At the start of the session we were posed the question as to whether or not it was possible to measure more than one variable at a time in an experiment. The majority of the class, comprised mostly of engineers and scientists, gave the answer that seemed the most logical to them… no. However, the Psychologist in the room started to explain examples of their work and how multivariant experimentation was inescapable in their line of research due to the sheer amount of variables when examining human behaviour.
This was a strange concept to the others in the class who had spent a large part of their undergraduate education trying to minimise all variables apart from one. Our lecturer then proceeded to take out small wooden catapults from black bags and explained that he would explain this concept using toys! The catapults had 3 different variables that you could change: tension, height of cup and launch angle, all of which had an effect on the distance that the projectile travelled.
Before we started playing with the catapult, the lecturer explained to us that to get results as accurate as possible, we should minimise the error in our experiment. To do this we taped down the catapult and the ruler we used to measure to stop them from moving too much, which could interfere with the results. Then we systematically changed the three variables one by one from their minimum to maximum values and measured the distance the ball travelled with these parameters.
Once all the data was collected, our lecturer took us through the theory of using multivariant regression analysis to be able to create an equation that would allow us to set the variables to get our catapults to shoot a certain distance. After this explanation, he then set us a task to obtain specific distances by setting the different variables to values we could obtain from the use of this analysis technique.
As always, the lecturer ended up being right. Pretty much all the catapult shots landed where the regression analysis equation predicted them to land. But then, the inevitable happened when you introduce catapults to students, open combat!
This was nevertheless a valuable lesson that will be taken to the next level in our course where we shall be employing this concept of analysing multiple variables on a much bigger scale in the form of testing a Toyota Prius.