Here's a graph of the retention rates of new editors in English Wikipedia.
https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_ retention.survival_proportion.svg
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the *12 hour *new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_ retention.survival_proportion.svg
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
From: Analytics [mailto:analytics-bounces@lists.wikimedia.org] On Behalf Of Neil Patel Quinn Sent: Friday, March 31, 2017 23:06 To: A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. Subject: Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the 12 hour new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.surviv...
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
_______________________________________________ Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Erik,
I appreciate pushing back on just looking for bigger metrics, but there's something more important when it comes to measuring people who contribute at least a little bit. Our licensing asserts that they must be attributed. After all, they have contributed something.
Also, for your astronomy comparison, this would be more like saying that anyone who contributes to publicly recorded astronomy observations is an astronomer -- even if they have only done so once. In my estimation, that doesn't sound crazy. Your comparison to "looking at the night sky" is a lot more like reading Wikipedia.
-Aaron
On Tue, Apr 11, 2017 at 6:35 AM, Erik Zachte ezachte@wikimedia.org wrote:
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Neil Patel Quinn *Sent:* Friday, March 31, 2017 23:06 *To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the *12 hour *new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
*https://commons.wikimedia.org/ https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.survival_proportion.svg*
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
--
Neil Patel Quinn https://meta.wikimedia.org/wiki/User:Neil_P._Quinn-WMF, product analyst Wikimedia Foundation
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Aaron,
Yeah my analogy is arguably imprecise.
And for your analogy, you assume that the public astronomy database is guarded Nupedia style, with credentials. Could be, explicit mention of this assumption would resolve ambiguity ;-)
Our licensing asserts that they must be attributed.
Sure these people who did one edit must be attributed whenever the page they edited is published somewhere else.
But do we ever do that for real these days? Seems like a dead clause from a distant past, expect for our onwiki history page.
Also giving credit is something else than counting, and publishing that count as some meaningful metric (not saying that you want to do that, but others will find the factoid and run with it)
We can discuss semantics. But when a person writes one word a year we wouldn't call that person a 'writer', do we?
Words lose their meaning if their definition is stretched in extremo, beyond common sense, beyond what any audience assumes those words mean.
Long ago we found that a huge amount of registered users made not even one edit.
One explanation might be that many people habitually sign up, just out of habit. Or that they want to tweak the UI (e.g. red links in preferences).
My point: count as you like, but could we avoid using a term with so many connotations for these edge cases, so as not to confuse people even more about our metrics?
Erik
From: Analytics [mailto:analytics-bounces@lists.wikimedia.org] On Behalf Of Aaron Halfaker Sent: Tuesday, April 11, 2017 16:55 To: A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. Subject: Re: [Analytics] Fwd: follow-up on editors
Erik,
I appreciate pushing back on just looking for bigger metrics, but there's something more important when it comes to measuring people who contribute at least a little bit. Our licensing asserts that they must be attributed. After all, they have contributed something.
Also, for your astronomy comparison, this would be more like saying that anyone who contributes to publicly recorded astronomy observations is an astronomer -- even if they have only done so once. In my estimation, that doesn't sound crazy. Your comparison to "looking at the night sky" is a lot more like reading Wikipedia.
-Aaron
On Tue, Apr 11, 2017 at 6:35 AM, Erik Zachte ezachte@wikimedia.org wrote:
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
From: Analytics [mailto:analytics-bounces@lists.wikimedia.org] On Behalf Of Neil Patel Quinn Sent: Friday, March 31, 2017 23:06 To: A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. Subject: Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the 12 hour new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.surviv...
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
_______________________________________________ Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
If we're going to have a conversation about terminology, I would like to drop the terms "active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month". There are multiple ways of measuring productivity, and I'm wary of the amount of prominence that's given to the number of edits as the primary metric of productivity. Also, I don't think it's clear to analytics nebwbies that "active editor" is a term with a specific definition rather than a general description of people who edit "actively" (whatever that means). I'm fine with using 5+ edits per month and 100+ edits per month as measures of productivity, but I would prefer to drop the terms "active editor" and "very active editor". I'd also like to see more prominence given to other metrics such as bytes changed and logged non-edit actions.
Pine
On Tue, Apr 11, 2017 at 10:22 AM, Erik Zachte ezachte@wikimedia.org wrote:
Aaron,
Yeah my analogy is arguably imprecise.
And for your analogy, you assume that the public astronomy database is guarded Nupedia style, with credentials. Could be, explicit mention of this assumption would resolve ambiguity ;-)
Our licensing asserts that they must be attributed.
Sure these people who did one edit must be attributed whenever the page they edited is published somewhere else.
But do we ever do that for real these days? Seems like a dead clause from a distant past, expect for our onwiki history page.
Also giving credit is something else than counting, and publishing that count as some meaningful metric (not saying that you want to do that, but others will find the factoid and run with it)
We can discuss semantics. But when a person writes one word a year we wouldn't call that person a 'writer', do we?
Words lose their meaning if their definition is stretched in extremo, beyond common sense, beyond what any audience assumes those words mean.
Long ago we found that a huge amount of registered users made not even one edit.
One explanation might be that many people habitually sign up, just out of habit. Or that they want to tweak the UI (e.g. red links in preferences).
My point: count as you like, but could we avoid using a term with so many connotations for these edge cases, so as not to confuse people even more about our metrics?
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Aaron Halfaker *Sent:* Tuesday, April 11, 2017 16:55
*To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Erik,
I appreciate pushing back on just looking for bigger metrics, but there's something more important when it comes to measuring people who contribute at least a little bit. Our licensing asserts that they must be attributed. After all, they have contributed something.
Also, for your astronomy comparison, this would be more like saying that anyone who contributes to publicly recorded astronomy observations is an astronomer -- even if they have only done so once. In my estimation, that doesn't sound crazy. Your comparison to "looking at the night sky" is a lot more like reading Wikipedia.
-Aaron
On Tue, Apr 11, 2017 at 6:35 AM, Erik Zachte ezachte@wikimedia.org wrote:
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Neil Patel Quinn *Sent:* Friday, March 31, 2017 23:06 *To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the *12 hour *new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
*https://commons.wikimedia.org/ https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.survival_proportion.svg*
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
--
Neil Patel Quinn https://meta.wikimedia.org/wiki/User:Neil_P._Quinn-WMF, product analyst Wikimedia Foundation
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Maybe looping back a little to Aaron's original question which I'm guessing is from Shannon: we don't have a way to measure " Number of editors who contribute 1 edit per month" as we don't have ways to accurately identify people who use multiple accounts, IPs, etc. We do have ways to track number of unique accounts, but that's different from number of unique editors. Although it would be nice to know how many people edit the projects on any given month, that's impossible to know, although maybe Nuria and the other analytics folks would have some ideas on how to get an approximation.
Pine
On Tue, Apr 11, 2017 at 9:48 PM, Pine W wiki.pine@gmail.com wrote:
If we're going to have a conversation about terminology, I would like to drop the terms "active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month". There are multiple ways of measuring productivity, and I'm wary of the amount of prominence that's given to the number of edits as the primary metric of productivity. Also, I don't think it's clear to analytics nebwbies that "active editor" is a term with a specific definition rather than a general description of people who edit "actively" (whatever that means). I'm fine with using 5+ edits per month and 100+ edits per month as measures of productivity, but I would prefer to drop the terms "active editor" and "very active editor". I'd also like to see more prominence given to other metrics such as bytes changed and logged non-edit actions.
Pine
On Tue, Apr 11, 2017 at 10:22 AM, Erik Zachte ezachte@wikimedia.org wrote:
Aaron,
Yeah my analogy is arguably imprecise.
And for your analogy, you assume that the public astronomy database is guarded Nupedia style, with credentials. Could be, explicit mention of this assumption would resolve ambiguity ;-)
Our licensing asserts that they must be attributed.
Sure these people who did one edit must be attributed whenever the page they edited is published somewhere else.
But do we ever do that for real these days? Seems like a dead clause from a distant past, expect for our onwiki history page.
Also giving credit is something else than counting, and publishing that count as some meaningful metric (not saying that you want to do that, but others will find the factoid and run with it)
We can discuss semantics. But when a person writes one word a year we wouldn't call that person a 'writer', do we?
Words lose their meaning if their definition is stretched in extremo, beyond common sense, beyond what any audience assumes those words mean.
Long ago we found that a huge amount of registered users made not even one edit.
One explanation might be that many people habitually sign up, just out of habit. Or that they want to tweak the UI (e.g. red links in preferences).
My point: count as you like, but could we avoid using a term with so many connotations for these edge cases, so as not to confuse people even more about our metrics?
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Aaron Halfaker *Sent:* Tuesday, April 11, 2017 16:55
*To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Erik,
I appreciate pushing back on just looking for bigger metrics, but there's something more important when it comes to measuring people who contribute at least a little bit. Our licensing asserts that they must be attributed. After all, they have contributed something.
Also, for your astronomy comparison, this would be more like saying that anyone who contributes to publicly recorded astronomy observations is an astronomer -- even if they have only done so once. In my estimation, that doesn't sound crazy. Your comparison to "looking at the night sky" is a lot more like reading Wikipedia.
-Aaron
On Tue, Apr 11, 2017 at 6:35 AM, Erik Zachte ezachte@wikimedia.org wrote:
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Neil Patel Quinn *Sent:* Friday, March 31, 2017 23:06 *To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the *12 hour *new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
*https://commons.wikimedia.org/ https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.survival_proportion.svg*
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
--
Neil Patel Quinn https://meta.wikimedia.org/wiki/User:Neil_P._Quinn-WMF, product analyst Wikimedia Foundation
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
I would like to drop the terms "active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month"
I totally agree.
On Wed, Apr 12, 2017 at 12:07 AM, Pine W wiki.pine@gmail.com wrote:
Maybe looping back a little to Aaron's original question which I'm guessing is from Shannon: we don't have a way to measure " Number of editors who contribute 1 edit per month" as we don't have ways to accurately identify people who use multiple accounts, IPs, etc. We do have ways to track number of unique accounts, but that's different from number of unique editors. Although it would be nice to know how many people edit the projects on any given month, that's impossible to know, although maybe Nuria and the other analytics folks would have some ideas on how to get an approximation.
Pine
On Tue, Apr 11, 2017 at 9:48 PM, Pine W wiki.pine@gmail.com wrote:
If we're going to have a conversation about terminology, I would like to drop the terms "active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month". There are multiple ways of measuring productivity, and I'm wary of the amount of prominence that's given to the number of edits as the primary metric of productivity. Also, I don't think it's clear to analytics nebwbies that "active editor" is a term with a specific definition rather than a general description of people who edit "actively" (whatever that means). I'm fine with using 5+ edits per month and 100+ edits per month as measures of productivity, but I would prefer to drop the terms "active editor" and "very active editor". I'd also like to see more prominence given to other metrics such as bytes changed and logged non-edit actions.
Pine
On Tue, Apr 11, 2017 at 10:22 AM, Erik Zachte ezachte@wikimedia.org wrote:
Aaron,
Yeah my analogy is arguably imprecise.
And for your analogy, you assume that the public astronomy database is guarded Nupedia style, with credentials. Could be, explicit mention of this assumption would resolve ambiguity ;-)
Our licensing asserts that they must be attributed.
Sure these people who did one edit must be attributed whenever the page they edited is published somewhere else.
But do we ever do that for real these days? Seems like a dead clause from a distant past, expect for our onwiki history page.
Also giving credit is something else than counting, and publishing that count as some meaningful metric (not saying that you want to do that, but others will find the factoid and run with it)
We can discuss semantics. But when a person writes one word a year we wouldn't call that person a 'writer', do we?
Words lose their meaning if their definition is stretched in extremo, beyond common sense, beyond what any audience assumes those words mean.
Long ago we found that a huge amount of registered users made not even one edit.
One explanation might be that many people habitually sign up, just out of habit. Or that they want to tweak the UI (e.g. red links in preferences).
My point: count as you like, but could we avoid using a term with so many connotations for these edge cases, so as not to confuse people even more about our metrics?
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Aaron Halfaker *Sent:* Tuesday, April 11, 2017 16:55
*To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Erik,
I appreciate pushing back on just looking for bigger metrics, but there's something more important when it comes to measuring people who contribute at least a little bit. Our licensing asserts that they must be attributed. After all, they have contributed something.
Also, for your astronomy comparison, this would be more like saying that anyone who contributes to publicly recorded astronomy observations is an astronomer -- even if they have only done so once. In my estimation, that doesn't sound crazy. Your comparison to "looking at the night sky" is a lot more like reading Wikipedia.
-Aaron
On Tue, Apr 11, 2017 at 6:35 AM, Erik Zachte ezachte@wikimedia.org wrote:
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Neil Patel Quinn *Sent:* Friday, March 31, 2017 23:06 *To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the *12 hour *new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
*https://commons.wikimedia.org/ https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.survival_proportion.svg*
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
--
Neil Patel Quinn https://meta.wikimedia.org/wiki/User:Neil_P._Quinn-WMF, product analyst Wikimedia Foundation
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
. We do have ways to track number of unique accounts, but that's different
from number of unique editors. Although it would be nice to know how many people edit >the projects on any given month, that's impossible to know, although maybe Nuria and the other analytics folks would have some ideas on how to get an >approximation. On my opinion (knowing little about ecosystem) I think this matters little. Any online service that relies on authentication (facebook, gmail) deals with accounts rather than individual people. I does not seem that counting "physical beings that edit" versus "accounts that edit" will make the meaning we decipher from numbers such us "5+ edits per month" significantly different.
On Wed, Apr 12, 2017 at 8:28 AM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
I would like to drop the terms "active editor" and "highly active editor"
and replace them with "5+ edits per month" and "100+ edits per month"
I totally agree.
On Wed, Apr 12, 2017 at 12:07 AM, Pine W wiki.pine@gmail.com wrote:
Maybe looping back a little to Aaron's original question which I'm guessing is from Shannon: we don't have a way to measure " Number of editors who contribute 1 edit per month" as we don't have ways to accurately identify people who use multiple accounts, IPs, etc. We do have ways to track number of unique accounts, but that's different from number of unique editors. Although it would be nice to know how many people edit the projects on any given month, that's impossible to know, although maybe Nuria and the other analytics folks would have some ideas on how to get an approximation.
Pine
On Tue, Apr 11, 2017 at 9:48 PM, Pine W wiki.pine@gmail.com wrote:
If we're going to have a conversation about terminology, I would like to drop the terms "active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month". There are multiple ways of measuring productivity, and I'm wary of the amount of prominence that's given to the number of edits as the primary metric of productivity. Also, I don't think it's clear to analytics nebwbies that "active editor" is a term with a specific definition rather than a general description of people who edit "actively" (whatever that means). I'm fine with using 5+ edits per month and 100+ edits per month as measures of productivity, but I would prefer to drop the terms "active editor" and "very active editor". I'd also like to see more prominence given to other metrics such as bytes changed and logged non-edit actions.
Pine
On Tue, Apr 11, 2017 at 10:22 AM, Erik Zachte ezachte@wikimedia.org wrote:
Aaron,
Yeah my analogy is arguably imprecise.
And for your analogy, you assume that the public astronomy database is guarded Nupedia style, with credentials. Could be, explicit mention of this assumption would resolve ambiguity ;-)
Our licensing asserts that they must be attributed.
Sure these people who did one edit must be attributed whenever the page they edited is published somewhere else.
But do we ever do that for real these days? Seems like a dead clause from a distant past, expect for our onwiki history page.
Also giving credit is something else than counting, and publishing that count as some meaningful metric (not saying that you want to do that, but others will find the factoid and run with it)
We can discuss semantics. But when a person writes one word a year we wouldn't call that person a 'writer', do we?
Words lose their meaning if their definition is stretched in extremo, beyond common sense, beyond what any audience assumes those words mean.
Long ago we found that a huge amount of registered users made not even one edit.
One explanation might be that many people habitually sign up, just out of habit. Or that they want to tweak the UI (e.g. red links in preferences).
My point: count as you like, but could we avoid using a term with so many connotations for these edge cases, so as not to confuse people even more about our metrics?
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Aaron Halfaker *Sent:* Tuesday, April 11, 2017 16:55
*To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Erik,
I appreciate pushing back on just looking for bigger metrics, but there's something more important when it comes to measuring people who contribute at least a little bit. Our licensing asserts that they must be attributed. After all, they have contributed something.
Also, for your astronomy comparison, this would be more like saying that anyone who contributes to publicly recorded astronomy observations is an astronomer -- even if they have only done so once. In my estimation, that doesn't sound crazy. Your comparison to "looking at the night sky" is a lot more like reading Wikipedia.
-Aaron
On Tue, Apr 11, 2017 at 6:35 AM, Erik Zachte ezachte@wikimedia.org wrote:
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Neil Patel Quinn *Sent:* Friday, March 31, 2017 23:06 *To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the *12 hour *new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker < ahalfaker@wikimedia.org> wrote:
*https://commons.wikimedia.org/ https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.survival_proportion.svg*
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker < ahalfaker@wikimedia.org> wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
--
Neil Patel Quinn https://meta.wikimedia.org/wiki/User:Neil_P._Quinn-WMF, product analyst Wikimedia Foundation
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Hmm. A few points of note:
1. In many circumstances, we don't require users to create accounts and edit with those accounts (unlike Facebook and Gmail. although people may create multiple Facebook and Gmail accounts and may share Facebook and Gmail accounts) 2. Users may have multiple Wikimedia accounts and may contribute many combinations of logged-in and not-logged-in actions. 3. Multiple users may use identical IPs and UAs. 4. Users are less likely to, and are prohibited by policy from, sharing registered accounts.
Maybe instead of asking for " Number of editors who contribute 1 edit per month", perhaps a better metric for Shannon to use would be "Number of unique devices (as far as WMF can determine) that contribute at least 1 edit per month". What do you think, Aaron and Shannon?
Pine
On Wed, Apr 12, 2017 at 8:35 AM, Nuria Ruiz nuria@wikimedia.org wrote:
. We do have ways to track number of unique accounts, but that's
different from number of unique editors. Although it would be nice to know how many people edit >the projects on any given month, that's impossible to know, although maybe Nuria and the other analytics folks would have some ideas on how to get an >approximation. On my opinion (knowing little about ecosystem) I think this matters little. Any online service that relies on authentication (facebook, gmail) deals with accounts rather than individual people. I does not seem that counting "physical beings that edit" versus "accounts that edit" will make the meaning we decipher from numbers such us "5+ edits per month" significantly different.
On Wed, Apr 12, 2017 at 8:28 AM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
I would like to drop the terms "active editor" and "highly active editor"
and replace them with "5+ edits per month" and "100+ edits per month"
I totally agree.
On Wed, Apr 12, 2017 at 12:07 AM, Pine W wiki.pine@gmail.com wrote:
Maybe looping back a little to Aaron's original question which I'm guessing is from Shannon: we don't have a way to measure " Number of editors who contribute 1 edit per month" as we don't have ways to accurately identify people who use multiple accounts, IPs, etc. We do have ways to track number of unique accounts, but that's different from number of unique editors. Although it would be nice to know how many people edit the projects on any given month, that's impossible to know, although maybe Nuria and the other analytics folks would have some ideas on how to get an approximation.
Pine
On Tue, Apr 11, 2017 at 9:48 PM, Pine W wiki.pine@gmail.com wrote:
If we're going to have a conversation about terminology, I would like to drop the terms "active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month". There are multiple ways of measuring productivity, and I'm wary of the amount of prominence that's given to the number of edits as the primary metric of productivity. Also, I don't think it's clear to analytics nebwbies that "active editor" is a term with a specific definition rather than a general description of people who edit "actively" (whatever that means). I'm fine with using 5+ edits per month and 100+ edits per month as measures of productivity, but I would prefer to drop the terms "active editor" and "very active editor". I'd also like to see more prominence given to other metrics such as bytes changed and logged non-edit actions.
Pine
On Tue, Apr 11, 2017 at 10:22 AM, Erik Zachte ezachte@wikimedia.org wrote:
Aaron,
Yeah my analogy is arguably imprecise.
And for your analogy, you assume that the public astronomy database is guarded Nupedia style, with credentials. Could be, explicit mention of this assumption would resolve ambiguity ;-)
Our licensing asserts that they must be attributed.
Sure these people who did one edit must be attributed whenever the page they edited is published somewhere else.
But do we ever do that for real these days? Seems like a dead clause from a distant past, expect for our onwiki history page.
Also giving credit is something else than counting, and publishing that count as some meaningful metric (not saying that you want to do that, but others will find the factoid and run with it)
We can discuss semantics. But when a person writes one word a year we wouldn't call that person a 'writer', do we?
Words lose their meaning if their definition is stretched in extremo, beyond common sense, beyond what any audience assumes those words mean.
Long ago we found that a huge amount of registered users made not even one edit.
One explanation might be that many people habitually sign up, just out of habit. Or that they want to tweak the UI (e.g. red links in preferences).
My point: count as you like, but could we avoid using a term with so many connotations for these edge cases, so as not to confuse people even more about our metrics?
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Aaron Halfaker *Sent:* Tuesday, April 11, 2017 16:55
*To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Erik,
I appreciate pushing back on just looking for bigger metrics, but there's something more important when it comes to measuring people who contribute at least a little bit. Our licensing asserts that they must be attributed. After all, they have contributed something.
Also, for your astronomy comparison, this would be more like saying that anyone who contributes to publicly recorded astronomy observations is an astronomer -- even if they have only done so once. In my estimation, that doesn't sound crazy. Your comparison to "looking at the night sky" is a lot more like reading Wikipedia.
-Aaron
On Tue, Apr 11, 2017 at 6:35 AM, Erik Zachte ezachte@wikimedia.org wrote:
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
*From:* Analytics [mailto:analytics-bounces@lists.wikimedia.org] *On Behalf Of *Neil Patel Quinn *Sent:* Friday, March 31, 2017 23:06 *To:* A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. *Subject:* Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the *12 hour *new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker < ahalfaker@wikimedia.org> wrote:
*https://commons.wikimedia.org/ https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.survival_proportion.svg*
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker < ahalfaker@wikimedia.org> wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
--
Neil Patel Quinn https://meta.wikimedia.org/wiki/User:Neil_P._Quinn-WMF, product analyst Wikimedia Foundation
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics
"active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month"
Putting number in is what Wikistats did all the time, see e.g. https://stats.wikimedia.org/EN/Sitemap.htm
It might be helpful to stick to that on new stats pages.
Still the more generic terms stuck over so many years, mainly in prose/discussions.
The terms with figures in it are somewhat more explanatory, yet a bit harder to remember.
In general people seem to prefer simple names for core metrics,
This also signals there isn't also a "200+ edits per month" core metric.
Accuracy in naming is always a compromise at best.
Or else we could end up with "5+ edits per month on content namespaces, while logged in"
Erik
From: Analytics [mailto:analytics-bounces@lists.wikimedia.org] On Behalf Of Pine W Sent: Thursday, April 13, 2017 5:40 To: A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. Subject: Re: [Analytics] Fwd: follow-up on editors
Hmm. A few points of note:
1. In many circumstances, we don't require users to create accounts and edit with those accounts (unlike Facebook and Gmail. although people may create multiple Facebook and Gmail accounts and may share Facebook and Gmail accounts)
2. Users may have multiple Wikimedia accounts and may contribute many combinations of logged-in and not-logged-in actions.
3. Multiple users may use identical IPs and UAs. 4. Users are less likely to, and are prohibited by policy from, sharing registered accounts.
Maybe instead of asking for " Number of editors who contribute 1 edit per month", perhaps a better metric for Shannon to use would be "Number of unique devices (as far as WMF can determine) that contribute at least 1 edit per month". What do you think, Aaron and Shannon?
Pine
On Wed, Apr 12, 2017 at 8:35 AM, Nuria Ruiz nuria@wikimedia.org wrote:
. We do have ways to track number of unique accounts, but that's different from number of unique editors. Although it would be nice to know how many people edit >the projects on any given month, that's impossible to know, although maybe Nuria and the other analytics folks would have some ideas on how to get an >approximation.
On my opinion (knowing little about ecosystem) I think this matters little. Any online service that relies on authentication (facebook, gmail) deals with accounts rather than individual people. I does not seem that counting "physical beings that edit" versus "accounts that edit" will make the meaning we decipher from numbers such us "5+ edits per month" significantly different.
On Wed, Apr 12, 2017 at 8:28 AM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
I would like to drop the terms "active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month"
I totally agree.
On Wed, Apr 12, 2017 at 12:07 AM, Pine W wiki.pine@gmail.com wrote:
Maybe looping back a little to Aaron's original question which I'm guessing is from Shannon: we don't have a way to measure " Number of editors who contribute 1 edit per month" as we don't have ways to accurately identify people who use multiple accounts, IPs, etc. We do have ways to track number of unique accounts, but that's different from number of unique editors. Although it would be nice to know how many people edit the projects on any given month, that's impossible to know, although maybe Nuria and the other analytics folks would have some ideas on how to get an approximation.
Pine
On Tue, Apr 11, 2017 at 9:48 PM, Pine W wiki.pine@gmail.com wrote:
If we're going to have a conversation about terminology, I would like to drop the terms "active editor" and "highly active editor" and replace them with "5+ edits per month" and "100+ edits per month". There are multiple ways of measuring productivity, and I'm wary of the amount of prominence that's given to the number of edits as the primary metric of productivity. Also, I don't think it's clear to analytics nebwbies that "active editor" is a term with a specific definition rather than a general description of people who edit "actively" (whatever that means). I'm fine with using 5+ edits per month and 100+ edits per month as measures of productivity, but I would prefer to drop the terms "active editor" and "very active editor". I'd also like to see more prominence given to other metrics such as bytes changed and logged non-edit actions.
Pine
On Tue, Apr 11, 2017 at 10:22 AM, Erik Zachte ezachte@wikimedia.org wrote:
Aaron,
Yeah my analogy is arguably imprecise.
And for your analogy, you assume that the public astronomy database is guarded Nupedia style, with credentials. Could be, explicit mention of this assumption would resolve ambiguity ;-)
Our licensing asserts that they must be attributed.
Sure these people who did one edit must be attributed whenever the page they edited is published somewhere else.
But do we ever do that for real these days? Seems like a dead clause from a distant past, expect for our onwiki history page.
Also giving credit is something else than counting, and publishing that count as some meaningful metric (not saying that you want to do that, but others will find the factoid and run with it)
We can discuss semantics. But when a person writes one word a year we wouldn't call that person a 'writer', do we?
Words lose their meaning if their definition is stretched in extremo, beyond common sense, beyond what any audience assumes those words mean.
Long ago we found that a huge amount of registered users made not even one edit.
One explanation might be that many people habitually sign up, just out of habit. Or that they want to tweak the UI (e.g. red links in preferences).
My point: count as you like, but could we avoid using a term with so many connotations for these edge cases, so as not to confuse people even more about our metrics?
Erik
From: Analytics [mailto:analytics-bounces@lists.wikimedia.org] On Behalf Of Aaron Halfaker Sent: Tuesday, April 11, 2017 16:55
To: A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. Subject: Re: [Analytics] Fwd: follow-up on editors
Erik,
I appreciate pushing back on just looking for bigger metrics, but there's something more important when it comes to measuring people who contribute at least a little bit. Our licensing asserts that they must be attributed. After all, they have contributed something.
Also, for your astronomy comparison, this would be more like saying that anyone who contributes to publicly recorded astronomy observations is an astronomer -- even if they have only done so once. In my estimation, that doesn't sound crazy. Your comparison to "looking at the night sky" is a lot more like reading Wikipedia.
-Aaron
On Tue, Apr 11, 2017 at 6:35 AM, Erik Zachte ezachte@wikimedia.org wrote:
About 'Number of editors who contribute 1 edit per month?'
I'm hoping we're not going that use that number for our next fundraiser ;-)
The more inclusive our numbers are, the less meaningful, bordering on alternative facts.
A person with one edit in any given month is as much an editor as a person who looks at the night sky a few times a year is an astronomer.
We have billions of those on this planet!
Erik
From: Analytics [mailto:analytics-bounces@lists.wikimedia.org] On Behalf Of Neil Patel Quinn Sent: Friday, March 31, 2017 23:06 To: A mailing list for the Analytics Team at WMF and everybody who has an interest in Wikipedia and analytics. Subject: Re: [Analytics] Fwd: follow-up on editors
Funny story: I noticed that Aaron's graph has the 1-month new editor retention on enwiki at about 7%, while I had recently done some queries https://github.com/wikimedia-research/2017-New-Editor-Experiences/blob/master/analysis.ipynb that put it a little under 4%.
It turns out I made an error in my Unix timestamp math, and I was looking at the 12 hour new editor retention rate. It'll be interesting to see if the ranking of wikis by retention changes significantly when I correct that.
On Wed, Mar 29, 2017 at 2:15 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
https://commons.wikimedia.org/wiki/File:Enwiki.monthly_user_retention.surviv...
On Wed, Mar 29, 2017 at 4:14 PM, Aaron Halfaker ahalfaker@wikimedia.org wrote:
Here's a graph of the retention rates of new editors in English Wikipedia.
_______________________________________________ Analytics mailing list Analytics@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/analytics