Hehe, thanks. Glad that these items are interesting enough to keep your good mood when you were completing the survey. Yes, definitely we will share results, when the analysis is done. :)

Quote: admiralbob77Glad I filled it out when I was in a good mood this morning, and not the crappy “my bike broke down on the road” mood I’m still in tonight. ;-)

Seriously, there were some thoughtful questions, and I hope you’ll share your thesis and summary results back here with us. :)

Like admiralbob says, it would be interesting to see the collated results. BTW, I’d preferred if you had defined what “interaction” was (or maybe you did, and I skimmed!) I also wondered why you had a “I prefer not to answer that” as an option for some questions like educational level and income (which is a good thing) but not for the “Are you married” question :)

Hi Ivan, the definition of interation is a good point. And it actually bothered me a lot. But at the end, Giorgos and I decided not to define it. We were afraid that providing a specific definition would restrain thinkings. So just the general idea of interaction, like talking to each other on forum, emailing each other, or becoming friends over social networking sites, etc. I hope this clarification is not confusing/annoying. :)

The other quetions is quite interesting. I would suggest, if anyone does not wanna answer the item about amarital status, maybe ‘other’ is a good option. :)

to clarify: I’m fine with you (colab) saying: I have made X personal friends on ccM with whom I share everything.

But if 100 ccM people say: I share personal life stuff with spinmeister, that is divulging something personal about me (spinmeister) without me having any control over that. And I’m uncool with that.

Hi spinmeister. I see. Yes, I agree, the privacy is a concern. When we were designing this survey, we also had such worries. You are right with remix, cos all information is public. We actually collected such public data of remix network before, which could to some extent reflect the relationship between remixers. But we are not sure whether such relations based on remix would become(or overlap with) some more personal relations (well, maybe some users don’t care, maybe some would), and that is why the self-identified people with whom you often interact are needed. If the privacy is an issue, there is an option going back to the previous page and click ‘no’. Then you can skip the following question. (Hope this is not going to motivate people to skip.) :)

my concern is that the survey essentially asks people about 3rd parties, rather than just themselves.

I have no influence about people mentioning my name in your survey. You are specifically asking people to do that. I think there is little justification to ask about names of others in personal relationship contexts in a survey of this nature.

I think you likely don’t mean anything evil. But consider the following example: if a Chinese dissident’s name gets mentioned as a personal friend by a few remixers, who are politically outspoken or who maybe have contributed to the winter remixing project in support of the Dalai Lama foundation - it could get that person into big trouble with the Chinese authorities.

That person might get into trouble without having participated in your survey, just because your questions were answered by some of their ccMixter friends.

I’m assuming that you are not working for the Chinese government, but how does anyone here really know?

And I hope you’re not working for ASCAP either :-)

Quote: spinmeisterI’m assuming that you are not working for the Chinese government, but how does anyone here really know? And I hope you’re not working for ASCAP either :-)I think we can trust snowflake on this :) The other way is to email Dr. Giorgos Cheliotis to verify the legitimacy of the study.

BTW, I do not know Rong at all. And I didn’t verify the study with Dr Cheliotis (he and I are fellow volunteers of CC-SG) before completing the survey. But I do know the information about the university department checks out, and that Dr Cheliotis has an on-going research interest in this area.

Quote: spinmeisterBut consider the following example: if a Chinese dissident’s name gets mentioned as a personal friend by a few remixers, who are politically outspoken or who maybe have contributed to the winter remixing project in support of the Dalai Lama foundation - it could get that person into big trouble with the Chinese authorities.

That person might get into trouble without having participated in your survey, just because your questions were answered by some of their ccMixter friends.I would think the PRC government would have been able to trawl ccMixter and obtain that sort of information long before this study. :)

p.s. I’m not saying spinmeister’s take on the privacy matter isn’t valid. But the foremost and strongest validation, for me, was snowflake’s endorsement of the study. Cheers!

Hey Ivan,

Thank you for your trust. It is interesting how we come to rely upon each other. The main reason we’ve tried to support this survey is because Mike Linksvayer made the introductions.

I don’t necessarily endorse or not endorse the survey itself, but promised I would pass it along to all of you, as requested by Rong and Giorgos, by way of Mike.

I do trust they will keep this information confidential…but I also trust each Mixter will determine whether they want to participate or not, and whether they feel right about it or not.

Peace.

Hi spinmeister, I would take your point as, you are trusting me for not working for Chinese government or ASCAP. And I am glad that I am not involved with either of the two. :)

I really can understand your concern about dragging people into this research project, even without letting them aware of this. But I promise all data would be kept only for academic analysis. And only me and Giorgos can get access to the database. The results we are going to share with ccMixters later are based on anonymous data, and no account names will be revealed.

http://www.nus.edu.sg/irb/i...

This is the link to NUS IRB (institutional review board), basically an institution for reviewing ethical merits of all research projects at NUS (national university of Singapore). IRB’s approval of projects requires that all researcher must have to keep all data strcitly confidential. The code number for this project is 10-146. Giorgos and I are listed as co-investigators for this project. There is an inquiry email address listed in the website, just in case it is needed. :)

Thanks for bringing this issue here. I am sorry for not making any clarification earlier. :)

Rong

Rong,

Admittedly, I did find it unusual that the survey asked for the personal names of individuals, as well as the mixter itself. I have become accustomed to surveys being fairly anonymous.

Can users take the survey while staying anonymous themselves and not divulging other names?

My concern has nothing to do with anonymity. And I never said that.

My concern has to do with someone creating a personal friendship diagram. That has nothing to do with what we do here at ccM and it should be off-limits in a questionnaire.

Hi spinmeister and everyone else, i’m Giorgos, the professor supervising Rong (laurawr), and i reactivated my long dormant ccmixter account to add some words of my own to the discussion.

I understand your concerns. Some of us, myself included, are quite sensitive when it comes to our privacy. And we have every right to be. Nobody can tell you how protective you should or should not be with respect to your own privacy. But we can try to reason with you on this topic and explain ourselves. I see that other members on the forum as well as Rong brought up a few reasons why it is not as serious a threat to your privacy as you may think. Let me make it clear here first of all that we really don’t care at all about the handles that users use or about the real identities behind the handles. But we do like building maps of relationships, that’s true. Yet, we only analyze and publish results on anonymized networks of relationships.

The user handles are needed just in the first stage, to be able to construct a composite network from the various responses of the different users. After we have such an (always incomplete) network, we just look at it as nodes with certain properties, the user handles being entirely irrelevant and, actually, uninteresting in our study, as we focus more on communities, i.e. collective behavior, rather than individuals. If we find that specific individuals play, say, key roles, and are interesting to perhaps interview, we will ask them whether they wish to be interviewed in person, at a different stage. And of course we will not publish anything that identifies you specifically as a user and discusses your relationships to other users.

Still, this leaves the question you rightly ask here: why do we even bother with this? Rong tries to explain it in her own words in her posting. From my side, I can say that I am particularly interested in understanding the networks of relations that people build. Why so? Because as a social scientist I find that a study that examines the workings of an online community can only then be truly grounded in reality when it considers the community as embedded in a network of both online and offline relationships, that may often extend beyond music-making and remixing.

This is an important trend in social science in general and in the way we try to understand the role of the Internet in human affairs. Early studies would ask things like: “Why do you go online”, or “can you make friends online”? These are still important questions, but as the Internet has long become an integral part of our lives, it is now becoming increasingly important that we look at any online community in the social (online and offline) context in which it operates. Online is not cut off from offline and I would argue music making, whether online or offline, is not something where one could say personal relationships don’t matter. On the contrary perhaps.

So, when we analyze networks of remixing activity, we want to compare them to the underlying relationships that people have to each other. Do friends remix each other more? Are people with many friends on the community more likely to be more active members? Do strong relationships even exist in the community or is it, like theory would sometimes suggest, a community that consists mostly of loosely coupled individuals? Or maybe there is a mixed picture, where there exist clusters of users with stronger relationships to each other, perhaps the more ‘senior’ users, while others are more weakly connected. And if that is so, we can then ask: where does the bulk of remixing activity come from? From the core, the periphery, from networks of friends, or from quasi-random others? To answer all these in my view fascinating questions we need to know as much as users are willing to divulge about the relationships they have to other users.

I hope this gives some idea about our motives and why we need to ask for user handles just at the first stage of data collection, while at the same time we are really not interested in using these handles in any way that would compromise the real identity or privacy of the individual, in any analysis or publication.

As a last point, and if it helps to reassure you of our intentions: NUS, our host institution, is one of the top universities in Asia and globally, with strict standards of research ethics. I have also personally worked with Creative Commons many times before and you can find a lot about my work if you google me. It would be insanely stupid of us to publicly divulge any personal information that ccmixter users may hand us through this survey without checking with them first. Our reputation is important to us, in fact for academics it is our most precious currency.

Also, on a personal note, I have done a fair share of remixing myself and because of this I have personally a lot of respect for people like you and for communities like ccmixter. You won’t find my remixes on ccmixter because most of the material I use is ripped of ARR works, but that’s another story. So, anyway, sorry for the long post, please have no doubt that we take this seriously and that’s why CC and some good ccmixter folks have supported us in bringing this survey to you. Indeed you can’t help that others may be asked about their relationship to you and they may want divulge some information about it, that is true always and in any situation, not just in this survey. And you can’t prevent them from sharing some basic info about the relationship. But I think you can be reasonably certain that we ask for good reason and that we have every reason not to divulge any of this publicly.

Thanks for your extensive, articulate and warm reply, Giorgos. I appreciate it very much that you have personally joined this conversation.

Just to be sure, that I say it: I do not doubt the excellent reputation you have earned and that you are working within top class organizational framework. And I have no doubt that the research you’re doing is scientifically interesting.

However I think this may very well be a question of science ethics beyond trusting you and Rong.

I’m NOT concerned about the questions where individuals divulge stuff about themselves. You may find in the survey data, that I even participated in the beginning part of the survey. I’m ok if you ask me some stuff about me.

However I stopped, when I encountered questions about my personal relationship network - specifically asking me to identify individuals (via their ccM handles, which for many people are NOT anonymous - even if the handle is somewhat abstract).

I believe, this is where the survey is possibly running into an ethics problem, because it is essentially collecting private information about individuals without their consent.
The US government defines Private Information this way in §46.102 (f): ” … Private information includes information about behavior that occurs in a context in which an individual can reasonably expect that no observation or recording is taking place. …” (emphasis mine)

I have a reasonable expectation that the fact I’m personal friends with some ccMixter community members is private between each one of those friends and me. Each one of those “friendship” nodes is private between that individual and me. However the survey is asking on a large scale many others if they are friends with me, specifically asking them to reveal our personal friendship. By asking my friends, to individually name me as a friend you are collecting data about me: the network diagram of MY personal relationships. But I have not given you permission to collect that data about me.

This may sound rather abstract, but maybe it takes more color via my previously mentioned example:

“… if a Chinese dissident’s name gets mentioned as a personal friend by a few remixers, who are politically outspoken or who maybe have contributed to the winter remixing project in support of the Dalai Lama foundation - it could get that person into big trouble with the Chinese authorities. …”

The problem doesn’t start if you would pass on your collected data to the wrong authorities. (I have no reason to believe you would, but that’s not my point.)

Again, I believe a science ethics problem starts when you collect this kind of private data without the consent of the individual.

I appreciate that ccMixter (and other social networking sites) are tempting fields for interesting scientific research. But when I signed up to ccMixter, I signed up to be in a public artistic relationship experiment, not to be the subject of a personal relationship science experiment.

When the Creative Commons handed over the operations of ccMixter to ArtisTech Media (ATM), it did not even give ATM the email addresses for ccMixter participants as became obvious to everyone who tried to log in after the transition. That’s how sensitive the Creative Commons organization is to privacy issues. Since you are mentioning your close relationship with the Creative Commons, wouldn’t you want to be as sensitive about privacy?

I honestly believe the right thing for you to do would be to remove from the survey those questions which ask people to divulge individuals with whom they have personal relationships.

And maybe Rong can write a research paper on the science ethics difficulties of conducting research on the relationship between public social networks and personal relationships. She would get an A from me! ;-)

More seriously - if you want to find out something about personal friendships, some data can be found quite publicly via things like twitter, facebook, people’s blogs etc. It’s hard work, but the best science always is!

Thank you for the long explanation spinmeister. I see that my email has not convinced you that this is a ‘necessary evil’ of sorts, for lack of a better term, in order to study the role of relationships that form in an online community, or in any community for that matter.

As a social scientist I must say I have a hard time accepting that asking about the nature and networks of relationships, even those that are 1-1, is some sort of ‘taboo’ topic and I think you will agree with me that generally it isn’t. But now, after thinking about your arguments some more, I think it would have been better if we had more specifically asked about people’s individual and subjective perceptions of closeness/affinity to other people, perhaps not asking at all about frequencies of interaction and not about intimate levels of interaction, but just limiting the options to two-three levels of perceived closeness, asking for example whether you think that the people you remix actually know you, or care about you, or whether you think you know them. And vice-versa.

Maybe these can be asked in the aggregate, not mentioning specific user namers, though I would of course prefer to give the users the option to mention usernames, as I am pretty fascinated by networks and what they can tell us about social behavior.

The point is that even if you want to stick to your objection to what you see as an indirect and perhaps ethically questionable disclosure of private data, I guess you would not object to us asking users about how they feel personally (at a very basic level) about (specific) others that they know or remix. Their feelings and subjective perceptions of closeness to others are surely their own, so they can decide on their own whether they should disclose them, wouldn’t you say so? Anyway, that’s a thought for now…

Rong has already made some changes to this survey and I’ve tried to pitch in with some thoughts. Maybe you can also take this up with her offline, so as not to burden this thread with too many messages. I think we can find a middle ground that addresses your concerns. I am away these days and on the road in rural areas, but will try to connect again sometime in the next days to check where we stand. I think through this interaction we will be able to craft something good and meaningful.


PS. About finding out stuff through blogs and the like, it’s not harder, it’s easier, at least if you can write a script that will collect and mine the data. I’ve done that in the past sometimes. But none of this can tell you how people really feel about their relationships. Only the can tell you. My wish was therefore to learn more about how ccmixter users feel.

Quote: … this is a ‘necessary evil’ of sorts, for lack of a better term, in order to study the role of relationships that form in an online community …
I believe that anyone doing something potentially “evil” has the obligation to explain in considerable detail why it is “necessary” (typically a “greater common good” kind of explanation). This would include your (or any) study.

Quote: As a social scientist I must say I have a hard time accepting that asking about the nature and networks of relationships, even those that are 1-1, is some sort of ‘taboo’ topic and I think you will agree with me that generally it isn’t.
Maybe I should ask you the question, if as a social scientist, you consider your research “ Human Subjects Research”? If not, why not? If yes, do you see yourself strictly bound to ethical guidelines like the “ informed consent” principle? If not, why not?

I believe, that as humans, we are a highly social species. Interacting with that part of our existence isn’t really that different than interacting with our physical bodies. And therefore I would propose, that social research isn’t all that different than medical research. On the positive side, better understanding of the social parts of our human existence can (eventually) create better lives, a lot like medical research into diseases. On the negative side, social research - just like medical research - can do harm. Therefore it should be subject to the same ethical principles - one of the most fundamental one’s being the “informed consent” principle.

Quote: But now, after thinking about your arguments some more, I think it would have been better if we had more specifically asked about people’s individual and subjective perceptions of closeness/affinity to other people, perhaps not asking at all about frequencies of interaction and not about intimate levels of interaction, but just limiting the options to two-three levels of perceived closeness, asking for example whether you think that the people you remix actually know you, or care about you, or whether you think you know them. And vice-versa.

Maybe these can be asked in the aggregate, not mentioning specific user namers, …
Yes, I think leaving out the 3rd party user names, eliminates the need to get their consent, and therefore is alright.

Quote: … though I would of course prefer to give the users the option to mention usernames, as I am pretty fascinated by networks and what they can tell us about social behavior.
I fully appreciate your fascination - and it would fascinate me as well. But not all things we may find fascinating are ethical.

Quote: The point is that even if you want to stick to your objection to what you see as an indirect and perhaps ethically questionable disclosure of private data, I guess you would not object to us asking users about how they feel personally (at a very basic level) about (specific) others that they know or remix. Their feelings and subjective perceptions of closeness to others are surely their own, so they can decide on their own whether they should disclose them, wouldn’t you say so? Anyway, that’s a thought for now…
If I understand your question correctly, you are wondering between the difference between asking for disclosure of a one-way vs. a two-way relationship. I believe, that “personal friendship” is a “two way” data item, and thus should be out of bounds, unless you can get the informed consent of both parties before you ask. And of course that makes it complicated. “Personal admiration” would be more a “one way” thing and thus ok to ask, because you are asking only about the data of the person doing the answering, and by answering they are implicitly consenting. It’s rather similar to the difference between a “friend” on myspace and a “follow” on twitter. Even myspace appreciates that “friend” needs the consent of both parties. On twitter, a follow doesn’t require consent. However even twitter allows you to block people from following you, which I find surprisingly sensitive from a platform that’s so obviously public.

see part 2 of my answer below this…

I had to split my answer in two, because it was too long for the ccMixter software :-)

so here is part 2:

Quote: Rong has already made some changes to this survey and I’ve tried to pitch in with some thoughts. Maybe you can also take this up with her offline, so as not to burden this thread with too many messages. I think we can find a middle ground that addresses your concerns.
Yes, Rong has gotten in touch with me, and I’ve sent her my email address, so we can continue this dialog in the context of her project.

Quote: I am away these days and on the road in rural areas, but will try to connect again sometime in the next days to check where we stand. I think through this interaction we will be able to craft something good and meaningful.
I wish you safe and successful travels, and I also hope that we can figure out something that will produce interesting and good science while avoiding the pitfalls of ethical limits.


Quote: PS. About finding out stuff through blogs and the like, it’s not harder, it’s easier, at least if you can write a script that will collect and mine the data. I’ve done that in the past sometimes. But none of this can tell you how people really feel about their relationships. Only the can tell you. My wish was therefore to learn more about how ccmixter users feel.
Yes, if you’re comfortable with web spider/robot scripting, you can gleam quite a bit of data. However, the “feelings” are often hidden behind protected parts of profiles, and they also require a lot more interpretation of language, while a survey makes the respondent create measurable data items - and thus I would think that’s a lot easier as a data collection mechanism. Creating a good survey is very complex, though. Not only for the reasons we are discussing here. :-)

Hi spinmeister, I don’t wish to make this a much longer thread than it already is, but I would only like to point out that your frequent mentions of the need for “research ethics” and “informed consent”, though appreciated, do not necessarily support your argument, namely that a person cannot talk about their relationships and feelings towards others unless these others provide their consent.

It would seem to me that following through your argument in full would lead to two rather absurd situations: (a) people cannot talk freely about their personal feelings towards others, i.e. you require that an important part of their thoughts and feelings be censored, and (b) it becomes impossible as a researcher to ask about personal relationships, as it is impossible to get prior consent from all possible parties that the surveyed person might mention. You surely wouldn’t really say that (a) or (b) are sane outcomes, would you? Where do you begin to draw the line even? Or that researching friendship is generally out of bounds? Yet your logic, while well-intentioned, is pretty much rendering this part of the research impossible.

Unless, for point (b), you think that the survey subjects themselves should seek the consent of others before they include them. I, as a researcher, cannot do that for the subjects as I don’t know whom they are going to enter in their forms. Perhaps a custom software design which would send automatic consent forms could fix that, but that would have to be implemented separately, would raise new privacy concerns (e.g. would require at least an email address for each member) and would still leave us with point (a) above, i.e. why subjects cannot talk to a third party about their subjective feelings towards others (which may be shared by those others, i.e. be reciprocal, or not, and this is part of the research question).

I would ask you to kindly think about these issues, not so much to come up with a retort, but to consider where your logic may not be as solid as you make it out to be. I do not purport to have all the answers, but if I am countering your arguments it is precisely for ethical reasons, to defend our and others’ right to ask questions and also to defend subjects’ rights to talk about their feelings and social circles, and especially because I have a hard time buying into your argument that certain questions on basic properties of human interaction such as friendship are “out of bounds”.

In summary, I understand that there are objections, at least on the part of some members, you being clearly the most vocal one, but I remain skeptical about your argumentation and whether you have considered the negative consequences of your efforts to essentially render certain types of questions taboo in this community. I hope you will consider these thoughts and thank you for the candid exchange on this forum. We can remove that part of the survey and never mention it again. But it will not be for ethical reasons. It will be done in order to move ahead with you and with anyone else who may feel uncomfortable about these questions, and to salvage the rest of the survey. So, of course, my preferred outcome is still to keep this part of the survey, but rephrasing such that it is acceptable to both you and us. Maybe you can follow up on that with Rong offline.

By the way, I’m sure I saw a comment from colab here a moment ago that was congruent with what I was writing in my reply to spinmeister. But it seems to have disappeared in the meantime. Anyone know why?

Also, spinmeister, really do try to engage with laura and me off the forum as I think you can help us craft a better version, one that would not compromise the integrity of our research and would also address your concerns.

Just wanted to write a quick note to say thanks for modifying the survey. Just so you know, I had sent Rong my email address for any further offline discussion and clarification she might have desired. However she never ended up getting in touch with me after that. And thanks for engaging in the dialog with me - and as colab already stated, both Rong and you came across well in the conversation. I realize that sensitivity about privacy issues can be quite a bit cultural and also generational. I’m well aware, that my sense on this topic is not shared by everyone else. I think that’s quite ok. It’s one of the strengths of well intentioned people everywhere, that we can have such discussions about different perceptions and attitudes in an open and rational way. Thank you again!

i really appreciated your input and perspective spin, and believe that the way this situation was handled by all involved demonstrates the cooperative methodology of solution oriented problem solving we abide by here at ccM :)

Hi Snowflake,
In the survey, it is not asking for personal names. As indicated in the guideline of question9, it requires people to list users’ account names. As mentioned by ‘colab’, these accounts names usually cannot let us identify your real identity. But I understand that even as an account name, it is still personal to some extent and to some members. So I would do some explanation why we want to collect this data of relations.

I agree with spinmeister that, personal friendship has nothing to do with what we intend to do in ccMixter. What we do care here is our common interest in music. This is true! And this is also why we didn’t define interaction. If we follow a traditiaonal understanding, interaction would mean personal and direct contact, like talking over phone, chatting over MSN, leaving postings on your facebook wall, sending SMS, etc. More relevant to ccMixter, personal interaction might be sending message over the intra-system, reviewing one’s music track, replying one’s posting on this forum… But what makes it intereting here is, how about we taking remixng also as a type of interaction, in the form of indirect/impersonal relations. I mean maybe member A has remixed member B’s music for many times, but do they really know each other? Maybe not. But we don’t care. What we care is, this impersonal interaction makes us connetced. It can help me find nice resources to remix, can help me find fresh ideas of making music. Or it can help me connect to some other members who could also introduce me nice music for remixing (like if we review the same member’s work and we begin to notice each other’s existence). So basically when we ask who you are interacting in the ccMixter communutity, you would answer people who you interact by remixing, or who you interact with review, or with forum postings, and so on, in both way of direct and indirect. I am not going to mention what kind of results we are expecting (otherwise I am inducing our ccMixters into our intended outcomes, and this is definitely not the wise thing to do.) :)

Again, if this is an issue for some ccMixters, yo can click to get back to the previous page and choose ‘no’ to the question8. Then you don’t have to enter the third party’s name ( I mean their account names in ccMixter). But I am afraid your own account name is needed. I will dicuss this issue with Giorgos later, to see if we can set the question16 as optional.

Thanks all for the discussion. :)

Rong

Quote: snowflakeRong,

Admittedly, I did find it unusual that the survey asked for the personal names of individuals, as well as the mixter itself. I have become accustomed to surveys being fairly anonymous.

Can users take the survey while staying anonymous themselves and not divulging other names?

Thanks Rong.

We’ve decided to pull the header-callout for now. We look forward to your response to the issues raised and appreciate your patience.

Note that the links here in this forum post are still active and anyone who wishes to participate in the survey may do so.