Reminders from TA/LA Training on 10/14/2016, for week of 10/17 (Lab 3, Part 2)
Hello!
I hope your weekend was lovely. :-) Here are the notes from Friday's training (10/14) for the week of 10/17 (Lab 3, Part 2).
1. Recitation: Electrophoresis
For working with "Electrophoresis":
(And check out that cool image of gel structure linked below!)
(It might help students if you discuss, after the recitation has concluded, the power of a 'null' result. Finding out what doesn't work or that a conclusion can't be made is often as powerful (i.e., instructive, generative) as finding something that works!)
From Ben Geller, Oct. 2013: "So Vashti and I spent an hour talking about electrophoresis after I talked to you guys today. We were tying to figure out exactly what agarose looks like at the microscopic level. There's a pic on this page:
http://ocw.mit.edu/courses/biological-engineering/20-109-laboratory-fundamentals-in-biological-engineering-spring-2010/labs/module-1-day-2-purify-aptamer-encoding-dna/
So basically, the little stuff gets through this mesh-like gel more easily than the big stuff. :)
This might be a pic you want to show the students in recitation if, by the end of the hour, they are wondering what the gel actually looks like."
(Break--Pass back Lab 2, if you have it graded... If not, pass it back next week during the break...)
2. Lab 3, Part 2
Link to documents: http://umdberg.pbworks.com/w/page/68933700/NEXUS%20Physics%20Labs%2C%202013-2014
a) Comments still relevant from last week's reminders:
1. The students should all be using the 40x optic, and the 1024 resolution is best. Calibration slides for finding the distance-to-pixel ratio are in the 'my documents' folder, in a 'calibration' folder. For these calibration slides, the '0' and the '1' are separated by 100 microns.2. Tricks for speeding up data analysis (I added this stuff)
i. After the video has been capture by VirtualDub, students can copy the file to the other desktop computer. Now, analyzing this video file on TWO computers, students can track beads in different halves of the video (one computer does the agreed upon frames for beads on the LEFT side of the video and the other computer does the agreed upon frames for beads on the RIGHT side of the video). This cuts the data collection/'harvesting' time in half for the 30-50 beads needed to create good histograms for video #1. Students should not track 'stuck' or 'clumped' beads. Help the students make a judicious selection of equally-spaced frames for tracking from the total frames imported into ImageJ.ii. For tracking beads in videos #2 and #3, far fewer beads need to be tracked, 15-25 should suffice. These two videos can be tracked on different computers to speed the process. Automatic tracking may help. If the quality of the video is insufficient (not enough moving beads, ensemble drift due to convection, bead density too high to track, only clumps present, etc.), then students should take NEW videos. There is no sense in tracking an obviously BAD video. (I emptied the vials and resupplied them from the stock solutions on Friday, so the solutions should be usable now.)iii. For video #1, after the data has been 'harvested' using ImageJ, students should combine their data into a single Excel file. For ALL videos, they should make a BACK-UP COPY of their data before beginning to manipulate with Excel. They should also PLAN AHEAD before trying to do the analysis. Planning ahead will save them a lot of frustration and wasted effort. (You can practice the steps you think are necessary on the sample data file attached to last week's email. Remember, there is not one single way to perform the analysis--so don't lecture your method to the students once you decide what you think is best--but the students should think through a plan to get the from the raw data to the information they will need for the averages and histograms.)iv. Once the data file is at the delta x, delta y, and r stage for all beads, the file can again be split onto two computers so that the histograms for video #1 can be produced more rapidly.v. The analyses of <r^2> vs. t for videos 1, 2 and 3 can each be done individually, on different computers.
b) The plots of <r^2> vs t ought to have error bars, at least vertically. Seeing as the vertical quantity is an average, standard deviation seems an appropriate tool to determine the 'random' uncertainty in the vertical data (i.e., the size of the 'error' bars). Ask students to consider why the error bars get larger for larger time. [Note, this use of 'random' to classify the uncertainty in position is not just because the motion is called 'Random motion.' I can see students getting confused.]
c) What if Automatic Tracking isn't working (either due to a bad video file--too hard to 'clean'--or due to low bead density--can't get clean portions of the video with 10 or more beads that move properly (aren't stuck or clumped) and track cleanly)? In this case, have the students do some manual tracking (the way they have done in previous labs). They can track the beads in every frame or in only a few selected frames, as the data on <r^2> vs. t from video #1 should clearly establish the linearity of the relationship--which will hold true for videos 2 and 3 as well. I would have them track the initial frame, the final frame, and two or three frames in between. (Do these need to be evenly spaced? Think about it. Also, the more frames they choose, the better the value of D they can determine--but they can do a limited number if manually tracking.) They can use the initial position as both zero distance and zero time for these plots. This type of manual tracking (for fewer beads (15-25) and fewer frames (5 or 6 at most)), should be MUCH faster than what they did for video #1.
d) Note that the data file output by Automatic tracking (the multitracker plugin) is formatted differently than that produced by Manual tracking. Therefore, students will need to have a new 'attack plan' for analyzing this data in Excel.
e) I expect every group can accomplish the following in lab this week:
i) Finish histograms of random motion in Video 1--fully characterizing the behavior.ii) Harvest data from videos 2 and 3.iii) Begin to analyze videos 2 and 3 to find their diffusion constants.iv) Students may also have time to: Determine the diffusion constant, D, for the combination of parameters in video #1 (and 4 groups should get roughly the same number! Will the polystyrene people match, too? What does it say if they do?)
v) STUDENTS MUST TAKE THEIR DATA AND FILES WITH THEM WHEN THEY LEAVE LAB!! Every student in the group must have a copy of ALL of their data (excel and word). I will wipe the computers after our training on Friday to make room for other data analysis in the following week.
f) Next week, we will:
i) Finish analyzing all three videos to the <r^2> vs. t stage, finding D for each solution.ii) Compare with other groups to determine which parameters affect the diffusion constant, D, and what the form of D might be. (Making arguments for the form of D from: 1) the experimental data; 2) the physical mechanisms; and 3) dimensional analysis.)iii) Prepare and present their posters to their classmates.iv) Finalize and submit their lab reports for Lab 3.
3) Other Logistics Stuff
a) Don't forget to clean the lab up and leave it in neat shape for the next lab group. You can make your students do most of the work--just give clear instructions!
b) Collect the recitation documents during the break between recitation and lab.
c) Make sure the lab documents don't get all mixed together.
d) Before you leave the lab, check that all the microscopes are turned off (they aren't likely to be used, and so should remain off throughout the lab, but things can get bumped by accident).
e) Get your Lab 2 report scores and your recitation and lab participation scores into ELMS promptly.
Have fun!