Experimental Design B/C

[butter side up]

What does it mean that variables must be "operationally defined"?

We always took operationally defined to mean not only describing the units and whatnot, but also to explain exactly how we measured the variable. For instance, if we were measuring the time it took for a car to stop moving, we would say "...time, measured in seconds from the moment of release until the moment the car came to a standstill, using a stopwatch" or something similar. This is also relevant when measuring distances- as in, "measured in cm from the bottom end of the ramp to the forward-most point of the car." Just describe exactly how you are measuring the variable. Then make absolutely sure your procedure and diagrams match this.

On another note, has anyone ever had live animals given to them as a mandatory part of their experiment? We got mealworms at Solon, and it kind of threw us a curveball. [Mainly because they, being organic creatures, didn't want to cooperate and crawl in a straight line or start moving when they were released.]

Also, if a section is marked for 4 points, and you have answers written that were indicated as good, just not quite enough, how does one get 5 points taken off? [We didn't get a rubric, just the scorers notes on the test.]

[deezee]

What does it mean that variables must be "operationally defined"?

We always took operationally defined to mean not only describing the units and whatnot, but also to explain exactly how we measured the variable. For instance, if we were measuring the time it took for a car to stop moving, we would say "...time, measured in seconds from the moment of release until the moment the car came to a standstill, using a stopwatch" or something similar. This is also relevant when measuring distances- as in, "measured in cm from the bottom end of the ramp to the forward-most point of the car." Just describe exactly how you are measuring the variable. Then make absolutely sure your procedure and diagrams match this.

On another note, has anyone ever had live animals given to them as a mandatory part of their experiment? We got mealworms at Solon, and it kind of threw us a curveball. [Mainly because they, being organic creatures, didn't want to cooperate and crawl in a straight line or start moving when they were released.]

Also, if a section is marked for 4 points, and you have answers written that were indicated as good, just not quite enough, how does one get 5 points taken off? [We didn't get a rubric, just the scorers notes on the test.]

Yup.

Also, if your are measuring, say, the flexibility of wood, your definition would be "flexible is defined as being able to bend..."

Stuff like that. Basically, you define your variables.

[BoldlyGoingNowhere]

What exactly is regression analysis? And how do you do it? This is a first time event for both of my partners and I and we're really confused -.-

[AlphaTauri]

Regression analysis, as I understand it, is finding the least squares regression line (aka line of best fit) for your data. A TI-30X IIS will do linear regression if you plug your data into 2-VAR STAT - you want a, b, and r.

y = ax+b, where:
a is your slope
b is your y-intercept

r is your correlation coefficient. Squaring it will give you r^2, the coefficient of determination (basically how much of the variation in y-values over different x-values is explained by your regression line; the better your regression, the closer r^2 is to 1).

As far as I know, TI does not have a non-graphing calculator that will do non-linear regressions, so if your data is exponential (something to the xth) or a power function (x to the somethingth), you might have to play around with raising your x or y values to an exponent or log-ing them, then running a linear regression. If you do this, be very careful to keep straight what your data values mean - x might not be linearly related to y, but log(x) or x^2 may be.

[BoldlyGoingNowhere]

Thanks! That makes way more sense.
Does anyone know what is/how to do measure of central tendency?
I have quite a lot of questions on this topic

[AlphaTauri]

Measure of central tendency is basically mean, median, and/or mode.

Means are good for generally symmetrically distributed data, and even data that's a little skewed one way or the other. However, large outliers in one direction will skew your mean (because of the way it's calculated), so I'd caution against using it if you see large outliers.
Median can also be used for roughly symmetrical and skewed data, and it CAN be used with data sets with a couple large outliers (since median is less affected by outliers than mean).
Mode may or may not be useful, depending on your data. In my opinion, it doesn't really do much.

And don't worry about asking too many questions - that's what this forum is here for, anyways. :)

[Shad160]

For Applications and Recommendations for Further Use, we always trip up at "Suggestions for other ways to look at hypothesis given", and "Suggestions for future experiments given"

I've always interpreted the latter to be other experiments we could do using the supplies provided, but what about the former? Or am I completely wrong about the latter, too? Any help would be appreciated

[foreverphysics]

We did none of the things that you guys have been talking about...we got fourth.

[bookworm2110]

I'm not sure how to delete messages....

[EpicFailure]

For Applications and Recommendations for Further Use, we always trip up at "Suggestions for other ways to look at hypothesis given", and "Suggestions for future experiments given"

I've always interpreted the latter to be other experiments we could do using the supplies provided, but what about the former? Or am I completely wrong about the latter, too? Any help would be appreciated

I might be wrong but I've always thought "Suggestions for other ways to look at the hypothesis given" meant listing other possible experiments to examine the same hypothesis. So if the experiment is testing the area of plastic bag parachutes vs. time of descent, a possible experiment to examine the same hypothesis would be testing the area of cloth parachutes vs. time of descent. I think "Suggestions for future experiments given" just means related experiments, not necessarily with the same materials. With the parachute example, a related experiment would be testing mass of object in the parachute vs. time of descent.