Lab Exercise Rules of Operation
The rules of operation for the labs below will help you get the most out of the lab exercises for this class. In summary, the rules are:
The primary rule for these labs is Work Independently. You can ask simple technical questions of each other (e.g., "what Vista command do I use to read in a FITS image"), but the actual work in the lab, and the various judgments that need to be made, should be your own and developed independently.
There is more to these lab exercises than just getting the answers. The exercises have been tailored to include such important but oft overlooked skills as finding fundamental data in catalogs (either in books or on the Web), identifying fields from finder images, researching alternative sources of information (library, web, etc.), and so forth.
For example, in the Lunar Craters lab given in past years, one of the steps was to identify the big central crater in our images using various Lunar Atlases in the department collection. I have seen cases where one person says something like, "The big crater is Gassendi", which then ruins the exercise for everyone else. I really do want to you struggle (a little) with pattern recognition. You don't learn field identification by consulting the local expert when you get frustrated, you learn it by getting frustrated and then finding your own way through it.
Besides, the crater wasn't Gassendi, it was Manilius, but that didn't stop about half the class. Oops!
[Index]
Once you get a result, check your work over from the start. Even if you are pretty sure of your results. If after doing this you find you cannot trust it (e.g., you get a result you think is bogus, like seeing on an image of CY Aquarii is 50 arcseconds), then check with the professor and see if you two can sort out where the problem lies (see Rule 4).
Beware, however, of falling into the trap of thinking that there is one, true "Right Answer" that you all must get to get a good grade (note the capitals). This is especially true of labs that require subjective judgment calls, like measuring the angular extent of some object with a cursor on a digital image display. There is a well-known effect in science whereby independent observations in an experiment tend to converge to the "Right Answer", even if that answer eventually turns out to be wrong. The most famous example is early measurements of the electron charge-to-mass ratio. After Millikan's landmark oil-drop experiment paper, others sought to refine the measurement of e/m, and after some initial discrepancies, their published results for e/m tended to converge to the original value published by Millikan. Then it was noticed that there was a systematic error in Millikan's original result, and a new, revised value was published. In many cases, many of the other physicists noted that they had originally gotten values of e/m similar to the later (correct) value, but that since their measurement was discrepant from Millikan's, they kept doing it until they got Millikan's [wrong] answer.
So, please don't compare you work with others (see Rule 1). I've seen N different people converge on the same [wrong] answer with the same uncertainty, even though N different (supposedly independent) measurements were involved. Besides, you might spend a lot of time chasing down errors in your work when the error was in theirs!
[Index]
Please show me the work that went into your various computational steps. If you used a spreadsheet, please include your worksheets as a supplement to, but not a substitute for, into your lab notebook to become part of your permanent records. It is not enough to hand in the printout and circle the answer, without comment, or units, etc., and expect me to know what the circle means. The writeup should be a narrative summary of your work, and should read like a brief scientific report, with a clearly defined beginning, middle, and end. The worksheets showing the gross details should be included as an appendix to this report, but are not the substance of the report itself. Imagine you are explaining it to someone other than me who hasn't actually done the analysis in parallel.
Showing your work also means estimating uncertainties where it is sensible or meaningful to do so. Do estimate uncertainties for any measured quantities, and either show how you got these values (e.g., show the error propagation), or briefly describe how you estimated them. Do not, however, go overboard assigning uncertainties to everything and anything. Sometimes it is inappropriate to make up uncertainties (e.g., when given the diameter of an instrumental aperture, take it at face value unless told to do otherwise). Use your judgment.
Another important aspect of "showing the work" is when making comparisons between two quantities. BE QUANTITATIVE. Think of this as the mantra for the lab exercises. Do not content yourself with saying "yes, A is bigger than B...", say by how much, and assess whether the difference is statistically significant given your estimated uncertainties.
Finally, please don't go crazy writing too much or making up uncertainties at random or making frivilous comparisons. Stick to the point and keep it brief (but not so brief as to contain little or no information). This requires judgment on your part, hence why I won't dictate page-length requirements for your reports (this ain't English Lit).
[Index]
The labs involve significant amounts of "live" data, and require you to develop new skills with unfamiliar equipment. As such, they will take you a little more time than you think. While the labs are self-paced, if you leave all the work to the last week of the quarter you are basically doomed.
Pace yourself. Get started right away with the measurement phases of the project (especially those that require use of the class computers), and then leave yourself plenty of time for analysis and re-measurement if necessary.
If you need extra time or help with the computers, please feel free to make the necessary arrangments with the professor.
[Index]
Following on Rule 4, the 5th and final rule is perhaps the most important for success in the labs.
In brief, don't let your time get wasted by spinning your wheels unnecessarily. I want you to work independently, but that is not to say I want you to work in a vacuum. If you get stuck, get help. The professor and TA are available any time (in person, email etc.) to answer your questions and help you complete the labs. You can also ask other grad students, but I'd rather you didn't because while they actually know quite a bit and are eager to help, they really don't know what it is you are doing for the labs, and in their zeal to be helpful they may lead you down the wrong path. This has happened in the past.
What this all means is that you should not hesitate to seek help if you get stuck. The sooner you get back on track, the more time you have to finish the analysis and write-ups of your labs.
[Index]
Copyright © Richard W. Pogge, All Rights Reserved.