Experimental Design B/C

[Phenylethylamine]

Just curious, what do your teams do for the line of best fit?

Even with a non-graphing calculator, you can enter data points into your calculator. I would suggest you consult a calculator manual to figure out if your calculator can do it or not. I use a TI-30X IIS and it gives me all the statistical analysis I need for Experimental Design. For the “line of best fit,” more accurately called the least-squares regression line, all you have to do is record the slope and y-intercept values your calculator spits out for regression line then plug in 3 or 4 points and draw them on your graph. Then you use a straight edge and connect the dots.

In B Division, I believe it is acceptable to just approximate the regression line by eye; however, for C, you definitely need to calculate and accurately represent it.

[andrewwski]

You can calculate regression by hand as well.

It's a pain, but entirely possible and not difficult.

The problem is when you don't know which regression line is the best. Not sure of the context of the event, if it's going to be a linear regression then it's easy. If it forms a non-linear correlation, you have to decide which type it forms, such as exponential or power, and calculate the r values for each type to determine which is a better fit. You can then make it linear and calculate an LSRL if you wish, or calculate the best-fit curve. If you want to calculate r values by hand though, you need to create a residual plot, that's a pain in the neck as well.

That may be beyond the scope of the event though. I'm guessing you probably have to just calculate a best fit line for linear data? In that case, it's easy.

[eyeball138]

Ok, thanks. I really just wanted to double check that you could just estimate for division B.

[Phenylethylamine]

That may be beyond the scope of the event though. I'm guessing you probably have to just calculate a best fit line for linear data? In that case, it's easy.

While you could theoretically choose an experiment that would produce nonlinear data, it's not likely. Almost anything that's possible to do with the given materials is going to give you a linear data set at the end, so in nearly every case, yes, you're working with linear data.

[Flavorflav]

That may be beyond the scope of the event though. I'm guessing you probably have to just calculate a best fit line for linear data? In that case, it's easy.

While you could theoretically choose an experiment that would produce nonlinear data, it's not likely. Almost anything that's possible to do with the given materials is going to give you a linear data set at the end, so in nearly every case, yes, you're working with linear data.

I think you will find that you may not be able to assume that in C division.

[Paradox21]

That may be beyond the scope of the event though. I'm guessing you probably have to just calculate a best fit line for linear data? In that case, it's easy.

While you could theoretically choose an experiment that would produce nonlinear data, it's not likely. Almost anything that's possible to do with the given materials is going to give you a linear data set at the end, so in nearly every case, yes, you're working with linear data.

I think you will find that you may not be able to assume that in C division.

I have definitely encountered non-linear data. You can either take the logarithm of the X values to linearize data if it the original data follows a square root function, or you can take the logarithm of the X values and the Y values to linearize a power function. Other than that I don't really know what you would do. One time I had data that looked more like a logistic function, but I just called it linear. It was close enough to be believable.

[nejanimb]

Actually, I think you'd take the log of the x values if you were looking for a logarithmic function, not a square root function (in which case, you'd just take the square root of the Xs!). You'd take the log of the Ys if you were looking for an exponential function, and both sides (as you suggested) if you were looking for power data. Also extremely common is inverse data, in which you'll just raise all of the X values to the negative first. You might also get inverse square, inverse root, etc.

Getting non-linear data is definitely a possibility. We also try to choose an experiment that'll give us linear data, but sometimes the experiment that clearly would be the best option involves a nonlinear relationship, in which case, you've gotta go for it. The transformations shouldn't be too bad, even without a graphing calculator.

[jacjackson]

My partner and I are total noobs. What are ways people prepare for Experimental Design lab?

[gyourkoshaven]

My partner and I are total noobs.

Way to give yourself credibility

What are ways people prepare for Experimental Design lab?

Obviously start by knowing the scientific method (If you don't already), and just practice doing a lab with 2 others. Write the lab report up and give it to a teacher. Practice is the best preparation.

[Paradox21]

Actually, I think you'd take the log of the x values if you were looking for a logarithmic function, not a square root function (in which case, you'd just take the square root of the Xs!). You'd take the log of the Ys if you were looking for an exponential function, and both sides (as you suggested) if you were looking for power data. Also extremely common is inverse data, in which you'll just raise all of the X values to the negative first. You might also get inverse square, inverse root, etc.

Getting non-linear data is definitely a possibility. We also try to choose an experiment that'll give us linear data, but sometimes the experiment that clearly would be the best option involves a nonlinear relationship, in which case, you've gotta go for it. The transformations shouldn't be too bad, even without a graphing calculator.

Oh my, I really messed up my post (I was doing homework). The square root function is a power function. So you log the X and Y. And if it is exponential, you log the Ys.