Reliability and Validity (1 Viewer)

[Paroissien]

Once and for all, what is the difference?

 

[Xayma]

About 2.6*10⁹ threads are on this.

DO A SEARCH!!!!!!!!

 

[helper]

Reliability is how consistent your data is. Has the experiment been repeated?

Validity is does the data represent what it is suppose to be. Does the data really answer your hypothesis. Is the experiment controlled etc. Is the thing being measured what you identified as the variable you were testing.

 

[Paroissien]

Cheers. I could've done a search but I was holding out for a nice simple response like this

 

[DAAVE]

Reliability is how consistent your data is. Has the experiment been repeated?

Validity is does the data represent what it is suppose to be. Does the data really answer your hypothesis. Is the experiment controlled etc. Is the thing being measured what you identified as the variable you were testing.

Yeah, so in summary:

Valid = experiment answers question
Reliable = experiment can be repeated X number of times with very similar results

?

 

[Plebeian]

I think validity also takes into account the suitability of your method. For example, in the molar heat of combustion experiment (using spirit burners), a lot of heat is lost to the surrounding environment. This means that the calculated value of the alkanol's molar heat of combustion is incorrect, so the experiment is not completely valid. If you repeat the experiment, you'll get the same (incorrect) answer so it is reliable, but not valid.

 

[Xayma]

Yes.

If I was to give you a solution of Ag⁺ and Pb²⁺ ions.

And you stuck in CO₃^2- ions it would be reliable (as you should get the same amount of Ag₂CO₃ and PbCO₃ each time) but not valid.

If you added chloride redissolved the lead chloride filtered and then added carbonate it would be both valid and reliabe.

 

[sneeble]

It's worth noting that an experimental procedure can be reliable (produces consistent results) without being valid (doesn't test proposed hypothesis).

We found this with our molar heat of combustion prac. All the groups in the class got very similar results, showing reliability, but they were all well below accepted values, showing an invalid procedure.

 

[googleplex]

http://www.cci.net.au/conqchem/ExamCommentsFrameset.htm

good analogy of accuracy reliability in the 2001 exam recap

 

[wind]

Reliable = method produces consistent results.

Valid = method is actually correct.