Could it be possible nowadays to observe a large number of objects in the universe, at various distances from us, and keep track of the redshift that we measure to see if the objects are actually accelerating?

 

I know that the acceleration will be difficult to measure, because in the course of a human lifetime, the relative change in velocity of the stars/galaxies will hardly be measureable. I know that we would be looking for tiny changes...

 

But the possible data set is nearly infinite. If the dataset is large enough, even small changes can be determined with some degree of certainty.

 

Imho, this would be a more direct method to support the theory of the accelerating expanding universe...

I think you need to first look at the change in redshift that will occur within a measurable amount of time - my gut-reaction would be that it was so far below the noise threshold that no statistical technique could uncover it

I think you need to first look at the change in redshift that will occur within a measurable amount of time - my gut-reaction would be that it was so far below the noise threshold that no statistical technique could uncover it

 

Statistics is not a strength of mine... are you saying that you just don't believe we can do it, or is there a fundamental mathematical problem?

(for example, if we would have both more time (a million years) and a much bigger dataset (all the stars), could it work then?

 

I thought that if the change is tiny, the noise is very high, but the dataset is immense, you will find a regression with a slope that has a reasonably narrow confidence interval... But I am not 100% certain about that, and that's why I ask.

Unfortunately Captain - it's the blind leading the blind on statistical methods. Also would we be encountering an uncertainty limitation - I can't think of one off hand (would time/energy impact on measurements?) but I think the accuracy one would need would be staggering. I am trying to get to grips with equations - and failing at present.