Understanding Systematic and Random Errors in Scientific Measurements

[gem0688]

Hi just a quick question.

What are systematic and random errors?

Are systematic errors, ones where you use equipment which is not 100% accurate. For example using a ruler which is accurate to +/- 0.1mm?

Are random errors ones where the experiment is done slightly different for example instead of taking readings every 30 seconds you take one at 31 seconds and another at 29 seconds?

Thanx

 

[Hootenanny]

Random errors are errors that are caused by unpredictable factors, often due to environmental factors (changes in atmospheric temperature). These errors produce a random effect on the data (sometimes the data will be higher than usual, other times it will be lower), hence the name random. Ideally, random errors should exhibit a normal Gaussian distribution.

Systematic errors are usually caused by measuring equipment and always affect the data in the same way. I.e. the data is offset always being higher or always being lower than the actaul value. The relationship between the varible is usually the same in systematic errors but not in random errors.

The significance of random errors can be reduced by increasing the number of data points. The significance of systematic errors is unaffected.

~H

 

[gem0688]

Ok thanx, is the Gaussian distribution the bell shaped one yea?!

 

[Hootenanny]

Yeah, where the peak is at the mean, 68% lies within one standard deviation, 95% within two and 99.5% (I think) lies within three.

~H

 

[gem0688]

ok i 'think' i understand! :)

 

[Hootenanny]

'think'

Well if your unsure of anything, I'll be happy to help.

~H

 

[gem0688]

ok, i do understand now!

Random: e.g, different room temperature / atmospheric pressure, might cause a sudden increase in results.

systematic: e.g a thermometer which allways reads 1oC higher than what the actual temperature is, but happens all the time and so will not affect the results.

Cheers! ;)

 

[Hootenanny]

so will not affect the results.

It will affect the results, they will be shifted, but the relationship between the results would be unaffected.

~H

 

[gem0688]

ohh ok. yea, all the results will be that 1oC higher, but the overall pattern/relationship will be the same.

 

[Hootenanny]

ohh ok. yea, all the results will be that 1oC higher, but the overall pattern/relationship will be the same.

Yup, you've got it

 

[gem0688]

At last! Thanx a lot!