- Alignment correction removed; made some improvement
- Looking at bias introduced by selection criteria or recon effect
- Considered possibility of scattering at different momenta introducing different bias; note that this is not dependent on the data set
- Considered possibility of dispersion transverse-pz correlations; suggest for example running MC with events off-axis and compare with events on-axis
- Rogers expressed scepticism about the "scattering angle acceptance" (VC slides) based on "back-of-the-envelope" type numbers; suggest getting higher statistics in the edge bins by e.g. putting in a higher Z scatterer
- Discussed algorithm; Ryan makes histogram of scattered vs transmitted scattered and takes ratio
- Note also tracker effects/efficiency
- Propose following a "permutation" argument for tracker efficiency, i.e. compare 5 point with 4 point to get Probability(misses station X); calculate implicit Probability(misses station X+Y) and use to deduce the number of 3 point tracks (which would not be reconstructed).
- Paul notes formulae in VC slides are incorrect; Ryan to check
Statistical Error Definitions¶
- Some discussion - Paul and Ryan talked about it before they left
Friday 6th January at 5 pm UK time 9 am Canada.