The figure below shows the refmult distribution for events with a central trigger (run-4 central data, triugger 15105) and various high-pt trigger particle selections (>2 GeV, >4 GeV and > 6 GeV). The multiplicity bias for central events when requiring a high-pt trigger particle is small. A small increase in refmult is seen when going from no requirement to > 2 or >4 GeV (2 times 1.5 per cent, see legend for values), but the additional effect of requiring > 6 GeV is only ~1 per mille.

The right-hand panel shows the ratio to the case with pt,trig > 6 GeV. Clearly the > 4 GeV requirement gives a flat ratio, while the other two give a slight slope to the distribution, leading to the slightly lower mean refmult.

This plot was made to address a question raised in

http://www.star.bnl.gov/HyperNews-star/protected/get/highpt/2394/1/1/1.html

Similar plots were made years ago by Jenn Klay, based on run-2 data:

http://www.star.bnl.gov/protected/highpt/jklay/doc/multBias/multBias.html