Galaxy Zoo Starburst Talk

What did you expect of Phase 2 of the Quench project? Hope? Wish? The 'doing' of science: perception vs reality ...

  • JeanTate by JeanTate

    There's no part of Quench Talk called 'discuss the nature of the project itself', or something similar. Yet that's what I'd like to discuss; if a mod thinks this thread is better somewhere else, please move it.

    Under the heading "Stage 2: Begins Now!", the current main page of the Quench project says:

    In this stage we will begin the 'Data Analysis and Discussion' portion of the project. Nobody has ever attempted an online science project quite like this before, so it is going to be an interesting few weeks!

    I'm very interested in knowing what you, zooites reading this (and presumably interested in this - and later - phases of the project) expected, when Stage 2 began; what were your hopes? fears? what did you expect of this stage, in terms of what you understood about how the science of astronomy is actually done?

    I'll be writing a few posts of my own expectations, hopes, fears, etc in a bit; for now, perhaps some more specific questions that might stimulate discussion:

    • how much 'number crunching' did you think might be involved at this stage?
    • did you expect/hope/wish that you'd have access to the 'raw data'?
    • did you even know what sort of thing the 'raw data' might be?
    • how well did you think you understood the scientific/astronomical context of the questions which this project might answer (to some extent)?
    • how daunting a prospect was it, for you, to try to figure out how Tools works?
    • what was your attitude towards Tools (or other other tools/techniques for astronomical analysis/data reduction)? For example, did you think you could apply what you already knew about databases or spreadsheets (to pick just two examples)?
    • what role(s) did you envision the SCIENTISTs would play in this stage?

    Posted

  • mlpeck by mlpeck

    I was expecting the credentialed scientists who are directly involved in the project to be actively engaged with the "citizen scientists," since that is what the project is ostensibly about.

    Posted

  • JeanTate by JeanTate

    A bit about my own expectations/wishes/hopes/fears/etc.

    Of all the stages/versions/projects/etc which were launched after the original Galaxy Zoo, the Quench project is the one which most excited me (though Letters comes a close second). Why? Because I expected to be an active participant in all stages of some astronomical research which explicitly aimed to result in the publication of a paper (possibly not in a peer-reviewed journal such as MNRAS, but good enough to submitted to such a journal, with a reasonable chance of being accepted).

    By actively participating in this project, I hoped to learn - by first-hand experience, by doing rather than reading about it - what's actually involved in data reduction, data analysis, formation and testing of hypotheses, deciding what to put into the paper (and what not), writing the first draft (collaboratively), revising and editing, doing a literature search, and making decisions - collaboratively - on the final draft.

    Sure, this wasn't going to be the same as what happens when a team of astronomers collaborate - this is an online, open project, and the participants are both professional astronomers (whether formally SCIENTISTs or not; whether formally members of the Galaxy Zoo team or not) and ordinary zooites - but I expected that the key aspects of the work/research would be essentially the same.

    Some parts of the actual 'number crunching' - the data reduction and analysis - I was particularly looking forward to; as is obvious from the numerous posts I've written, this includes the identification of outliers and anomalies, making decisions on what to do about them, and how they are presented and discussed in the paper. Also, what statistical tests should be considered, and why; particularly as some of the key data is not explicitly quantitative (the classifications).

    (more later)

    Posted

  • jules by jules moderator

    I was hoping more volunteers would be involved - volunteers with varying data analysis skills. I thought moderating might be more onerous as I expected Talk to be busier. As it is Talk is pretty much self moderating. I expected I would learn how to use Tools and do some basic plots. I wanted to learn more about colour-colour plots and BPT diagrams and to play with the data looking for unexpected patterns. I also expected volunteers with more skills than me to forge ahead enabling me to learn from them. To a degree I have done all of these things. I just didn't expect there would be only a handful of us. That makes the project so much more difficult for us volunteers as we have more work to do. Finally, this is very unfamiliar ground for me so feedback from the team is essential.

    Posted

  • mlpeck by mlpeck in response to JeanTate's comment.

    did you expect/hope/wish that you'd have access to the 'raw data'?
    did you even know what sort of thing the 'raw data' might be?

    What's your understanding of what the raw data might be?

    The really raw data are things like electron counts from CCDs, which unless you want to redo the photometric and spectroscopic data reduction are probably not what you really want.

    What I'd like to see are the data that were used to select the quench sample. I'm guessing it was a set of principal components scores from Chen et al.'s population models. As of DR10 those are available from the CAS jobs website -- there's a table named stellarMassPCAWiscBC03 and another one named stellarMassPCAWiscM11.

    I've spent some time trying to reverse engineer how the sample selection might have been made based on the assumption that it was done using the PC coefficients in one of these tables, but I haven't succeeded yet. I have determined that their models have no ability to predict the properties of emission lines in the spectra, which at least partially explains why so many "quenched" galaxies are forming stars.

    Posted

  • JeanTate by JeanTate in response to mlpeck's comment.

    Me: did you expect/hope/wish that you'd have access to the 'raw data'? did you even know what sort of thing the 'raw data' might be?

    mlpeck: What's your understanding of what the raw data might be?

    I was trying to foster a discussion; I have little idea what my fellow zooites think 'raw data' might be, and none as to how much of an expectation there was about access to it.

    For me, I know that the cleaned and calibrated SDSS photometric and spectroscopic data (plus the meta-data) - in the form of compressed FITS files - are available for download, so if anyone had expressed an interest in that, I'd've been only too happy to describe my understanding of how to get it.

    I had expected to be able to get a more fine-grained version of the classification data, at least to the level of that found in the GZ2 catalog (as described here for example).

    Posted

  • JeanTate by JeanTate in response to JeanTate's comment.

    re Tools: initially I hoped that they would prove useful, as a means of doing number crunching and sharing the results. Being new, I expected that there'd be problems, and that it would take a while - and possibly several versions/revisions - to get rid of the bugs that were obviously there.

    (more later)

    Posted