Dr. Shawn's Super Science Project Support Blog: How Many Measurements Per Trial?

Remember, doing science is all about not fooling yourself. And so rather than trust our crude and imprecise senses, scientists make measurements. In fact, nearly all science projects absolutely require them. If you're trying to find out how fast plants grow vs. the amount of fertilizer you are giving them, or discover how the temperature of a water bath changes as salt crystals dissolve inside it, or unravel the metabolism of an insect as its diet is varied, you are going to need to make measurements. And probably lots of them!

Each data point that you might plot on your graph represents one "trail." A trail is a mini-experiment within your whole experiment. Sometimes the method you have to measure something just isn't very precise and so you have to repeat the same measurement a number of times to get a number you can have confidence in to plot on your graph. The question here is, how do you know how many measurements you need to make for each trial and how to you find the number to plot?

Here are a couple of rules to guide you.

Simple Trial: You need to determine the value of some physical property of just one single thing: Its weight, or pH, or height--at some time during your experiment. The number of measurements you need to make depends on how accurately you can determine the value.

Rules of Thumb: For example, if you have high confidence that every single measurement you make is accurate to within about one percent or so, measure the quantity twice--just to be certain you get the value both times. That's a good check to make sure you didn't make a mistake If when you make the same measurement a few times in a row you find that find that you find that the answers are always close--say within about 5 percent of each other-- then measure the quantity three times and use the average of those three measurements as your data point. (To calculate an average you simply add the numbers to together and then divide by the number of measurements you added.) If successive values vary by more than five percent, then measure the quantity least five times and take the average.

(Exception: If you are looking for very small changes in something, changes on the order of one percent or less, you will need a different rule to follow. If so, then you're probably doing something pretty advanced. Since most science fair projects don't need highly precise data, I'll skip the subtle complications that can arise for now.)

Large Trials: You need to measure the same quantity for every member of your test and control group. For example, you are testing a new fertilizer and you have decided to grow 30 test plants and 30 control plants. If you measured the height of each plant three times you'd need to make 180 separate measurements for every trail in your experiment!

Now, I would advise a research scientist who intended to publish their data in a peer-reviewed journal to do just that. But I wouldn't advise a student to work that hard because that kind of drudgery can really kill a young person's budding interest in doing a science project. In fact, I strongly advise students to find the simplest possible way to get a reasonably good answer. If you must make a numerical measurement for each member of your test and control group, then make the measurement once for each member and average over the each group. If this is too much work, then cut your test and control groups down to a more manageable number. It is better that your science experiment to loose sensitivity than for you not finish it. Finally, see if you can find a fast visual way to estimate the average. For example, if you are measuring the heights of plants, try creating a set of horizontal lines along a wall, like what you see behind people in mug shots, and lining the plants along a wall so you can easily compare the height of each plant. With a little practice, you will be able to get a pretty good visual estimate of the average heights at a glance. Make sure you photograph these trials and include each in your science project report in case a judge wants to check your results. (Do you think your desire for a positive result might influence your judgment? [Say, YES!] Then you need to think about how you can control for this possible bias. Why not let a friend with no stake in the outcome make the judgment for you from the pictures?) If you can find a simple way to visually get the information you need, you can save a whole lot of time and get great results!

Dr. Shawn

For more FREE help, visit my Super Science Project Support Site. Check out my collection of killer science fair project downloads. Are you running out of time? Check out my Desperation Science Projects for complete science project instructions that can be carried out in just one day! Or, better, have a complete science experiment sent directly to your home for professional results fast! And, of course, you'll find plenty of science project ideas at my science project idea bank, or my student-driven science project idea exchange.

Dr. Shawn's Super Science Project Support Blog

This blog is devoted to helping students learn how to actually do real science, and to helping teachers, parents and homeschoolers inspire a love in science in their students and loved ones.

How Many Measurements Per Trial?

Comments

About

Recent Posts