Reliability (1 Viewer)

[crammy90]

can we say we could increase reliability by trying an entirely different approach to the experimental design (i.e. a new experiment) and if we get the same trends in reulst they are more reliable?
or is it only doing it more times?

 

[Undermyskin]

Only repetition is considered here, I suppose. Trying a different method should be to assess the validity. Two methods giving a similar results don't give you much anything. We then have to assess validity and reliability of the new one AND of the old one. Very troublesome.

To increase reliability:
- Repetition
- Large number of students involved with the same method in same conditions --> get average results and pick out the general trends