- the "just get the programmer to fix X" mentality -- but actually, changes to electronic entities require just as much planning as changes to physical institutions (if not more). Mike runs these sites. It's how he makes his living. And he spends all day every workday thinking about, creating, testing, and implementing features that improve the functionality of WR. It is a full-time job. He has many projects at many stages of development, including many on the back burner. It's not that improving archive searchability doesn't matter. It's just that I wouldn't want you to be frustrated by the absence of a quick fix or immediate changes in response to your comments. I don't think I have much more to say on the matter. :)
     
  21. EStjarn

    EStjarn Senior Member

    Spanish
    Jann, the reason I asked those questions was because it's difficult to be convinced by, or argue against, an opinion or belief when we don't know what it's based on. Thank you for taking your time to reply. I caught the message and will refrain from asking further questions. However, I'd like to comment on some of the points you're making.

    A single opinion will perhaps not make a difference here, or two, or even five. But if we get, say, ten opinions, a pattern should begin to appear, which eventually - as more opinions are added - can be understood as a consensus. Say we've received 100 opinions, such a pattern might be, for example, 50/50, or 5/95, or 95/5. Each ratio would represent a particular consensus, and that consensus would express more than a simple 'yes' or 'no'.

    You're implying that by giving our opinion regarding the usefulness of a thread, we would be "rewarding oversimplification and dumbed-down explanations." I had hoped we would be rewarding simplicity - the simplicity needed to pull a thing like this off - and more likely the completeness of explanations.

    I'd like to point out that the rating feature is meant as a complement to, not a substitute for, the present feature for determining whether a thread might be informative or not, i.e. its thread title. If the thread title isn't relevant, no rating figures will be able to convince users to open the thread. In this sense, the rating is secondary to the title.

    Another thing is that one could easily argue that a list of threads without ratings represents the epitome of oversimplification, since it gives the impression that all threads are of equal use, an opinion I think most users would hesitate to endorse.

    I cannot immediately see why 'dumbed-down explanations', i.e. explanations written for a less educated or less sophisticated audience, would get a higher rating than more serious or complete explanations. People interested in questions regarding language have by definition an intellectual vein, which in turn should generally mean an appreciation of informative texts.

    If the aim is to preserve every thread that is not unquestionably useless, then the relevance of a rating system should be even greater as the lists of threads will become increasingly longer with time, and users will have even greater difficulties in avoiding threads that don't contain the information they're looking for. (My personal threshold seems to be three such threads: if I can't find what I'm looking for within three tries, I seek another solution.)

    What I meant was, it's the moderators who end up doing these things - combining threads, improving thread titles, and removing threads - as no member has the authority to do that; they can only make suggestions. So, if we were to use these methods instead of a rating system to achieve the kind of searchability that I'm suggesting, the workload would probably overwhelm the moderators. In other words, I'm not speaking of how things are, but what they would be like.

    By opposing the idea you're contributing to a more thorough review of it, which I for one appreciate. I'm not asking for hasty decisions. Any foreseeable problem should be addressed before a decision is made. But each side of the argument has to be given a fair chance. By discussing openly like this, you are contributing to that.
     
    Last edited: Mar 14, 2011

Share This Page