1 1 STATE OF WASHINGTON DEPARTMENT OF LICENSING 2 3 4 5 ERIC ARNTSON 6 Plaintiff, 7 v. Case No. (*redacted information) 8 9 STATE OF WASHINGTON, 10 DEPARTMENT OF LICENSING, 11 12 Defendant. 13 ___________________________ 14 15 HEARING 16 OCTOBER 26, 2007 17 18 - - - 19 20 BE IT REMEMBERED THAT, pursuant to the Washington Rules of 21 Civil Procedure, this proceeding was taken before Valerie 22 Allard, a Certified Court Reporter, No. 3040, on October 23 26, 2007, commencing at the hour of 1 p.m., the proceedings 24 being reported at 320 North 85th Street, Seattle, 25 Washington. 2 1 APPEARANCES: 2 KATHRYN KOEHLER, HEARINGS OFFICER 3 Department of Licensing 4 Hearings and Interviews 5 320 North 85th Street 6 Seattle, Washington 98103 7 Phone (206)297-4525 8 Fax (206)706-4267 9 E-mail kkoehler@dol.wa.gov 10 ELLEN BARTON, HEARINGS OFFICER 11 Department of Licensing 12 Hearings and Interviews 13 320 North 85th Street 14 Seattle, Washington 98103 15 Phone (206)297-4525 16 Fax (206)706-4267 17 E-mail kkoehler@dol.wa.gov 18 JERALD R. ANDERSON, Senior Counsel 19 Assistant Attorney General 20 1125 Washington Street Southeast 21 Olympia, Washington 98504 22 Phone (360)753-6987 23 Fax (360)664-0174 24 E-mail jerrya1@atg.wa.gov 25 3 1 TED VOSK, ESQUIRE 2 The Bianchi Law Firm 3 705 Second Avenue, Suite 1000 4 Seattle, Washington 98101 5 Phone (206)622-3122 6 Fax (206)622-3129 7 E-mail georgebianchi@thebianchilawfirm.com 8 KEVIN TROMBOLD, ESQUIRE 9 The Law Offices of Kevin Trombold 10 720 Third Avenue, Suite 2015 11 Seattle, Washington 98104 12 Phone (206)382-9200 13 Fax (206)467-3152 14 E-mail kevin@tromboldlaw.com 15 MOSES GARCIA, ESQUIRE 16 Stafford frey cooper 17 601 Union Street, Suite 3100 18 Seattle, Washington 98101 19 Phone (206)667-8263 20 Fax (206)748-9047 21 E-mail mgarcia@staffordfrey.com 22 23 24 25 4 1 ALSO PRESENT 2 Cesar O. Velasquez, Esquire 3 The Law Offices of Cesar Velasquez 4 2315 112th Avenue Northeast, Suite 210 5 Bellevue, Washington 98004 6 Phone (425)889-5923 7 Fax (425)455-4354 8 E-Mail covcmv@msn.com 9 10 ROD GULLBERG, RESEARCH ANALYST 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 5 1 EXAMINATION INDEX 2 EXAMINATION BY PAGE 3 Mr. Vosk 8 4 Ms. Garcia 59 5 6 7 8 9 10 11 EXHIBIT INDEX 12 EXHIBIT NO. DESCRIPTION PAGE 13 63 Biographical Record 8 14 64 Batch No. 07007 19 15 65 Calculation Mean Schedule 19 16 66 Plots 22 17 67 Hand Calculations Batch No. 07007 27 18 68 Batch No. 07004 37 19 69 Hand Calculations Batch No. 07004 39 20 70 Hand Calculations Batch No. 06012 41 21 71 Batch No. 06012 42 22 72 Batch No. 07023 42 23 73 Corrected Calculations 07007 82 24 25 6 1 Seattle,Washington 2 Friday, October 26, 2007, 3 1 p.m., 4 HEARINGS OFFICER KOEHLER: This is an 5 administrative proceeding before the Department of 6 Licensing of the State of Washington. I'm Kathy Koehler, 7 Hearing Officer. And today's date is October 26, 2007. 8 This is a reconvened hearing in the matter of Eric 9 Arntson. Other hearings were held on September 18, October 10 5, October 12, October 16, October 17, as well. And this 11 is a recorded proceeding. There is a court reporter 12 present and also the proceeding is being recorded digitally 13 and by tape. 14 As I indicated, the Petitioner is Eric Arntson. 15 And Mr. Arntson is represented by George Bianchi and Ted 16 Vosk. Mr. Vosk is present. 17 HEARINGS OFFICER KOEHLER: And is Mr. Arntson 18 waiving his presence? 19 MR. VOSK: Yes, your Honor. 20 HEARINGS OFFICER KOEHLER: All right. And the 21 Department of Licensing is represented by Moses Garcia 22 who's also present. And to my right is Jerry Anderson, 23 Assistant Attorney General with the State of Washington. 24 And Mr. Anderson is legal advisor to the Hearing Officer on 25 issues of procedure. And then to my left is Ellen Baron, 7 1 another Hearing Officer with the Department of Licensing. 2 And perhaps for the record, we can again go around 3 the room and state your names so that we have a good 4 identification of who's present. 5 MR. TROMBOLD: Good afternoon. Kevin Trombold. 6 DR. EMERY: Ashley Emery. 7 HEARINGS OFFICER KOEHLER: Mr. Vosk is present. 8 DR. EMERY: Ashley Emery. 9 CESAR VELASQUEZ: Cesar Velasquez. 10 MR. DAHLQUIST: James Dahlquist, Department of 11 Licensing, not participating. 12 MR. GULLBERG: Mark Gullberg, Washington State 13 Patrol. 14 MR. GARCIA: And Moses Garcia for the Department 15 of Licensing. 16 HEARINGS OFFICER KOEHLER: All right. And we 17 convened here today for further testimony in the Arntson 18 matter. And Dr. Ashley Emery is present to testify; is 19 that correct, Mr. Vosk? 20 MR. VOSK: That is correct, your Honor, yes. 21 HEARING OFFICER KOEHLER: Okay. And you're ready 22 to proceed? 23 MR. VOSK: Yes, your Honor. 24 Can I get this marked real quick? 25 HEARINGS OFFICER KOEHLER: Okay. I believe that 8 1 we left off with Exhibit 62. So did you provide a copy to 2 counsel for the Department. 3 MR. VOSK: That's what I'm trying to do right now, 4 your Honor. I had two of them but somehow -- oh, we go. 5 Okay. 6 HEARINGS OFFICER KOEHLER: And this appears to be 7 Dr. Emery's CV? 8 MR. VOSK: Yeah. We'll go ahead and identify it 9 through the -- I just wanted to get it marked first, so I 10 could identify it for the record before I had him -- 11 HEARINGS OFFICER KOEHLER: Okay. I'm marking it 12 as Exhibit 63. 13 MR. VOSK: Thank you, your Honor. 14 ASHLEY EMERY was sworn by Hearings Officer 15 Koehler. 16 HEARINGS OFFICER KOEHLER: You may proceed Mr. 17 Vosk when you're ready. 18 MR. VOSK: Thank you, your Honor. 19 EXAMINATION 20 BY MR. VOSK: 21 Q Dr. Emery, can you please state your name, your 22 full name for the record, please. 23 A Ashley Emery. 24 Q Where are you employed? 25 A The University of Washington. 9 1 Q And what is your position there? 2 A Professor of mechanical engineering. 3 Q And how long have you been a professor of 4 mechanical engineering? 5 A 46 years. 6 Q Can you give us a summary of your educational 7 background, please? 8 A Bachelor's, master's and PhD from the University 9 of California at Berkeley in mechanical engineering. 10 Q Have you held any other positions at the 11 University of Washington? 12 A Been associate dean, chair of the department, 13 chair of the faculty senate, chair of the budget committee. 14 Q And do you belong to any professional or 15 scientific societies? 16 A Currently, I belong to the American Society of 17 Mechanical Engineers and the American Society for Heating 18 and Ventilating and Refrigeration Engineers. 19 Q Okay. Are you also a member of the Society of 20 Industrial and Applied Mathematics? 21 A No longer. 22 Q No loner. You were at one time? 23 A Yes, I was. 24 Q And what kind of classes do you teach? 25 A What? 10 1 Q What kind of classes to you teach? 2 A I teach Numerical Analysis. I teach laboratories. 3 I teach Heat Transfer, Fluid Mechanics, thermal analysis, 4 thermal stresses, fracture mechanics, bioengineering. 5 Q And have you taught statistics? 6 A Yes, I have. 7 Q And advanced math? 8 A Yes, I have. 9 Q And the Design of Experiments? 10 A Yes. 11 Q And have you published in peer review journals? 12 A Yes. 13 Q How many publications do you have? 14 A 250 -- 300 -- I don't know. 15 Q Okay. Did you write a chapter for a book entitled 16 "The Effect of Uncertainties and Correlations on the 17 Efficiency of Estimating and the Precision of Estimated 18 Parameters"? 19 A Yes, I did. 20 Q And what book was that in -- or let me ask you 21 this: Was that in the Inverse Engineering Handbook for CRC 22 Press, 2002? 23 A Yes, it was. Thank you. 24 Q And did you write a paper, "The Use of Kriging and 25 Nuisance Variables in Parameter Estimation" which was 11 1 submitted to Inverse Problems in Science and Engineering? 2 A Yes, I have. 3 Q Okay. And a paper named "Estimation of Errors in 4 a Complex Simulation Model When Applied at Conditions Far 5 From Where it was Calibrated"? 6 A Yes, I did. 7 Q Another article "Parameter Estimation for Noisy 8 Data and Nuisance Variables using Bayesian Inference"? 9 A Yes, I did. 10 Q Another article "Some Thoughts on Solving the 11 Radiative Transfer Equation in Media with Stochastic 12 Properties using Polynomial Chaos and Wick Products"? 13 A Yes, I did. 14 Q Another paper "The Use of Spatial Statistics in 15 Designing Experiments and Validating Models"? 16 A Yes, I did. 17 Q Another paper "Designing Optimal Heat Transfer 18 Experiments"? 19 A Yes. 20 Q Another paper "Parameter Estimation in the 21 Presence of Uncertain Parameters and with correlated data 22 errors"? 23 A Yes. 24 Q Another paper entitled "An Overview of Stochastic 25 Systems"? 12 1 A Yes. 2 Q Another one using "Using the Concept of 3 Information to Optimally Design Experiments with Uncertain 4 Parameters"? 5 A Yes. 6 Q Another one entitled "The Precision of Optimally 7 Designed Experiments with Uncertain Parameters"? 8 A Yes. 9 Q Another one entitled "Optimal Experiment Design"? 10 A Yes. 11 Q Another one entitled "Computing the Variance using 12 Different Approaches"? 13 A Yes. 14 Q Another one entitled "Parameter Estimation and 15 Optimal Experiment Design with Uncertainties in a-prior 16 Known Parameters"? 17 A Yes. 18 Q "Using the Modified Fisher Information Matrix to 19 Design Experiments in the Presence of Uncertain 20 Parameters"? 21 A Yes. 22 Q "Design of Experiments Using Uncertainty 23 Information"? 24 A Yes. 25 Q And these -- the publications that these have been 13 1 transferred in range from the ASME Journal of Heat Transfer 2 to Measurement Sciences and Technology referee journals? 3 A Yes. 4 Q Have you also done work for NASA? 5 A Yes. 6 Q And that was with respect to the space shuttle? 7 A Two different projects. One, radiative transfer 8 from the space shuttle and large space structures, and one 9 on the tiles. 10 Q Okay. And have you presented on the topics of 11 statistics and experimental design internationally? 12 A Yes, I have. 13 Q Do some of your papers include, with respect to 14 this, "Estimating Parameters using Griddy Markov Chain 15 Monte Carlo"? 16 A Yes. 17 Q Another one "Accounting for Heat Losses Using 18 Bayesian Inference"? 19 A Yes. 20 Q "The Use of Bayesian Inference, Monte Carlo, 21 Markov Chain Monte Carlo, and Gaussian Processes in 22 Designing Experiments and Analyzing Data"? 23 A Yes. 24 Q Did you in 2003 have one of your papers "Measured 25 and Predicted Temperatures of Automotive Brakes" chosen as 14 1 one of the outstanding papers of the year? 2 A Yes. 3 Q Now, you said that you were a professor of 4 mechanical engineering. Are you also a metrologist? 5 A Yes. 6 Q And what is a metrologist? 7 A Science and measurements. It's the study of the 8 science and measurements. 9 Q Okay. And as part of that, you need to have an 10 understanding of statistics, correct? 11 A Yes, I do. 12 Q A mastery of statistics? 13 A Yes. 14 Q But simply because someone knows statistics 15 doesn't make them a metrologist, does it? 16 A No, it does not. 17 Q To be a metrologist, is one of the things you 18 would have to know is an understanding of how to design 19 measurements to measure specific quantities of interest? 20 A Yes. 21 Q How to properly read and record data? 22 A Yes. 23 Q How to interpret data? 24 A Yes. 25 Q How to subject that data to proper statistical 15 1 analysis? 2 A Yes. 3 Q How to properly report -- I'm sorry -- how to 4 properly report the statistical characteristics of data? 5 A Yes. 6 Q I'm going to hand you what's been marked as 7 Exhibit 63. Can you identify that for The Court, please? 8 A It looks like my biographical record. 9 Q Okay. 10 MR. VOSK: Your Honor, at this time, I would ask 11 to have this admitted. 12 HEARINGS OFFICER KOEHLER: Mr. Garcia, any 13 objection? 14 MR. GARCIA: No objection. 15 HEARINGS OFFICER KOEHLER: All right. Exhibit 63 16 is admitted. 17 MR. VOSK: Your Honor, at this time I would also 18 like to ask The Court to take judicial notice of what the 19 Supreme Court's comment in City of Seattle v. Clark-Munoz 20 where I refer to Dr. Ashley Emery as an expert in the 21 science of metrology, which is the science of measurements. 22 And I'm going to ask this Court to allow him to testify as 23 an expert witness. 24 HEARINGS OFFICER KOEHLER: Mr. Garcia, do you have 25 any objections to Dr. Emery being qualified as an expert 16 1 witness in metrology? 2 MR. GARCIA: No. 3 HEARINGS OFFICER KOEHLER: All right. Dr. Emery 4 is so qualified. 5 MR. VOSK: Okay. Thank you, your Honor. 6 Thank you, Counselor. 7 Q (By Mr. Vosk) Now, when we're calculating the mean 8 of a set of numbers and we don't care about their origin or 9 anything, we can just take all those numbers at one time 10 and throw them in a pot and do it in a purely mathematical 11 way -- add them all up and divide by the total number that 12 are there. 13 A That's called the arithmetic mean and it's 14 commonly used. 15 Q Okay. And is -- that's what a math student might 16 use where you just walk into class and are given a set of 17 numbers? 18 A Typically, yes. 19 Q Okay. And if we were to do measurements on a 20 single instrument and over the period of that time the 21 instrument was stable, we'd be able to use that method to 22 determine the mean of the data generated, wouldn't we? 23 A Yes. 24 Q Now, if we try the same exercise and collect the 25 data on multiple different instruments, can we necessarily 17 1 do that anymore? 2 A Well, you can do it. If you're asking me, is that 3 the correct procedure -- 4 Q Then let me ask you that: Is that the correct 5 procedure? 6 A No. 7 Q Okay. And why is it different? 8 A Because different instruments have different 9 levels of precision. 10 Q Okay. 11 A And so you would normally weight the data 12 according to the precision of the instrument. 13 Q Okay. And is it -- is this important -- can it be 14 important? 15 A Well, you should weight them first. Yes, you 16 should always weight them. 17 Q Okay. How would one determine whether or not 18 there was a difference in the precision between 19 instruments? 20 A You look at their standard deviation of the 21 measurements that come from each instrument. It's 22 gathering the data -- assuming that the data has a set 23 value. During the calibration procedure, you'd be looking 24 at the precision of the instrument. 25 Q Okay. And so to be scientifically valid during 18 1 this process, we would have to take a look at that? 2 A Yes. 3 Q Okay. Now, is there a solution if the machines 4 have or show a different variance? 5 A You mean if the different instruments have 6 different levels of precision? 7 Q Correct. 8 A Then you would weight them and there is a formal 9 method for weighting them, yes. 10 Q Okay. And if it came out that the precision on 11 each was the same, we could again go back to the purely 12 mathematical method, couldn't we, in calculating a mean? 13 A Go back to the arithmetic mean, yes. 14 Q Okay. 15 A You'd get the same answer. 16 Q And with the differing precisions, we've got to do 17 this weighting? 18 A Yes. 19 Q So, in essence, using a purely mathematical 20 approach without weighting, it might be right in a given 21 context, it might be wrong, but you'd never know using 22 different instruments unless you calculated the precision? 23 A That's correct. 24 Q Okay. So the determination of a mean would be 25 scientifically invalid unless the precision across those 19 1 instruments was accounted for? 2 A It would -- the weight mean is the accepted way to 3 do it. I don't know what you mean by a scientifically 4 invalid. 5 Q Okay. 6 MR. VOSK: Your Honors, can we have this marked? 7 And can we have that marked? 8 HEARINGS OFFICER KOEHLER: All right. I've marked 9 the documents that you've given me as Exhibit 64 and 10 Exhibit 65. 11 MR. VOSK: Okay. And your Honor, pursuant to the 12 WAC, I would just ask The Court to take judicial notice off 13 of the website that these are, in fact, the certificates 14 from the Washington State Patrol website and I'd asked to 15 have them admitted at this point. 16 HEARINGS OFFICER KOEHLER: And they are the 17 documents off the website for what? 18 MR. VOSK: For the simulator solution 07007, and 19 they are the corrected -- let the record show I am using 20 quotation marks, the "corrected" simulator solution 21 certificates done by Mr. Gullberg and Trooper Dunn. 22 HEARINGS OFFICER KOEHLER: And have these not been 23 entered into evidence before? 24 MR. VOSK: I don't know that these particular ones 25 have, no. These were sent to me by Mr. Gullberg, which is 20 1 why I'm using them. 2 HEARINGS OFFICER KOEHLER: Mr. Garcia, do you have 3 any objection? 4 MR. GARCIA: Well, I guess my query is -- my 5 understanding was that all exhibits were to have been 6 disclosed by Monday so that I'd have an opportunity to 7 review these with the witness. And my query is what the 8 relevance of 07007 is to this proceeding, this was not the 9 batch in this case. 10 MR. VOSK: We're using data -- if your Honors 11 remember at the last hearing, we informed The Court that we 12 had just been provided information by Mr. Gullberg and we 13 had not had a chance to review it. The information he 14 provided was on all this solution 07007. 15 And so we figured if we were going to address the 16 issues that we had raised and they were trying to say 17 weren't issues, we'd go ahead and use the numbers that Mr. 18 Gullberg's already done calculations on to make it easier. 19 HEARINGS OFFICER KOEHLER: Are you going to have 20 Dr. Emery testify about these documents? 21 MR. VOSK: He'll testify to calculations -- 22 HEARINGS OFFICER KOEHLER: All right. 23 MR. VOSK: -- on these documents, yes. 24 HEARINGS OFFICER KOEHLER: Why weren't they 25 disclosed on Monday? 21 1 MR. VOSK: We disclosed the plots that -- the 2 plots that we disclose to The Court were based on this. 3 And we indicated that we would use all certificates, either 4 provided and hadn't been marked or were on the website. It 5 was in that e-mail that Mr. Bianchi sent. 6 HEARINGS OFFICER KOEHLER: All right. That was 7 indicated in your disclosure from Monday where you 8 indicated documents relied upon simulator solution 9 certification worksheets and associated gas chromatograms 10 available on WSP website and/or already marked as Exhibits 11 2. 12 MR. VOSK: No, no, no. As exhibits, and then 13 there's two. 14 HEARINGS OFFICER KOEHLER: Oh, I see it. Okay. I 15 see. All right. I'll allow it. 16 MR. VOSK: Thank you, your Honor. 17 So we would move to admit Exhibit 64 at this 18 point. 19 HEARINGS OFFICER KOEHLER: Exhibit 64 is admitted. 20 MR. VOSK: Okay. Did your Honors mark that? 21 HEARINGS OFFICER KOEHLER: 65. 22 MR. VOSK: I'm going to show this to the other 23 side. I only have one but when he's done, he'll pass it to 24 you. You'll see why there's only one. 25 Q (By Mr. Vosk) Dr. Emery, did I send you some 22 1 materials that I indicated had been sent to me by Mr. 2 Gullberg? 3 A Yes, you did. 4 Q Okay. And did those materials related to 5 Simulator Solution No. 07007? 6 A Yes. 7 Q And did I ask you to take a look at those -- at 8 those solutions and do some calculations? 9 A Yes, you did. 10 Q Okay 11 MR. VOSK: Your Honors, I'm sorry. Can we mark 12 this collectively as one exhibit? 13 HEARINGS OFFICER KOEHLER: And have you shown a 14 copy of this to counsel? 15 MR. VOSK: Yes. This is what was -- these are the 16 plots that were sent by Trooper Gullberg. So they've been 17 e-mailed to everybody already. 18 HEARINGS OFFICER KOEHLER: All right. I'm marking 19 the document that you handed me as Exhibit 66, and it 20 includes five page. 21 MR. VOSK: Okay. Thank you. 22 Q (By Mr. Vosk) Okay. Dr. Emery, I'm going to hand 23 you what's been marked as Exhibit 66. Are those materials 24 that I also sent to you that I indicated had been provided 25 by Trooper -- Mr. Gullberg? 23 1 A Yes. 2 Q Okay. And on page -- on Figure 1. Did you take a 3 look at Figure 1? 4 A I'm looking at it. 5 Q Okay. Now, first off, it indicates that there are 6 -- this is for three instruments. What does the plot 7 actually indicate, three instruments or four? 8 A Four. I presume the numbers below refer to 9 instruments, one, three, four, and five. 10 Q Okay. And on there it also refers to a reference 11 value, does that appear to be the mean value? 12 A Yes, it does. 13 Q Okay. And, if I might -- 14 MR. VOSK: Did your Honors print up the exhibits 15 that we sent? 16 HEARINGS OFFICER BARTON: I did. 17 MR. VOSK: You did. Okay. Do you have the plots 18 because I'll have him explain -- 19 HEARINGS OFFICER BARTON: Does he have a copy? 20 MR. VOSK: Yes, he does. 21 Q (By Mr. Vosk) Now, when we look at that, what do 22 the dots in that picture seem to indicate? 23 A They appear to indicate the average of a set of 24 readings. 25 Q Okay. And what do those bars indicate? 24 1 A The bars are indicative of a standard deviation. 2 Q Okay. Now, some of the bars are smaller than 3 others. What does that mean? 4 A That that instrument had less scatter in it's 5 readings. 6 Q Okay. Now does that mean the measurements agree 7 with each other more in those sets? 8 A Yes, it does, but at different -- less from the 9 mean than others. 10 HEARINGS OFFICER KOEHLER: Counsel, because I 11 don't want there to be any confusion -- 12 MR. VOSK: Yes. 13 HEARINGS OFFICER KOEHLER: -- when people read 14 this record, what I'd like to do is number the pages in 15 Exhibit 66 -- 16 MR. VOSK: Okay, your Honor, yes. 17 HEARINGS OFFICER KOEHLER: -- so you can refer to 18 those on the record. 19 MR. VOSK: Certainly. Then I guess -- why don't 20 we make the -- 21 HEARINGS OFFICER KOEHLER: All right. The cover 22 page is the narrative, that's page 1. And the graphs are 23 numbered pages 2 through 5. 24 MR. VOSK: Thank you, your Honor. 25 HEARINGS OFFICER KOEHLER: So the graphs and the 25 1 narrative is Exhibit 66. 2 MR. VOSK: Correct. 3 HEARINGS OFFICER KOEHLER: All right. 4 MR. VOSK: I will try to pick it up a little bit, 5 your Honor, I'm dragging a little bit. 6 Q (By Mr. Vosk) Taking a look at page 1, Figure 1 -- 7 HEARINGS OFFICER KOEHLER: That's page 2. 8 Q (By Mr. Vosk) Page 2, Figure 1. Can you explain 9 with respect to weighting when you look at that figure, how 10 your discussion of weighting values applies? 11 A Instruments 4 and 5 have a larger standard 12 deviation. It appears something in the order of three to 13 four times the standard deviation for Instruments 1 and 3. 14 Weighting is proportional to the square of a standard 15 deviation, so if they were -- if experiment -- if 16 Instrument 4 had three times the standard deviation of 17 Instrument 1, it would be weighted 1/9 of the weight 18 assigned in Instrument 1. 19 Q Okay. So when we're looking at that, in essence, 20 what you're saying is the small bars representing greater 21 precision are given more weight than the data sets with 22 large bars which have less precision? 23 A That's correct. 24 Q Okay. Now, did I have you go through and -- on 25 these four instruments that were identified here -- do 26 1 calculations -- calculate the mean based on a weighted 2 average? 3 A Yes, you did. 4 Q And I'm going to hand you what has been marked as 5 Exhibit 65. What do the numbers on Exhibit 65 represent 6 at this point? 7 A Exhibit 65 lists the individual readings for 8 Instruments 1, 3, 4, and 5. And then there are 9 placeholders for calculational results. 10 Q Okay. And did you go through and do the 11 calculations that you explained earlier for 07007? 12 A Yes, I did. 13 Q And did you bring those calculations with you? 14 A Yes, I did. 15 Q Can you fill in the spots on that blank indicating 16 the calculations you made and then also provide The Court 17 with a copy of your actual calculations? 18 A Yes, I can. 19 MR. VOSK: If we could just have a moment for him 20 to do that, your Honor? 21 HEARINGS OFFICER KOEHLER: Yes. 22 THE WITNESS: And you want me to fill those 23 numbers in -- 24 MR. VOSK: Yes. 25 THE WITNESS: -- that's what you're asking me to 27 1 do. 2 MR. VOSK: And that's the weighting factor. 3 THE WITNESS: Yes. 4 HEARINGS OFFICER KOEHLER: Were you handing me a 5 copy of 65? 6 MR. VOSK: I'm going to. As soon as he fills it 7 out, I'm going to hand it to you to admit. This is the 8 actual calculations he did. I'd like to have that marked. 9 HEARINGS OFFICER KOEHLER: Okay. I'm marking the 10 document you just handed me as Exhibit 67. It's a 11 three-page document. 12 THE WITNESS: And do you want the weighting 13 factors associated with these? 14 MR. VOSK: Sure. 15 And your Honors, as soon as he's done doing these, 16 I'm going to have him explain the calculations that he's 17 doing so your Honors understand what it is he's done. 18 Q (By Mr. Vosk) Okay, Dr. Emery. Now, can you 19 explain for The Court, what you just did, what those 20 calculations mean? 21 A I calculated the average of the readings for each 22 instrument, the standard deviation for each instrument -- 23 the standard deviation being the square root of the 24 variance for each one of the instruments -- for Instruments 25 1, 3, 4, and 5. And then I assigned a weight to the 28 1 average of each instrument being one over the variance. 2 And I weighted the means by those weights and I obtained a 3 weighted mean, and I obtained an arithmetic mean which 4 assumes that all instruments have equal weight. 5 Q And what was the value for the weighted mean? 6 A The value for the weighted mean was .10198. 7 Q Okay. And if we round that to four digits, what 8 is that? 9 A It would be 1020. 10 Q And what was the arithmetic mean rounded to four 11 digits? 12 A Arithmetic mean rounded to four digits, 1018. 13 Q And was that the mean that was reported for 14 solution 07007? 15 A I will have to double check. Yes, it is. 16 Q Okay. 17 A You're assuming -- I assume that you're asking off 18 of the document that came from the website? 19 Q That's correct. 20 A Yes. 21 Q And what is the difference between the two? 22 A .0002. 23 Q Okay. 24 MR. VOSK: Now, your Honors, I would ask at this 25 point to have Exhibit 65 and Exhibit 67 admitted, the 29 1 calculations. 2 HEARINGS OFFICER KOEHLER: Are 67 his handwritten 3 calculations? 4 MR. VOSK: That's correct. So The Court will have 5 the work before it so it can check to make sure he didn't 6 do something wrong. 7 HEARINGS OFFICER KOEHLER: Mr. Garcia, do you have 8 any objections? 9 MR. GARCIA: The objection would be that these 10 were not disclosed in advance so that we could check them 11 prior to the hearing. This is exactly the reason. This 12 was supposed to be disclosed by Monday. 13 MR. VOSK: Well, your Honor, I didn't have Mr. 14 Emery do this until last night, is the first point. The 15 second point, when we came in last week, we had just been 16 disclosed a bunch of documents by Mr. Gullberg. We had 17 asked for some extra time so we could take a look at what 18 they meant. The Court did not give us that time. At this 19 time, these calculations are exactly what we needed the 20 time to check. 21 When I called Ashley Emery yesterday, I requested 22 -- I asked him if he could do the calculations so The Court 23 would have them, and he said yes. There was no intent not 24 to provide anybody -- I just thought it would be better to 25 do this for The Court. 30 1 HEARINGS OFFICER KOEHLER: Has Mr. Gullberg been 2 able to check the calculations? I see that he's sitting 3 next to you, counsel. 4 MR. GARCIA: Yes, your Honor. I've asked him to 5 go through them. And he's already identified one are two 6 issues that we have with them. We'll have to discuss them 7 outside. My problem is, that's exactly -- the fact that it 8 causes problems for counsel, I don't think is relevant. 9 The Court made an order, and I think The Court should 10 enforce its order on the basis that counsel doesn't get to 11 decide, well, I really need the information. 12 Obviously, this is prejudicial to our ability to 13 cross examine Dr. Emery. I mean, we gave this information 14 to him over a week ago. It's now two weeks ago that he 15 received this information and last night he asked his 16 expert to review the data. So we don't think that it's 17 fair to suggest to The Court that somehow he need the 18 additional time of this week, eliminating a couple of days. 19 But aside from that, your Honor, there's the issue 20 of relevance. I still don't understand how Dr. Emery's 21 examination of the solution that doesn't have anything to 22 do with this case is relevant. So I guess ultimately when 23 we're done and we ask Mr. Emery at the end, how are they 24 going to be able to tie his theory to this case. And 25 that's the difficulty I'm having. I mean, aside from his 31 1 calculations and whether they're wrong or right, how is he 2 going to be able to say that this had any affect on the 3 breath-test result in this case. 4 MR. VOSK: This is how we're going to tie it in, 5 your Honor. A couple of things, first with respect to 6 having provided us materials last week, it is very 7 intellectually dishonest for counsel to try to put up any 8 wall of defense based on that, because when we came in here 9 last week and said, well, we'd like a little extra time to 10 consider it, he was pushing to go forward and was 11 unrelenting. So we've had a week for me to allow Mr. 12 Emery to take a look at these. We got together over the 13 weekend and spoke about them. 14 In your order -- what your order indicated was 15 that we were supposed to supply all materials that he was 16 going to be relying upon. He's relying upon the 17 certifications, the plots, the data. What is here is the 18 work that he produced in reliance upon these things. If 19 you would like -- and the reason I had him do these last 20 night was, I was going to have him come in here and do this 21 right here in front of you, and he can, but it was going to 22 take forever. 23 HEARINGS OFFICER KOEHLER: Okay. I'm holding up 24 my hand. Now, do I understand that you used 0007 because 25 those were -- that was the batch that Mr. Gullberg had 32 1 originally used when he was providing you information? 2 MR. VOSK: Correct. It was a batch provided by 3 Mr. Gullberg. And what this is -- and we're going to tie 4 this into specific batches in this case as well -- it was a 5 demonstration that their calculation of the mean is not 6 being done correctly. 7 I could have chosen any certification to do it; I 8 chose to rely on Mr. Gullberg's, because I would assume 9 that if he's sending me this this is his best example. So 10 I figured, okay, let's see if his best example gets around 11 the problems that we found. And when we did the 12 calculations, it didn't. That's why we used it. 13 HEARINGS OFFICER KOEHLER: Okay. 14 MR. VOSK: So it was relied upon by their expert 15 and provided to us. 16 HEARINGS OFFICER KOEHLER: All right. I am going 17 to allow the testimony and when Dr. Emery's testimony is 18 completed, I'll determine whether or not the exhibits will 19 be admitted. 20 MR. VOSK: Okay. 21 HEARINGS OFFICER KOEHLER: And I'm referring to 22 Exhibit 65 and 67. 23 MR. VOSK: I guess, your Honor, if that's going to 24 be the ruling, then I'll have him -- I can just have him 25 right now go through and do the calculations, but that's 33 1 going to -- I mean, it's going to bog us down for 20 to 25 2 minutes. I was honestly trying to facilitate the Court. 3 HEARINGS OFFICER KOEHLER: It is demonstrative 4 evidence and he can rely on it during his testimony and 5 then, as I indicated -- 6 MR. VOSK: Well, let me ask this then, your Honor, 7 the conclusions at least, because it is -- those are the 8 written conclusions of our expert witness based on the 9 material supplied by the State's expert witness -- the 10 conclusions at least should be admissible. If you don't 11 want to accept the calculations so that you can examine 12 them, I guess that's your choice. But his opinions -- his 13 expert opinion as he's written down on that paper -- you've 14 already called him an expert -- that is admissible. 15 HEARINGS OFFICER KOEHLER: But he's going to 16 testify to it. 17 MR. VOSK: Well, that -- well, he did just testify 18 to the numbers. He's given those to you and now he's going 19 to explain what they mean. If you want him to go through 20 and indicate each number, I just thought this would be an 21 easier way for you to have it right in front of you. I'm 22 not trying to play games. I'm trying to make it easier for 23 you. 24 HEARINGS OFFICER KOEHLER: Well, go ahead and have 25 him testify -- 34 1 MR. VOSK: Okay. 2 HEARINGS OFFICER KOEHLER: -- to what they mean. 3 Q (By Mr. Vosk) Can you go through this, Dr. Emery? 4 For each instrument indicate which analysts were included 5 for each instrument; read off each data figure that was 6 there; then give us the number of measurements that were 7 there, the mean, the standard deviation, and the weighting 8 factor for each instrument, please. 9 A Yes, I can. Do you want me to? 10 Q The Court does want you to. 11 Mr. Garcia: And I'm going to object. And I'm 12 going to indicate that it's a waste of The Court's time to 13 have him indicate the data. It is sufficient that he gives 14 The Court his opinion on the data. And we intend to cross 15 examine him on the basis for his opinion. I don't think 16 that it's necessary for The Court to hear each data point. 17 HEARINGS OFFICER KOEHLER: All right. I'm going 18 to reverse myself on Exhibit 65. I'm going to allow that 19 into evidence, because I do believe that it will be helpful 20 to the trier of fact and it will eliminate a lot of time 21 that we would otherwise have to spend. So 65 is admitted. 22 Q (By Mr. Vosk) Mr. Emery can you just then for each 23 instrument indicate what mean you found, what the standard 24 deviation was, and the weighting factor. Just those three 25 numbers for each instrument. 35 1 A For Instrument 1, 20 readings. The mean was 2 0.10285; the standard deviation was 0.67082 times 10 to the 3 minus third; the weighting factor was 2.222222. 4 MR. VOSK: And your Honors, do you want him to go 5 through and read that or are you willing to just accept 6 what's written. 7 HEARINGS OFFICER KOEHLER: Yes. 8 Q (By Mr. Vosk) Okay. Can you now -- did you -- 9 after you calculated the means, did you divide those by 10 1.23? Did you do that calculation? 11 A No, I did not. 12 Q Would you be able to do that calculation for The 13 Court here today? 14 A Yes. 15 Q Would you, please, using means rounded to four 16 digits? 17 A You want me to take the weighted mean and divide 18 it by 1.23? 19 Q Right. The weighted mean rounded to four digits. 20 A You want me to round it before dividing or after 21 dividing? 22 Q Before dividing. 23 A The first reading would be -- for the weighted 24 mean, it would be .08291. 25 Q Can you write that down? 36 1 A Yes, I can. 2 Q And what would the answer have been without 3 rounding, without -- 4 A .082911. 5 Q So just an extra one? 6 A Yes. 7 HEARINGS OFFICER BARTON: Counsel, can you just 8 say -- what's he calculating right now? 9 MR. VOSK: What he's doing -- I'm not sure if the 10 testimony was here or it was in Skagit -- when you take the 11 -- and I think -- I believe it was in Skagit -- when you 12 take the solution concentration, you divide it by 1.23 and 13 that will give you the vapor concentration. 14 So the number that we're looking at, for instance, 15 in the memo from August of 2007 where it talks about the 16 breath-test readings either being bumped up or knocked down 17 because there was an improper number in the fourth decimal 18 place, it's the vapor concentration that's he's calculating 19 now, which is that number. 20 HEARINGS OFFICER BARTON: Thank you. 21 Q (By Mr. Vosk) And so for the unweighted mean, what 22 did you come up with? 23 A 0.08279. 24 Q Okay. And what is the difference between those 25 two, the absolute difference? 37 1 A .0001. 2 Q So one ten-thousandth of -- one ten-thousandth? 3 A Yes. 4 Q Okay. 5 MR. VOSK: Your Honors, can I please mark this. 6 HEARINGS OFFICER KOEHLER: Okay. We're looking at 7 Exhibit 21 and it looks like part of this other exhibit 8 that you've handed me is in Exhibit 21. Is this the -- 9 MR. VOSK: Does 21 include the chromatogram -- I 10 didn't think -- so what this would be -- this is meant to 11 be then -- if 21 is 07004, this is meant to include all of 12 the chromatograms in addition to it. 13 This is -- Mr. Trombold printed this. Mr. 14 Trombold printed this with two pages -- I mean, subsequent 15 pages. So if it looks different, that's why. 16 HEARINGS OFFICER KOEHLER: All right. I'm marking 17 this document as Exhibit 68. 18 MR. VOSK: Okay. 19 MR. TROMBOLD: Thank you, your Honors. 20 Q (By Mr. Vosk) Now, Dr. Emery, if you take a look 21 at what's been marked as Exhibit 6. 22 A Yes. 23 Q Does it indicate what batch number that is for? 24 A This is not marked? 25 Q Oh, this is 68. 38 1 A Exhibit 68 indicates it's for batch 07004. 2 Q Okay. And for that batch, if you can take a look 3 at the gas chromatograms as we go back, were three 4 different instruments used? 5 A No, it was only one instrument. 6 Q Okay. What I'm talking about is a gas 7 chromatogram for each separate analyst. If you can turn to 8 page -- this first page for Sarah Swenson. What instrument 9 does it indicate that she used? 10 A Instrument No. 5. 11 Q Okay. And if we turn further back to analyst 12 Paige Long, what instrument does it indicate that she used? 13 A She used Instrument No. 4. 14 Q Okay. And if you can turn back to analyst 15 Estuardo Miranda, what instrument does it indicate that he 16 used? 17 A He used Instrument 1. 18 Q Okay. So each of these used a different 19 instrument? 20 A That's correct. 21 Q Okay. Can you then -- since we've only got three, 22 can you, for The Court, right now go ahead and do for this 23 certificate what you did for the longer one, right now? 24 A Yes, I could. I've done -- I also did that last 25 night. 39 1 Q Oh, you did that in advance. Okay. Can you 2 report then for -- 3 MR. VOSK: And at this point, your Honor, I would 4 ask to admit Exhibit 68, again from the WAC. It's judicial 5 notice extracted from the website. 6 HEARINGS OFFICER KOEHLER: Mr. Garcia, do you have 7 any objection? 8 MR. GARCIA: No objection. 9 HEARINGS OFFICER KOEHLER: All right. Exhibit 68 10 is admitted. 11 Q (By Mr. Vosk) For Instrument 1, did you come up 12 with a mean? 13 A Yes, I did. It's 0.1298. 14 Q And did you come up with a standard deviation? 15 A It's .4472 x 10 to the minus third. 16 Q Okay. And did you come up with a mean and a 17 standard deviation for the other two instruments as well? 18 A Yes, I did. 19 Q And are those calculations there before you? 20 A Yes, they are. 21 Q Can you hand those to me, please. 22 MR. VOSK: Your Honors, can we get this marked? 23 HEARINGS OFFICER KOEHLER: I'm marking these 24 hand-calculations regarding batch 07004 as Exhibit 69. 25 Q (By Mr. Vosk) And what's been marked as Exhibit 40 1 69, do they represent the same analysis on that solution as 2 you did to the previous one? 3 A Yes, they do. 4 Q And did you come up with a weighted mean for that? 5 A Yes, I did. 6 Q And what was the weighted mean? 7 A 0.1287. 8 Q Okay. And what was -- did you come up with an 9 arithmetic mean? 10 A Yes. And it was 0.1289. 11 Q Okay. And can you divide each of those numbers by 12 1.23? 13 A Yes, I can. The weighted mean divided by 1.23 is 14 .104634. 15 Q And can you indicate -- can you put that down on 16 your calculations? 17 A (Witness complies.) 18 And the arithmetic mean divided by 1.23 is 19 0.104821. And the difference, if that's what you'd look 20 for -- 21 Q I was going to ask you the difference. 22 A The difference is, rounded to four places, 0002. 23 Q Okay. So there the difference between the 24 weighted mean and the arithmetic mean is twice as much as 25 the last one, two ten-thousandths? 41 1 A Yes, it is. 2 Q Okay. 3 MR. VOSK: And your Honors, for the same reason as 4 the first one, I'd simply ask to have Exhibit 69 admitted. 5 MR. GARCIA: Same objections as before. 6 HEARINGS OFFICER KOEHLER: All right. Exhibit 69 7 is admitted. 8 Q (By Mr. Vosk) And did you do a similar set of 9 calculations for simulator solution batch 6012? 10 A Yes, I did. 11 Q And can you hand those to me, please. 12 MR. VOSK: Your Honors, we're handing you two 13 exhibits to be marked. 14 HEARINGS OFFICER KOEHLER: All right. I'm marking 15 some hand-calculations for batch 06012 as Exhibit 70. And 16 I'm marking what appears to be the Washington State Patrol 17 printouts for batch 06012 including chromatograms as 18 Exhibit 71. 19 MR. VOSK: Thank you, your Honor. 20 Q (By Mr. Vosk) Looking at Exhibit 70, did you do 21 the same calculations on that batch as you did on the prior 22 two? 23 A Yes, I did. 24 Q And the mean, the standard deviations, and the 25 weighting factors indicated there, they have the same 42 1 meaning that those done in your prior calculations did? 2 A Yes, they do. 3 Q And did you compute a arithmetic mean? 4 A Yes, I did. 5 Q And what was the arithmetic mean? 6 A 0.126667. 7 Q And did you compute a weighted mean? 8 A Yes, I did. That is 0.12702. 9 Q And did you calculate the equivalent vapor 10 concentration for each of those? 11 A Yes, I did. 12 Q And what is that? 13 A .102981 for the arithmetic average; .103268 for 14 the weighted mean. 15 Q So rounded to four digits? 16 A The absolute difference is .0003, rounded to four 17 decimal places. 18 Q Okay. So now that's three times greater than the 19 first one we saw. We're now up to three ten-thousandths? 20 A Yes. 21 Q Okay. 22 MR. VOSK: Your Honor, I would move to admit 23 Exhibit 70 and 71. 24 HEARINGS OFFICER KOEHLER: Mr. Garcia? 25 MR. GARCIA: 71 is the calculations? 43 1 HEARINGS OFFICER KOEHLER: 71 is the 2 hand-calculations for batch 06012. 3 MR. GARCIA: I'll just request standing objection 4 to newly discovered, newly disclosed evidence that we 5 haven't had an opportunity to review. 6 HEARINGS OFFICER KOEHLER: I'm sorry. Exhibit 70 7 is the hand-calculations. 71 is the documents from the 8 Washington State breath-test site. 9 Did you have any objection to 71? 10 MR. GARCIA: I'd request a standing objection to 11 any newly disclosed exhibits that Defense is offering. I 12 understand The Court's intention is to admit them. 13 HEARINGS OFFICER KOEHLER: All right. Exhibits 70 14 and 71 are admitted. 15 MR. VOSK: And we have one outstanding exhibit, 16 Your Honor, the calculations that Dr. Emery did for 07007. 17 You've got the results in front of you, the three pages of 18 actual calculations. I would move to admit. If your 19 Honors wanted to examine them to see what the process is -- 20 I think they are relevant if -- 21 HEARINGS OFFICER KOEHLER: 67 was the hand-written 22 calculations for batch 07007. 23 MR. VOSK: Yes. And you've got the results on a 24 sheet on 65. And those are just what was used to calculate 25 those. 44 1 HEARINGS OFFICER KOEHLER: All right. And exhibit 2 -- same standing objection, Mr. Garcia? 3 MR. GARCIA: Yes. 4 HEARINGS OFFICER KOEHLER: All right. Exhibit 67 5 is admitted. 6 MR. VOSK: Thank you, your Honor. 7 Q (By Mr. Vosk) When we talked -- you were here last 8 week, Dr. Emery. Did you hear Mr. Gullberg speak about 9 consensus values? 10 A Yes, I did. 11 Q And the weighted means that you just calculated, 12 are those consensus values? 13 A Yes. 14 Q And are those the correct way to calculate a 15 consensus value? 16 A Yes. 17 Q And is it true that generally speaking, neither a 18 grand average of all measurements nor the average of 19 measurements of an individual set will necessarily be the 20 same as the consensus value? 21 A A consensus value defined to be the weighted 22 average, the arithmetic average, need not be the same as 23 the weighted average. 24 Q Okay. Now when we make measurements -- 25 MR. VOSK: I'm sorry, your Honors, just give me a 45 1 second. 2 Q (By Mr. Vosk) When we make measurements, do we 3 often have a preconceived notion of what the value we're 4 measuring should be? 5 A Yes, frequently but sometimes not. 6 Q Okay. Well, in those occasions when we know or we 7 have this notion of what the value should be then, why do 8 we do any measurements? 9 A Well, because typically you only have an 10 approximate idea of what it should be. If I give an 11 example, your heighth. I might guess that you're 5 foot 12 nine or something like that, but I'm a pretty poor judge of 13 heighth. 14 Q Okay. And so -- 15 A So I measure you. 16 Q And so we do measurements because we might be 17 wrong? 18 A Yes. 19 Q Okay. 20 A Because your preconceived ideas might be wrong. 21 Q Okay. Now, if we simply blindly reject any data 22 that doesn't conform to those preconceived ideas, is that 23 an appropriate scientific procedure methodology? 24 A Usually you establish a criterion by which you 25 reject data. 46 1 Q Okay. 2 A It's probably the most difficult aspect of 3 laboratory measurements. 4 Q Okay. And why -- why is that difficult? 5 A Well, because a data point which is far from the 6 average or far out of your expectation can be the result of 7 simply an inaccurate measurement, an instrument which is 8 not precise enough, a mistake recording data, or it may 9 actually be telling you something about the readings that 10 you're making. 11 And so ignoring it, you may lose an opportunity to 12 understand more fully what's happening; but on the other 13 hand, there are statistical techniques for making a 14 decision as to when to ignore something, but they all take 15 risk that you're throwing away something that's valuable. 16 Q Okay. Now, if we don't utilize those statistical 17 techniques, if we just throw it away -- if we just discard 18 any number that disagrees with our initial preconceived 19 ideas, don't we almost guarantee that we're going to get 20 the answer we wanted to get in the first place? 21 A Well, if you carry it out long enough, yes, but 22 then you might ask yourself, why do the experiment. 23 Q Okay. So then when we talk about throwing out 24 data, are we talking about something called an outlier? 25 A Yes, that's the typical term used. 47 1 Q Okay. Now, in general terms, can you 2 qualitatively tell us what an outlier is? 3 A An outlier is one or more -- outlier or outliers 4 are one or more data points which on first inspection 5 appear to be unusual and either the result of a 6 transcribing error or a variation in the result. 7 Sort of like if you get up and weigh yourself in 8 the morning, and you've been weighing yourself all week, 9 and every morning you weigh within one pound and suddenly 10 one time you stand up there and it weighs four pounds 11 different. It may be that you're actually four pounds 12 different weight or it may be you stood on the scales in a 13 funny way. And so what you do is that you do the 14 experiment again. 15 Q Okay. Now, just because the measurement deviates 16 though, doesn't mean it's a bad measurement? 17 A No, it does not. 18 Q And so before discarding that measurement, how do 19 we determine whether or not it is an outlier? 20 A Well, there are a number of different statistical 21 criteria which are used. The one that's taught most often 22 to students is Chauvenet's criteria. And that one says 23 that you're willing to take a risk of one over two times 24 the number of readings when you throw it out. So if you 25 have 10 readings, 1/10 is 1/20 or 5 percent. So if you 48 1 throw -- if that one is judged to be an outlier by 2 Chauvenet's criteria, you're taking a 5 percent risk of 3 throwing away something that really matters. 4 Q Okay. 5 A Or, another way to look at it -- this way you're 6 95 percent sure it's an unusual reading that just occurred 7 by accident. 8 Q Okay. And when you're computing that number, are 9 you using any of the quantities that we've previously 10 looked at -- a mean, a standard deviation? 11 A Yes. If you're looking at the distance from the 12 mean, the deviation from the mean in terms of number of 13 standard deviations. So if the standard deviation is one 14 and the reading is three from the mean, that means it's 15 three standard deviations. If the standard deviation were 16 1 1/2, it would be two standard deviations away from the 17 mean. And then there are tables which tell you -- give 18 different levels of probability when you can throw 19 something out. 20 Q Okay. So then I'm going to try to -- I'm going to 21 try to put this in a -- make this a little bit clearer. So 22 if we've got a number that we think is an outlier, what 23 you're saying is the first step in the process is, we'll 24 take that number and determine the difference between it 25 and the mean of all the measurements? 49 1 A Yes. 2 Q Okay. And then we divide that difference by the 3 standard deviation? 4 A Yes. 5 Q And then depending on what kind of criteria we 6 want to use, it's that ratio which is going to determine 7 whether or not the value is an outlier? 8 A Yes. 9 Q Okay. Now, with respect to that ratio, could 10 somebody use 1.5 as a criteria to throw something out as an 11 outlier? 12 A Sure. 13 Q Could they use 2? 14 A Yes. 15 Q Could they use 2.5? 16 A Yes. 17 Q Could they use 4? 18 A Yes. 19 Q Essentially, when you're trying to establish the 20 criteria then, you can define the criteria you want to use? 21 A Yes. You'll have to argue for its support, but... 22 Q Okay. And different authorities recommend 23 different numbers? 24 A Yes. There are three. Three of the most common 25 criteria are Naire's, Pierce's and Chauvenet's. And they 50 1 come up with slightly different criteria. 2 Q Okay. 3 A And they're all -- they're generally expressed in 4 terms of what probability are you willing to accept that 5 you're throwing something out which is really a good value. 6 Q Okay. But regardless of which one you're going to 7 use for a set of measurements, we have got to ensure that 8 we're always using the same criteria, correct? 9 A Yes. And you can only apply in once. 10 Q Okay. 11 HEARINGS OFFICER KOEHLER: We're looking for 12 Exhibit 64. 13 MR. VOSK: I apologize to The Court. 14 Q (By Mr. Vosk) I'm going to hand you what's been 15 admitted as Exhibit 64. 16 Can you look at the readings for analyst 14? 17 A Yes. 18 Q Can you report to The Court what those five 19 readings are? 20 A In order they're .106, .100, .099, .100, and .099. 21 Q And would you be able to determine based on what 22 you've just testified to whether or not the .106 in that 23 set of data is an outlier or -- let me put it another way: 24 Can you calculate the ratio? 25 A Yes, I could. 51 1 Q And will you please do that for The Court? 2 A So I have to calculate everything. Okay -- I 3 guess I better turn it on. 4 Q And can you write your calculations down for The 5 Court as well, please. 6 A The average is .1008. The standard deviation is 7 0.00295. And I just want to double check that, if you 8 don't mind. Yes, the standard deviation is 0.00295. 9 And you're asking me how far the 106 is away? 10 Q Yes. I'm asking for that ratio you spoke about 11 earlier, the ratio between the difference between the 12 suspected outlier and the mean to the standard deviation. 13 A And the ratio is 1.763. And this is for analyst 14 14 and batch 07007. Okay. 15 Q Okay. Now I'm going to hand you what been marked 16 as Exhibit 72. Can you identify what batch number that is? 17 A That's batch 07023. 18 Q And in that -- Kelly Gross, on the test that she 19 discarded -- previously testified about -- can you tell The 20 Court -- 21 HEARINGS OFFICER KOEHLER: Counsel, what -- you 22 said that she previously discarded. Are you referring to a 23 particular exhibit? 24 MR. VOSK: This is batch number 07023. In Skagit 25 County, I think the only thing that was admitted as the 52 1 initial spreadsheet and Ms. Gross' discarded data. We're 2 just submitting the whole thing for completeness. 3 HEARINGS OFFICER KOEHLER: Okay. So when you're 4 referring to what was previously discarded, that's what -- 5 MR. VOSK: Yes, that's what I'm referring to. The 6 values were .101, .101, .108, and .101, and .102. 7 HEARINGS OFFICER KOEHLER: Which analyst is she? 8 MR. VOSK: That's Kelly Gross. She is analyst 15. 9 HEARINGS OFFICER KOEHLER: Thank you. 10 MR. VOSK: And so it won't be the values that 11 appear in the spreadsheet, they are actually the values 12 that were discarded by her. 13 Oh, I'm sorry, your Honor. Previously, if you 14 recall last week, we spoke about data that instead of being 15 entered into the spreadsheet, they were stuck into a folder 16 and not entered, and they would do a second run. 17 And if you read the Skagit County decision, this 18 was where the -- this was one of the tests the Judges 19 discussed that indicated it caused them some concern. We 20 had testimony on it last week, and I'm just going to have 21 Dr. Emery calculate the ratio for this one as well. 22 HEARINGS OFFICER BARTON: So I'm just seeing one 23 that was discarded. Is there more than one? 24 MR. VOSK: Right. What she did was she discarded 25 her data set and then reran a new five. 53 1 HEARINGS OFFICER BARTON: Well, just the one 2 chromatogram. I see that she -- 3 MR. VOSK: If you look, she's got two sets of five 4 in here. And what the testimony showed in Skagit County 5 was that when they get -- when they get something that they 6 folder, they rerun the whole thing. And so when she 7 crosses this out and says rejected because of the .108, 8 what that means is she rejected all five data -- 9 HEARINGS OFFICER BARTON: Aliquots? 10 MR. VOSK: All five aliquots that were collected 11 at that time and redid it with a new five. 12 HEARINGS OFFICER BARTON: Okay. But you agree 13 with me that just one has been crossed out? 14 MR. VOSK: That's correct, but the other one -- 15 the rest of the batch wasn't included in her data here. 16 What she did was she reran a new five and you have -- 17 because you'll find ten actual aloquets measured in there. 18 HEARINGS OFFICER BARTON: In Exhibit 72, do we 19 have 10 chromatograms? 20 MR. VOSK: Yes. 21 HEARINGS OFFICER BARTON: Okay. 22 MR. VOSK: Actually, I think you'll probably have 23 12 because they'll be a control and a blank. 24 HEARINGS OFFICER BARTON: Thank You. 25 MR. VOSK: They'll be 14, because they will be two 54 1 controls and two blanks for each one. Each time they run 2 the five aloquets, there's a control that goes along with 3 it to make sure that the -- they try to make sure the 4 machine is reading the control properly and then a blank so 5 it can try to get a zero. 6 HEARINGS OFFICER BARTON: Thank you. 7 Q (By Mr. Vosk) Now with those numbers, can you go 8 ahead and do with this one -- I'm sorry. The numbers we 9 have are .101 -- 10 A Okay. 0.1 -- 11 Q -- .101 -- 12 A -- well, .101 -- 13 Q .101. 14 A 101. 15 Q .108. 16 A 108. 17 Q .101. 18 A 101. 19 Q And .102. 20 A Okay. 21 HEARINGS OFFICER BARTON: And where do those 22 numbers come from? 23 MR. VOSK: They come from the chromatograms that 24 were discovered. 25 HEARINGS OFFICER BARTON: Where? Where do those 55 1 particular numbers come from? 2 MR. VOSK: Okay. If you'll notice, rejected due 3 to a .108. Where that's crossed out that's the control, 4 because at the top right-hand corner, you'll see a .10 5 control. 6 HEARINGS OFFICER BARTON: Okay. 7 MR. VOSK: The next one is a blank. Then you 8 begin getting the aloquets -- .101 -- right down here, 9 that's your measurement. 10 HEARINGS OFFICER BARTON: How do you know that 11 those were not included? 12 MR. VOSK: They're not on the front and they're 13 from Instrument 1 here, all run together. You can tell by 14 the times and the instrument. The data that she collected 15 that was included was from Instrument 3. And because of 16 her testimony -- she didn't testify, no. So it's just from 17 the chromatographs. These are from Instrument 1. These 18 are from Instrument 3. The ones from Instrument 3 were the 19 ones that were not discarded. 20 HEARINGS OFFICER BARTON: Thank you. 21 MR. VOSK: And that was, as a matter of fact, the 22 testimony up in Skagit is that that entire set was 23 discarded. 24 HEARINGS OFFICER BARTON: Did somebody else 25 testify to that? 56 1 MR. VOSK: I believe it was Dr. Logan or somebody 2 else because we spoke to them about it and they said yes, 3 this is what happens and it was admitted in Skagit. 4 HEARINGS OFFICER BARTON: Okay. Thank you. 5 MR. VOSK: And we just moved to admit Exhibit 72. 6 HEARINGS OFFICER KOEHLER: Counsel, same 7 objection? 8 MR. GARCIA: No, your Honor. I mean, this is on 9 the website, so I'm assuming that this is complete and it 10 is what it appears to be. 11 HEARINGS OFFICER KOEHLER: All right. Exhibit 72 12 is admitted. 13 Q Now, what did you get for a mean of those five 14 measurements? 15 A .1026. 16 Q What did you get as a standard deviation? 17 A 0.00305. 18 Q And what did you get as the ratio of the 19 difference between your suspected outlier and the mean to 20 the standard deviation? 21 A 1.77. 22 Q Okay. And so the difference between those two 23 when we're taking a look at it, just the ratio not as an 24 absolute number, 1.76 to 1.77; is that correct? 25 A Yes. 57 1 Q Now, we've had testimony that the second set of 2 data was discarded. Based on the number there, could 3 somebody using appropriate criteria have decided to discard 4 that set of data because the 108 was an outlier? 5 A Depending on what criterion they wanted, yes. 6 Q Okay. Now in the first number, with respect to 7 07007 that .106, that was not discarded. Now depending on 8 what criteria somebody wanted to use, it would be 9 appropriate not to discard that, wouldn't it be? 10 A Yes. 11 Q But in looking at the data, does there appear to 12 be a uniform criteria that's been applied to these two 13 tests? 14 A Well, first of all, I know nothing about the 15 decision to accept or to reject. And I don't know if they 16 apply any criteria to make this decision. 17 Q Would you have treated these numbers differently? 18 A That 176 and 177, I would probably have thought 19 that they were the same -- 20 Q Okay. 21 A -- but they're not substantially different from 22 each other. 23 Q So then if one had been rejected, in your 24 opinion, the other should have been rejected? 25 A If the criteria for rejection rejected one, the 58 1 other one, I would think so. If the criteria were to 2 accept one, the other one should have been accepted. 3 Q Okay. And so from that, do you see apparent 4 there any type of criteria that was utilized? 5 A I don't know if they utilized any criteria. I 6 don't have any access to what they do in the lab. Plus 7 there's no documentation that I was given that said that 8 they evaluated this ratio. 9 Q Okay. Would it have been -- in order to determine 10 whether or not it was an outlier however, the correct -- 11 the scientifically accepted procedure would have been what 12 you've just done. 13 A It would have been performed -- the ratio of the 14 deviation from the mean to the standard deviation and then 15 compare it to whatever criterion you want. Some criteria 16 might be 1 1/2 standard deviations and they're both 17 rejected, other criterions would be 2 and they'd both be 18 accepted. You just have to justify the criterion that you 19 want. 20 Q Okay. So just to quickly summarize what we've 21 done today -- 22 MR. GARCIA: Objection; cumulative. 23 MR. VOSK: I just want to make clear what he's 24 testified to, because I think there's been a lot of 25 techno-speaks and I just to try to make it clear for the 59 1 Judges. 2 HEARINGS OFFICER KOEHLER: Sustained. 3 Can you ask a question that's close to -- 4 MR. VOSK: Yeah. That's what I'm going to do. 5 HEARINGS OFFICER KOEHLER: All right. 6 Q (By Mr. Vosk) With respect to the calculation of 7 the mean, the scientifically appropriate method to use in 8 this context, is to use a weighted mean? 9 A Yes, it is. 10 Q And with respect to the determination of an 11 outlier, a scientifically appropriate method is to utilize 12 an established criteria? 13 A Yes, it is. 14 MR. VOSK: No further questions, your Honor. 15 HEARINGS OFFICER KOEHLER: Mr. Garcia, cross 16 examination? 17 MR. GARCIA: Yes, please, your Honor. 18 CROSS EXAMINATION 19 BY MR. GARCIA: 20 Q Dr. Emery, you said "in this context." What 21 context were you talking about when you said in this 22 context they should be using a weighted mean? 23 A Well, consensus value, as indicated by the 24 document that Mr. Gullberg provided and by the standard and 25 accepted practice, it is a weighted mean. 60 1 Q But my question was, you said "in this context." 2 A I cannot remember when I used the word context. 3 I'm sorry. 4 Q Okay. You would agree that in different contexts, 5 different methods of calculating the mean might be 6 appropriate? 7 A No. I am saying that sometimes the arithmetic 8 mean and the weighted mean will give you the same result. 9 Q Okay. Did you determine whether not the 10 differences that you found between the arithmetic mean and 11 the weighted mean made any difference in this context? 12 A I don't know. I can't speak to that. All I can 13 do is report what the differences are. Somebody else has 14 to make a judgment as to whether those differences are 15 important or not. 16 Q So you can't offer an opinion as to whether or not 17 the differences between using weighted mean and arithmetic 18 mean had any significant affect upon the breath test in any 19 case. 20 A No, I cannot. 21 Q Did you compare any of the values that were 22 actually applicable in this case, the Arntson case? 23 A I don't know what you mean by "applicable". 24 Q Did you examine the batch number for the 25 particular case in which we are involved 070 -- 61 1 A I have the information one, yes. I was provided 2 with information of this form, yes. 3 Q Okay. And you ran the calculation -- 4 A Yes. 5 Q -- for a weighted mean? 6 A Yes, I did. 7 Q And what was the difference between -- in other 8 words, what was the T value for -- 9 MR. VOSK: I'm sorry, your Honors, can Counsel 10 identify which solution we're discussing at this point 11 because I did not identify -- I didn't tell Mr. Emery which 12 solutions applied to which case. 13 HEARINGS OFFICER KOEHLER: Okay. You might be 14 discussing apples and oranges here. 15 By this solution, Mr. Garcia, do you mean 07010, 16 that batch number. 17 MR. GARCIA: Correct. 18 MR. VOSK: I don't believe I gave that -- provided 19 that to Mr. Emery. 20 HEARINGS OFFICER KOEHLER: Okay. 21 MR. VOSK: The ones that I provided to him were 22 the ones that he did the calculations on. 23 MR. GARCIA: All right. Well, my question was 24 that he had performed calculations on the solution 25 applicable to this case. He indicated he had. 62 1 Q (By Mr. Garcia) So you didn't perform any 2 calculations on 0701. I apologize. 3 A You have me at a disadvantage. I don't know what 4 this case is. 5 Q Yes, I understand. 6 A I performed the calculations on 07007. 7 Q Exhibit No. 64. 8 A And on 07004. 9 Q Yes. 10 A And on 06012. 11 MR. VOSK: And your Honors, for the record, I just 12 want to indicated I purposely didn't let him know whether 13 or not any specific solutions were applicable to this case 14 because I didn't want to take the chance of biasing any 15 results. So I just gave him the numbers and had him do the 16 calculations. 17 HEARINGS OFFICER KOEHLER: Did he answer your 18 question, Mr. Garcia? 19 MR. GARCIA: He did. He indicated that he had not 20 done -- according to the numbers of the batch list that he 21 identified, none of those was identified with this quality 22 assurance test. 23 MR. VOSK: And your Honor, I would -- I'm going to 24 make an objection that that's not what he indicated because 25 one of the solution batches he tested does have to do with 63 1 this case, the QAP solution. What he said is he doesn't 2 know and he read the numbers that he did. 3 HEARINGS OFFICER KOEHLER: Okay. I think the 4 question is, did he look at the batch numbers for 07010. 5 Dr. Emery, did you look at the batch numbers for 6 batch 07010? 7 THE WITNESS: No, I did not. 8 HEARINGS OFFICER KOEHLER: Okay. 9 Q (By Mr. Garcia) And you indicated previously that 10 you had run your weighted average upon several of the 11 solutions that counsel provided to you. And in each of 12 those circumstances you ended up with differences between 13 the arithmetic mean and the weighted mean in the fourth 14 place -- 15 A Yes. 16 Q -- the fourth digit? 17 A That is correct. 18 Q You also testified about outliers. One of the 19 things you said was that this was a difficult area of lab 20 management? 21 A It's one that almost all students have difficulty 22 with, all people involved in making measurements. It's a 23 judgment call. And it is probably the most contentious 24 issue when making measurements, yes. 25 Q Would you agree it is wise to give trained 64 1 experienced personnel discretion in making these 2 determinations? 3 A Well, let me say how I would approach it in class. 4 I would say there are criteria that you may use to accept 5 or reject an outlier. You perform the calculation 6 according to those criteria and you make a decision based 7 upon that criteria. You cannot just say a number looks 8 bad, I'll throw it out. You have to justify doing it. 9 You have to state your criteria, you have to do 10 the calculations, and then draw the conclusion. 11 Q And, again, do you have any idea if any outlier 12 issue arises in this case, the Arntson case? 13 A I don't know what case we're talking about. I'm 14 very sorry. So I have no idea. 15 Q Okay. So you can't offer an opinion as to whether 16 that issue had any impact on any particular case? 17 A No, I cannot. 18 Q Do you have any idea how often that issue arises 19 generally at the state toxicology lab? 20 A I have no idea. 21 Q Now, you indicated earlier that your area of 22 specialty is mechanical engineering? 23 A That's correct. 24 Q And that's one of the reasons you're indicating 25 that you can't give an opinion upon whether or not the 65 1 variations that you've described between mathematical mean 2 and weighted mean have an impact on breath-test results is 3 because this is not your field of study? 4 A Well, number one, breath testing is not my field 5 of study. Number two, the differences appear to be judged 6 on the basis of legal terms, and I'm not a lawyer. 7 Q You'd agree it would be important to look at the 8 overall quality assurance program for any result before 9 attributing significance to any one figure, specifically 10 the mathematical mean versus the weighted mean that you 11 describe? 12 A No. If you're asking me, is the difference 13 important, I can't make a statement about that. But are 14 you asking me, should you have done the weighted mean 15 calculation, the answer is yes. 16 Q Are you aware if that statement is followed 17 anywhere within the area of forensic science? 18 A It's followed in all forms of science. 19 Q I'm asking specifically about forensic science. 20 Are you aware -- well, let me ask it a different way. 21 Have you ever had experience with a forensic 22 science lab? 23 A Other than the state toxicology lab? 24 Q Correct. 25 A No. 66 1 Q Have you ever performed an audit on any kind of 2 forensic lab? 3 A No. 4 Q Have you ever published any articles in any of the 5 national or international organizations with specialized 6 forms in either Forensic Toxicology or Forensic Laboratory 7 Science? 8 A No, I have not. 9 Q Do you regularly read any of the journals or 10 magazines related to -- specifically, Journal of Forensic 11 Science Society, Forensic Science International, The 12 Journal of Studies on Alcohol, The Journal of Analytical 13 Toxicology, Medicine Science & the Law, Journal for Quality 14 Comparability and Reliability in Chemical Measurement, 15 Journal of Forensic Sciences, Journal of Traffic Medicine, 16 Journal of the Alcohol Testing Alliance, Canadian Society 17 of Forensic Sciences Journal. 18 A I read some of them sporadically, only because of 19 my involvement here. 20 Q Are you a member of any of those organizations? 21 A No, I'm not. 22 Q Ever attended any of the forums? 23 A No. 24 Q Do you have any specialized knowledge in the 25 operation of the DAC DataMaster? 67 1 A Definitely not. 2 Q Dr. Emery, in terms of your calculation of the 3 weighted mean, did you use "N" in that calculation? 4 A Yes. I assume by "N," you mean the number of 5 readings? 6 Q Correct. 7 A Yes. 8 MR. GARCIA: I have no further questions. 9 Thank you, your Honor. 10 HEARINGS OFFICER KOEHLER: And you don't need any 11 time to consult with Mr. Gullberg regarding any additional 12 questions? 13 MR. GARCIA: Not at this time. 14 MR. VOSK: I've got re-direct, your Honor. 15 HEARINGS OFFICER KOEHLER: All right, Mr. Vosk. 16 REDIRECT EXAMINATION 17 BY MR. VOSK: 18 Q Dr. Emery, are the laws of physical -- of physics 19 applicable pretty much everywhere on the planet? 20 A Yes. 21 Q Do they change when we step into a forensic's lab? 22 A I would hope not. 23 Q So the principles that you were discussing 24 earlier, you indicated that your specialty -- one of your 25 areas of expertise was in metrology, you're a metrologist. 68 1 A Yes. 2 Q That relies on the physical principles related to 3 measurement, correct? 4 A Yes. 5 Q And those physical principles, those physical laws 6 are the same in every lab we walk into? 7 A Yes. 8 Q So you don't need to read a forensic's magazine to 9 know whether or not gravity works in a forensic's lab, do 10 you? 11 A No. No, I don't. 12 Q You don't need to read a forensic's magazine to 13 know that the laws of electromagnetism apply to a 14 forensic's lab, do you? 15 A No. 16 Q And do you need to enter a forensic's lab to 17 understand how the measurement process works? 18 A No. 19 Q Like other physical laws, the rules with respect 20 to the laws governing the science of metrology are uniform 21 around the planet. 22 A Yes. 23 Q Now, Counsel asked you whether or not you used "N" 24 in your calculation of the weighted means. When you 25 determined the standard deviation, you used N minus 1 in 69 1 your calculation, correct? 2 A That's correct. 3 Q Okay. So "N" was just a way of keeping track of 4 the total number of measurements we have? 5 A I wouldn't say that. It's an important ingredient 6 in calculating the mean of each set of readings and in the 7 standard deviation set of readings. 8 Q Okay. So we used "N" in determining the mean? 9 A Yes. 10 Q And -- of each set of readings, and N minus 1 to 11 determine the standard deviation? 12 A Yes. 13 Q And it was the standard deviation that we used in 14 determining our weighting factors, correct? 15 A That's correct. 16 Q Okay. Now, counsel also asked you about whether 17 or not -- determining whether or not something was an 18 outlier, it was a judgement call, was based on discretion. 19 The criteria you're talking about when you tell 20 your students they can use it -- if it's an individual 21 student working in your class, he could probably chose a 22 criteria he wanted to rely upon, couldn't he, amongst those 23 that are available? 24 A If he can justify, yes. 25 Q Okay. But now in your lab when you're trying to 70 1 measure a particular quantity, can each of the scientists 2 use their own different method of determining what an 3 outlier is? 4 A You mean in a classroom laboratory? 5 Q No, in your lab when you do real science. 6 A I think the students believe it's real science. 7 The two most common are Chauvenet's and Pierce's. 8 Chauvenet is probably the one that is used most often, but 9 Pierce's is a better one. 10 Q Okay. 11 A So I will typically tell them to use Chauvenet. 12 Q Okay. So when you're all working together to 13 measure one specific quantity -- for instance in this 14 case, we're talking about reading the value of a solution. 15 Everybody needs to use the same criteria to make sure your 16 results, each separate person's data are comparable? 17 A Yes. 18 Q Okay. So they can't just switch haphazardly back 19 and forth? 20 A No, they can't. 21 Q Now -- 22 MR. VOSK: No further answers, your Honor. 23 HEARINGS OFFICER KOEHLER: How about questions? 24 MR. VOSK: I'm sorry. No further questions. I'm 25 sorry, your Honor. 71 1 HEARINGS OFFICER KOEHLER: Anything further from 2 you? 3 MR. GARCIA: Just very briefly, your Honor. 4 REDIRECT EXAMINATION 5 BY MR. GARCIA: 6 Q Dr. Emery, could you tell me what equation did you 7 use for the weight? Was it N over -- N over variance? 8 A No. It was just one over the variance. 9 Q Why not N? 10 A N's already embodied in the variance. If you look 11 on the article by Paul and Mendel, the weights are listed 12 right there under equation one as one over the variance. 13 Q So the number of times that I used an instrument 14 in that circumstance wouldn't matter? 15 A I suppose what you're trying to lead me into or 16 you are leading me into, is the question of whether I'm 17 going to use the variance of the mean or variance of the 18 readings? 19 Q Well, I guess I'm more concerned about -- there 20 are different ways of making this calculation. The 21 differences between them might be subtle, you'd agree? 22 A Yes, there are differences. I'd say you have to 23 do the calculation and see the result. About the only 24 thing I could say is that you've got to have a criteria and 25 you've got to use it. 72 1 Q Okay. So you're just saying, pick a path and then 2 apply that path? 3 A Yes, uh-hum. 4 Q Okay. You indicated earlier -- or counsel asked 5 you a number of questions about the physical laws. And I'm 6 assuming -- when he talks about the law of gravity, et 7 cetera. 8 He then asked you a question saying, the laws 9 governing measurement are uniform around the planet, and 10 you said yes. It seemed like an overstatement. Do you 11 agree that that statement is not entirely correct? 12 A (No audible response.) 13 Q Well, we just talked about some variations -- 14 A Well, if you want to argue -- 15 Q -- in ways of doing that. 16 A If you want to argue Einstein's theory of 17 relativity versus Newtonian mechanics. 18 Q Well, we're talking about measurement. 19 A Uh-hum. 20 Q And counsel was trying to equate that to some form 21 of law. Wouldn't you agree that the science of measurement 22 is an evolving science, it has differences of opinion even 23 among the scientists? 24 A All scientists disagree to a certain point. 25 Q Certainly. 73 1 A Otherwise, they wouldn't be doing research. 2 Q And you have been in courtrooms where there have 3 been experts on the other side that disagreed with you on 4 issues of measurement? 5 A Yes. 6 Q So, I mean, a more fair way of saying it, is 7 there's a consensus, but it's not uniform? 8 A Yes. 9 Q All right. And it would be fair to say that you 10 don't know what the consensus is among scientists in the 11 alcohol-testing field? 12 A That's true. 13 Q All right. 14 MR. GARCIA: Thank you. No further questions. 15 MR. VOSK: I have just a couple follow-ups, your 16 Honor. 17 HEARINGS OFFICER KOEHLER: All right. 18 REDIRECT EXAMINATION 19 BY MR. VOSK: 20 Q With respect to the laws of measurement, the 21 physical principles involved in measurement, the base -- 22 the way chemicals combine, the way electromagnetic forces 23 work, the physical principles involved in measurement, 24 those are the same in every lab anywhere on this planet, 25 aren't they, the physical laws of the universe don't change 74 1 discounting relativity versus Newtonian mechanics? 2 A I would say yes. 3 Q Okay. And when counsel asked you, it's okay, as 4 long as you pick a path and follow it, that's fine. You 5 were referring to alternate methods of weighting data in 6 determining a mean; is that correct? 7 A Yes. 8 Q You weren't saying that it was okay to go ahead 9 and just use the arithmetic mean without choosing some 10 weighting criteria? 11 A No, that's -- you've got to have a criteria. 12 Q Okay. So differences between methods of weighting 13 is okay, but it's not okay to just jump in and use the 14 arithmetic mean? 15 A That's true. 16 Q We've been talking about differences in the 17 ten-thousandth place of these numbers. And these are 18 things that we can detect in the lab. You mentioned the 19 difference between Newtonian mechanics and general 20 relativity. 21 Are the differences that would be present in a 22 typical lab between those two theories going to be apparent 23 -- 24 A No. 25 Q -- at that level? 75 1 A No. 2 Q You'd need far more precise instruments for that? 3 A Yes. 4 Q So we don't need to worry about the distinction 5 between classical gravity and relativistic? 6 A No. 7 MR. VOSK: No questions, your Honor. 8 HEARINGS OFFICER KOEHLER: Mr. Garcia? 9 MR. GARCIA: No. 10 HEARINGS OFFICER KOEHLER: In terms of 11 measurement, Dr. Emery, is there a difference between the 12 terms reliability and accuracy? 13 THE WITNESS: Between precision and accuracy? 14 HEARINGS OFFICER KOEHLER: No. Between 15 reliability and accuracy. 16 THE WITNESS: Yeah, I thinks there's a fundamental 17 difference. But in terms of measurement -- are you saying 18 reliability in the sense that you can believe a 19 measurement? 20 HEARINGS OFFICER KOEHLER: Well, I guess I'm 21 asking you that question. How would you define that term? 22 THE WITNESS: Well, accuracy is the degree to 23 which the average of a great number of readings agrees with 24 the true value. 25 I don't know that there is a statistical 76 1 definition in terms of measurements as to reliability. I 2 would assume that what you're thinking of is how confident 3 am I in the results. 4 HEARINGS OFFICER BARTON: Since we're asking you 5 to define terms, can you tell me what you mean by 6 precision? 7 A Okay. It's easier to give you an example. If you 8 were -- the one that we usually give students is we're 9 shooting a bow and arrow at a target. Okay. After you got 10 through all shooting, if you -- you have all of these 11 points around your -- if you looked at the average, that 12 distance of the average from the bull's-eye would be a 13 measure of its accuracy. It's precision would be the size 14 of a cloud of impact points that you have. 15 So something could be very precise, meaning that 16 everytime you shot the arrow, you came to the same point on 17 the target, but it could be terribly inaccurate. Okay. Or 18 it could be very accurate, which means that on the average, 19 you were hitting the bull's-eye every time; but each 20 individual reading was all over the place. So precision is 21 related to the spread of the data. Accuracy is related to 22 how far the average of the data agree with the true value. 23 So you can have inaccurate precise measurements, 24 you can have accurate and precise measurements. Inaccurate 25 and precise and accurate and precise. That sort of covers 77 1 everything. 2 HEARINGS OFFICER BARTON: Right. And can you tell 3 me what -- you talked about consensus value. 4 THE WITNESS: A weighted average. 5 HEARINGS OFFICER BARTON: Okay. 6 HEARINGS OFFICER KOEHLER: Mr. Vosk, do you have 7 any follow-up questions based on the Hearing Officers' 8 questions? 9 Mr. Vosk: Just real quick, your Honor, yes. 10 REDIRECT EXAMINATION 11 BY MR. VOSK: 12 Q The two qualities you were talking about, accuracy 13 and precision, is that -- are those typically the two 14 things, the primary two things you're worried about when 15 you're reporting data, a scientific measurement? 16 A Yes. 17 Q And if a statute or a regulation or a Court used 18 the term as accuracy and reliability, you don't know 19 whether or not The Court's referring to your scientific 20 terms or not, do you? 21 A No I don't. 22 Q And so if those terms were defined in a WAC, in a 23 regulation, or in some administrative protocol, and they 24 identified reliability with precision, then your definition 25 of precision would apply to reliability? 78 1 A Yes. 2 Q Okay. 3 HEARINGS OFFICER KOEHLER: Mr. Garcia, do you have 4 any other follow up questions? 5 Mr. Garcia: No, questions. 6 HEARINGS OFFICER KOEHLER: All right. Thank you 7 very much for your testimony today, Dr. Emery. 8 THE WITNESS: Thank you, very much. 9 HEARINGS OFFICER KOEHLER: And you're excused and 10 free to go. 11 Mr. Garcia, will you be calling any witnesses? 12 MR. GARCIA: Just briefly, your Honor. 13 HEARINGS OFFICER KOEHLER: Would you like to take 14 a break? 15 MR. GARCIA: Please. 16 (Whereupon, a recess was taken.) 17 HEARINGS OFFICER KOEHLER: Mr. Vosk, do I 18 understand correctly that you wanted to call Dr. Emery 19 briefly? 20 MR. VOSK: Just very briefly to correct some 21 testimony that he gave, your Honor. 22 HEARINGS OFFICER KOEHLER: All right. 23 MR. VOSK: But it will be very quick. 24 HEARINGS OFFICER KOEHLER: Just go ahead. 25 MR. VOSK: Can I hand this to The Court and get 79 1 this marked, please? 2 HEARINGS OFFICER KOEHLER: Has counsel seen this? 3 MR. VOSK: Oh, I'm sorry. 4 HEARINGS OFFICER KOEHLER: The only thing that 5 looked like I was missing was the admitted copy of Exhibit 6 66. 7 So I remember correctly, there was no writing put 8 on Exhibit 66. 9 THE WITNESS: Well, you numbered the pages 10 afterwards. 11 HEARINGS OFFICER KOEHLER: All right. It's fine. 12 All right. Was there some confusion about another exhibit? 13 MR. VOSK: No, no. 14 HEARINGS OFFICER KOEHLER: All right. Then I've 15 marked the document that you just gave me. That will be 16 Exhibit 73. 17 MR. VOSK: Okay. 18 FURTHER EXAMINATION 19 BY MR. VOSK: 20 Q (By Mr. Vosk) Dr. Emery, earlier you did some 21 calculations of the weighted mean with respect to solution 22 07007. 23 A That's correct. 24 Q Were your calculations wrong? 25 A Yes. 80 1 Q Okay. Could you explain to The Court what you 2 did? 3 A I used the variance of the -- let me back up a 4 little. 5 Q Please. 6 A The correct variance is the variance in the mean, 7 not the variance of the individual readings. And in the 8 tests 07004 and 06012, the same number of readings were 9 used for each test and so the number of readings didn't 10 matter, it divided out. 11 But in 07007, they have -- we have a test with 20 12 readings and a test with 30 readings and a test with 10 13 readings and another test with 20. So a number does make a 14 difference there. 15 Q Okay. And so in that case, what is your -- what 16 is your weighting factor? 17 A The weighting factor should have been multiplied 18 by the number of readings. 19 Q And over the standard deviation? 20 A The standard deviation of the mean -- the root and 21 over the number -- over the standard deviation -- the 22 standard deviation -- the variance of the mean is the 23 variance of the individual readings divided by the number 24 of readings. 25 Q Okay. And did you redo those calculations? 81 1 A Yes, I did. 2 Q And can you tell The Court what you got for your 3 weighted mean? 4 A My weighted mean is now .1021. 5 Q Okay. And what did you get for the arithmetic 6 mean? 7 A .1018. 8 Q And did you divide each of those by 1.23? 9 A Yes, I did. 10 Q What did you get for the weighted mean? 11 A For the weighted mean, .0830; and for the 12 arithmetic mean, .0828. 13 Q Okay. Now, originally you said that the 14 difference between the arithmetic and weighted mean was 15 only one ten-thousandth, correct? 16 A Yes, I did. 17 Q And you were wrong, right? 18 A That's correct. I was wrong. 19 Q What is the difference between the two equivalent 20 -- 21 A .0002. 22 Q So two ten-thousandths -- 23 A Yes. 24 Q -- greater than before. 25 MR. VOSK: Your Honor, at this point, I would ask 82 1 to admit Exhibit 73, the corrected calculations. 2 HEARINGS OFFICER KOEHLER: Mr. Garcia, do you have 3 the same objection? 4 MR. GARCIA: No, your Honor. I think it was my 5 cross that eliminated any problem with the calculation. 6 HEARINGS OFFICER KOEHLER: All right. Then 7 Exhibit 73 is admitted. 8 MR. VOSK: No further questions, your Honor. 9 HEARINGS OFFICER KOEHLER: Mr. Garcia, do you have 10 any questions based on the additional questions? 11 MR. GARCIA: I do not, your Honor. 12 HEARINGS OFFICER KOEHLER: All right. Thank you 13 very much. 14 MR. GARC