Science Buddies: "Ask an Expert"

Hi, for my Science fair project I made a homemade lie detector test. The meter moves when your lying by measuring the amount of sweat on your fingers. My question is what kind of questions should I ask my subjects? Im going to ask the same questions to all my subjects. Should I blind fold them so they cant see what the meter is doing? Thanks Hunter.

Hi unactress35,

This sounds like an interesting project. You'll want to think carefully about what questions you choose--you don't want to ask anyone something personal that might make them feel uncomfortable, for example. You may need to have your questions approved by your teacher or science fair SRC. And, you need to make sure you have informed consent from each of your participants. Your teacher will be able to help you know what you need to do to meet the rules of your science fair. Some questions might be:

- Do you have a dog?
- Do you like playing in the snow?
- Have you ever been to the beach?

These are somewhat inane questions with yes/no answers. You'll want to make sure that you don't already know the answer because that might bias your results. For example, if you know that your best friend has a dog, you wouldn't want to ask your friend that question if s/he were one of your test subjects.

To assess how well your meter works, you'll want to have some other way of knowing when people are lying. One way to do this would be to have the person taking the test purposely lie when they answer one question. If you have five questions, they might decide to lie on question number 2, for example. They could then write down that number on a piece of paper that also has their number as a test subject. Let's say the 4th person taking your test decided to lie on question 2. That person might write "test subject #4, lied on #2" on a slip of paper and then put that paper in a box, so that you can't see what they wrote. Then, you could have a data table where you write down what question you think the person lied about based on the data you get from the meter. It would be a good idea to have the meter and your data table somewhere where the person can't see them. At the end of your testing, you can open the box where the subjects put their slips of paper and compare the results of the lie detector with the actual lies told by your test subjects.

Hi there,

Some other questions that you could also ask are:
Do you like scary movies?
Have you ever pulled a prank on someone?
Have you ever cheated on a test?
Have you ever told your parents that you were sick when you really just wanted to stay home?

Those would be some good questions to ask. Like the person before me stated, you don't want to ask anything too personal like " have you ever stolen something?". Make sure that you absolutely get informed consent from those that participate, you will need to have them saying that they are aware that their participation is voluntary and they may refuse to answer a question and stop the questioning (since this is for an experiment and not the subject of a criminal investigation, participants are allowed to do this). The form must also state that their names will not be used. If you're going to compare which question got the most "lies" then ask the same questions to everyone. If you're just looking at the overall percentages of those who lied and how many questions they lied on, then feel free to mix up the questions.

There are a number of online resources that talk about lie detector tests and how you can calibrate them. I would suggest googling for these.

I will have to disagree with one of the other commenters. You're going to have to make you subjects at least a little bit uncomfortable in order to get any real results from a lie detector. The entire basis of calibrating them is to ask questions that you know people can be made to lie about but which will make them uncomfortable, like "are you completely 100% honest" and then convincing them that to pass the test they have to say "yes".

And, yes, informed consent is important. Make sure that the subjects know the kinds of things you will be asking them and that they agree.