|
Maybe including a weighted standard deviation of the vote could be added in. If someone has a high weighted standard deviation on their vote, then it could possibly indicate a wide range of votes, or if the weighted standard deviation is small, then the range is smaller and possibly a more true indicator.
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
That is nonsense to me. Standard deviation is used to measure the agreement between many people
on one issue, and not the agreement of opinions of one person on many different issues.
If I were to vote 5 (or 1 as the Univoter seems to do) on every article, then that does not offer
any information, hence it has no value; whereas if I vote 1 on half of the articles, and 5 on
the other half, I am offering a lot of information, and I am clearly telling I like some and
dislike some, so there is then no reason to moderate my votes.
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
Luc Pattyn wrote: That is nonsense
Not so. I think it would be helpful in casting out the junk votes that the Univoter likes to cast. If you know the standard deviation, then you have an idea of the distribution of the votes.
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
Paul Conrad wrote: ... if the weighted standard deviation is small, then the range is smaller and possibly a more true indicator. <blockquote>>
The univoter's deviation is zero, you can't get it any smaller; I am pretty sure his votes
don't give a true indication of the articles' values.
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
That's true. I was just trying to get some ideas spinning to find a way to get a decent sample of votes that could disregard univotes. Maybe some feature the author can see a breakdown of each vote by the vote and membership level ( no revealing of the actual member being allowed )?
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
I just got done fiddling around with the idea in Excel, and it's probably best to leave the voting system the way it is 
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
eliminating the extremes (known as "modified Bucholtz" to some), what I explained in the message
that got you started on standard deviations, is the best remedy known to me for the current
anomalies and complaints, but it only works when there is a sufficient number of votes (at least 5 or so).
There is no solution for avoiding the first vote, whether considered too favorable or too
unfavorable.
I'll add another suggestion to this thread in a separate message.
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
Luc Pattyn wrote: at least 5 or so
Yes. You couldn't really use the modified Bucholtz when there are too few votes. On the other hand, when you get articles where there are 50+ votes, it's pretty pointless.
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
Do you mean something like tossing out the outliers? Throw out one 1 vote and one 5 vote? There is always the possibility of getting a 1 vote or a 5 vote from a person who just doesn't know what they are voting, just some random moron 
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
I agree with you. Badly voting without an apparent reason is really weird.
If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler.
-- Alfonso the Wise, 13th Century King of Castile.
[my articles]
|
|
|
|
|
The ANZAC wrote: vote <= 3 for an article you should have to give a reason why your voting this low
This has been suggested many times over and over.
The ANZAC wrote: How can you improve something if people don't tell you whats wrong?
I think that is what the forum at the bottom of the article should be used for. Maybe there could be a way to take the low vote and allow the voter to be forced to leave a message in the forum without revealing who he or she is, since leaving a message doesn't necessarily mean you voted.
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
I'll try another suggestion here:
1. add a short radio-buttoned list (with mutually exclusive items, unchecked by default),
asking for the most relevant comment:
- bad formatting
- too short
- unintelligable
- irrelevant
- plain wrong
- fine
- great
2. disable the vote button as long as nothing got checked.
3. show the percentages of the selected comments (optimization: only to the author, and the
people that already voted on the article).
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
I agree on that one. I don't see why people would object. There is nothing overly revealing about the voter with this approach. The radio buttons/checkboxes can then be a tool for the author to improve him/herself.
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
Yes this could be displayed much like the weekly poll, as a sort of article profile. That way you would know where you have to improve your article. I only suggest one more item, Other with a textbox to write in.
|
|
|
|
|
The ANZAC wrote: I only suggest one more item, Other with a textbox to write in.
Only problem about that, is some idiot could fill it in with gibberish...
"I guess it's what separates the professionals from the drag and drop, girly wirly, namby pamby, wishy washy, can't code for crap types." - Pete O'Hanlon
|
|
|
|
|
I am in favor of keeping it simple, predictable and easy to fill in. So my layout would just
double the row of radio buttons: the first row would contain "Most relevant remark: " followed
by the radio buttons and texts, the second row would be the actual voting row as it is now.
When the viewer has voted (or is the author), I would replace the radio buttons in the first
row by the current percentages. And I would not provide an "other" category; if none seem
appropriate, the voter would still have to select an existing one, but he always can add a
message.
The net result is:
- it takes almost no extra space
- it does not disrupt the current layout by much
- hence it should have a high acceptance
High acceptance is essential, I do not want to reduce the probability a vote is cast.
Voting is the primary goal, getting info as to why a low vote is cast is secondary to that.
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
|
![Rose | [Rose]](https://codeproject.global.ssl.fastly.net/script/Forums/Images/rose.gif)
Luc Pattyn [Forum Guidelines] [My Articles]
This month's tips:
- before you ask a question here, search CodeProject, then Google;
- the quality and detail of your question reflects on the effectiveness of the help you are likely to get;
- use PRE tags to preserve formatting when showing multi-line code snippets.
|
|
|
|
|
I disagree. You don't always vote on an article intending it as feedback to the author. You can vote for an article as an attempt at telling others that it's good or bad.
So if I see an article that is flawed (in my opinion), I might give it a 1 or a 2 so that other people are forewarned. I may not feel inclined to spend 3-4 minutes explaining why I voted low.
Similarly I might see an article I really like and I might give it a 5. But there's no reason for me to post a big message praising the article and the author.
The other thing is that a lot of people vote blindly. Sometimes you might find a not so useful article from a popular author that's got two dozen 5-votes within a day of posting and you can be sure that most of the votes were from people who did not read the article.
It can work the other way too. If you are an unpopular Soapboxer for instance, you can expect some low votes. If you are from a commonly despised country (or your name suggests it), you might get voted down.
Too many factors go into how people vote.
I understand your pain, but insisting that people need to justify their votes with a message is plain ridiculous.
|
|
|
|
|
But that's exactly where the current system comes unstuck. The voting system, in my opinion, should be geared to cater for both author and reader. A reader wants to have a good article and a writer wants to be able to provide it. If an author is providing bad or flawed articles then they deserve to know. In this way there is collective improvement, the quality of articles is increased. If the author then chooses to do nothing it is at his detriment. Correct me if i'm wrong but the whole purpose of this website is collective improvement. An online community where everyone can benefit, including authors.
I don't know if you read the rest of the posts or just the first one but you should, the current system we are all agreed on does not require a message just a simple radiobutton click.
You Click 2 then you click 'bad formatting' or something. I don't think that's terrible difficult, it's certainly not gonna waste any precious time, and its a simple and effective way to indicate to the author where he could improve. Frankly, if an uthor has gone to the trouble of trying to provide other people with useful information, he deserves to know if, and where, he's going wrong.
The worst way for people to vote is when they are voting on something negatively just because its not what they were looking for.
Nishant Sivakumar wrote: But there's no reason for me to post a big message praising the article and the author
Agreed, but praise should be given where its due as should constructive criticism.
Nishant Sivakumar wrote: If you are from a commonly despised country
As for this, i have personally never judged anyone on a forum based on their name or provenance and believe it is utterly detestable, yet i'm sure it occurs.
|
|
|
|
|
Nishant Sivakumar wrote: I may not feel inclined to spend 3-4 minutes explaining why I voted low.
Also, 3-4 minutes might not be enough, if the article really rates a 1 vote. Lately I have seen several articles where the author has not even replaced the boilerplate text. A 1 is too good for these articles - they deserve some large negative number.
|
|
|
|
|
Voted: 1
Reason: asdf
Otherwise [Microsoft is] toast in the long term no matter how much money they've got. They would be already if the Linux community didn't have it's head so firmly up it's own command line buffer that it looks like taking 15 years to find the desktop.
-- Matthew Faithfull
|
|
|
|
|
As a matter of fact, a rating of 1-2 on a *decent* article can't be justified, and it's done anyway. Accordingly, people won't give you positive feedback, because it's not about that at all.
Maybe it's just my rotten idea of man, but I think people vote in that area (let's forget the very few horrible articles here on CP) for 3 reasons:
- they asked too many stupid questions, expecting to get everything spoon-fed, and cannot deal with the fact that you don't see yourself as a 24-7 support center.
- they have a competing article - possibly just published -, that is about to drop out from the front page because of your work.
- they have severe problems that are not all related to your article. 
I'd say the *valuable* feedback comes from people that vote in the area of upper 3s to 5s. These readers have read your article, are interested in your work and appreciate that you've put a lot of energy into your project. The intention here is not to hurt you, but to help you to improve your article (accordingly, they might well wait with their final rating if the there are a few flaws).
Just my 0.02$
Cheers!
modified on Monday, February 04, 2008 2:57:15 PM
|
|
|
|
|
There is another reason i've seen pele vote bad. They go to the article expecting one thing and because the article isn't exactly what they wanted suddenly they think its a crap article.
|
|
|
|
|
Here's one [^]worthy of being turfed. Thanks.
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
