[Transcript: Understanding Sample Size – Science Fair Club Video #3]

Jack:
Hello, my name is Jack, and I am a STEM instructor at Creative Encounters with Science, conducted through the University of Guelph.

In our last video, we reviewed how to implement safety and ethics into your science fair project, which we discussed in our last Science Fair Club meeting. If you missed the last video, click the link in the description or visit our website to learn more about our free virtual Science Fair Club.

In today's video, we'll be talking about samples—and not the kind you get for free in the grocery store!

Put simply, sample size is the number of times you run your experiment. In order for your experiment to be a fair test—which means all variables are controlled for except your dependent variable—your sample size has to be reasonable.

More specifically, your sample size is the number of trials or tests you do with each condition of your independent variable, or simply the number of participants in your experiment.

If you would like to review the role of independent and dependent variables in your science fair project, click the link in the description to watch our recent video on the scientific method.

First, let's talk about the radish plant experiment discussed in our previous videos.

This experiment contained four plants, each placed in a different environment, exposing them to different amounts of light. In this experiment, we are testing the growth of the plant under certain conditions over time.

In the first condition, the plant was placed under an ultraviolet light for the whole experiment.
In the second condition, another plant alternated between ultraviolet light and the lower intensity light in the classroom.
In the third condition, the plant stayed under the classroom lights at all times.
And in the final condition, the plant was permanently placed underneath a cup in complete darkness.

As you can see, you need to expose your test subjects to a spectrum of different conditions of your independent variable.

In this example, we had two plants on the extremes of the spectrum: one received high intensity ultraviolet light every single day, and the other was not exposed to any light.

We could have done this experiment with just these two plants, but that doesn't really tell us anything about what would happen to a plant which received an intermediate or medium amount of light. So we added two more plants to see exactly where a change in plant growth occurs.

In this case, the sample size is represented by the number of conditions. Since there is only one plant for each condition we are testing, we have a sample size of four.

For another example, let's say you're testing whether the size of a paper airplane's wings will affect how far it travels.

First, you make a paper airplane with larger wings, and one with smaller wings. Then you throw each of them once, and the plane with smaller wings went a further distance than the plane with longer wings.

This must mean that smaller wings make a glider travel farther, right?

Well... maybe not.

How do you know that the glider with small wings will always travel farther than the glider with larger wings—if you only tested both of them once?

If you pick a really small sample size like this, your results will likely be inaccurate. It's a lot like flipping a coin. You can't assume that because you got tails once, you will get tails every single time. Because if you flip it, say, 29 more times, you'll land on tails about 50% of the time.

You should come up with a sample size large enough that you can see a specific trend and variation in your results.

That being said, you shouldn't throw each glider 100 times—because that would take too long, and you'd probably get bored of it.

A good number of trials for each condition of your independent variable would be around 30 in total. In this example, that means 30 throws for each plane.

Some of you might be wondering: How do I come up with a sample size for an experiment with multiple participants in each condition?

When coming up with a survey, you have to consider both the number of questions it will ask and how many people you want to fill it out.

It's important that you only ask questions that are related to what your project is studying.

For example, if you are conducting a survey with a participant over video chat, remember to stay focused on getting the information you need—without getting off track with your experiment.

Keep your hypothesis or research question in mind when you're coming up with your survey.

Now, when considering the number of responses you need, remember that the fewer people you ask, the bigger the chances are that your results may not correctly reflect the truth.

This is why you will likely not get accurate results if you only conduct one trial of the paper airplane experiment.

But what if your survey is comparing the responses from two different groups?

Let's say: adults versus kids.

In this case, you would want to have an equal number of kids and adults doing your survey. Because if there are more kids than adults, or vice versa, the group with more participants will give you more data, and you'll get more accurate results for the larger group than the smaller group.

To categorize your participants into the two groups, like adults and kids, make sure that you ask each participant to specify which age category they fall into in your survey.

In this video, we covered:

Using the number of experimental conditions as a sample size

Using the number of trials of each condition as a sample size

And using the number of participants in your experiment as a sample size

If you are having trouble with identifying an appropriate sample size for your experiment—or any other component of your science fair project—please email us at ces.scifair@uoguelph.ca, and I or another member of our team will be happy to help you.

We will be releasing one more video about logging data in the next few days, so keep an eye out for that.

Also, our virtual Science Fair Club meetings will take place every other Monday in the new year, starting on January 4th.

Visit our website to sign up for our free Virtual Science Fair Club on Zoom video chat, and follow our Instagram at @cesclubs to get frequent updates on all of our club initiatives.

Creative Encounters with Science is a proud member of the Actua Network, and we would like to thank our parent company as well as all of their sponsors for making Science Fair Club possible.

I hope you enjoyed this video, and thank you for watching!