Students are always looking for a “right answer” in science, which leads students to be myopic when it comes to analyzing data. “What am I supposed to get” or “What is supposed to happen?” are common lines I hear from students as a result of the quest for a right answer. In labs, the expectation of a right answer results in students expecting that there needs to be an effect in an experiment. An experiment should always result in an increase or decrease in something, right? And when that doesn’t happen, there must be something wrong with the experiment or data, right? That’s not the right way to approach a lab when analyzing data.
The Big Idea
Students need to realize that no effect/answer is sometimes a right answer too. In the late 19th century, scientists believed that space was filled with a so-called “ether” that allowed light to travel between the sun and the earth. Albert A. Michaelson and Edward W. Morley devised a method to measure the effect of the ether on the speed of light. However, when they were analyzing data, they were not able to detect the ether, scientists realized there was no ether at all. That was an amazing discovery – a result of a lab that produced no effect. This is known as a null hypothesis. We teach students all about writing hypotheses and testing hypotheses, but rarely do we ever talk about identifying a null hypothesis. So, how can we have students practice this useful science skill?
To have students practice identifying null hypotheses, we give students a lab where there is a null effect. It’s even better if the null effect conflicts with what students believe (because students will already be looking for a specific result). We use a pendulum lab to illustrate null hypotheses. And, at the end of the post, you can download our handouts.
Life sized example of null hypothesis
The inspiration for our lab comes from a fantastic video of former MIT professor Walter Lewin’s last lecture (where he features his “best of” segments). In it, he swings on a massive pendulum (circa Miley Cyrus’ Cannonball video) to make a point about uncertainty and error. That is, the mass at the end of a pendulum has no effect on the period of a pendulum’s swing. To check out the video, click video clip below (demo starts at 10 minutes in)
In our 1st version of this lab, students create a pendulum (using a ring stand, ring clamp, and string) and hang different masses from the end (starting from 50, 100, 150, 200, and 250g weights). Students start swinging their pendulums from the same height and measure the time it takes to complete 10 periods (ie. cycles). Then, they divide the time by 10 to get the time for 1 period. This version uses items that are already present in most labs. But, a problem that arises is that larger masses can affect the length of the entire pendulum. The length of the pendulum is from the point of rotation to the bottom of the weight. Since 250g weights are longer than 50g weights, the pendulum using a 250g weight is also a longer pendulum. We want to control for length in this lab (because the length of a pendulum does affect period)..
Thus, in our most recent version of this lab, students tie washers at the end of the pendulum instead of using weights (this idea came from a similar lab we saw on Stanford’s website). Students start by tying 1 washer at the end of the pendulum and then adding 1 more for each trial until a maximum of 5 washers. This version requires less mass. And, we can keep the length of the pendulum at a consistent length too (which makes analyzing data easier). By making sure the washers are tied together and that the length between the tops of the washers and the point of rotation is kept consistent, the lengths of the pendulum for all 5 trials are kept consistent.
- Pendulums need to be swung no more than 10 degrees away from rest position. Have students use a protractor to make sure they are not swinging from a greater angle.
- Find washers that are the same size. This way, the increases in mass are the same each time a student adds a washer.
- Have students leave plenty of slack at the end of the pendulum. Students will use the slack to add washers while keeping the length between the washers and pivot constant.
- Remind students that they will find the time of 1 period of rotation by taking their time for 10 periods and dividing by 10. Many students forget and end up measuring one period.
- A lot of students will ask if it is correct that their results fluctuate with no apparent trend. I usually say tell them to run the trial again carefully. If they already have, then I say, “well, those are results then.”
- There is no effect of mass on period. Only length of pendulum and acceleration due to gravity have an effect. But, don’t tell students that.
Putting it all together
A null hypothesis is an underrated yet important part of doing science. Students need to be able to identify it when it happens instead of always expecting a result or trend during an experiment. Ultimately, we want students that are not only good at doing an experiment but also good at analyzing data and concluding what their results mean. And, sometimes, their result may indicate nothing happened. And, that’s fine. That still puts students one step further of where they started. If you want a copy of our lab handouts, click on the link below and enter your email address. You’ll also be added to our email list (if you’re already on our list, awesome!).
Until next time, keep it REAL.