Experimental Design B/C

[Slarik]

This brings up a good point: sadly, the goal of Experimental Design is not to do good science, but to do a good write-up. It's often to your advantage to intentionally introduce some nice, controlled form of error that you can point to and say, "Look, here's [error] that we could fix by doing [procedure step] differently!"

Too true.

My parents, both scientists, hate this event, because its short timeframe and rigid rubric encourage data falsification and doing experiments where you know what results to expect

To the first bolded: I think I disagree with your parents a bit. I find it a lot of fun, despite the not-so-great aspects.

However, the first couple ED contests I was at, we were given some topics that were a lot of fun to design an experiment on (I love the design aspect, coming up with an experiment is so much fun). If though, my first contest was this year's regionals, I don't think I would have enjoyed this event quite as much -- we were given water, salt, dye, a paper towel, a spoon, a stopwatch (and a couple other things) and told to design an experiment dealing with capillary action. To me, an experiment like that doesn't interest me very much because they seem to have given you materials with an experiment in mind (dye various concentrations of salt water, measured with the spoon, and record the time it took the water to flow up a given distance of the paper towel) and have given a topic that basically enforces you do that experiment.

To the second bolded, while true, I've found a lot of times (like at least half) I *don't* get the results I expected.

[Slarik]

I guess a problem with multiple trials of paper airplanes will have human error because they can't be folded exactly the same way, but that could go into the errors.

Not if you do all the trials with the same plane.

[Phenylethylamine]

This brings up a good point: sadly, the goal of Experimental Design is not to do good science, but to do a good write-up. It's often to your advantage to intentionally introduce some nice, controlled form of error that you can point to and say, "Look, here's [error] that we could fix by doing [procedure step] differently!"

Too true.

My parents, both scientists, hate this event, because its short timeframe and rigid rubric encourage data falsification and doing experiments where you know what results to expect

To the first bolded: I think I disagree with your parents a bit. I find it a lot of fun, despite the not-so-great aspects.

They don't argue that it's not fun; I originally enjoyed this event quite a lot. They just believe it's encouraging scientific practices that are sloppy at best and dishonest at worst, among students who are likely to go into the sciences and do actual research in the future.

However, the first couple ED contests I was at, we were given some topics that were a lot of fun to design an experiment on (I love the design aspect, coming up with an experiment is so much fun). If though, my first contest was this year's regionals, I don't think I would have enjoyed this event quite as much -- we were given water, salt, dye, a paper towel, a spoon, a stopwatch (and a couple other things) and told to design an experiment dealing with capillary action. To me, an experiment like that doesn't interest me very much because they seem to have given you materials with an experiment in mind (dye various concentrations of salt water, measured with the spoon, and record the time it took the water to flow up a given distance of the paper towel) and have given a topic that basically enforces you do that experiment.

This is how the event is usually run. Alternately, they'll give you something completely open-ended that really doesn't give you much to work with. The right balance between specificity and freedom is hard to achieve, and event writers are busy people.

To the second bolded, while true, I've found a lot of times (like at least half) I *don't* get the results I expected.

Yes, and if you can't explain why, that's a problem for your write-up, isn't it? The event requires you to do an experiment where you know what trend you should get, so you know when something has gone wrong and can explain the departure from the appropriate trend in your conclusion.

[Slarik]

They don't argue that it's not fun; I originally enjoyed this event quite a lot. They just believe it's encouraging scientific practices that are sloppy at best and dishonest at worst, among students who are likely to go into the sciences and do actual research in the future.

I see your point, and I think it is a good one; I just think ED is encouraging many good things as well, enough to outweigh the negative aspects.

This is how the event is usually run. Alternately, they'll give you something completely open-ended that really doesn't give you much to work with.

How does open-ended leave you not much to work with? Wouldn't a broader, more open-ended topic leave you more possible experiments?

The right balance between specificity and freedom is hard to achieve, and event writers are busy people.

Yes, they are busy, and I'm very thankful to those that take the time to write and grade tests. But, I do like topics where the experiment you do isn't "given to you."

To the second bolded, while true, I've found a lot of times (like at least half) I *don't* get the results I expected.

Yes, and IF you can't explain why, that's a problem for your write-up, isn't it? The event requires you to do an experiment where you know what trend you should get, so you know when something has gone wrong and can explain the departure from the appropriate trend in your conclusion.

That's a big if. I think usually, the experiments are simple enough that you can figure out why (even if you fail to identify it correctly, as sometimes happens in real life, you still get points for it.)

[Phenylethylamine]

This is how the event is usually run. Alternately, they'll give you something completely open-ended that really doesn't give you much to work with.

How does open-ended leave you not much to work with? Wouldn't a broader, more open-ended topic leave you more possible experiments?

"Doesn't give you much to work with" as in "doesn't give you any particular direction to go in designing your experiment." I have seen events where, despite the rules, you are given just the materials and no topic, or a topic like "physics" that's totally unhelpful.

To the second bolded, while true, I've found a lot of times (like at least half) I *don't* get the results I expected.

Yes, and IF you can't explain why, that's a problem for your write-up, isn't it? The event requires you to do an experiment where you know what trend you should get, so you know when something has gone wrong and can explain the departure from the appropriate trend in your conclusion.

That's a big if. I think usually, the experiments are simple enough that you can figure out why (even if you fail to identify it correctly, as sometimes happens in real life, you still get points for it.)

My point is, you still have to know in advance what the trend should be in order to even recognize that something has gone wrong.

[mnstrviola]

At states we were given a large variety of objects.
Chemicals such as H20, HCl, and stuff with sodium and calcium is what I can remember, there were about 10 bottles.
6 Dice
3 Toy Cars
1 Ramp
1 container
1 basket
2 bouncy balls
1 thermometer that was really big
a test beaker
... and some other stuff

We were told to do an experiment on rate, which is pretty open as any good experimental design experiment is based on a rate

We did "how does the height a ball is dropped affect how long it takes for it to stop bouncing"

EDIT: the result: 24th out of 25th

[deezee]

I guess a problem with multiple trials of paper airplanes will have human error because they can't be folded exactly the same way, but that could go into the errors.

Not if you do all the trials with the same plane.

oh I see.

[Slarik]

EDIT: the result: 24th out of 25th

Did you forget to write your school's name on some of the papers?

[mnstrviola]

No, actually we would have gotten 4th place. We were put in second tier because we needed to have 5 materials in our experiment. We only listed the ball, ruler, and timer. I so mad

[deezee]

No, actually we would have gotten 4th place. We were put in second tier because we needed to have 5 materials in our experiment. We only listed the ball, ruler, and timer. I so mad

Does the five item rule only apply to your particular experiment, or is it a general rule for the whole event?

[mnstrviola]

to be honest I didn't read the rules they gave us XD . I rather skimmed it. However it must have been in their rules for everyone at the Norcal State tournament. It's not an official rule though.

[deezee]

to be honest I didn't read the rules they gave us XD . I rather skimmed it. However it must have been in their rules for everyone at the Norcal State tournament. It's not an official rule though.

oh alright thanks : )

[EpicFailure]

What would be an example of "Results not relating to the DV"?

[butter side up]

What would be an example of "Results not relating to the DV"?

That would be anything that you noticed that could affect your results, but wasn't actually your dependent variable.
For example, using the generic 'change the height of a ramp and measure the distance a car will travel off it' some observations would include:
At the 30 cm ramp, the car had a pronounced tendancy to veer sharply to the right.
At the 10 cm ramp, the car was jolted each time it crossed the transition between the ramp and the floor.
There were some irregularities on the floor that caused slight changes in the direction and speed of the car.

These observations are often closely related to the experimental errors.

[XJcwolfyX]

FOURTH IN THE NATION YES!!!!!!!!!