Test of analysis method for top-antitop production and decay events

We have carried out Monte Carlo calculations on two sets of randomly genera ted QCD events due to p (p) over bar --> t (t) over bar with top mass m(t) = 170 GeV, one set leading to e(+)e(-) or e(+/-)mu(-/+) or mu(+)mu(-) 2-jet s (dilepton) and the other leading to e(+/-) or mu(+/-) 4-jets (unilepton) configurations, in order to test the likelihood methods we have proposed fo r determining the top mass by analyses of these two sets of configurations. For the set of unilepton events, our method gives a very efficient and qui te sharp measure of the top mass lying several GeV below the input mass. Fo r the dilepton set, our method gives a much broader and markedly asymmetric distribution for the top mass estimates, 75% of them lying below 170 GeV, but the dilepton data will have much lower background than unilepton data. We then illustrate these methods by applying them to the data available fro m CDF in 1995 and discuss the results obtained in relation to the results f or the sets of Monte Carlo events. The dilepton events from CDF and DO more recently yield masses spread widely, from 130 to 180 GeV, generally lower than the CDF unilepton events, which cluster around 175 +/- 8 GeV. In an ap pendix, we discuss the nature of the additional 'slow' mu(+) observed in on e CDF dilepton event, concluding that it is most probably a 'tertiary lepto n' resulting from the decay sequence b --> c + hadrons, followed by c --> s mu(+)nu(mu).