Below is a first attempt at listing all of the systematic errors we need to calculate and proposed techniques for calculating them.

• QCD rate – maximum variation found in CDF 6636 is 26% for N(jets)>3.  This may be overly conservative.
• Lepton & trigger ID – can be obtained from the R-ratio as done by Sasha.
• MC modelling for soft lepton ID – Some possibilities are
  • Use top selection. Take delta(1 mu) as a starting point for the uncertainty.  See if the difference changes as a function of N(mu) and blow up the error by this slope.
  • Do the single lepton fit using pTrel & S(d0).  Use this as a starting point for the uncertainty.  Blow up by Delta(Nmu) slope obtained from the top selection.
  • Take the maximum variation in N(taggable), pT(taggable), eta(taggable) in the single lepton bin and then feed the difference through the tag rate/mistag matrices.
  • There is expected to be a 5% tracking over efficiency in the MC.  Depending on how we estimate this uncertainty, it may be covered.  If it is not, I can throw out 5% of the tracks (possibly depending on pT) and rerun the prediction.
• Soft muon tag rate –
  • propagate fit errors from J/psi.
  • fit epsilon(pT) in 2 bins of eta and vary fit function.
• Soft muon mistag rate –
  • take largest (Observed-Predicted)/Observed from various jet samples.

Last week, I produced “final” 4 panel plots combining all the triggers for W and Z selections.  Here is the 4 panel Z plot I showed last week

There are 2 issues with this plot:

1. The jet overlapping the 2nd loose lepton is not being removed from the jet list.  This causes the funny bump in the Et(jet) plot in the bottom right.
2. All of the plots are for dilepton events across the whole mass spectrum.  We should be plotting the Z region, since this is “W/Z + additional leptons”.

Both of those issues were fixed this week to give the following plot. Note that the mass plot still shows the whole range.

The other plot produced last week was the N(muons) plot, shown below.

There are 2 issues with this plot as well:

1. The count of muons in the Z plot is over the entire mass region, as above.
2. We are counting muons coming from the W and Z as well.  This causes weird shapes because we are combining electron and muon triggers.

We fix these issues by, as above, only counting inside the Z window and by ignoring leptons coming from the W or Z.  In addition, the signal expectation is added.  The new plot is below.

The absolutely normalized prediction for the heavy component of the single muon final state and the fit results differ widely.  Below are pTrel and d0 significance plots obtained from absolute normalization.

And here is a corresponding fit result for mu+.

The fit returns N(W+heavy)=1861 +- 65 while the absolute prediction is 897.  The reason for this may become apparent when we plot some of the absolutely normalized components below.

Continue reading Discrepancy Between Absolute Normalization and Fit Result in Single Muon Bin →

Below is the plot of the muon multiplicity distribution in the TCE trigger with the Dark Higgs signal model overlaid.  Note that this distribution is the sum of soft and hard muons.

The number of events with >2 muons predicted and observed are:

N(SM bg)=0.688; N(Dark Higgs)=7.6; N(data)=1

We want to compare the taggable/tagged populations obtained from absolute normalization to that obtained from the ratio scheme.  Below are the met fits used for absolute normalization.

Continue reading Absolute prediction using EW scaling and heavy flavor overlap removal →

Following Scott’s discovery of the discrepancy in the single lepton population between the W+jets & W+heavy samples, I below used the heavy ratios obtained from the W+jets samples to obtained the predicted background and compared to the observed yield.

TCE mu+ mu+:

#mu^{+}_{l}#mu^{+}_{l}=0.443923

#mu^{+}_{l}#mu^{+}_{h}=0.106633

#mu^{+}_{h}#mu^{+}_{h}=1.36741

nbg=1.91797

data=9

TCE mu+ mu-:

#mu^{+}_{l}#mu^{-}_{l}=3.10351

#mu^{+}_{l}#mu^{-}_{h}=0.461862

#mu^{+}_{h}#mu^{-}_{l}=0.258485

#mu^{+}_{h}#mu^{-}_{h}=10.282

nbg=14.1059

data=23

TCE mu- mu-:

#mu^{-}_{l}#mu^{-}_{l}=0.971373

#mu^{-}_{l}#mu^{-}_{h}=0.365418

#mu^{-}_{h}#mu^{-}_{h}=0.549428

nbg=1.88622

data=8

CMUP mu+ mu+:

#mu^{+}_{l}#mu^{+}_{l}=0.399164

#mu^{+}_{l}#mu^{+}_{h}=0.0449528

#mu^{+}_{h}#mu^{+}_{h}=2.32589

nbg=2.77001

data=5

CMUP mu+ mu-:

#mu^{+}_{l}#mu^{-}_{l}=1.91507

#mu^{+}_{l}#mu^{-}_{h}=0.573502

#mu^{+}_{h}#mu^{-}_{l}=0.173787

#mu^{+}_{h}#mu^{-}_{h}=4.48139

nbg=7.14375

data=16

CMUP mu- mu-:

#mu^{-}_{l}#mu^{-}_{l}=0.634605

#mu^{-}_{l}#mu^{-}_{h}=0.0522837

#mu^{-}_{h}#mu^{-}_{h}=2.67786

nbg=3.36475

data=2

CMX mu+ mu+:

#mu^{+}_{l}#mu^{+}_{l}=0.168544

#mu^{+}_{l}#mu^{+}_{h}=0.201404

#mu^{+}_{h}#mu^{+}_{h}=0.294008

nbg=0.663956

data=0

CMX mu+ mu-:

#mu^{+}_{l}#mu^{-}_{l}=1.2084

#mu^{+}_{l}#mu^{-}_{h}=0.0506138

#mu^{+}_{h}#mu^{-}_{l}=0.011716

#mu^{+}_{h}#mu^{-}_{h}=3.07523

nbg=4.34596

data=6

CMX mu- mu-:

#mu^{-}_{l}#mu^{-}_{l}=0.341453

#mu^{-}_{l}#mu^{-}_{h}=0.292227

#mu^{-}_{h}#mu^{-}_{h}=0.382209

nbg=1.01589

data=2

We have gone from having an observed excess everywhere to having an observed deficit almost everywhere.

After speaking with Henry  on Friday, I’m using the W+jets instead of W+heavy samples to get the 2-to-1 soft lepton ratios.  Here’s the newest table:

Category	Expected	Observed
DILEP_TCE_eP_eP	339.107	378
DILEP_TCE_eP_eN	702.108	737
DILEP_TCE_eN_eN	357.002	342
DILEP_CMUP_eP_eP	192.871	242
DILEP_CMUP_eP_eN	402.9	479
DILEP_CMUP_eN_eN	205.709	259
DILEP_CMX_eP_eP	119.375	126
DILEP_CMX_eP_eN	236.311	301
DILEP_CMX_eN_eN	114.012	152
DILEP_TCE_eP_mP	34.9529	19
DILEP_TCE_eP_mN	34.5579	26
DILEP_TCE_eN_mP	33.7819	34
DILEP_TCE_eN_mN	38.8574	27
DILEP_CMUP_eP_mP	18.4859	12
DILEP_CMUP_eP_mN	17.1124	23
DILEP_CMUP_eN_mP	19.8476	15
DILEP_CMUP_eN_mN	19.6184	13
DILEP_CMX_eP_mP	10.6777	13
DILEP_CMX_eP_mN	12.6213	7
DILEP_CMX_eN_mP	8.30193	10
DILEP_CMX_eN_mN	10.0776	9

The numbers are much closer than they were before, but there are still issues to fix:

• The e+mu predictions seem too high, (at least in the TCE trigger) and I just fixed a problem that will, I expect, increase them further.
• I believe the expected contribution from two conversion electrons is still two low:

• The soft electron parameterization needs to be tweaked (or I could just cut off the soft electrons at 1.5 GeV):

-Scott

I’ve found the reason the predictions are high in the 2-lepton categories.  The problem is in the 2-lepton to 1-lepton ratios, but I’m not sure how to fix it.

In order to calculate the expected contribution from a particular 2-lepton process, say e+ from a heavy jet and e- from a light jet (abbreviated ePh and eNl) we find N(ePh)_data from the pTrel/d0 fit, and multiply that by the ratio found in the MC: N(ePh + eNl)/N(ePh)

N(ePh)_data * N(ePh + eNl)_MC / N(ePh)_MC = N(ePh + eNl)_predicted

Currently, we’re using the W+heavy MC to calculate N(ePh + eNl)_MC / N(ePh)_MC.  Naively, I would expect that using the W+jet MC would give the same answer, but with larger error bars, since there is some W+heavy in the W+jet sample.  However, as it turns out, the two samples give a ratio that differs by a factor of two.  They agree on the number of expected ePh + eNl events, but the W+jet has significantly more ePh events. (remember that the categories are exclusive)

In the following plots, the X-axis is the number of e+ from heavy jets in the event, and the Y-axis is the number of e- from light jets.  Look at the (1,1) and (1,0) bins.  The top plot is W+jets, and the bottom one is W+heavy.

The N(ePh) is twice as large in the W+jets sample as in the W+heavy sample.  We’re currently using the W+heavy numbers in our calculations, (on the assumption that the error bars would be smaller) but the W+jets numbers seem to agree better with the data.  Is there a reason the two should be so wildly discrepant?

-Scott

We have successfully generated a test sample of Dark Higgs MC according to the prescription given by Matt Reece.  The decay chain is WH->e nu chi0 chi0, where chi0 is the SUSY LSP.  The chi0’s then decay through the dark sector as chi0->chiD N(aD) where chiD is the dark photon, aD is the dark pseduscalar Higgs, and N can range from 2 to 4.  The dark Higgseses then decay to two electrons or two muons.

The model parameters used in this generation are as follows:

M(H)=120 GeV

M(chi0)=10 GeV

M(chiD)=1 GeV

M(aD)=300 MeV

Br(chi0->chiD aD aD)=0.333

Br(chi0->chiD aD aD aD)=0.333

Br(chi0->chiD aD aD aD aD)=0.333

Br(aD->ee)=0.525

Br(aD->mu mu)=0.466

Br(aD->pi pi)=0.009

There are other SUSY model parameters that are not listed above.

The plots below compare the shape of the electron and muon multiplicity distributions for the Dark Higgs with respect to Standard Model W+heavy jet production.  These plots are normalized to unit area.

Clearly, the lepton multiplicity in this Dark Higgs model is much higher than in SM W+heavy jets.

Here are the new expected and observed numbers in e+e and e+mu:

Old Track Quality:

Selection	BG Prediction	Data Observed
TCE_eP_eP	550.536	469
TCE_eP_eN	1122.09	901
TCE_eN_eN	583.46	428
CMUP_eP_eP	331.307	302
CMUP_eP_eN	675.02	563
CMUP_eN_eN	342.74	315
CMX_eP_eP	189.794	170
CMX_eP_eN	366.203	361
CMX_eN_eN	185.485	201
TCE_eP_mP	65.4135	19
TCE_eP_mN	63.8633	26
TCE_eN_mP	58.6098	36
TCE_eN_mN	64.9233	30
CMUP_eP_mP	32.5401	12
CMUP_eP_mN	30.2969	24
CMUP_eN_mP	31.4779	16
CMUP_eN_mN	32.2473	14
CMX_eP_mP	16.3119	14
CMX_eP_mN	20.3123	8
CMX_eN_mP	15.8709	12
CMX_eN_mN	22.3682	13

New Track Quality: (tighter silicon requirements for electrons)

Selection	BG Prediction	Data Observed
TCE_eP_eP	460.208	378
TCE_eP_eN	1011.86	737
TCE_eN_eN	537.621	342
CMUP_eP_eP	297.91	242
CMUP_eP_eN	612.326	479
CMUP_eN_eN	306.243	259
CMX_eP_eP	177.752	126
CMX_eP_eN	350.248	301
CMX_eN_eN	171.667	152
TCE_eP_mP	55.1458	19
TCE_eP_mN	59.1393	26
TCE_eN_mP	54.0227	34
TCE_eN_mN	59.9562	27
CMUP_eP_mP	29.1257	12
CMUP_eP_mN	28.783	23
CMUP_eN_mP	29.4332	15
CMUP_eN_mN	28.9681	13
CMX_eP_mP	14.4311	13
CMX_eP_mN	20.0945	7

The most dramatic change between these two was the reduction in the predicted contribution from (e+e-, both from conversions).  That contribution was reduced by a factor of two, but that’s the one that we aren’t currently getting a good estimate for. (a reduction from 12 to 6 doesn’t show up very well)

-Scott