(GZ1) When I go into SDSS, the galaxy is the other way round! What's going on?

[Alice]

This is the most frequently asked question by a long way on this forum!

Bear with me. It's quite a long story, and it is from something begun in December, called "The Bias Study".

One of the things Galaxy Zoo was trying to find out was whether or not a study of galaxies done by Michael Longo held true across a larger area of the Universe. He had surveyed well over 1000 spiral galaxies and there seemed to be more rotating anticlockwise than clockwise in his sample. (This is from our point of view. You might very correctly point out that from their other side they would be rotating clockwise. But it is still strange if more galaxies are rotating in any one direction than another. We wondered why it might be . . .)

Anyway, we found the same thing in our much larger sample. There seemed to be more anticlockwise than clockwise spirals in our sample, too. But we had to check one thing: was this human bias? To find out, zookeeper Kate designed a program to deliberately flip and rotate some images. For example, if a galaxy is really clockwise, we might see it as anti-clockwise in Galaxy Analysis, and click the anti-clockwise button.

Guess what we found? Well, if there really are more anti-clockwise galaxies in the Universe, the bias study should have found an excess of clockwise galaxies. In fact we found the opposite. People are still clicking anti-clockwise more often. The zookeepers have started to look for psychologists to work with to find out why this might be. It may have a deep psychological bearing, or it may simply be the web design layout. Who knows?

In any case, please always classify what you see in Galaxy Analysis, not what you see in the SDSS page.

Hope that helps.


Update: We're getting a lot of questions about black and white galaxies as well. In short, this was another test that was done at the same time - but it didn't show any interesting results so it didn't get talked about much. Again just classify what you see in Galaxy Analysis, and if you really want to see it in colour you can go to the SDSS pages.

[Mirja]

I've also noticed that in 'mygalaxies' the photos sometimes seem rotated 45 degrees and mirrored. To my shock it appeared that I had clicked often anticlockwise when the pic in mygalaxies showed the opposite (plus a rotation) and vice versa. But, going back to the pic in galaxyanalysis, I seemed to have done it correctly!

Strange...

[waveney]

This was the whole point of the bias study, to see if people clicked consistently independent of the way the image was presented.  They didn't they seem to have a bias towards anti-clockwise.  (There are left right inversions, and rotations by 90 degrees but none by 45 degrees).

The "My Galaxies" was never updated to reflect the bias study, so this always shows the SDSS original image.

[Mirja]

Whoops! I must have written down '45 degrees' before my first coffee this morning instead of 90...

Thanks!

[FermatsBrother]

...There are left right inversions, and rotations by 90 degrees..

Hi waveney - "rotations by 90 degrees." by themselves do not elicit the required transformation, unless you mean "followed by rotations by 90 degrees."

There are two transformations used by GZ; it's either a reflection across the vertical or a reflection across the NW-SE diagonal.
(I appreciate that your some of your definitions fit these transformations).

Fermats Brother

[GendoIkari]

So far I count 47 anti-clockwise spirals, and 75 clockwise spirals that I've rated out of 270+.  So in other words, I'm backwards from the norm.  I wonder if the fact that I'm left handed has anything to do with it?

Not including me, it may be because there are image artifacts that tend to run diagonally across the images.  It may cause people to see arms spiraling in.  If I see anything that looks like spiral arms, I'm including it as a spiral, because IMO ellipticals shouldn't have shape irregularites other than dark dust lanes.  The one exception would be if an elliptical had recently had a close encounter with another galaxy.  Then it would become deformed.  I've used the Colliding Galaxies simulator for several years, so I'm always looking for the fainter stuff in these shots to provide clues to the galaxies actual structure.  I think many spirals are being checked as elliptical because most people probably aren't looking for subtle features.

[daniel]

I seem to remember that most right-handed people will draw a circle or make an "O" using an anti-clockwise motion (try it). There are all sorts of possible reasons for this - instruction, hand anatomy, hemisphere dominance. It may be that it is easier to perceive that a galaxy is anti-clockwise so those are more likely to be correctly classified. That is, when a mistake is made it is more likely to be one of classifying a clockwise galaxy as anti-clockwise than the opposite. This would lead overall to an excess of anti-clockwise classifications.

[simona72]

Thank you Alice! I was  about to send out a messagge asking clarifications as I found the same problem...but you clarified this to me!
Ciao!

[Nicky666]

Glad I found this thread instead of making another one about the topic 

The outcome of that "bias study" is pretty interesting.

[snowstargazer]

I seem to remember that most right-handed people will draw a circle or make an "O" using an anti-clockwise motion (try it). There are all sorts of possible reasons for this - instruction, hand anatomy, hemisphere dominance. It may be that it is easier to perceive that a galaxy is anti-clockwise so those are more likely to be correctly classified. That is, when a mistake is made it is more likely to be one of classifying a clockwise galaxy as anti-clockwise than the opposite. This would lead overall to an excess of anti-clockwise classifications.

Very interesting. I tried it and I went anti-clockwise.

[Alice]

That is really cool. I'd love to see how classifications compare to hand dominance. I always write O's anti-clockwise. And I used to make circling motions with the mouse to determine the galaxy's direction . . .

I also read an interesting piece of research - on confabulation, of all things - in New Scientist. People were asked to select their favourite shirt out of a bunch of identical shirts. The point of the article was that, when asked why they liked the one they'd picked, they'd waffle about the texture or whatever when it was identical (I bet this wasn't true, personally. I've worked in a charity shop and all clothes feel different). What I noticed, being a zooite, is that they tended to pick the one on the right. I wonder if the ACW button being on the right (or simply in the middle, to be pedantic) has anything to do with it?

[Alice]

Do we still need this topic? What do people think?

[fluffyporcupine]

i guess its not really relevant any more

[Alice]

Thanks both

On the other hand, people are still being pointed to Zoo 1, too. What do you zooites think? Let me know if this topic is still useful.

[snowstargazer]

Thanks both

On the other hand, people are still being pointed to Zoo 1, too. What do you zooites think? Let me know if this topic is still useful.

It's interesting. Why not leave it? If you are in need of space, I'd suggest removing 6,000 Hi and Goodbyes from chat.