2019 Golden Gate Invitational

Rossyspsce
Member
Member
Posts: 210
Joined: June 3rd, 2018, 5:32 pm
Has thanked: 8 times
Been thanked: 4 times

Re: 2019 Golden Gate Invitational

Post by Rossyspsce »

primitive_polonium wrote:Post Competition Blurb

Hello everyone! I'm the Protein Modeling event supervisor. It was a great experience; really glad to have come out. The GGSO directors board was crazy organized with everything and super nice as well. Protein can be a lot to run, but they made it manageable and breezy (or as breezy as it can be). I talked to a few of you competitors after the competition, and y'all were also very sweet. It's clear you're really invested and want to learn stuff, which warms my heart tremendously.

Anyway, some general comments. I can't divulge much about raw scores, and I most certainly can't divulge anything about MSOE-released materials.

Stats
Here's an unlabeled-axes-graph of overall scores. I know there's a bit of mystique about onsites, so here's an unlabeled-axes-graph of onsite build scores.

The average on the written exam was ~45%, and there was a good spread from ~15% to ~75%. I generally aim for averages of 50%, and for the highest score to be in the 70-80% range. I'm fairly satisfied with my spread. I just wish I had less ties, but the chances of teams getting the same overall score on all three sections is next to none so I didn't have to use any tiebreakers.

Comments about Performance
As you all know, the scoring rubric was recently changed from 40/30/30 to 20/40/40. This happened 3 days before GGSO, so the exams had already been printed and retained the original breakdown. The onsite rubric was out of 30, but it had to be normalized to 40 for the new system.

I did not grade the onsite section (I trained a volunteer to do it for all the entirety of the day), so I can't comment too specifically about this. Looking at all of them at the end of the day, and generally watching over the kids during the testing periods, my first impression was that at least a few teams had some kind of time crunch with these. There was one onsite that was basically a swirl with endcaps, and I noted that some teams waited until the last 10 or so minutes to go get their endcaps. However, most teams did get the general shape down. Onsite scores were fairly close together.

As for the written exam, I was generally satisfied with performance. Most teams were able to at least make an attempt at every question. I don't think anyone got a perfect score on the MC, but one or two teams came very very close. A good handful broke 10/15. The exam was originally harder, having more advanced protein topics (cold denaturation, thermo). While writing the test, I was debating between an exam which had a deeper testing of concepts and did not conform to MSOE's format, and one that was in the MSOE format. The former is better at gauging understanding and less prone to ties since it isn't constrained to 30 points, but it also take practice to ace the MSOE format. I ended up settling for something that was MSOE format, but had a few more MC questions than MSOE would normally give. The exam was made easier after the MIT Invite. I was one of the people involved in organizing Protein Modeling, and the distribution of scores and overall performance gave me a sense of how much high schoolers would realistically know.

The prebuild...I graded all the prebuilds. A good number of teams came afterwards to ask for prebuild tips. Now, I can't divulge the specifics of the rubric. What I will say is this: the most important part of prebuilding is making sure that the shape of your prebuild isn't going to change during transport, grading, and general handling.

One thing I noted was that some teams would use things like stands or string their prebuild to a box top. Nothing wrong with presentation. However, if you're going to suspend your prebuild from something, you're going to have to secure it from all angles. You really don't want gravity coming in and possibly deforming your build; aim for as few degrees of freedom as you possibly can. It's also a pain (on the part of the grader) to have to maneuver and untangle while grading, and you can and will lose points for that. If you are going to suspend it from a display, make sure that you have strings attached from the bottom and the top so that we can rotate it without too much harm to the prebuild or inconvenience to us. I saw a really nice suspended prebuild. It was attached to two pieces of wood, and the strings were pulled so taut I could freely rotate it without worrying at all about harming the protein. (The same idea goes if you decide not to use a display; make sure you have supports and things. I handled a few lone toobers that bounced whenever I picked them up.)

Most prebuilds were decent. Most common "room for improvement" was basically refining the tertiary structure and making sure the model looks like the jmol from all angles.

I also got some questions about creative additions, or specifically ideas for them. Very important note: "hydrogen bonds", "N/C termini", "alpha helices", "beta sheets", etc do not count as creative additions. When you think about creative additions, think about things unique to AcrII4A. Creative additions are kind of like extra credit. Every protein has hydrogen bonds, and you're expected to model secondary structures in your model to get the non-extra credit points. They are not creative additions.

Conclusion
Not much. I welcome any questions/comment about my exam, GGSO PMod, or just PMod in general.

Mandatory reminder that 1) there is a new scoring rubric (20/40/40) and that 2)you are allowed only 1 piece of cheat sheet for this exam.
Appreciate the score distributions, although with unlabeled axes, anyone know if we can get some more, specifically builds?
primitivepolonium
Member
Member
Posts: 53
Joined: August 3rd, 2013, 9:00 am
Division: Grad
State: CA
Has thanked: 0
Been thanked: 0

Re: 2019 Golden Gate Invitational

Post by primitivepolonium »

Rossyspsce wrote:
primitive_polonium wrote:Post Competition Blurb

Hello everyone! I'm the Protein Modeling event supervisor. It was a great experience; really glad to have come out. The GGSO directors board was crazy organized with everything and super nice as well. Protein can be a lot to run, but they made it manageable and breezy (or as breezy as it can be). I talked to a few of you competitors after the competition, and y'all were also very sweet. It's clear you're really invested and want to learn stuff, which warms my heart tremendously.

Anyway, some general comments. I can't divulge much about raw scores, and I most certainly can't divulge anything about MSOE-released materials.

Stats
Here's an unlabeled-axes-graph of overall scores. I know there's a bit of mystique about onsites, so here's an unlabeled-axes-graph of onsite build scores.

The average on the written exam was ~45%, and there was a good spread from ~15% to ~75%. I generally aim for averages of 50%, and for the highest score to be in the 70-80% range. I'm fairly satisfied with my spread. I just wish I had less ties, but the chances of teams getting the same overall score on all three sections is next to none so I didn't have to use any tiebreakers.

Comments about Performance
As you all know, the scoring rubric was recently changed from 40/30/30 to 20/40/40. This happened 3 days before GGSO, so the exams had already been printed and retained the original breakdown. The onsite rubric was out of 30, but it had to be normalized to 40 for the new system.

I did not grade the onsite section (I trained a volunteer to do it for all the entirety of the day), so I can't comment too specifically about this. Looking at all of them at the end of the day, and generally watching over the kids during the testing periods, my first impression was that at least a few teams had some kind of time crunch with these. There was one onsite that was basically a swirl with endcaps, and I noted that some teams waited until the last 10 or so minutes to go get their endcaps. However, most teams did get the general shape down. Onsite scores were fairly close together.

As for the written exam, I was generally satisfied with performance. Most teams were able to at least make an attempt at every question. I don't think anyone got a perfect score on the MC, but one or two teams came very very close. A good handful broke 10/15. The exam was originally harder, having more advanced protein topics (cold denaturation, thermo). While writing the test, I was debating between an exam which had a deeper testing of concepts and did not conform to MSOE's format, and one that was in the MSOE format. The former is better at gauging understanding and less prone to ties since it isn't constrained to 30 points, but it also take practice to ace the MSOE format. I ended up settling for something that was MSOE format, but had a few more MC questions than MSOE would normally give. The exam was made easier after the MIT Invite. I was one of the people involved in organizing Protein Modeling, and the distribution of scores and overall performance gave me a sense of how much high schoolers would realistically know.

The prebuild...I graded all the prebuilds. A good number of teams came afterwards to ask for prebuild tips. Now, I can't divulge the specifics of the rubric. What I will say is this: the most important part of prebuilding is making sure that the shape of your prebuild isn't going to change during transport, grading, and general handling.

One thing I noted was that some teams would use things like stands or string their prebuild to a box top. Nothing wrong with presentation. However, if you're going to suspend your prebuild from something, you're going to have to secure it from all angles. You really don't want gravity coming in and possibly deforming your build; aim for as few degrees of freedom as you possibly can. It's also a pain (on the part of the grader) to have to maneuver and untangle while grading, and you can and will lose points for that. If you are going to suspend it from a display, make sure that you have strings attached from the bottom and the top so that we can rotate it without too much harm to the prebuild or inconvenience to us. I saw a really nice suspended prebuild. It was attached to two pieces of wood, and the strings were pulled so taut I could freely rotate it without worrying at all about harming the protein. (The same idea goes if you decide not to use a display; make sure you have supports and things. I handled a few lone toobers that bounced whenever I picked them up.)

Most prebuilds were decent. Most common "room for improvement" was basically refining the tertiary structure and making sure the model looks like the jmol from all angles.

I also got some questions about creative additions, or specifically ideas for them. Very important note: "hydrogen bonds", "N/C termini", "alpha helices", "beta sheets", etc do not count as creative additions. When you think about creative additions, think about things unique to AcrII4A. Creative additions are kind of like extra credit. Every protein has hydrogen bonds, and you're expected to model secondary structures in your model to get the non-extra credit points. They are not creative additions.

Conclusion
Not much. I welcome any questions/comment about my exam, GGSO PMod, or just PMod in general.

Mandatory reminder that 1) there is a new scoring rubric (20/40/40) and that 2)you are allowed only 1 piece of cheat sheet for this exam.
Appreciate the score distributions, although with unlabeled axes, anyone know if we can get some more, specifically builds?
You mean prebuilds? Ask, and ye shall receive. https://imgur.com/a/EPT5jmE
Div D! I really like chem, oceanography, and nail polish--not in that order.

Troy HS, co2016.

Feel free to PM me about SciOly or college or whatever! I really enjoy making online friends.
Rossyspsce
Member
Member
Posts: 210
Joined: June 3rd, 2018, 5:32 pm
Has thanked: 8 times
Been thanked: 4 times

Re: 2019 Golden Gate Invitational

Post by Rossyspsce »

primitive_polonium wrote:
Rossyspsce wrote:
primitive_polonium wrote:Post Competition Blurb

Hello everyone! I'm the Protein Modeling event supervisor. It was a great experience; really glad to have come out. The GGSO directors board was crazy organized with everything and super nice as well. Protein can be a lot to run, but they made it manageable and breezy (or as breezy as it can be). I talked to a few of you competitors after the competition, and y'all were also very sweet. It's clear you're really invested and want to learn stuff, which warms my heart tremendously.

Anyway, some general comments. I can't divulge much about raw scores, and I most certainly can't divulge anything about MSOE-released materials.

Stats
Here's an unlabeled-axes-graph of overall scores. I know there's a bit of mystique about onsites, so here's an unlabeled-axes-graph of onsite build scores.

The average on the written exam was ~45%, and there was a good spread from ~15% to ~75%. I generally aim for averages of 50%, and for the highest score to be in the 70-80% range. I'm fairly satisfied with my spread. I just wish I had less ties, but the chances of teams getting the same overall score on all three sections is next to none so I didn't have to use any tiebreakers.

Comments about Performance
As you all know, the scoring rubric was recently changed from 40/30/30 to 20/40/40. This happened 3 days before GGSO, so the exams had already been printed and retained the original breakdown. The onsite rubric was out of 30, but it had to be normalized to 40 for the new system.

I did not grade the onsite section (I trained a volunteer to do it for all the entirety of the day), so I can't comment too specifically about this. Looking at all of them at the end of the day, and generally watching over the kids during the testing periods, my first impression was that at least a few teams had some kind of time crunch with these. There was one onsite that was basically a swirl with endcaps, and I noted that some teams waited until the last 10 or so minutes to go get their endcaps. However, most teams did get the general shape down. Onsite scores were fairly close together.

As for the written exam, I was generally satisfied with performance. Most teams were able to at least make an attempt at every question. I don't think anyone got a perfect score on the MC, but one or two teams came very very close. A good handful broke 10/15. The exam was originally harder, having more advanced protein topics (cold denaturation, thermo). While writing the test, I was debating between an exam which had a deeper testing of concepts and did not conform to MSOE's format, and one that was in the MSOE format. The former is better at gauging understanding and less prone to ties since it isn't constrained to 30 points, but it also take practice to ace the MSOE format. I ended up settling for something that was MSOE format, but had a few more MC questions than MSOE would normally give. The exam was made easier after the MIT Invite. I was one of the people involved in organizing Protein Modeling, and the distribution of scores and overall performance gave me a sense of how much high schoolers would realistically know.

The prebuild...I graded all the prebuilds. A good number of teams came afterwards to ask for prebuild tips. Now, I can't divulge the specifics of the rubric. What I will say is this: the most important part of prebuilding is making sure that the shape of your prebuild isn't going to change during transport, grading, and general handling.

One thing I noted was that some teams would use things like stands or string their prebuild to a box top. Nothing wrong with presentation. However, if you're going to suspend your prebuild from something, you're going to have to secure it from all angles. You really don't want gravity coming in and possibly deforming your build; aim for as few degrees of freedom as you possibly can. It's also a pain (on the part of the grader) to have to maneuver and untangle while grading, and you can and will lose points for that. If you are going to suspend it from a display, make sure that you have strings attached from the bottom and the top so that we can rotate it without too much harm to the prebuild or inconvenience to us. I saw a really nice suspended prebuild. It was attached to two pieces of wood, and the strings were pulled so taut I could freely rotate it without worrying at all about harming the protein. (The same idea goes if you decide not to use a display; make sure you have supports and things. I handled a few lone toobers that bounced whenever I picked them up.)

Most prebuilds were decent. Most common "room for improvement" was basically refining the tertiary structure and making sure the model looks like the jmol from all angles.

I also got some questions about creative additions, or specifically ideas for them. Very important note: "hydrogen bonds", "N/C termini", "alpha helices", "beta sheets", etc do not count as creative additions. When you think about creative additions, think about things unique to AcrII4A. Creative additions are kind of like extra credit. Every protein has hydrogen bonds, and you're expected to model secondary structures in your model to get the non-extra credit points. They are not creative additions.

Conclusion
Not much. I welcome any questions/comment about my exam, GGSO PMod, or just PMod in general.

Mandatory reminder that 1) there is a new scoring rubric (20/40/40) and that 2)you are allowed only 1 piece of cheat sheet for this exam.
Appreciate the score distributions, although with unlabeled axes, anyone know if we can get some more, specifically builds?
You mean prebuilds? Ask, and ye shall receive. https://imgur.com/a/EPT5jmE
Appreciate it but I meant the builds like mousetrap, weigh stuff, mission, etc
Rossyspsce
Member
Member
Posts: 210
Joined: June 3rd, 2018, 5:32 pm
Has thanked: 8 times
Been thanked: 4 times

Re: 2019 Golden Gate Invitational

Post by Rossyspsce »

Sorry for double post, but has anyone recieved pictures/know where to get them
User avatar
onoga17
Exalted Member
Exalted Member
Posts: 13
Joined: February 5th, 2019, 10:32 pm
Division: C
State: CA
Has thanked: 0
Been thanked: 2 times
Contact:

Re: 2019 Golden Gate Invitational

Post by onoga17 »

Rossyspsce wrote:Sorry for double post, but has anyone recieved pictures/know where to get them
Pictures were uploaded recently to the GGSO website under the "Announcements" tab. Here's a direct link to the folder containing the images: https://drive.google.com/drive/folders/ ... XMLduDc9iD
Post Reply

Return to “2019 Invitationals”

Who is online

Users browsing this forum: No registered users and 1 guest