| Pages in topic: [1 2] > |
'Your confidence level' - can we get rid of it? Thread poster: Kirill Semenov
|
|---|
Kirill Semenov 
Ukraine
Local time: 22:25
Member (2004)
English to Russian
+ ...
Ah, yet another case when I send my answer just to see this reminder that the `confidence level' is required...
Am I confident? Never! There is always a chance that I'm wrong.
Can I evaluate my confidence level? It's more than two years ago I started to input "3". I can't evaluate it, sorry.
I see no sense in it. Even if I "5" points confident, does it help the asker anyhow? Do we need it? How specifically it helps? For me, "5"-rated indicator means the ask... See more Ah, yet another case when I send my answer just to see this reminder that the `confidence level' is required...
Am I confident? Never! There is always a chance that I'm wrong.
Can I evaluate my confidence level? It's more than two years ago I started to input "3". I can't evaluate it, sorry.
I see no sense in it. Even if I "5" points confident, does it help the asker anyhow? Do we need it? How specifically it helps? For me, "5"-rated indicator means the asker is over-confident, that's the only use.
And it's a real hindrance when answering to kudoZ, an extra field which has nothing to do with translating skill or the relevance of my answer.
[Edited at 2006-09-12 12:43]
[Subject edited by staff or moderator 2006-09-12 15:00] ▲ Collapse
| | | |
Samuel Murray
Netherlands
Local time: 21:25
Member (2006)
English to Afrikaans
+ ...
| I used to say "4" | Sep 12, 2006 |
Kirill Semenov wrote:
Am I confident? Never! There is always a chance that I'm wrong.
I used to say "4" when I'm very sure, but then I saw other translators go for "5" even if they're wrong, so why bother being honest about it? I'll say "2" when I'm not sure, though.
| | | |
I fail to see any practical use for the "confidence level." It is an utterly subjective thing that merely reflects the answerer's attitude to his/her own translation skill. Let the proof be in the pudding.
| | | |
| I think it can be useful | Sep 12, 2006 |
There was a time when I thought the same way, but now I've changed my mind and would rather keep it. I have a tendency not to use confidence levels of above 3 unless I'm *really*, *really* sure, but I think they can be useful.
If someone wants to make it clear that their answer is more of a suggestion than a definite "yes, I know this one for a fact", then I think it's a good idea for them to be given the chance to indicate this with a lower confidence level. This applies particular... See more There was a time when I thought the same way, but now I've changed my mind and would rather keep it. I have a tendency not to use confidence levels of above 3 unless I'm *really*, *really* sure, but I think they can be useful.
If someone wants to make it clear that their answer is more of a suggestion than a definite "yes, I know this one for a fact", then I think it's a good idea for them to be given the chance to indicate this with a lower confidence level. This applies particularly to questions where there may not be a specific "correct" answer or term, e.g. marketing/literary questions, but there may also be other reasons why you can't be too confident: lack of context, maybe you suspect there's an error/typo in the source text, etc. So this way, the asker knows it's nothing more than a suggestion and can take it at face value. I know that when I ask questions and get an answer from someone with a lowish confidence level, I appreciate their honesty (and these answers often turn out to be helpful). If an answer comes with a confidence level of 5 but no real explanation, it can sometimes be difficult to know how to take it, especially when the answerer "specialises in field": should I believe them because they're a specialist and they're sure, or are they just assuming it's right? On the other hand, if you see an answer with a lowish confidence level and little or no explanation, at least you know it's just a suggestion and can then treat it as such.
Conversely, if someone is actually very familiar with the term and has translated it before many times, as opposed to someone who has never seen it before, I think it's fair enough for them to be able to indicate this with a higher level of confidence - although admittedly I think it would also be nice if they could actually say this somewhere in their answer too.
The other thing is that although I've sometimes seen people using very high confidence levels even for answers which are manifestly wrong - and that certainly isn't helpful to anyone - before long, you begin to spot the people who make a habit of doing this and can learn to watch out for it. In a way, then, I would argue that the confidence level does tell you something more than just an answer on its own would.
If we're looking for fields we can do without, how about the "target dialect" one for English answers (which is always United Kingdom/United States - why not Australian/NZ/Irish etc., since some clients are based in those countries?) I have to select one of these every time and now and then I have seen answers rejected for being (and I quote) "too British" when the asker hasn't bothered to say they want US English (in fact, askers rarely seem to do so). In many cases it makes no difference to the answer anyway, but what I really don't understand is why this information is required at all when it seems to disappear immediately into a void. And what about the "I think this is a good question" button - what's the point of it? I've tried using it, but nothing ever seems to happen.
[Edited at 2006-09-12 16:45] ▲ Collapse
| | |
|
|
|
| It's not a Bad Thing in itself, it's just misused | Sep 12, 2006 |
I think it's useful when it's used properly by the Answerer.
The trouble is that it is used as criterion for ranking the answers at the top of the page. So the temptation exists for people to misuse it to get their answer to appear at the top of the list. The trouble is, that in Fr-Eng at least, these answers often leave a little to be desired.
I would suggest that the ranking in the list at the top of the question page is based purely on the number of agrees and then t... See more I think it's useful when it's used properly by the Answerer.
The trouble is that it is used as criterion for ranking the answers at the top of the page. So the temptation exists for people to misuse it to get their answer to appear at the top of the list. The trouble is, that in Fr-Eng at least, these answers often leave a little to be desired.
I would suggest that the ranking in the list at the top of the question page is based purely on the number of agrees and then the time posted when net agrees are equal.
Furthermore, it might be useful if "5" was disabled for people who have not said that they work in the field (and possibly if the target language is not declared as a mother tongue?).
But in and of itself, it's not a bad idea - it saves those who actually use it properly having to explain that a suggested answer is just a shot in the dark, or a possible answer but they're not certain, or indeed, that they are fairly sure they're right... ▲ Collapse
| | | |
Monika Coulson
Local time: 13:25
Member (2001)
English to Albanian
+ ...
SITE LOCALIZER | Agree with Peter and Charlie above | Sep 12, 2006 |
I also think it can be useful. IMO, the asker has a right to know whether the answerer is guessing or not (or even how much is guessing.) Putting a level 5 there does not make an answer correct; it only shows that the answerer thinks he is sure or not so sure and that's it. If used correctly, it can be a helpful tool IMO.
Monika
[Redaktuar më 2006-09-12 14:39]
| | | |
KathyT
Australia
Local time: 05:25
Japanese to English
| Don't really use it... | Sep 12, 2006 |
I just put a "1" for every answer by default, even when I'm really sure I'm right. Meh!
I think I started doing this when I found myself getting annoyed by "serial 5ers," who repeatedly posted 5's for every single answer, even when they were clearly guessing or fabricating the answer, sometimes even stating in the body of their answer "just guessing!!"
| | | |
Kirill Semenov
Ukraine
Local time: 22:25
Member (2004)
English to Russian
+ ...
TOPIC STARTER | Subjective plus objective | Sep 12, 2006 |
Yes, this was another reason why I had critisized the idea of `confidence level' in the past. It's adding a subjective opinion of your own to your peers' opinion which sounds strange to me. I would leave only the peer comments, seems enough to me, personally.
[Edited at 2006-09-12 14:58]
| | |
|
|
|
Heinrich Pesch
Finland
Local time: 22:25
Member (2003)
Finnish to German
+ ...
instead of the subjective confidence level a figur that expresses, how "good" the answerer is, how many times his/her answers have been chosen before. I know, this could be misused too.
I do not seem to trust answerers who do not have their portret uploaded, regardless of the confidence level.
But alas, my questions usually get one answer, if any at all.
[Bearbeitet am 2006-09-12 14:58]
| | | |
craigs
Local time: 15:25
English to Portuguese
+ ...
| Peer agreement is stronger than Confidence level | Sep 12, 2006 |
A person's confidence level doesn't matter so much as peer agreement.
| | | |
Özden Arıkan
Germany
Local time: 21:25
Member
English to Turkish
+ ...
| Depends on how one views Kudoz | Sep 12, 2006 |
If it's viewed as a competition, it might look weird, of course, as if competitors rating themselves. But if it's a help network -and I hope it is- then stating your confidence level is part of helping the asker. True, it's not appropriately used by some, and even abused by others. But the solution is at taking the more difficult way, as in everything else: educating -if this word is appropriate- people of its use and function, as others have told. We don't get rid of cars because we have accide... See more If it's viewed as a competition, it might look weird, of course, as if competitors rating themselves. But if it's a help network -and I hope it is- then stating your confidence level is part of helping the asker. True, it's not appropriately used by some, and even abused by others. But the solution is at taking the more difficult way, as in everything else: educating -if this word is appropriate- people of its use and function, as others have told. We don't get rid of cars because we have accidents, do we?
Or, was your suggestion rather rhetorical, Kirill, to draw attention to the issue of misused confidence level?
[Edited at 2006-09-12 16:26] ▲ Collapse
| | | |
Kirill Semenov
Ukraine
Local time: 22:25
Member (2004)
English to Russian
+ ...
TOPIC STARTER | A default setting | Sep 12, 2006 |
Özden Arıkan wrote:
Or, was your suggestion rather rhetorical, Kirill, to draw attention to the issue of misused confidence level?
Not, in fact, if someone finds the option useful, it's OK, but I propose to set it by default on 3 or 2, or to a customized value. It's just not ergonomic, as for me, and it's annoying to select the value more and more again for each answer, and when you forget it, the system warns you - and so on. Really annoying. I've lost some 2-page long answers because of it.
| | |
|
|
|
Monika Coulson
Local time: 13:25
Member (2001)
English to Albanian
+ ...
SITE LOCALIZER | Saving the answer, even when we forget to set the confidence level | Sep 12, 2006 |
Kirill Semenov wrote:
Really annoying. I've lost some 2-page long answers because of it.
Ouch! Now, this is upsetting. I think this should be programmed in a way that the answer should still be saved even if we forget to set the confidence level.
Monika
| | | |
Kirill Semenov
Ukraine
Local time: 22:25
Member (2004)
English to Russian
+ ...
TOPIC STARTER | No, of course :) | Sep 12, 2006 |
Monika Coulson wrote: Kirill Semenov wrote:
Really annoying. I've lost some 2-page long answers because of it. Ouch! Now, this is upsetting. I think this should be programmed in a way that the answer should still be saved even if we forget to set the confidence level.
Monika, no, I cannot blame the ProZ system, it keeps the answer allright in cases like this, but it was like: typing an answer, clicking "Send", then closing my Web browser. After that the answer was lost because of the confidence menu.
| | | |
Özden Arıkan
Germany
Local time: 21:25
Member
English to Turkish
+ ...
| This should no longer be a problem with the added Preview function | Sep 12, 2006 |
Kirill Semenov wrote:
Monika, no, I cannot blame the ProZ system, it keeps the answer allright in cases like this, but it was like: typing an answer, clicking "Send", then closing my Web browser. After that the answer was lost because of the confidence menu.
[Edited at 2006-09-12 17:16]
| | | |
| Pages in topic: [1 2] > |
Login to reply/comment 
