Data Working Group/Results of Organised Editing Survey 2017

From OpenStreetMap Foundation

Results of the “organised editing” survey

DWG conducted a survey to inform a potential OSMF policy on corporate editing between 20 September 2017 and 08 October 2017. The survey was announced on major mailing lists and forums, and also mentioned in OSM blog and weekly news.

The welcome message on the survey explained the rationale as follows:

When OpenStreetMap started, it was largely a project of hobbyists contributing to OSM in their spare time. They chose freely what to map and which tools to use, and they took individual responsibility for their contributions.

The continuing growth and popularity of OSM have also brought more and more organised mapping efforts, mostly in the form of companies setting up paid data teams to improve OSM data in specific regions or for specific use cases, but also unpaid groups like school classes that are directed to work on OSM. These organised mapping efforts are an integral part of today's OSM contribution landscape and, when done well, help make OSM better and more widely known.

In order to ensure good communication, and a level playing field, between individual community members and organised editing groups, the OSMF Data Working Group has been tasked with developing guidelines for organised groups. These guidelines will above all set out some transparency requirements for organised groups – things that are already voluntarily followed by most groups today, like informing the mapping community about which accounts edit for the team.

We have prepared the following survey with a few questions about such a policy to give us a better understanding of what the mapping community expects from such a policy. The survey is aimed at everyone editing (or planning to edit) in OSM, whether as individual mappers or as part of a team, and your answers will help us in fleshing out a draft policy.

Within the scope of this survey, and the policy to be written, we define paid mapping (or paid editing) as any editing in OSM performed by someone who is told by a third party what to map (and potentially also how to map it) and who receives money in exchange. We define other organised mapping (or editing) as any editing that is also steered by a third party, but where no money is paid.

The following is a textual description of the main survey results. We invite you to peruse the detailed spreadsheet (see links at the bottom), and draw your own conclusions from the responses.

There were many questions where you could choose one option between “strongly agree” and “strongly disagree”. For the purpose of this document, “strongly agree” is +2, “neutral” is 0, and “strongly disagree” is -2, and for the textual representation we have also introduced the labels “somewhat agree” and “somewhat disagree” for the 0.25 to 0.5 range. So everything below 0.25 is neutral, 0.25-0.49 is “somewhat agree”, 0.5-1.49 is “agree”, and 1.5 up is “strongly agree”. But we have the exact number in parentheses so you can ignore the label if you wish.

Where we specify percentages, we have rounded to the nearest integer; where “no answer” was among the options you could check, we counted that just as if you hadn't actually given any answer.

Overall, we had 668 completed questionnaires (“completed” means they didn't drop out before the end – it doesn't mean they answered all questions).

Scope of Policy

Our question was:

What types of mapping activities should be covered by a policy?

  • A policy should only apply to paid editing.
  • A policy should apply to paid editing as well as other organised editing.
  • I don't think we need a policy at all.

20% of respondents said that a policy is needed for paid mapping. 63% said that a policy should apply to both paid and other organised mapping, while 17% said we shouldn't have a policy at all.

If we look only at those respondents who said they are involved with an organisation that does paid mapping, these results change to 5% “policy needed for paid mapping”, 52% “policy should apply to both” and 43% “no policy needed at all”.

If we look only at those respondents who said they are involved with an organisation that does other organised mapping (but not with one that does paid mapping), these results change to 18% “policy needed for paid mapping”, 57% “policy should apply to both” and 25% “no policy needed at all”.

Quality

Our questions were:

In OpenStreetMap, anyone can map and we don't expect newbies to attend training first. The community will fix problems and help new mappers learn the trade. Should this also apply to paid or organised groups?

  • Yes, paid/organised mappers can learn as they go just like individual mappers.
  • No, the leader, instructor, teacher, or trainer of a team should make sure that the mappers they bring to OSM have the proper training.

68% of overall contributors said that “the leader, instructor, teacher, or trainer of a team should make sure that the mappers they bring to OSM have the proper training“, while 32% said that paid or organised mappers can learn as they go along just like everybody else. Among those involved with an organisation that does paid mapping, only 37% say that proper training needs to be ensured by the lead, and 63% say it is not necessary. We also looked at the group of users who said they weren't involved with paid or organised mapping, but that they had interacted with such mappers through changeset comments and other means. Among this group, the notion that paid/organised groups should receive training from their organisation had the highest support, at 81%.

Transparency

Our questions were:

Should paid/organised mappers seek community buy-in for their projects?

  • Before starting to work towards a specific goal, paid/organised mappers must inform the mapping community of that goal and the mapping techniques used.
  • Mappers have the right to say no towards a project or goal.

(These questions had an agree/disagree scale as outlined above.)

Respondents agreed (0.62) that paid or organised mappers should inform the community of the goal and the mapping techniques used before starting to work on a specific goal. Respondents however were neutral (0.06) on whether the community should have a “veto right” concerning paid or organised mapping projects.

How should paid/organized mappers document their group?

  • A paid/organised mapper can use as many accounts as they want to protect their privacy.
  • A paid/organised mapper needs to disclose who is directing them.
  • A paid/organised mapper must disclose their identity (real name).

(These questions had an agree/disagree scale as outlined above.)

Respondents somewhat disagreed (-0.35) with allowing paid or organised mappers to use as many accounts as they want to protect their privacy. They agreed (0.91) that paid or organised mappers should disclose who is directing them, and disagreed (-0.52) with forcing paid or organised mappers to disclose their identity (real name).

Questions about instructions to members of group and performance metrics

Our questions were:

  • If the paid/organised team use OSM-related criteria to measure the performance of their mappers (e.g. "objects mapped per hour"), these criteria must be made known.
  • Instructions to the paid/organised team must be available for inspection by the mapping community.

(These questions had an agree/disagree scale as outlined above.)

Respondents somewhat agreed (0.25) that performance metrics need to be disclosed, and somewhat agreed (0.47) that instructions given to the paid/organised team should be available for inspection by the community.

Communication

Our questions were:

Individual mappers communicate among themselves using direct messages and changeset comments. Mappers who ignore changeset comments can be forced to reply. How should the community expect to interact with paid/organised mappers?

  • Individual mappers can expect to communicate with paid/organised mappers just like with anyone else
  • Individual mappers should be able to discuss problems affecting the entirety of a project with the group leader
  • How to contact the leader, instructor, teacher, or trainer of the paid/organized mapping group must be clear from each member's profile page

(These questions had an agree/disagree scale as outlined above.)

The three questions in the communication section had the highest agreement rate of all.

Respondents strongly agreed (1.53) that individual mappers should expect to communicate with paid/organised mappers just like with anyone else, and agreed (1.29) that problems affecting the entirety of the project should be discussable with the group leader. Respondents agreed (1.05) that a way of contacting the group leader should be clear from each member's profile page. Among those involved with an organisation that does paid mapping, agreement with the first two questions was also high (1.52 and 1.17 respectively), but they only somewhat agreed (at 0.29) about the contact info having to be clearly stated on each member's profile page.

Personal Information

We also asked a few questions about respondents' backgrounds:

  • What is your OpenStreetMap username? (optional)
  • How long have you been editing OpenStreetMap?
  • What country are you from? (optional)
  • Are you affiliated with an organization that does paid mapping?
  • Are you affiliated with an organisation that does other organised mapping?
  • Have you interacted through changeset discussions, direct messages, or notes with people you knew to be paid/organised mappers?

78% of respondents were with OSM for 3 years or more; a further 16% for over one year. Among those affiliated with organisations that do paid mapping, only 55% were with OSM for 3 year or more, and a further 35% for over one year.

A total of 98 people (15%) said they were involved with organisations that do paid mapping. 172 (26%) said they were involved with organisations that do other organised mapping. 77 people (of these) said yes to both. For our evaluation we made three groups: The 98 who said they were involved with an organisation that does paid mapping; the 95 who said they were involved with other organised mapping but not already counted in the first group, and the 475 who checked neither of the two boxes. In programming speak, the groups are “paid”, “organised && !paid”, “!organised && !paid”.

We further split the third group into two according to their response to “have you interacted with people known to be organised or paid mappers”. This gave us 179 “not organised, not paid, but interacted” people, and 264 “neither organised, nor paid, nor interacted”. (The difference to 475 is those who haven't answered the “interacted” question.) We hoped to gain from this some insights about how interactions with paid or organised mappers shape the views of our community members; but it turned out that the views of both these groups rarely differ in a significant way.

Regional distribution

We also did an analysis by continent. We had 417 results from Europe (which for our purpose includes Russia), 187 from North America, 35 from Asia, 15 from Australia and Oceania, 13 from South America, and 1 from Africa. (We used the submission IP number to determine the continent where people hadn't given their country.) We're publishing the results for all continents except Africa (where the one contributor is too easy to de-anonymise), however South America and Australia/Oceania have too few submissions to draw any conclusions there. Overall, Europe is more in favour of regulation than North America and Asia (11% in Europe say no policy is needed, while 31% and 30% say so in North America and Asia, respectively). This is, however, likely to be a reflection of where the paid/organised mapping organisations are. Only 6% of respondents in Europe are affiliated with an organisation that does paid mapping, but 32% in North America and 31% in Asia are; only 16% of respondents in Europe are affiliated with an organisation that does other organised mapping, but 47% in North America and 37% in Asia are.

Detailed analysis

Download the spreadsheet in OpenOffice format or as a PDF file.