Power 2.0 Opinion Gathering

    The Judge

    Kill me please
    Joined
    Aug 12, 2014
    Messages
    409
    Reaction score
    176
    • Purchased!
    • Legacy Citizen 3
    This isn't just a simple poll, but instead an alternative method that should hopefully eliminate bias that might have possibly existed with previous polls. The current stage of this poll is gathering a population (a group that I wish to gather data from), however I can't just grab a list of all SMD users since some users are banned, many are inactive, and nonresponse will be a difficulty, so instead I am attempting to form a list of all semi-active individuals. Filling out this form won't guaranteed you will be represented, as I can only use about 10% of the results from the initial form. (Its a lot of Statistics nonsense that would take forever to explain)

    For now I just need as many people answering this as possible, so just fill it out regardless of your opinion on anything Starmade related.

    Population Gathering Form

    The poll will close next Friday.
     
    Joined
    Jan 25, 2014
    Messages
    95
    Reaction score
    53
    • Community Content - Bronze 1
    • Purchased!
    • Legacy Citizen 6
    Unfortunately, asking for people to be polled will introduce response bias. Some might be active, but not really care, or some might miss this poll. I applaud the effort, but I don't think it'll be accurate enough to make a firm conclusion on (maybe a rough idea though).

    Ideally, the best way would be polling a random sample from all people whose profiles show them being online often/recently and pestering them until you get a response to reduce bias. Unfortunately, I'm not sure if SMD even tracks user activity enough to make this possible, and that would probably be too much of a pain for any sane person to do.

    Once you've finished it, please make your data available for us to see. Good luck.
     

    The Judge

    Kill me please
    Joined
    Aug 12, 2014
    Messages
    409
    Reaction score
    176
    • Purchased!
    • Legacy Citizen 3
    Unfortunately, asking for people to be polled will introduce response bias. Some might be active, but not really care, or some might miss this poll. I applaud the effort, but I don't think it'll be accurate enough to make a firm conclusion on (maybe a rough idea though).

    Ideally, the best way would be polling a random sample from all people whose profiles show them being online often/recently and pestering them until you get a response to reduce bias. Unfortunately, I'm not sure if SMD even tracks user activity enough to make this possible, and that would probably be too much of a pain for any sane person to do.

    Once you've finished it, please make your data available for us to see. Good luck.
    Yeah thats the thing there isn't a good way to do this
     
    Joined
    Jun 27, 2013
    Messages
    896
    Reaction score
    166
    I just need as many people answering this as possible, so just fill it out regardless of your opinion on anything Starmade related.
    Frankly I'm not going to register, however well your intentions may be.
    Should you be interested in my opinion, ask away here via forum post or PM.
     
    • Like
    Reactions: MacThule
    Joined
    Jan 31, 2015
    Messages
    1,696
    Reaction score
    1,199
    • Thinking Positive
    • Likeable
    Unfortunately, asking for people to be polled will introduce response bias. Some might be active, but not really care, or some might miss this poll. I applaud the effort, but I don't think it'll be accurate enough to make a firm conclusion on (maybe a rough idea though).

    Ideally, the best way would be polling a random sample from all people whose profiles show them being online often/recently and pestering them until you get a response to reduce bias. Unfortunately, I'm not sure if SMD even tracks user activity enough to make this possible, and that would probably be too much of a pain for any sane person to do.

    Once you've finished it, please make your data available for us to see. Good luck.
    Also it's sign up by name to be 'randomly' selected for non-anonymous responses. Unsure how that is more useful than simply soliciting opinions here. Taking it out of the community, allowing selection by name pretty much ensures that even if it is totally random as stated, it retains an appearance of bias. I signed up, but probably won't respond after thinking about it, because it would only lend credibility to a process that seems meant to brand a formal, collective "opinion" 'about Starmade', which would be occurring without any real possibility of oversight.

    Like why ask usernames on common platforms instead of doing it via email (which won't be recognizable like user names are)? Why not have people take the poll immediately, anonymously, then randomizng the selection of responses?

    Also, what sorts of opinions are being solicited and what bias will determine what is asked and how that is phrased?

    What is the goal meant to be achieved by soliciting a bunch of "opinions" about (all of? part of?) StarMade?

    What could it possibly show?

    What could statistics formed from such a poll even be relevant to or used for?

    Apologies! I hope signing up then not responding doesn't cause any problem, but on reading the responses and some further consideration I have changed my mind.
     
    Joined
    Jan 25, 2014
    Messages
    95
    Reaction score
    53
    • Community Content - Bronze 1
    • Purchased!
    • Legacy Citizen 6
    Also it's sign up by name to be 'randomly' selected for non-anonymous responses. Unsure how that is more useful than simply soliciting opinions here.
    The point is to reduce response bias by being sure each person surveyed responds, but it introduces response bias at the same time by over representing those who feel strong enough about it to want to take the survey. Like I said in my earlier post, there isn't really a way to do this properly. The best that we can do is either randomly poll and calculate a margin of error that takes into account those who didn't respond (which would be huge if a lot don't respond or if the sample size is small like I expect it to be), or we can ask for people to participate and make it clear that this is a biased survey because we don't want to lose our sanity trying to be accurate. With the first option, I'd be surprised if we got a margin of error within 10%. (In order to use a survey as evidence, the margin of error should be at most 5% or less with most statisticians not being satisfied unless it's 1% or less.)
     
    • Like
    Reactions: MacThule
    Joined
    Jan 31, 2015
    Messages
    1,696
    Reaction score
    1,199
    • Thinking Positive
    • Likeable
    The point is to reduce response bias by being sure each person surveyed responds, but it introduces response bias at the same time by over representing those who feel strong enough about it to want to take the survey. Like I said in my earlier post, there isn't really a way to do this properly. The best that we can do is either randomly poll and calculate a margin of error that takes into account those who didn't respond (which would be huge if a lot don't respond or if the sample size is small like I expect it to be), or we can ask for people to participate and make it clear that this is a biased survey because we don't want to lose our sanity trying to be accurate. With the first option, I'd be surprised if we got a margin of error within 10%. (In order to use a survey as evidence, the margin of error should be at most 5% or less with most statisticians not being satisfied unless it's 1% or less.)
    If we could poll with true randomness and an acceptable margin of error (and an acceptable sample size), what would we even get? What would be the application of the resulting statistics?
     
    Joined
    Jan 25, 2014
    Messages
    95
    Reaction score
    53
    • Community Content - Bronze 1
    • Purchased!
    • Legacy Citizen 6
    If we could poll with true randomness and an acceptable margin of error (and an acceptable sample size), what would we even get? What would be the application of the resulting statistics?
    We would get an idea of how many players like/hate the new system. I think the reason this is being attempted is because some people keep throwing around random numbers: "PVPers who hate it are only 10%..." "Half of the dedicated players stopped playing because of power 2.0." etc. etc. It would give us a real percentage instead of baseless claims about how many of us are on each side. Aside from that, it wouldn't really serve a purpose, except maybe spark more arguments.
     
    • Like
    Reactions: MacThule