Howdy, y’all! I’m at the American Astronomical Society meeting in Austin, Texas. Many of you had said you wished you could come to this meeting. Zookeeper Chris and I are here, and we’ll try to give you a vicarious experience of what it’s like here. We’ll be posting more often than usual this week, as well as later – Austin is 6 hours behind the U.K. Next week, we’ll resume our normal Monday-Thursday schedule.

We’re also coordinating with a couple other astronomy blogs to better cover the meeting. We’re working with Fraser Cain and Pamela Gay from Astronomy Cast, and Phil Plait from Bad Astronomy.

You can follow all our blogging at the Astronomy Cast Liveblogging page:

http://www.astronomycast.com/LIVE/

Whenever any of us posts, we’ll mirror it on that site for everyone to see. If this works, we’ll try to do it again at future meetings. Hope you enjoy it!

Following Anze’s post we’ve had a number of requests from people to see their weighting. Indeed we did think about this for a bit, but alas I reckon it’s a bad idea to publish users weights for a number of reasons – but mainly because it is very unclear what the weights really say about a user. For example, a user might be really upset to see that they have a low-weighting, indicating that they often disagree with the majority of people (if we use a democratic weighting scheme). They might then try to alter their classifying behaviour so to ‘improve’ their weighting. This would be disastrous, because perhaps they are actually excellent at classifying, and much more meticulous than the majority of other users. In this case a low weighting would be good! The majority ain’t always right, right?!

If providing this kind of feedback was to have any effect then this would be bad – because it would mean our data becomes correlated in complicated ways that we can’t trace i.e. the results from one week can affect the next week. Ultimately everyone will want to up their weighting, and I can imagine a horrible situation where everyone just clicks elliptical all the time! Because this would give everyone great weightings – but completely ruin the project!

Therefore, you see that knowing your weight can only have a bad effect (if it isn’t going to have an effect then there’s no need to know 😉 ). Ideally we don’t want anything to have an effect on you – we want everything to be as unbiased and open and transparent as possible when it comes to analysing the results.

Plus the weights change all the time, as more classifications are made and it is computationally very intensive to compute them. Further, there’s an infinite number of ways of working out the weightings! We could see how well you agree with each other for just the bright galaxies, or how well you agree with an ‘expert’. But the point of all this is that we do not know the true morphology of these galaxies – and therefore we cannot give you a true weight (ie. how well you are classifying).

I hope that helps to explain our situation a bit! I appreciate that it must be a bit frustrating not to get more feedback on your classifications. Perhaps when this phase of the project is wrapped up then we can feedback more… and ultimately all this data is probably going to be made public! Cheers, Kate

Today I delivered the final pre-bias testing data to the collaboration. In other words, for some time now, the website is serving only images for the purpose of bias testing – the mirrored, black and white, etc images. Therefore, the standard data are as complete as they will ever be. The process of getting the data into a form suitable for processing for individual science projects is beautifully inefficient and convoluted! Below is a somewhat technical description of what I do.

First I login into a database server at the Johns Hopkins University and perform an SQL query that dumps the entire live database into a text file, which I then compress and FTP over to my computer workstation at the Lawrence Berkeley Laboratory. This is to bridge the gap between computer science world (pretty ASP.NET code and SQL backend) and science world (spaghetti FORTRAN code on UNIX and binary files).

The data is then reduced in a series of steps. First, the data is organized and sorted by galaxies, and usernames are converted into consecutive numbers (so that the usernames are anonymous in the final database). Second, the data from various downloads are combined into one big dataset. Third bad data are weeded out (misconfigured browsers, bots and similar). Finally, the reduced “histograms” for each galaxy are produced. These correspond to our final state of knowledge about each galaxy.

There are four ways of doing these: spirals can be combined or separate and users can be reweighted or not (and two times two makes four). In the combined spirals sample, we combine all three spiral subsamples (clockwise, anti-clockwise, and edge-on) into a single spiral category: science projects that are interested purely in the galaxy evolution don’t care about orientation of a given galaxy. In the reweighted sample, we try to improve the sample by essentially comparing the agreement between users: the idea is that if ten users claim that a certain galaxy is a spiral and the eleventh users says it is an elliptical, it is likely that the 11th user got it wrong. Users who commonly disagree with everyone else gets down-weighted and those who always agree get up-weighted.

It is a purely statistical exercise meant to remove pranksters that click randomly and up-weight careful users. In practice, we can check how well it works. We do this (well, Steven does it) by looking at galaxies that have the same absolute luminosity and size and shouldn’t evolve over the small redshift range probed by the SDSS. The upshot is that it doesn’t work as well as initially anticipated: as an old english proverb goes: if one million French believe in something, it doesn’t make it right. And so we also produce the unweighted sample in which all users are given the same weight. It is up to individual science projects to decide which combination to use.

Finally, the reduced data is uploaded to a super-secret web server where other collaborators can download it.

The final datasets contain 34,617,406 clicks done by 82,931 users. Hooray for all of you! However, the previous downloads already went over 30 million, and hence this will make only small improvements to our science results. Now, the important task is to gather enough information about biases in our datasets and so keep clicking, please!

Happy New Year from Galaxy Zoo! 2007 has already left for some of you, and for others, it will be leaving soon. 2007 has been a great year for the study of galaxies, thanks to the galactic classifications that you have completed. I thought that in today’s post, I’d tell you a little more about what we are doing right now, as 2007 turns to 2008. Chris originally posted about this in the forum, and Kate has been providing a status update.

In last Thursday’s blog post, Chris gave an introduction to projects that the team is now working on. When he talked about Kate and Anze’s cosmology study (finding the rotation directions of spirals), he mentioned that a key to this project was completing the bias study. That “bias study” includes the rotated, mirrored, and black-and-white galaxy images that we’re asking you to classify now.

For those of you that aren’t familiar with what scientists mean by “bias,” it’s one of those funny words that has a specific meaning in science that is different from its meaning in everyday life. When we say “bias” in daily life, we often mean that someone has an agenda that means they can’t be trusted. For example, we might say that a news organization has an [insert political viewpoint] bias that means you just don’t trust that they’re giving you an accurate picture of the news.

That’s not what we mean here, though! There are several related meanings of the word “bias” in science, but they all come down to the question of whether results were influenced by some factor that the scientists didn’t think about. The scientist might ask himself or herself: Did I really study everything I could have studied? Did I do the right type of analysis on the data? Did I have some preconceived idea that kept me from interpreting the data with an open mind?

The question that we are asking is about how people see galaxies. When you see a fuzzy galaxy, is it easier to see it as an anticlockwise spiral than a clockwise spiral?

We’ve talked with some colleagues from psychology, and there is no reason to think that should be the case – but we want to test the idea just the same. That’s why we’ve introduced the rotated and mirrored galaxy images. The mirror image of a clockwise spiral should appear anticlockwise, and vice versa. So, all the clockwise galaxies in our original data should now appear anticlockwise in the “bias study data,” and vice versa.

The specific type of bias we’re looking for is called sample bias. Scientists often take a sample of a thing by measuring properties of some number of that thing. We’ve asked you to classify many galaxies, and the sample consists of your classifications. If, as mentioned above, humans classify anticlockwise galaxies easier, then the classification sample will have more anticlockwise galaxies than the real universe. If our sample of galaxy classifications is not the same as the classifications of galaxies in the real universe, then any conclusions we draw about the universe from our sample will be wrong.

And we don’t want to be wrong. Scientists are, by nature, very careful. We want to think of every possible effect, and every possible interpretation of our data, before deciding on the right one. By being as careful as we can, we hope to increase our understanding of the real universe in which we live.

As always, we need your help to do this. See Kate’s status update on the forum, and keep classifying those galaxies!

Welcome to the brand new Galaxy Zoo blog!

To start with, here’s a page with a bit of info about the project and what we hope to do with this blog.

Stay tuned for more!