Cross-examination of Dr. Robin Cotton (part 2)

📄 Cross-examination of Dr. Robin Cotton (part 2) — Thursday, November 14, 1996

Address:

C:\DEPT103\CIVIL\1996\NOV\14\CROSS-EXAMINATION-OF-DR-ROBIN-.DOC

TRIAL

▲ Day 15 of 57

◀ Sidebar: RFLP match meaning Request to approach bench ▶

Witness: Dr. Robin Cotton
Examiner: Robert Blasier
Called by: Plaintiff • Date: Thursday, November 14, 1996 • Utterances: 391

Defense attorney Robert Blasier cross-examines Cellmark's Dr. Robin Cotton on the fragility of her DNA statistical conclusions, pointing out that only 3 of 10 RFLP bands were computed as identical lengths yet all counted equally toward a 1.2-billion-to-1 match frequency. Blasier then methodically walks through PCR contamination risks, degradation mechanics, and Cellmark's documented error history — two false positives in roughly 100 proficiency-test samples — to argue that a lab error rate of 1-in-50 swamps the astronomical statistical frequencies Cotton testified to on direct.

# 1 THE BAILIFF:

Jury walking in.

# 2 (Jurors resume their respective seats.)

# 3 (The following proceedings were resumed in open court in the presence of the jury:)

# 4 THE COURT:

Okay.

# 5 MR. BLASIER:

Thank you, Your Honor.

# 6 Q:

(BY MR. BLASIER) Dr. Cotton, one of the items you testified to was item number 12, a Rockingham drop, correct?

# 7 A:

Yes.

# 8 Q:

And the frequency that you gave for that five probe match was 170 million to 1 in 1.2 billion, correct?

# 9 A:

Yes.

# 10 Q:

And as we said before, a five probe match means you're looking at 10 bands, correct?

# 11 A:

Yes.

# 12 Q:

And you determined that under the way you used the term "match," all of those matched, correct?

# 13 A:

That's right.

# 14 Q:

Now, you use a computer to estimate how long the fragments are of those bands, correct?

# 15 A:

Yes.

# 16 Q:

And that's from your Autorad, correct?

# 17 A:

Yes.

# 18 Q:

Of those ten bands, tell me how many of them your computer said were the same length as the evidence band, identical lengths?

# 19 A:

Okay. But let -- give me just a minute.

# 20 Q:

So what you want to know is based on our estimated band sizes. Were any of those identical in length between the DNA banding pattern in the foyer and the DNA banding pattern from Mr. Simpson?

Your computer comes up with an estimate that this band is 10,120 base pairs, right?

Isn't that what your computer comes up with?

Where is it?

# 21 A:

There's no size on here. That's what you just said. Okay.

# 22 Q:

So it doesn't measure the number of base pairs, does it?

# 23 A:

Yes, the -- your estimating the number of base pairs that makes up each one of those bands. I was just trying to clarify and make sure I was answering the question that you asked me.

# 24 Q:

Okay.

# 25 A:

Then you didn't say anything so I -- that helped me with that so --

# 26 Q:

Well --

# 27 A:

-- I didn't know what to do.

# 28 Q:

When you get something from your computer, does it say I estimate this band to be a specific number of base pairs?

# 29 A:

Right. Those are the numbers on these two sheets.

# 30 Q:

Okay. So it looks at the suspect bands, looks at the evidence bands and it estimates each of those bands, correct?

# 31 A:

Yes.

# 32 Q:

Tell me, of the ten bands that you looked at for item number 12, how many of those did your computer say had the same number of base pairs?

KEY QUOTE

# 33 A:

Three.

# 34 Q:

So seven of them, your computer said had a different number of base pairs, right?

# 35 A:

Yes. The sizes are different.

# 36 Q:

And if you have a different number of base pairs, it's a different person; isn't it?

# 37 A:

If you knew that then, the answer would be yes.

# 38 Q:

Okay. And your 1 and 1.2 billion doesn't take into account the fact that you are just estimating fragment length. You cannot say they are the same, can you? Can you, Doctor?

# 39 A:

I can't answer your question because those two things aren't exactly related. Your question doesn't make sense to me.

# 40 Q:

In terms of the formula that gets you to the 1.2 billion, does it make any difference whether your computer says two bands are very close together or two percent apart?

# 41 A:

No.

# 42 Q:

So you score a band, two different bands. If there's a difference in the evidence of the suspect of two percent, you still call it a match and it counts such -- just as much as if your computer says they're the same, right?

# 43 A:

Yes it does.

# 44 Q:

I'm going to switch to PCR for a second.

PCR is a much less discriminating system than the RFLP system we've been talking about, correct?

# 45 A:

In terms of looking at difference between people, what we have now, the answer is yes.

# 46 Q:

There's less information that you can glean from, to put into a formula, correct?

# 47 A:

Right, for the PCR, things that we're testing.

# 48 Q:

That's because there is not as much variations among people in the segment of DNA that you look at with PCR type tests, right?

# 49 A:

Yes.

# 50 Q:

Now, PCR is not exactly a type of test. That's the process by which you start with a little amount of DNA and make it into a big amount?

# 51 A:

Yes, exactly.

# 52 Q:

And that process is called amplification, correct?

# 53 A:

Yes, it is.

# 54 (Counsel displays chart.) (The instrument herein referred to as a chart entitled "PCR Amplification" was marked for identification as Defendants' Exhibit No. 988.)

# 55 MR. BLASIER:

This is number 788.

# 56 Q:

(BY MR. BLASIER) And what happens in this process is that you don't actually make copies of the whole DNA. You cut out a small section like we talked about, for instance, for the DQ Alpha system, each segment that you look at is about 254 base pairs, correct?

# 57 A:

Yes.

# 58 Q:

It's a very, very small segment?

# 59 A:

It's 254 base pairs or whatever it comes to be.

# 60 Q:

Right. And what you do is if you have a very small amount of DNA in your evidence, you have to use this process to evaluate it at all, correct?

# 61 A:

Yes.

# 62 Q:

The RFLP system is not sensitive enough to evaluate small amounts of DNA, correct?

# 63 A:

That's right.

# 64 Q:

And what you do is you take those fragments and you go through a process which we don't really have to understand but essentially it goes through cycles and it doubles the amount of DNA with each cycle, correct?

# 65 A:

That's right.

# 66 Q:

So if you started out, just hypothetically, with one fragment, 254 base pairs long and you went through and you put it in your little machine and it goes through 32 cycles, theoretically, if it doubled each time, you wind up with a whole heck of a lot of fragment?

# 67 A:

Yes.

# 68 Q:

If you had this many fragments, that's enough to do some analysis on, correct?

# 69 A:

That's right.

# 70 Q:

Whereas the smaller amount is not?

# 71 A:

That's right.

# 72 Q:

Now, it is also true, is it not -- this is 989.

# 73 THE CLERK:

I think that last exhibit is 988.

# 74 MR. BLASIER:

I'm sorry 988 that makes sense.

# 75 (The instrument herein referred to as a chart entitled "PCR Amplification with Three Copies" was marked for identification as Defendants' Exhibit No. 989.)

# 76 Q:

(BY MR. BLASIER) It is also true, if you have a tiny amount of DNA from more than one source in a sample, when you put it through these cycles conceivably every component also multiplies and doubles each time you do it, correct?

# 77 A:

Yes.

# 78 Q:

So if you have a couple of fragments that came from one source and let's say a contaminant, which can be a DNA from another source is -- if there's a small amount, when you amplify it the contaminant gets amplified?

# 79 A:

Assuming there's enough of it.

# 80 Q:

The process by which you do this amplification is what gives you the great sensitivity with PCR test that you don't have with RFLP tests, right?

# 81 A:

That's right.

# 82 Q:

By the same token, it's also the biggest weakness in the sense a contaminant gets multiplied also, correct?

# 83 A:

Yes.

# 84 Q:

So that you have to take many extra precautions while you're doing PCR work that may not be as necessary with RFLP work to protect genes against contamination, giving you the wrong results, correct?

# 85 A:

Yes.

# 86 Q:

Mr. Lambert asked you about degradation. I'm going to to put up --

MR. P. BAKER: This is 1034.

# 87 MR. BLASIER:

We're going to go through a series of slides here as an exam -- Can you focus that a little better?

# 88 (The instrument herein referred to as illustration of blood degradation slide was marked for identification as Defendants' Exhibit No. 1034.)

# 89 Q:

(BY MR. BLASIER) This is just a hypothetical situation, you have DNA from person number one. And just -- my example is you have four different pieces of DNA and the type is one. You got me -- you with me so far?

# 90 THE COURT:

I think he's defocusing it.

# 91 Q:

Doesn't have to be clear to understand what we're doing here?

# 92 A:

We have four pieces of DNA and they're all A1.

# 93 Q:

You with me so far?

# 94 A:

Yes.

# 95 Q:

Now, the process of degradation is the DNA actually breaking up into pieces, right?

# 96 A:

Yes.

# 97 Q:

And as a piece degrades, for instance, if you have DNA wet, a wet blood sample for instance, that's out, not sealed into plastic for a period of time, wet, that can cause degradation, can it not?

# 98 A:

It might.

# 99 Q:

And that would mean that you would perhaps, less -- DNA as it breaks into pieces, you'd have less DNA to analyze, correct?

# 100 A:

You don't physically have less DNA. The DNA isn't going away but the pieces that are there are getting smaller.

# 101 Q:

Right. The portion of it that you can analyze is getting smaller, right?

# 102 A:

Yes.

MR. P. BAKER: 1035 is displayed.

# 103 (The instrument herein referred to as illustration of blood degradation slide was marked for identification as Defendants' Exhibit No. 1035.)

# 104 MR. BLASIER:

Now 1036.

# 105 (The instrument herein referred to as illustration of blood degradation slide was marked for identification as Defendants' Exhibit No. 1036.)

# 106 Q:

(BY MR. BLASIER) Heat is another mechanism by which DNA can degrade, correct?

# 107 A:

Yes.

# 108 Q:

And so, if we add heat to our mixture here, you might wind up with less DNA you can analyze, it's still there in pieces but there's less you can analyze, correct?

# 109 A:

Yes.

# 110 Q:

If we add moisture to that --

MR. P. BAKER: 1037.

# 111 (The instrument herein referred to as illustration of blood degradation slide was marked for identification as Defendants' Exhibit No. 1037.)

# 112 Q:

(BY MR. BLASIER) 1037. Moisture's also factor that can cause DNA to break down even further?

# 113 A:

Yes.

# 114 Q:

If we add all those things together, we might take a sample of DNA, and we add all those things and you wind up --it's so broke up that you have too little to test?

# 115 A:

That's possible.

# 116 Q:

It's not going to change the type, but you don't get any type, right?

# 117 A:

Right.

MR. P. BAKER: 1038.

# 118 (The instrument herein referred to as illustration of blood degradation slide was marked for identification as Defendants' Exhibit No. 1038.)

# 119 Q:

(BY MR. BLASIER) 1038. We might end up at the end of the chain with broken up DNA that you can't tell anything about?

# 120 A:

Yes, that's possible.

# 121 Q:

And if you tested that, you wouldn't get a result?

# 122 A:

Right.

# 123 Q:

I want to talk about contamination. The term "contamination" can mean several things?

# 124 A:

Sure.

# 125 Q:

One of the things it can mean is a little DNA from another source getting into a sample, correct?

# 126 A:

Yes.

# 127 Q:

And let's say you start it, hypothetically, with a sample that had been subjected to plastic, heat and moisture and had degraded to the point where there wasn't enough there to really analyze. And you add blood from a second person, small amount of blood from the second person. You with me so far?

# 128 A:

Well sort of, but you said there wasn't enough to analyze, but you still got one there.

# 129 Q:

Okay. Well, let's -- you can -- you can -- there is a lower amount beyond which you can't analyze. It's still there, but you can't analyze it?

# 130 A:

Right. On your last thing, when you said there wasn't anything, there was no one there so --

# 131 Q:

Okay. Assume there's no one there.

MR. P. BAKER: 1041 displayed.

# 132 (The instrument herein referred to as illustration of blood degradation slide was marked for identification as Defendants' Exhibit No. 1041.)

# 133 Q:

(BY MR. BLASIER) Let's say a small amount of DNA from another person gets in that sample through whatever mechanism. Okay?

# 134 A:

Okay.

# 135 Q:

And one mechanism might be that if there was a small amount of blood on the hand of the person examining a cloth swatch, for instance, it might be transferred from the person's hand into the swatch, correct.

# 136 MR. LAMBERT:

Objection. Assumes facts not in evidence. Misstates the evidence, improper hypothetical.

# 137 THE COURT:

Overruled on the hypothetical.

# 138 Q:

(BY MR. BLASIER) That's one mechanism you can get a contaminant into a sample, correct?

# 139 A:

If you have -- you had blood on your -- let's say glove, which would be -- and that blood was wet and you physically touched the swatch, could you transfer some?

# 140 Q:

Yeah.

# 141 A:

Possibly so.

# 142 Q:

And again we're talking about PCR. We're talking about extremely small amounts that you can amplify, correct?

# 143 A:

Yes. You can amplify extremely small amounts and then you also have to worry about, does your test detect that, so . . .

# 144 Q:

Right.

# 145 A:

So there are limitations there.

# 146 Q:

So when you amplify this particular sample and we're assuming there's not enough here in the one to come up with a type (referring to exhibit 1041), I have four -- I have four times as much as one.

Let's say there's no one there when you amplify this, you're going to amplify all these twos each cycle, correct?

# 147 A:

Yes. There's no one there and you add those four twos of -- you'll amplify the twos since there's no one left over.

# 148 Q:

Okay. And at the end, if you test that because you're not -- Because you've made much, much more, we'll assume there's no --

# 149 MR. BLASIER:

What slide is this?

MR. P. BAKER: 1043.

# 150 (The instrument herein referred to as illustration of blood degradation slide was marked for identification as Defendants' Exhibit No. 1043.)

# 151 Q:

(BY MR. BLASIER) 1043.

That's going to look like when you test the blood, consistent with person number 2; isn't it?

# 152 A:

Yes.

# 153 Q:

Now, the amounts of DNA and the various samples you tested, 23 samples I think you said?

# 154 A:

Yes.

# 155 Q:

Most of those samples had an -- a very small amount of DNA; isn't that correct?

# 156 A:

No. Some of them had fairly substantial amounts of DNA and some of them didn't.

# 157 Q:

Okay. You were only able to do RFLP tests, I think, on two or three?

# 158 A:

We did RFLP tests, got results for RFLP tests on the sock. On the Bundy drop, number 52 --

# 159 Q:

52 and 12?

# 160 A:

And 12 from the foyer.

# 161 Q:

That's out of the 23, three of them had enough DNA for an RFLP test?

# 162 A:

Yes.

# 163 Q:

What's the minimum amount of DNA that you need to get an RFLP result?

# 164 A:

You can generally get an RFLP result with about 25 nanograms.

# 165 Q:

Okay. And if a drop of blood has, hypothetically, 1 to 2 thousand nanograms, 50 nanograms is a very small amount?

# 166 A:

Is 50 a small amount of 2,000?

# 167 Q:

Yeah.

# 168 A:

Yeah.

# 169 Q:

Okay. And sample number 52, the Bundy drop, you were able -- 52 -- able to test that to see approximately how much DNA was there before you did the RFLP test?

# 170 A:

We did some test that give you an indication of amounts of DNA, yes.

# 171 Q:

And for 52, it was a very small amount; wasn't it?

# 172 A:

Actually 52 is in such a small amount.

# 173 Q:

What was your estimate in terms of nanograms?

# 174 A:

The estimate, when you go back and you look at how much -- how much you get on the tests that's designed to do this, you come up with about 200, 250 nanograms in that -- in that sample.

# 175 Q:

Now, you also did, on the other Bundy drops, you did a similar analysis to try and find out how much DNA was in the other Bundy drops, correct?

# 176 A:

Yes, we did.

# 177 Q:

And do you -- that by means of a test called a slot blot?

# 178 A:

That's right.

# 179 Q:

We don't have to understand how that works. They just have a way of measuring quantities, correct?

# 180 A:

That's a way of --

# 181 Q:

Estimate?

# 182 A:

-- Estimating how much human DNA you have present.

# 183 Q:

And you did that for item 47, which was the first Bundy drop and you didn't find any human DNA, did you.

# 184 (Pause for witness to review documents.)

# 185 A:

You just asked me about 47.

# 186 Q:

Correct.

# 187 A:

That's right. We did not. On that slot blot test, we're not detecting any DNA.

# 188 Q:

And isn't it true that for all of the other Bundy drops, other than 52, and 47 which has none that you detected, the amounts were extremely small?

# 189 A:

For the other Bundy drops, the amounts were in the range of 50 and less.

# 190 Q:

And some of them were down. One of them was like 1.8 nanograms, approximately. These are rough estimates; aren't they?

# 191 A:

I actually have redone these estimates and I can't remember where I put the paper where I wrote those down right this second. But there were estimates in the range of 55. I don't remember if there's one, one specifically at 1.8. That wouldn't surprise me.

# 192 Q:

We're talking about small amounts on the Bundy drops with the exception of 52?

# 193 A:

With the exception of 52.

# 194 Q:

Which is a small amount of RFLP. It was enough to get a result.

# 195 A:

Yes.

# 196 Q:

And the kit that you use for taking DQ Alpha testing is warranted to amounts no smaller than 2 nanograms, correct?

# 197 MR. LAMBERT:

Objection. Irrelevant. It's all been admitted.

# 198 THE COURT:

Sustained.

# 199 MR. BLASIER:

What grounds?

# 200 THE COURT:

What's the relevance of the warranty?

# 201 MR. BLASIER:

If their using amounts less than what the test is warranted for, that's relevant.

# 202 MR. LAMBERT:

It's been admitted.

# 203 MR. BLASIER:

What's been admitted?

# 204 THE COURT:

Is there a request for admissions?

# 205 MR. BLASIER:

This goes to it's weight.

# 206 THE COURT:

Excuse me?

# 207 MR. BLASIER:

This goes to it's weight.

# 208 THE COURT:

You said you were going to develop some numbers. Fine, develop the numbers. I don't think it would go to the warranty of the testing.

# 209 Q:

(BY MR. BLASIER) Did you follow the instructions in the manual that comes with the kit with respect to the minimum quantities that it's warranted for?

# 210 A:

No, we don't.

# 211 MR. LAMBERT:

Same objection, Your Honor.

# 212 Q:

(BY MR. BLASIER) Now I'm going ask you a couple questions about accreditation. There is -- there's an organization in the country that evaluates forensic labs, crime labs to accredit them?

# 213 A:

Yes.

# 214 Q:

What's the name of that organization?

# 215 A:

It's the American Society of Crime Laboratories, Crime Laboratory Directors Laboratory Accreditation Board.

# 216 Q:

And there's a fairly complicated process that you have to go through to show that your lab is up to speed in terms of following all the required protocols and doing good work before you can be accredited, correct?

# 217 A:

Yes.

# 218 Q:

And your lab is accredited; is it not?

# 219 A:

Yes.

# 220 Q:

In fact, it's one of the very few. There's only five in the country that's accredited?

# 221 A:

There are many labs in the country, but we are the only private lab in the country that's currently accredited.

# 222 Q:

Police Department SID lab is not?

# 223 MR. LAMBERT:

Objection. Irrelevant.

# 224 THE COURT:

Sustained.

# 225 Q:

(BY MR. BLASIER) Now, you indicated when you were talking about your background that you have some connection to the Association of Blood Banks I think you said.

# 226 A:

American Association of Blood Banks.

# 227 Q:

And so you're familiar with the way blood banks do their jobs, correct?

# 228 A:

Actually, I'm not at all familiar with the way blood banks do their jobs.

# 229 Q:

Are you familiar with the proficiency testing that is used in that industry?

# 230 A:

Only as it applies to paternity testing.

# 231 Q:

Okay.

Now, proficiency testing is a way to kind of monitor yourself to see how well your people are doing, correct?

# 232 A:

Yes.

# 233 Q:

And the form of proficiency testing that is the most desirable in terms of giving you the most helpful information is called external blind proficiency testing, correct?

# 234 A:

That's one form of testing, but I don't agree that it's necessarily the most desirable.

# 235 Q:

Well, that's external, means somebody other than you makes up the test, right?

# 236 A:

Yes.

# 237 Q:

And blind means you don't know you're being tested, right?

# 238 A:

Yes.

# 239 Q:

And there are lots of studies that show that if you know you're being tested, you may perform differently than when you don't know you're being tested.

# 240 MR. LAMBERT:

Objection. Lack of foundation, irrelevant, outside the scope.

# 241 THE COURT:

Overruled. Direct examination inquired about proficiency tests.

# 242 Q:

(BY MR. BLASIER) Would you agree with that?

# 243 A:

I don't know about the proficiency tests that are done in DNA typing labs. I don't know anything about studies that may be done about what kind of proficiency tests show what --

# 244 Q:

Isn't that -- your in charge of all the people that work in your lab, aren't you?

# 245 A:

Yes. Of course, I am.

# 246 Q:

Isn't that something that you would want to be familiar with?

# 247 A:

The area of proficiency test in general, and whether they are blind or not blind and how that mentally affects the person that is taking them, is outside my area of expertise.

# 248 Q:

So you don't care about that?

# 249 A:

I didn't say that.

# 250 MR. LAMBERT:

Objection. Argumentative.

# 251 MR. BLASIER:

You don't monitor any studies about that?

# 252 MR. LAMBERT:

Objection. Argumentative.

# 253 THE COURT:

Sustained.

# 254 Q:

(BY MR. BLASIER) You're not subjected to external blind proficiency testing, are you?

# 255 A:

We are not. I don't know any lab that is.

# 256 Q:

External blind proficiency testing is used all the time within the medical field, with blood banks, that sort of thing; isn't it?

# 257 MR. LAMBERT:

Objection. Beyond the scope, irrelevant.

# 258 THE COURT:

That's irrelevant.

# 259 Q:

(BY MR. BLASIER) Now Mr. Lambert asked you about a couple of proficiency tests that you -- that you took. One was in 1988 and that was put on by the California Association of Crime Lab Directors, correct?

# 260 A:

Yes.

# 261 Q:

And in that test you were given, your people knew they were being tested, correct?

# 262 A:

Yes, we did.

# 263 Q:

And you were given 44 different samples, correct?

# 264 A:

49.

# 265 Q:

49? And you had one error, one false positive out of 49, correct?

# 266 A:

Yes.

# 267 Q:

And that error was a, what's called, a "false positive," correct?

# 268 A:

It was an incorrect match.

# 269 Q:

In other words, if that had been a forensic sample, you would have said the sample matched the suspect and you would have been wrong, correct?

# 270 A:

That's right.

# 271 Q:

And what was the frequency when you applied your formula to that one test that you got wrong? What was the frequency that your match revealed, using your statistics?

# 272 A:

I have no idea.

# 273 Q:

It was 1 and 1.8 billion people; wasn't it, Dr. Cotton?

KEY QUOTE

# 274 A:

Mr. Blasier, I have not reviewed those numbers in many years and I do not know what the number is.

# 275 Q:

Would you like to review that study? It was a very rare number; wasn't it?

# 276 A:

Any RFLP match is generally a rare number. I'm sure there was a number and I'm sure it wasn't a common one.

# 277 Q:

And it was wrong, correct?

# 278 A:

Well, the number was okay but the match was wrong.

KEY QUOTE

# 279 Q:

It doesn't come from the two samples -- didn't come from the same person?

# 280 A:

It's numbers are saying how common that pattern was that we had. What was important was that the match was not correct.

# 281 Q:

So the number says nothing about the accuracy of the test that you're doing, does it?

# 282 A:

It doesn't tell you whether that match is correct --

# 283 Q:

Thank you.

# 284 A:

-- No.

# 285 Q:

Now, you made some changes in your procedure as a result of that error, did you not?

# 286 A:

Yes, we did.

# 287 Q:

And what procedure did you change?

# 288 A:

We changed our labeling procedure and we bought an additional piece of equipment that allowed us to handle large samples, physically large stains.

# 289 Q:

And you were able to determine, were you not, that what happened in that particular test is somebody mixed up a sample and you didn't know, correct?

# 290 A:

Somebody mixed up two twos which were portions of a sample, yes.

# 291 Q:

And now, as a result of that error, you don't want to have that error occur again. So you now have two people witness. Every time you put a sample from one tube to another, you have two people witness it?

# 292 A:

That was in place at the time. So that's why we changed our labeling procedures.

# 293 Q:

So there were two people looking at it?

# 294 A:

That's right.

# 295 Q:

Now, in 1989 you were tested again by the same organization, correct?

# 296 A:

Yes.

# 297 Q:

And that one had 50 samples, correct?

# 298 A:

That's right.

# 299 Q:

And you made one false positive in that test as well, correct?

# 300 A:

Yes, we did.

# 301 Q:

One out of 50, correct?

# 302 A:

Yes.

# 303 Q:

And you made changes in your procedures as a result of that, correct?

# 304 A:

Yes, we did.

# 305 Q:

What changes did you make?

# 306 A:

We changed the procedure so that the DNA extraction from the evidence samples was done at a separate time as any of the known standards so that they were not handled for DNA extraction at the same time.

# 307 Q:

And by known samples, you're talking about such things as reference samples from a suspect or from a victim?

# 308 A:

That's exactly what I mean.

# 309 Q:

That's because a reference sample from a suspect or victim is blood that might come from the person's arm or from an autopsy that's rich in DNA, correct?

# 310 A:

Well, it didn't. It didn't have anything to do with whether it was rich in DNA. It had to do with -- if you didn't have them out at the same time in the same location, side by side, you couldn't mix them up.

# 311 Q:

But the problem is because you've got so much DNA in a reference sample, very tiny amount of it, if it gets into your evidence, which may be a small amount to start with, is going to compromise your results; isn't it?

# 312 A:

Of course.

# 313 Q:

So you now don't do your reference samples at the same time and place as your evidence, do you?

# 314 A:

That's right.

# 315 Q:

LAPD does it the opposite way?

# 316 MR. LAMBERT:

Objection. Irrelevant, no foundation.

# 317 THE COURT:

Sustained.

# 318 Q:

(BY MR. BLASIER) Now, in your first two years on those two tests, you got two errors out of roughly 100 or 1 in 50?

# 319 A:

Yes.

# 320 Q:

There's a concept in DNA forensic technology known as error rates, correct?

# 321 A:

I'm sure you could apply that concept to anything, not just DNA.

# 322 Q:

Sure. And error rates refers to how often does a lab make a mistake?

# 323 A:

Yes.

# 324 Q:

And one way to measure that is with these kind of proficiency tests to determine how often do you screw it up, right?

# 325 A:

Yes.

# 326 Q:

And in these two years, in the two tests that your lab took, you screwed it up once every 50 times, right?

# 327 A:

Absolutely not.

# 328 Q:

I'm sorry?

# 329 A:

You in those --

# 330 Q:

Those two tests?

# 331 A:

In those two tests, we have 2 errors. That comes down to a fraction of 1 in 50. That doesn't say anything about all the other work that was done in the lab during those two years.

# 332 Q:

Right. You don't know one way or another when you made a mistake.

And you can assign a number to that, it's called an error rate. Is -- there's a lot of controversy about how you do this; isn't there?

# 333 A:

Yes, there is.

# 334 Q:

But if you -- if you have error rates in the neighborhood of 1 in 50 or 1 in 1000; isn't that a much more significant figure than 1 in 530 billion?

# 335 MR. LAMBERT:

Objection. Argumentative.

# 336 THE COURT:

Sustained. You can argue that to the jury.

# 337 Q:

(BY MR. BLASIER) Doctor, You use, in your procedure, you use for PCR tests what's called a laminar flow hood, correct?

# 338 A:

Yes, we do.

# 339 Q:

And that is a device -- it's kind of like a work station that has a fan that creates a wall of air that circulates from the bottom to the top, correct?

# 340 A:

Yes.

# 341 Q:

And it allows your sample to be inside this wall of air and you kind of put your hands through the wall of air to work on it?

# 342 A:

Right.

# 343 Q:

And the purpose of that is to keep possible contaminants from the outside coming in to your sample or contaminants from your sample going to the outside, correct?

# 344 A:

Yes.

# 345 Q:

And that's considered to be required, correct?

# 346 A:

No.

# 347 Q:

That -- it's good practice; isn't it?

# 348 A:

It's a very good practice but there are many labs who do not have laminar flow hoods who do perfectly acceptable work. You just have to have another precaution to be clean.

# 349 Q:

It's a procedure that you feel is necessary in your lab, correct?

# 350 A:

We had the hoods and we use them. If we didn't have them, we would have to do something else.

# 351 Q:

Now, you also require, when you do testing, that you change the paper on the bench from one sample to the next, correct?

# 352 A:

Paper inside the laminar flow hood.

# 353 Q:

No. I'm talking about when you were working with evidence samples on a work bench.

# 354 A:

When we're working with evidence samples on a work bench, you might -- you would put a clean piece down and then when you're done with that sample, you would put another piece down.

# 355 Q:

That's because you know, particularly with dried blood stains, for instance, there can be flaking. You can get very tiny flecks of something from an evidence sample. You might not see it and you want to protect against that contaminating the next piece of evidence.

# 356 MR. LAMBERT:

Objection. All irrelevant in light of the admissions.

# 357 THE COURT:

Sustained.

# 358 MR. BLASIER:

It goes to the weight, Your Honor. This is a procedure that they described.

# 359 THE COURT:

Well, I'm having a little problem with respect to your request for admissions and this course of examination. Didn't seem to have any relevance at all.

# 360 MR. BLASIER:

The procedure that the lab used doesn't have any relevance?

# 361 THE COURT:

You stipulated to the results.

# 362 MR. BLASIER:

I'm talking about the procedure they used by which they got the results.

# 363 MR. BAKER:

We stipulated to the numbers, not how they got the numbers.

# 364 THE COURT:

Well, your argument was that the meaning of the results are debatable in terms of when windows of errors, et cetera; not as to the results.

# 365 MR. BLASIER:

This is a completely different area, Your Honor.

# 366 THE COURT:

No. I sustained the objection.

# 367 Q:

(BY MR. BLASIER) Do you wipe down the area that you work with samples with bleach between each sample?

# 368 MR. LAMBERT:

Same objection.

# 369 THE COURT:

Sustained.

# 370 Q:

(BY MR. BLASIER) Now, you testified on direct that you processed some of the control or substrate controls sent to you by LAPD, correct?

# 371 A:

Yes.

# 372 Q:

Those substrate controls, by the way, you didn't process a substrate control for every -- for every sample, did you?

# 373 A:

No. We don't have all the substrate controls.

# 374 Q:

In fact, when LAPD sent them to you, they had been separated from the evidence samples and they didn't even send you the substrate controls. You asked for them, correct?

# 375 A:

Well, actually none of what you said is correct. We didn't ask for them.

# 376 Q:

They weren't sent at the same time, were they?

# 377 A:

They weren't sent at the same time and we simply analyzed the samples that were sent to us. We didn't request any particular samples.

# 378 Q:

Now, when you prepare bindles from samples, you know what a bindle is?

# 379 A:

The little wax paper thing.

# 380 Q:

Yeah. You always put your initials on it, don't you?

# 381 MR. LAMBERT:

Objection. Irrelevant, outside the scope.

# 382 THE COURT:

Overruled. You can answer yes or no.

# 383 DR. ROBIN COTTON:

Well, nobody in my lab is preparing any bindles. We're opening bindles, but we're not preparing them, since we're not ever getting evidence as it comes in directly from the crime scene.

So when we open things, when we close them up, we put our initials on the outer packaging, not on the immediate thing that the evidence is contained in.

# 384 Q:

Okay.

# 385 THE COURT:

Mr --

# 386 Q:

(BY MR. BLASIER) The unbindled drops you got from LAPD --

# 387 THE COURT:

Mr. Blasier.

# 388 MR. BLASIER:

I'm sorry.

# 389 THE COURT:

The scintillating examination is having an affect on our jurors.

KEY QUOTE

# 390 MR. BLASIER:

Well --

# 391 THE COURT:

I -- one of them. So I think we better take a recess. Okay. Ten minutes, please.

Temperature

tense

Key Quotes (5)

Robert Blasier

Tell me, of the ten bands that you looked at for item number 12, how many of those did your computer say had the same number of base pairs?

Sets up the reveal that only 3 of 10 bands matched by the computer's own estimate, undermining the precision implied by the 1.2-billion statistic.

Dr. Cotton

Three.

Admission that 7 of 10 bands were scored as different lengths by the computer, yet all counted toward the match frequency.

Robert Blasier

It was 1 and 1.8 billion people; wasn't it, Dr. Cotton?

Reveals that the 1988 Cellmark false positive — a wrongly declared match — itself carried a 1-in-1.8-billion frequency, directly attacking the meaning of large statistical frequencies.

Dr. Cotton

Well, the number was okay but the match was wrong.

Concedes the core point: a statistically rare frequency does not guarantee the underlying match is correct.

Hiroshi Fujisaki

The scintillating examination is having an affect on our jurors.

Judge's dry aside that the highly technical testimony was putting jurors to sleep, prompting a recess.

Evidence (10)

Defendants' Exhibit No. 988

Chart titled 'PCR Amplification'

introduced and discussed to explain PCR amplification process

Defendants' Exhibit No. 989

Chart titled 'PCR Amplification with Three Copies'

introduced to illustrate multi-source contamination scenario

Defendants' Exhibits No. 1034–1038

Series of blood degradation illustration slides

introduced sequentially to walk through heat, moisture, and combined degradation effects on DNA

Defendants' Exhibit No. 1041

Blood degradation slide depicting depleted original source with introduced contaminant

introduced to demonstrate how a contaminant can appear as the sole DNA source after amplification

Defendants' Exhibit No. 1043

Blood degradation slide showing amplified contaminant result

used to show final PCR result would type as the contaminant, not the original source

Informal

Item 12 (Rockingham foyer drop) — five-probe RFLP result, 1.2-billion frequency

challenged on band-size matching accuracy

+ 4 more

Notable Exchanges (4)

Robert Blasier•Dr. Cotton

Blasier establishes that of the 10 RFLP bands used to generate the 1.2-billion frequency for item 12, the computer estimated only 3 as having identical base-pair lengths. Cotton confirms 7 differed, then agrees a two-percent difference still scores as a match with equal statistical weight.

strategic

Robert Blasier•Dr. Cotton

Blasier reveals that Cellmark's 1988 false positive match carried a statistical frequency of 1-in-1.8-billion. Cotton cannot recall the number but confirms the match was wrong. Blasier then extracts the admission that the frequency number 'doesn't tell you whether that match is correct.'

devastating

Robert Blasier•Dr. Cotton

Blasier argues that Cellmark's two false positives across approximately 100 proficiency samples amounts to a 1-in-50 error rate, which he contends dwarfs the significance of billion-to-one statistical frequencies. Cotton pushes back that proficiency samples don't reflect all lab work; Judge Fujisaki sustains an argumentative objection before the argument is fully developed.

heated

Robert Blasier•Hiroshi Fujisaki•Tom Lambert

Extended bench colloquy over whether Blasier can question lab procedures (bench paper changes, bleach wipe-downs) given the defense stipulated to DNA results. Fujisaki sustains Lambert's objections, distinguishing the stipulated numbers from the underlying procedures.

procedural

Light Moments (1)

Hiroshi Fujisaki

Judge Fujisaki interrupted a question about bindles to announce 'The scintillating examination is having an affect on our jurors' — a dry reference to a juror nodding off — and called a ten-minute recess.

Credibility Attacks (3)

⚔ Dr. Cotton / Cellmark

prior lab error / proficiency test failures

Blasier surfaces two consecutive false positive matches in Cellmark's 1988 and 1989 proficiency tests (1 error per ~50 samples), then argues this empirical error rate is more meaningful than the astronomical statistical frequencies Cotton cites.

⚔ Dr. Cotton / Cellmark

internal inconsistency in match criteria

Blasier demonstrates that only 3 of 10 RFLP bands for item 12 were computed as identical by Cellmark's own software, yet all 10 were scored as matches contributing equally to the 1.2-billion frequency, suggesting the matching criterion obscures real differences.

⚔ Dr. Cotton / Cellmark

bias / knowledge gap on blind proficiency testing

Blasier presses Cotton on whether she is aware of studies showing performance differences under blind versus non-blind testing; Cotton denies familiarity, and Blasier implies this represents indifference to quality control.

Witness Demeanor

(Pause for witness to review documents.)

Witness uncertain and defensive when Blasier quotes the 1-in-1.8-billion figure from the 1988 false positive: 'Mr. Blasier, I have not reviewed those numbers in many years and I do not know what the number is.'

Objections

13 objections (10 sustained, 3 overruled)

Proceeding 8251 • 391 utterances • Plaintiff witness

Civil Trial

Department 103