Posts: 1
Joined: Wed Oct 28, 2020 11:31 am
Occupation: Parent

Does this kids experiment have too many variables?

Postby wilsonbeth » Wed Oct 28, 2020 11:44 am

This is the research problem that was submitted but I'm not sure if I should let him (7th grader) continue.

I recently read a book called Grit by Angela Duckworth. The author develops a questionnaire to test how gritty a person is. I was wondering if I could use this proven grit scale to test my own method of testing people's grit. I was wondering if I could use this proven grit scale to test my own method of testing people's grit. The reason why I wanted to make my own grit scale is all the questions in Mrs. Duckworth's grit scale are so dry and repetitive for example, “Do you have perseverance?” The problem with these questions is they seem very obvious that they're testing for grit so I was wondering if I could create a different way to test people's grit.
His plan is to create a puzzle box that is unsolvable but seems solvable and test how long people work on the puzzle. Then administer Mrs. Duckworth proven grit scale questionnaire and see if the people who scored high on her test also spent the most time working on the puzzle.

I love the experiment idea but not sure if it will be good for a first science fair project. Let me know your thoughts

Posts: 2
Joined: Fri Sep 25, 2020 8:11 pm
Occupation: Other Adult

Re: Does this kids experiment have too many variables?

Postby Cherylreif » Wed Oct 28, 2020 1:51 pm

Hi there! I love your student's science fair project idea--it's very original and asks an interesting question. I'm glad that you're thinking about the number of variables involved now, though. Too many variables could potentially make this a complex project. However, I think your student may be able to limit the number of variables in order to make the experimental results easier to interpret.

The most important first step in this experiment would be to make a list of the potential variables, or factors, that might contribute to the experiment's results. For example, things that might influence the length of time a person spends working on the puzzle box might include: age, gender, time of day, and the person's level of "grit". Ideally, you want to keep all of these variables constant while changing only one--the individual person's time spent on the puzzle. You would then measure how grit (the dependent variable) changes with respect to how long an individual worked on the puzzle (the independent variable).

To learn more about variables, see the Science Buddies guide on Variables in Your Science Fair Project:


And Variables for Beginners - Doing a Fair Test:


Another useful resource is the Science Buddies guide on Experimental Design for Advanced Science Projects:


The thing that makes the project tricky is that you're actually changing the whole person, not just the amount of time they spend on the puzzle, from one test to the next. You can keep some things the same between tests--for example, the age and gender of the different people--but obviously you can't keep everything the same. Personally, I think you could choose a limited number of factors to try to keep the same, which would allow you to design an experiment that's not too complicated for a science fair project. It would be worth getting the opinions of other experts as well, though.

I hope this is helpful. Good luck with your science fair project!


Student Expert
Posts: 93
Joined: Tue Aug 27, 2019 9:33 am
Occupation: Student

Re: Does this kids experiment have too many variables?

Postby cnoonan180 » Sun Nov 01, 2020 7:01 pm


This is a very interesting project, and well-thought-out so far!

This project could be categorized as a survey project. You can read more about what this means here: https://www.sciencebuddies.org/blog/sur ... hypothesis

Also check out this resource about designing experiments to test certain aspects of human behavior, like the aim of this project. https://www.sciencebuddies.org/blog/sur ... hypothesis

The table in the second article link (directly above) should help you decide whether to keep your project to strictly a "questionnaire" type project, or an "observation-based" project, or both. For your project, I recommend the "observation" option to start and potentially build on (If you would like to keep the project at a simpler level, stick to administering Duckworth's questionnaire and have this be a survey project only). One idea would be for you to establish a list of parameters or characteristics that you can observe as your subjects (people who have agreed to participate in your experiment) work on the puzzle that you believe would demonstrate their level of "grit." These could be characteristics adopted from Duckworth's questionnaire if you would like. For example, you could test perseverance based on the time each subject spends on the puzzle with more time spent meaning higher perseverance. Also, you could look at facial expressions to see if certain subjects exhibit frustration after working on the puzzle for a certain amount of time as compared to other subjects who may not exhibit this behavior. After recording observations about the way each subject behaved as they worked on the puzzle box, you could ask them to complete Duckworth's questionnaire. This way, you can implement both a "laboratory" portion of the project, which would be having subjects solve the puzzle, as well as a "survey" or questionnaire portion of your project and compare the data you collected from each portion of the project to come to a conclusion. By separating your project into parts (naming them A and B for example) you will be able to control more aspects of each portion of the project, thus having fewer variables. Also, this will make it easier to find correlations in your data.

I recommend using Duckworth's tested method of testing a person's "grit" if it has, in fact, been proven. Science becomes more "reliable" as it is tested through time, so this project's data, in theory, could contribute to helping scientists determine the efficacy of Duckworth's method in the future. (Or help you formulate your own scientific opinions regarding Duckworth's method as you compare the data you collected about people as they attempt to solve the puzzle and their answers as part of Duckworth's questionnaire!).

If you'd like, feel free to share a list of the parameters or variables you come up with for the "laboratory" portion of your project so I and the other mentors can further assist you! Some variables and parameters to think about to help you get started could be the environment in which subjects are working on the puzzle box, age, gender, grades in school or SAT scores, education level, the type of puzzle (should be the same for all subjects).

One important point to think about before beginning this project is how you will define "grit" since this will set the tone for what you are trying to test with this experiment. Note that you may use Duckworth's definition or do some more research.

Hope this helps and good luck with your project!

Return to “Grades 6-8: Life, Earth, and Social Sciences”