Review of EXO-22-012

Title: Search for dark matter produced in association with a Higgs boson decaying to four leptons using the full Run II data

*Color Code

Preapproval notes

Question: Clarify if the trigger efficiency is correctly implemented for the low pT (down to 5 GeV) leptons. The trigger efficiency plots show a turn-on behavior for these pT ranges. The analysis exploits the trigger efficiency calculated in the context of Higgs ZZ 4l analysis in HIG-19-001 where the current analysis is inherted from. The trigger efficiency has been calculated using the TnP -like method that is capable of estimating 3-lepton trigger efficiency and 1 lepton as a leg. From data to MC comparison there is a good agreement and the small observed discrepancies are propageted as uncertanties in lepton selection rather than SFs.

Question: Check if there would be a need for applying a MET based HEM veto in the analysis. This can be done by looking at the met phi distribution in the 10% of the SR (as a first step of unblinding when the analysis receives green light to unblind)

The HEM effect has been already checked in a control region (side band of m4l). The HEM veto is excluding the events with Jet pT > 30 GeV in the region of eta -3.0 to -1.3 and phi -1.57 to -0.87. The following plots show the ratio for data events in the CR before and after applying the HEM veto for PFMET (left) and MET phi (right) distributions. We find no noticeable change in the above-mentioned distributions before and after applying the veto, we didn't didn't apply veto for the HEM issue.

Question: Show the numerator distribution for the Z+X estimation. Clarify what the uncertainties are in this measurement. Naively, once the WZ background is subtracted, one would expect to be left with a small number of events in the tails of the distribution. This would yield to large uncertainties in the fake rate measurement. It was surprising not to see this effect in this measurement.

Those plots show the muon fake rate numerator before (left) and after (right) subtracting WZ background for barrel (top) and endocarp (bottom) regions, Indeed in high pT bins the number of events are decreased after subtracting WZ background and we left with higher uncertainties. The uncertainty values can be shown in the 2D plots shown in AN figure 9 and 10.

Similarly the following plots show the electron fake rate numerator before (left) and after (right) subtracting WZ background for barrel (top) and endocarp (bottom) regions.

Question: Do you understand why the Z+X background is very important in the most sensitive region of the analysis whereas the BDT analysis does not have any sizable contributions from these fakes in the sensitive bins. It is not clear to us how the two analyses can have such large differences in the backgrounds in the sensitive regions.

The question is not clear on which plots.

Question: Please provide the MET and the BDT results in the PAS/Paper. Also please provide impact plots for the MET fit and the BDT fit. In the AN, there is an impact plot corresponding to BDT. In this plot, we notice that met nuisances are one sided. Please show the shape of these nuisances so we understand the impact of up and down variations.

The MET plots have been added to the PAS with the BDT plots. The impact plot for MET fit has been added to AN. The shapes of the nuisances are shown in the AN in figures 31 - 36.

Question: Lastly, if you would like to make the HL-LHC projections public as well, we need to follow up on that more closely. These results were not presented today, therefore we will take a few more days to go over these results and iterate on them with you early next week.

Object review status

Physics Object	Status	hypernews link
photon	Done	https://cms-pub-talk.web.cern.ch/t/photon-object-review/6398

Electron	Done	https://cms-pub-talk.web.cern.ch/t/electron-object-review/6551/2
Muon	Done	https://cms-pub-talk.web.cern.ch/t/muon-object-review/6450/6
Jet	Done Done	https://cms-pub-talk.web.cern.ch/t/jet-object-review/6562/2

MET	Done	https://cms-pub-talk.web.cern.ch/t/met-object-review/6258/4
Combined expert	Done	https://cms-pub-talk.web.cern.ch/t/green-light/7448
Green light	Done	https://cms-pub-talk.web.cern.ch/t/green-light/7448

Muons

Question:

As the muon contacts, we are reviewing the muon usage described in the latest AN(v4). We don’t have much concerns as this is an extension of published Higgs analyses, and shares much of the object usage already.

We have a few questions: (1) Referring to AN2019_139, it described the tracking and IP scale factors were measured as well as the ID and Iso SFs. Is this the case for this analysis as well? If then, can you please add more lines about this? (2) In L832, it only describes the effect of the lepton reco and ID efficiency. But at least the effect of Iso SF should be included, and tracking and IP might be also included. Could you update this sub-section as well?

(1) We applied the same scale factor as in standard model HZZ4l analysis AN2019_139. (2) Actually this is an overall effect for muon tracking, reconstruction, identification, impact parameter and isolation requirements as shown in figure 24 of AN 2019_139. The sentence will be updated to “Overall Lepton identification, reconstruction, Isolation, tracking and IP efficiency that results in a lnN contribution ranging from 0.7 to 1.2% and from 11 to 15.5% on the overall yields, in the 4mu and 4e channels, respectively” in the next version of the AN.

Question: when the AN is updated, we will happily give you the green light. The AN has been updated and new version “V5” now available in CADI.

Question: we checked the GL for the muon usage.

Electrons

Question: I have reviewed AN-2020/013 as EXO electron contact. The analysis shares samples, electron selection and ID (and the related SFs ) with [1], for which electron usage has been extensively reviewed and approved by EGamma, therefore you have my GL.

Good luck with the rest of the approval process,

Missing Energy

Question: As the EXO MET contact, I have reviewed AN2020_013_v4. Thank you for the nice document. I think it is very clear. I only have a few questions. Correct me if I am wrong, but the XY corrections are also applied here. Is this correct? It should be added to the note. I also have one question about the distributions that you presented in figures 19-21 showing the distributions for variables used in the BDT training. The final bin for MET, which is blinded for the SR, has a similar excess in all three years. Has this been investigated any further to see where this disagreement is coming from? Actually, looking closely at these plots, the hashes for the uncertainty looks strange to me (off the MC distribution and not seeming to match the ratio). Figure 19 specifically is a bit suspicious. Can you please confirm that everything is healthy with these plots and if the excess is there/understood.

Yes that is true, the XY correction is applied. It has been added to the AN in section 10.2 Concerning the MET plots in the CR, I removed the hashes for the uncertainty it was taken from old root file and not updated . Concerning the last bin, I checked the plots and nothing weird in the last bin, the excess is there and within the statistical uncertainty.

Question: Thank you for the follow up. You have been marked as good from the MET side. Best of luck!

Photons

Question: As the photon contact, I am reviewing the AN2020_013_v4.pdf. Thank you for very well written analysis note. I have just one doubt. For FSR photons, I see you have considered old analysis method and the selections. But my doubt is did you perform the optimization again in your case for the selections on #DeltaR and E_{T}? Another small concern is about the references 49 and 50. Can you please provide the link to the analysis for reference 49 there(https://cms.cern.ch/iCMS/jsp/db_notes/noteInfo.jsp?cmsnoteid=CMS%20AN-2015/277). And the reference 50 is not present there.

The analysis method is the same one used in the standard model Higgs to ZZ to 4 leptons analysis for full Run II. As our analysis is inherited from HZZ4l analysis, we relay on this analysis in our object selection. I added the latest reference for HZZ4l AN and updated Ref 50. The link of Ref. 49 is given in [1]. A new version of the AN attached.

Question: I see the suggestions are made in note. You have my green light. Good luck for pre-approval!!

Jets

Question: I have had some email exchanges with Rehem and we agreed that the jet section includes a statement that jet is not the main object of the analysis, no selections applied based on jets, and that it is only used for diagnostic purposes. This is reflected in the late version of AN. I notice that on line 455 there is a “<” sign missing before 50GeV. Otherwise, the jet section looks good to me. I give GL to the jet usage in this analysis.

Thanks a lot for the GL. Just to confirm that I already included the sentence in the AN “V6” which already available in CADI. I will fix ““<”” in the next version. Thanks a lot.

Combine

All object reviews including the datacard review are complete, we therefore think this analysis can proceed to the pre-approval with the caveat that the various sub-sources included in the uncertainty of the MET should be understood better in the spirit of solving the one-sided impacts observed on these.

After having read through the latest version (v7) of the documentation and verified that the minor changes produced are addressing the comments emerged from the object review, I confirm the above stated (green light), thus wishing all the best to the authors in next stages of the review.