Showing if there is statistical significance between two data sets (data analysis) Hey guys so I have been given a homework problem where there is a table of video views with one column being views(in millions) before an event and then the other column is views after the event. The question asks to make meaningful data analysis insights. I made some graphs and such comparing the two and also did a t test showing if it was statistically different or not. Can anyone think of any other tests or anything I could do for insights? Thanks

 A: You could show and maybe compare distributions of the data. You can show it with histograms, density curves, kernel density etc. And you can assess the normality of distributions, skewness, homoscedasticity and potential outliers. 
A: Comment: It is not clear exactly what additional input you are looking for. Here
are a few thoughts.
I put the first six Before/After pairs into Minitab. (Typing
them all would have been too tedious.) Dif is Aft - Bef so that all of these differences are positive.
Data Display 

Row   Bef   Aft   Dif
  1  1.19  4.36  3.17
  2  1.60  4.01  2.41
  3  1.28  5.08  3.80
  4  1.59  4.46  2.87
  5  1.77  4.55  2.78
  6  1.51  4.18  2.67

There is no reason to suspect that these differences are other than normal.
A t test of the differences rejects the null hypothesis that
there the differences center around $0$ with a P-value less than $0.0005,$ shown in the printout as 0.000.
One-Sample T: Dif 

Test of μ = 0 vs ≠ 0

Variable  N   Mean  StDev  SE Mean      95% CI          T      P
Dif       6  2.950  0.485    0.198  (2.441, 3.459)  14.90  0.000

Just scanning down the entire dataset, it seems that the 'After' values
are mainly larger then corresponding 'Before' values---except for a few
pairs at the very end. 
Roughly speaking, this is confirmed by the paired t test on the first six pairs.
It might be interesting to investigate why the last
few pairs are different from the rest in this regard.
(If you plot After (y) vs Before (x), do you see a trend?
If so, a regression might show that 'After' can be roughly
predicted by 'Before'.) 
I would not
expect any further insights from a paired t-test on the whole dataset; the P-value may be a little larger (but I guess still significant at the 5% level).
If you have further questions please post the entire dataset
in computer-readable format (actual text, not a picture). And ask specific questions by editing your original post.
For example:
Bef = (1.19, 1.60, 1.28, 1.59, 1.77, 1.51, ..., 5.06)
Aft = (4.36, ...                                3.38)

