I think you should probably read up on basic electronics first, understanding and using sensors, displaying data, and theory.

A MAJOR problem with using a consumer grade ohmmeter is sensitivity and lack of calibration.

If you have five subjects, each of those five will sweat in different rates naturally, before lying. You would have to take into account room temperature, body temperature, How clean the persons hand are, if they used a hand sanitizer, lotion, creams, drank gatoraid, battery level in your meter, and many many other variables.

With out the ability to CONTROL your experiment, your results WILL be flawed. woah INACCURATE.

You would have to repeat the tests with the same people five times or more. And if they are friends, you will envoke many other emotional resposnes like laughter, envy, love, which can all lead to increased persperation. So you will never tell if it was a lie that invoked a response on your meter. Sorry.

 

not to mention you could ruin friendships... You will have to explain your results during the fair, and your friends may not like there lies exposed to everyone in the school.

 

I think you should probably read up on basic electronics first, understanding and using sensors, displaying data, and theory.

A MAJOR problem with using a consumer grade ohmmeter is sensitivity and lack of calibration.

If you have five subjects, each of those five will sweat in different rates naturally, before lying. You would have to take into account room temperature, body temperature, How clean the persons hand are, if they used a hand sanitizer, lotion, creams, drank gatoraid, battery level in your meter, and many many other variables.

With out the ability to CONTROL your experiment, your results WILL be flawed. woah INACCURATE.

You would have to repeat the tests with the same people five times or more. And if they are friends, you will envoke many other emotional resposnes like laughter, envy, love, which can all lead to increased persperation. So you will never tell if it was a lie that invoked a response on your meter. Sorry.

Thank You sooo much, that was really helpful, my mark will go up because of that, and now I am more in the right track to knowing the control variables, if you have any other notes about it that would be helpful to the project please post it.

Thank You!!!!

 

A MAJOR problem with using a consumer grade ohmmeter is sensitivity and lack of calibration.

Doesn't this one use the meter and work http://www.aaroncake.net/circuits/lie.asp ? What is the difference? And for the calibration I think that the average skin resistance of a healthy person is between 1.5 and 2.3M ohms, and I would have control questions to see the specific person's regular skin resistance (without lying).

Thank You!!!!

 

On that site, those that tried it complain that it doesnt work....well as expected for a primitive circuit like that.
And yes the avarage human skin resistance is between a few ohms and a few MEGA ohms. Do you even realize how big that difference is ? Its on the order of million percent !.
People usually don't have "specific" skin resistance. Its a function of so many things including hygiene, dryness of the skin, etc.

 

So how can I measure it quatitately, and still be accurate?


Thank You!!!!

 

But wouldn't the control questions give me a parameter?

Thank You!!!!

 

Since the resistance depends on temperature around you for example. You'll propably experience constantly lowering resistance if the room you're taking the test in is too hot. Your subject will start to swet and your reading lowers despite it not lying.
Another aspect is if the subject gets tense, simply because they are sat down and hooked to a lie detector and start swetting. Then again your resistance reading would drop down simply because the subject is swetting.

Lie detectors today rely on more then just skin resistance such as heart rate and blood pressure and so on because the skin resistance factor alone is VERY inaccurate. When subjects are asked questions they rate the answers by how likely they are to be correct or not based on many variables. Modern lie detectors can only imply if a person is lying or not. So basicly every conclusion you draw with a lie detector has a possibility of being false. Hence they are never used in trials because they are essentially unreliable.

 

If you have a decent video camera, a better way to do your lie detection would be with pupil response. That way you can control the light in the room ( use the same room for all subjects ) and video tape the eye. Have them look into the lens. Then you can measure the pupil size change with the questions. The skin resistance test is WAY to huge. Like steinar96 said, the numbers are so huge, you will end up with nothing. Plus, you have to think about how you are going to explain it to the judges. You just cant rule out all of the options. ALSO lie detecting is about 99.9% inaccurate if performed by someone you know. Your friends will feel to comfortable lying to you. use the pupil response. tell them it works 100% every time. that will get them worried. and the results will be more accurate.

 

If you have a decent video camera, a better way to do your lie detection would be with pupil response. That way you can control the light in the room ( use the same room for all subjects ) and video tape the eye. Have them look into the lens. Then you can measure the pupil size change with the questions. The skin resistance test is WAY to huge. Like steinar96 said, the numbers are so huge, you will end up with nothing. Plus, you have to think about how you are going to explain it to the judges. You just cant rule out all of the options. ALSO lie detecting is about 99.9% inaccurate if performed by someone you know. Your friends will feel to comfortable lying to you. use the pupil response. tell them it works 100% every time. that will get them worried. and the results will be more accurate.

The bolded are my biggest questions.
Thank you, that is a good idea, but unfortunately, I can't change what I am doing it on, after I have closed the topic with my teacher :-(. But what I don't understand is that if I have control questions, that is the parameter, I can see if the skin perspiration changed from the control questions, when the person lies, what is wrong with that? Also, I am not doing this to figure out when someone is lying, I am doing this to figure out if skin resistance changes when one lies. So I don't care if skin perspiration changes when someone lies for now (when building it), that is what the purpose of my experiment is. I just really need to know what to build first, and it has to measure skin perspiration quantitatively and accurately. All of the posts were helpful, but now I really need to find out how to build a device that measures skin perspiration quantitatively and accurately and how it works. How can I?

Thank You!!!!

 

Since the resistance depends on temperature around you for example. You'll propably experience constantly lowering resistance if the room you're taking the test in is too hot. Your subject will start to swet and your reading lowers despite it not lying.

The room will be room temperature (which is one of my control variables).

Another aspect is if the subject gets tense, simply because they are sat down and hooked to a lie detector and start swetting. Then again your resistance reading would drop down simply because the subject is swetting.

That's a good point, so we would do a few tests in a row with each person and the same questions, which would make the chances of that happening very small.

Thank You!!!!

 

That might work but its like counting sheeps by counting their legs then divide by four. Why you don't count by counting the heads? Similarly, if the output is plotted on a recorder it could provide far more analyzable results than watching an ohm meters pointer on monitor.

Thank you, but I need to measure it quantitatively how can I? And also, what would be the difference between this circuit http://www.aaroncake.net/circuits/lie.asp and just the ohm meter directly? How is that one better?

Thank You!!!!

 

I think your project is to ambitious for your skill level, and it seams that your teacher also do not have the needed qualifications to guide you through this project. You should get another project. Your project is more a university level project. As you need both electronic and computer skills. The topic skin conductance is actually quite complex. In this thread I have seen some quite humourous misunderstanding and wrong statements about skin conductance.
Your project is interesting, correct and doable. But as both you and your teacher do not have the needed resources or competence, your project will most probably end in nothing.

 

I think your project is to ambitious for your skill level, and it seams that your teacher also do not have the needed qualifications to guide you through this project.

My teacher is not allowed to help me in this project, only in some parts, but not all.

You should get another project. Your project is more a university level project.

I now feel better because now I know that it is hard, university level, and that is why it is hard for me.

As you need both electronic and computer skills. The topic skin conductance is actually quite complex.

I have computer skills, I know C#, some VB.NET, some C++, and some C, and I have a little bit of electronics knowledge, but I am willing to learn more.

In this thread I have seen some quite humourous misunderstanding and wrong statements about skin conductance.

Which ones? just so I know what is right and what is wrong.

Your project is interesting, correct and doable. But as both you and your teacher do not have the needed resources or competence, your project will most probably end in nothing.

I have time, and I am willing to learn, but before I do, I really, urgently need a circuit that will measure accurately, and quantitatively the skin conductance, not something that will make a graph, because I do not want to include time, but something that outputs in a meter, or something like that. I will have to understand it after, but first and for now, I urgently need to find one, if you can find, or make a diagram that does that, I will be very, very grateful.

Thank You!!!!

 

I have a skin conductance project using the sound card of a PC if you are interested

 

I have a skin conductance project using the sound card of a PC if you are interested

Yes , but does it display the result quantitatively? I need something that displays it quantitatively.

Thank You!!!!