Check out some similar questions!

Yusf · May 22, 2016, 09:11 PM

It must, because it reduces the effect of random errors. So the final value is closer to the correct value. Right?

Fr_Chuck · May 23, 2016, 02:03 AM

Not always, it can also prove your experiment invalid or incorrect.

If for example, you did the first one, and got an answer of 4, then the next time 10, then the next time 2. A average would be fairly worthless (in general)

Repeating or having others repeat your experiment proves the validity of it.

Yusf · May 23, 2016, 02:19 AM

I get it. But say I got values like 8.1, 8.3, 8.3, 8.4, 8.1 and did the average of all values. Now is my result more accurate or jist has improved reliability?
They often ask us what we can do to improve accuracy if this experiment. So is repeating amd averaging a valid option or not?

Fr_Chuck · May 23, 2016, 02:56 AM

terms like correct value or more accurate needs to be terms you lose from your vocabulary in research.

Accurate assumes there is a right and/or wrong answer already known.
In experiments, we are looking to find what a result is. The answer you get is always the right answer for that experiment, Although it may or may not fit a pattern of other tests.

You can show a reliability that the hypotenuse or your experiment is true. Such as speed of fall, by dropping an object, over and over and over.
So your idea is to prove your assumption for the reason of the experiment.

This is done by doing experiments over and over, and getting same or simulacra results. It is the answer that is getting more accurate, it is the results proving your experiment valid. (ok, sort of the same thing, but professionally very different.)

ebaines · May 23, 2016, 05:48 AM

There are two sources of errors in experiments:

1. Random errors, which are effects that create randomness in results. This can come from things like minor variations in experimental conditions, or uncertainty in measurements. As you noted, repeating experiments can help average-out these types of random effects.

2. Systemic errors, which are built into the way the experiment is conducted or measured, and cause consistent errors. Examples would include recording data with a piece of equipment that is not properly calibrated (such as using a thermometer that consistently reads too low), or failing to account for the effects of some influencing factor - such as neglecting air resistance in calculating acceleration due to gravity. Repeating the experiment does not help eliminate this type of error, because it influences each trial run the same way.

So the answer is: in general repeating experiments does reduce some errors, but doesn't eliminate all errors so cannot always guarantee an accurate result.

Yusf · May 23, 2016, 07:35 AM

Thank you very much.

talaniman · May 23, 2016, 08:26 AM

Originally Posted by Yusf

I get it. But say I got values like 8.1, 8.3, 8.3, 8.4, 8.1 and did the average of all values. Now is my result more accurate or jist has improved reliability?
They often ask us what we can do to improve accuracy if this experiment. So is repeating amd averaging a valid option or not?

Maybe framing your results more accurately as in rather than seeking exact value through averaging, accept the RANGE of the results, and account for any variables in conditions, or other factors. Sometimes there are no absolutes that can be quantified.