Cross-examination of Dr. Terence Speed (afternoon)

📄 Cross-examination of Dr. Terence Speed (afternoon) — Tuesday, August 8, 1995

Address:

C:\DEPT103\CRIMINAL\1995\AUG\8\CROSS-EXAMINATION-OF-DR-TERENC.DOC

TRIAL

▲ Day 131 of 167

◀ Jury preparation Redirect examination of Dr. Terence Speed ▶

Witness: Dr. Terence Speed
Examiner: Rockne Harmon
Called by: Defense • Date: Tuesday, August 8, 1995 • Utterances: 162

Prosecutor Rockne Harmon cross-examined defense DNA statistics expert Dr. Terence Speed, pressing him on the contradiction between his testimony that laboratory error rate 'plays a more important role' than population frequency estimates while simultaneously admitting he had produced no statistical evidence that any errors actually occurred in this case. Harmon repeatedly cornered Speed on the point that without quantified error statistics, the only statistical evidence before the jury remained Dr. Bruce Weir's population frequency estimates, which Speed had not directly criticized on direct examination.

# 1 THE COURT:

All right. Thank you, ladies and gentlemen. Please be seated. All right. Dr. Speed, would you resume the witness stabbed, please. All right. The record should reflect that we have been rejoined by all the members of our jury panel. Dr. Terence Speed is again on the witness stand undergoing cross-examination by Mr. Harmon. And Mr. Harmon, you may proceed.

# 2 MR. HARMON:

Thank you, your Honor. Your Honor, may we mark the two pages of Professor Weir's testimony that I had shown to Professor Speed?

# 3 THE COURT:

That would be 578.

# 4 MR. HARMON:

78, yes, your Honor.

# 5 THE COURT:

578.

# 6 (Peo's 578 for id = 2-page document)

# 7 MR. HARMON:

Professor Speed, I want to come back to where we left off and I have blocked out a couple of the pertinent passages that you referred to yesterday. You had a chance to read that before the break?

# 8 PROF. SPEED:

Yes.

# 9 MR. HARMON:

Now, isn't it true that Professor Weir, at the conclusion of the series of questions and answers, his final statement on the subject of laboratory error rate, with respect--with regard to the frequency estimates that he presented was: "I reject the notion that it," the laboratory error rate, "Should be incorporated"?

# 10 PROF. SPEED:

That's correct.

# 11 MR. HARMON:

Is that correct?

# 12 PROF. SPEED:

That is what it says.

# 13 MR. HARMON:

And you understand from having read Professor Weir's testimony, that the only statistics that he presented in his testimony were the population frequency estimates for the biological matches in this case?

# 14 PROF. SPEED:

That's correct.

# 15 MR. HARMON:

Is that true? And you don't take issue with Professor Weir's statement that he rejects the notion that laboratory error rate and the population frequencies should be incorporated, do you?

# 16 MR. NEUFELD:

Objection. That is not the testimony and there is no--

# 17 THE COURT:

Overruled. Overruled.

# 18 MR. NEUFELD:

Objection, assumes facts not in evidence.

# 19 THE COURT:

Overruled.

# 20 PROF. SPEED:

Could you restate the question, please?

# 21 MR. HARMON:

Sure. Professor Weir, in the end of that series of questions and answers stated, did he not: "I reject the notion that it should be incorporated"?

# 22 PROF. SPEED:

Yes.

# 23 MR. HARMON:

And from having read his testimony you know that he--that he was referring to the "It" was the laboratory error rate?

# 24 PROF. SPEED:

Yes.

# 25 MR. HARMON:

Okay. And the--what it should be incorporated or should not be incorporated with was the population frequency estimates, which is the only estimate that he presented? That is what you understood that answer to mean, did you not?

# 26 PROF. SPEED:

Yes.

# 27 MR. HARMON:

Yesterday one of the slides that we saw was a quote from the national research council on page 88. Do you recall?

# 28 PROF. SPEED:

Yes.

# 29 MR. HARMON:

And that quote was: "Coincidental identity and laboratory error are different phenomena so the two cannot and should not be combined in a single estimate; however, both should be considered." Do you recall that being up on the screen?

# 30 PROF. SPEED:

I do indeed.

# 31 MR. HARMON:

You endorse that statement yesterday from page 88 of the NRC report?

# 32 PROF. SPEED:

I still do.

# 33 MR. HARMON:

And you agree with Professor Weir's statement that it should be incorporated, the laboratory error rate should be incorporated with the population frequency estimate; isn't that true?

# 34 PROF. SPEED:

That is true.

# 35 MR. HARMON:

And the reality is Professor Weir didn't agree that you could quantify laboratory error rate. That is what he discussed in that segment, didn't he?

# 36 MR. NEUFELD:

Objection, assumes--that is not the testimony.

# 37 THE COURT:

Overruled.

# 38 PROF. SPEED:

May I read the sentence that we are talking about? He says: "You are asking me in essence to quantify something I don't think exists, so I reject the notion of an error rate."

# 39 MR. HARMON:

"I reject the notion that it should be incorporated"?

# 40 PROF. SPEED:

Well, I'm sorry, he rejects the notion it should be incorporated, but the "It" that we are talking about he has just previously said doesn't exist. That is what I thought--that is the quote that I was agreeing to yesterday. Incorporation of something that doesn't exist seems a moot point. He is stating very clearly here, as far as I read it, you are asking me in essence to quantify something I don't think exists. The "Something" in question there is error rates, so "I reject the notion of error rate." I don't see how that contradicts what I was saying yesterday. I didn't go on to comment on the quote that he rejects the notion that the "It" that he has just said doesn't exist should be incorporated, but I think he said very clearly that he doesn't think error rate exists. That is my reading of this document.

# 41 MR. HARMON:

And you read his entire testimony?

# 42 PROF. SPEED:

I do, yes.

# 43 MR. HARMON:

And he did not acknowledge that it is possible to make errors in DNA testing? Is that your testimony?

# 44 PROF. SPEED:

No, he--

# 45 MR. HARMON:

I'm sorry.

# 46 PROF. SPEED:

That is the curious thing. He acknowledges that errors are possible, but here, as we have just read, he rejects the notion of an error rate. He says it doesn't exist so it can't be quantified. To me that is a very paradoxical point of view.

# 47 MR. HARMON:

Well, you haven't made any attempts to quantify it yourself independent of his statement, have you?

# 48 PROF. SPEED:

I have spoken--the answer is no to that, if I may elaborate.

# 49 MR. HARMON:

Well, it is?

# 50 PROF. SPEED:

I have spoken to the--

# 51 MR. HARMON:

Objection. It is nonresponsive, your Honor.

# 52 THE COURT:

Ask another question.

# 53 MR. HARMON:

Now, yesterday at the very close of your direct examination Mr. Neufeld posed to you the dilemma when one has a coincidental match where the estimate is one in billions. Do you recall that at the very close of your testimony yesterday?

# 54 PROF. SPEED:

Well, I think--I know roughly what you are talking about, but I don't think you have stated it very precisely.

# 55 MR. HARMON:

Okay. Well, I will do my best. And that furthermore, that there could be a match by mistake. Do you recall that phrase that you used, "A match by mistake"?

# 56 PROF. SPEED:

Yes.

# 57 MR. HARMON:

And the numbers that you refer to were 1 in 50 to 1 in 200. Do you recall that series of questions?

# 58 PROF. SPEED:

I do. I do.

# 59 MR. HARMON:

And that is the very point that the NRC addresses on page 88, the quote that I just read to you about not combining laboratory error rate and chance of a match. Do you recall that?

# 60 PROF. SPEED:

Well, I recall you reading it, yes.

# 61 MR. HARMON:

And that is what was up on the screen yesterday?

# 62 PROF. SPEED:

That is what was up on the screen.

# 63 MR. HARMON:

How many biological samples were tested in this case?

# 64 PROF. SPEED:

I really--I really don't know the answer to that, but I would say in the range 50 to a hundred, but I'm just guessing, so perhaps I shouldn't even bother.

# 65 MR. NEUFELD:

Objection. I move to strike since he said he was guessing.

# 66 THE COURT:

Overruled. The jury can put that into appropriate context.

# 67 MR. NEUFELD:

Okay.

# 68 MR. HARMON:

Let me give you a definition so that we can go a little bit further with this question. Let's define a sample as any biological sample which undergoes the DNA testing process, okay?

# 69 PROF. SPEED:

Yes.

# 70 MR. HARMON:

Let's assume just for a hypothetical, since I don't want you to guess about how many, that there have been two run samples analyzed in this case, according to the definition I just gave you. Okay?

# 71 PROF. SPEED:

Yes.

# 72 MR. HARMON:

If those numbers that you threw out yesterday, 1 in 50 to 1 in 200 were real laboratory case work, laboratory error rate--are you with me so far?

# 73 PROF. SPEED:

I am.

# 74 MR. HARMON:

--what that means is that the chances--or that if those estimates are correct that 196 to 199 of the results that were--that might have been presented, if 200 samples were tested, were in fact the correct answers, doesn't it? Isn't that what it means?

# 75 PROF. SPEED:

That is one way of interpreting it, yes. I'm not sure--

# 76 MR. HARMON:

That is a correct way to interpret it?

# 77 PROF. SPEED:

Well, error rate is a rather vague term. To address the issue you are talking about one has to split it up into false positives and false negatives and I think in terms of the sample you are referring to, not just the number of samples, but during the testing process the number of opportunities for making a false positive, the number of opportunities for making a false negative, and then assess the expected number of each of those kinds. That would be assuming you had an estimate of the rate. I don't think it is enough just to count samples.

# 78 MR. HARMON:

Okay. Well, those are your numbers 1 in 50 to 1 in 200?

# 79 PROF. SPEED:

I mention those as the sort of numbers one sees coming from blind external proficiency tests in related areas, for example, the bone marrow program where they do conduct these sort of tests and they get results of that order.

# 80 MR. HARMON:

Okay. So one correct way to interpret the numbers that you threw out yesterday, 1 in 50 to 1 in 200, if 200 samples were tested, as I have defined samples, would be that 196 to 199 of them were the correct answers; isn't that true?

# 81 PROF. SPEED:

Well, in my last answer I think I just said I think it is actually important in this context to split--

# 82 MR. HARMON:

Objection, your Honor. That is nonresponsive.

# 83 PROF. SPEED:

The answer is it is not true then.

# 84 MR. HARMON:

It is not true?

# 85 PROF. SPEED:

No.

# 86 MR. HARMON:

Yesterday you said that at the end of that series of questions that of the two, the coincidental match of one in billions and the match by mistake of 1 in 50 to 1 in 200 you said matched, the match by mistake plays a more important role. Do you recall giving that answer at the very end of your direct examination?

# 87 PROF. SPEED:

I do.

# 88 MR. HARMON:

And when you said that that plays a more important role, weren't you violating the very caution that the NRC prescribed in 1992 when they said: "Coincidental identity and laboratory error are different phenomena so the two cannot and should not be combined in a single estimate; however, both should be considered"? Weren't you violating that?

# 89 PROF. SPEED:

I don't see how. I was making a statement about the relative magnitude of the components. I wasn't combining them and I was considering both.

# 90 MR. HARMON:

Professor Speed, you have not criticized any of the frequency estimates which have been presented to this jury in this case, have you?

# 91 MR. NEUFELD:

Objection, argumentative and beyond the scope.

# 92 THE COURT:

Sustained. Beyond the scope.

# 93 MR. HARMON:

Professor Speed, have you criticized any of the frequency estimates which have been presented in this case?

# 94 MR. NEUFELD:

Objection, beyond the scope.

# 95 THE COURT:

Sustained.

# 96 MR. HARMON:

Professor Speed, you do not mean by any of your discussions about laboratory error and the match by mistake plays a more important role, that if the jury decides that there was not a mistake made in any of the testing in this case that the population frequencies are incorrect?

# 97 PROF. SPEED:

I haven't previously said that, but if you asked me do I believe the population frequencies, the answer is no.

KEY QUOTE

# 98 MR. HARMON:

You have not criticized in your direct examination any of the population frequency estimates in this case, have you?

# 99 MR. NEUFELD:

Objection, beyond the scope of the direct examination.

# 100 THE COURT:

Sustained. Sustained.

# 101 MR. HARMON:

You have simply alerted the jury to consider the possibility of certain errors which may have occurred in this case; isn't that true?

# 102 MR. NEUFELD:

Objection, argumentative, the form of the question.

# 103 THE COURT:

Overruled.

# 104 PROF. SPEED:

I think the answer simply is inappropriate there. I spoke on one aspect. I have not approved or disapproved. At least I did not approve or disapprove of the frequencies in my direct examination. I didn't address them, so--

# 105 MR. HARMON:

Some of the possible errors that might have occurred in this case you listed as the chance of a sample mix-up?

# 106 PROF. SPEED:

Yes.

# 107 MR. HARMON:

Do you remember that? You just generalize with errors in DNA testing. Do you recall that?

# 108 PROF. SPEED:

Yes.

# 109 MR. HARMON:

And then you describe generally errors in all stages of handling. Do you recall that?

# 110 PROF. SPEED:

I do.

# 111 MR. HARMON:

Now, you have relied on the evaluation and testimony of Dr. Gerdes in recognizing the possibility of sample handling cross-contamination in your case, haven't you?

# 112 PROF. SPEED:

Not solely, no. I also relied on that flow chart that we had up there which I believe is a Prosecution flow chart which shows the possibilities of errors before the samples are split and sent to the separate labs. And I am just relying on my general experience that nobody can say with certainty no errors occurred in that initial phase.

# 113 MR. HARMON:

You have not made any attempt to quantify the likelihood that any of those errors actually occurred in this case, have you?

# 114 PROF. SPEED:

That is not my role. That is the role of external blind proficiency tests. I'm simply pointing out their possibility. And in fact I was saying there should be tests to quantify these error rates.

# 115 MR. HARMON:

Professor Speed, I want to ask you, you yourself, are you saying you are not capable of quantifying the likelihood of errors having occurred in a case such as this, a complex case such as this?

# 116 MR. NEUFELD:

Objection, asked and answered and the word "Complex."

# 117 THE COURT:

Overruled.

# 118 PROF. SPEED:

I am not able to on the basis of my knowledge at the moment in this specific context. I have said how I think they should. It is not possible to look at a single case and come up with error rates. They require--sorry to keep saying this--external blind proficiency tests.

# 119 MR. HARMON:

And the weak part is your knowledge at the moment in deciding how to approach this problem, isn't it?

# 120 PROF. SPEED:

Well, "Weak" in what sense?

# 121 MR. HARMON:

It is deficient?

# 122 PROF. SPEED:

I'm not attempting to produce figures.

# 123 MR. HARMON:

You have not attempted to produce figures?

# 124 PROF. SPEED:

I have not attempted to produce figures.

# 125 MR. HARMON:

In fact, nothing you have said constitutes statistical evidence that any of those possible errors actually occurred; isn't that true?

KEY QUOTE

# 126 PROF. SPEED:

That is true.

# 127 MR. HARMON:

If the jury decides that none of these possible errors actually occurred, then the only statistical information which they have to evaluate the match evidence in this case is the frequency estimates produced by Dr. Weir; isn't that true?

# 128 MR. NEUFELD:

Objection.

# 129 THE COURT:

Sustained.

# 130 MR. HARMON:

Professor Speed, would you be willing to allow the jury to decide whether any of these possible errors actually occurred?

# 131 MR. NEUFELD:

Objection. Objection.

# 132 THE COURT:

Sustained. Sustained.

# 133 MR. HARMON:

Professor Speed, since you have not provided any statistical information about the possibility that any of these errors occurred, what information is left to this jury to decide the significance of the biological matches that have been presented in this case?

# 134 MR. NEUFELD:

Objection. May we approach, your Honor?

# 135 THE COURT:

Overruled.

# 136 PROF. SPEED:

What I have spoken to is the way in which errors can be assessed with correct testing.

# 137 MR. HARMON:

Objection, your Honor. That is nonresponsive.

# 138 THE COURT:

It is. Answer the question, professor.

# 139 PROF. SPEED:

I'm sorry, could you repeat the question?

# 140 MR. NEUFELD:

I'm sorry. Were you sustaining the objection, your Honor, or overruling it?

# 141 THE COURT:

No, I'm directing him to answer the question.

# 142 MR. NEUFELD:

Thank you.

# 143 MR. HARMON:

# 144 PROF. SPEED:

I thought I was starting to answer that, but I was speaking generally to the possibilities. I mentioned some figures that I know of in my experience as a statistician and in reading literature on error rates in similar biological context. I believe that is information that I've provided the jury. I don't understand why--

# 145 MR. HARMON:

What about Dr. Weir's frequency estimates?

# 146 PROF. SPEED:

Excuse me. What about them?

# 147 MR. HARMON:

Well, is that left for the jury to decide since you have not presented any statistical evidence of cross-contamination or any of these errors?

# 148 MR. NEUFELD:

Objection, argumentative and beyond his expertise.

# 149 THE COURT:

Argumentative. Rephrase the question.

# 150 MR. HARMON:

Do you mean by the generalizations that you have presented to this jury that the jury should not be free to consider the population frequency estimates that were presented by Dr. Weir?

# 151 PROF. SPEED:

I'm not telling the jury what they should or should not consider. I was not speaking about population genetics frequencies and I have made no statement to date, except a very brief one earlier in this cross-examination, about those issues.

# 152 THE COURT:

All right. Let's wind this up.

# 153 MR. HARMON:

I'm trying to, your Honor.

# 154 MR. HARMON:

If the jury decides that none of those possibilities--

# 155 THE COURT:

Counsel, rephrase that. The jury gets to decide all issues of fact in this case. They understand that.

# 156 MR. HARMON:

Since the jury gets to decide whether any possible errors occurred--

# 157 THE COURT:

No. Let's start again.

# 158 MR. HARMON:

Since you have presented no statistical evidence that demonstrates the likelihood that any errors which might have occurred actually occurred, the only estimates, statistical estimate which they have to evaluate the significance of the match evidence in this case is Professor Weir's population frequency estimate; isn't that true?

# 159 PROF. SPEED:

I don't believe that is a correct description of my testimony. I believe, without actually giving numbers and saying these are error rates on the basis of certain sorts of tests, I have still given them statistical information.

# 160 MR. HARMON:

Given them estimates?

# 161 PROF. SPEED:

I have not given them numbers, but I believe my testimony has been informative.

KEY QUOTE

# 162 MR. HARMON:

Thank you. I have no further questions.

Temperature

tense

Key Quotes (4)

Dr. Terence Speed

He acknowledges that errors are possible, but here, as we have just read, he rejects the notion of an error rate. He says it doesn't exist so it can't be quantified. To me that is a very paradoxical point of view.

Speed identifies an internal contradiction in Weir's position — acknowledging errors exist while rejecting the concept of an error rate — but this cuts both ways since Speed himself cannot quantify error rates.

Dr. Terence Speed

In fact, nothing you have said constitutes statistical evidence that any of those possible errors actually occurred; isn't that true? That is true.

The key concession of the cross — Speed admits his entire testimony about error possibilities carries no statistical weight as evidence that errors actually occurred in this case.

Dr. Terence Speed

I have not given them numbers, but I believe my testimony has been informative.

Speed's final position — a defensive retreat that Harmon leaves hanging as his last point, underscoring the gap between raising possibilities and providing proof.

Dr. Terence Speed

I haven't previously said that, but if you asked me do I believe the population frequencies, the answer is no.

An unexpected admission volunteered by Speed that he does not believe Weir's frequency estimates — despite never having criticized them in direct examination, drawing a sustained objection.

Evidence (3)

People's 578

Two pages of Professor Bruce Weir's prior testimony, highlighted passages regarding laboratory error rate

Marked for identification, used to impeach Speed's characterization of Weir's position

Informal

NRC (National Research Council) 1992 report, page 88, on not combining laboratory error rate with population frequency estimates

Discussed — Harmon used the NRC quote Speed endorsed to argue Speed's testimony violated the NRC's own caution

Informal

Prosecution flow chart showing sample handling chain of custody stages

Referenced by Speed as basis for his error possibility opinions, independent of Dr. Gerdes

Notable Exchanges (3)

Rockne Harmon•Dr. Terence Speed

Harmon repeatedly tried to get Speed to confirm that without statistical error evidence, Weir's population frequencies were the only estimates left for the jury. Speed resisted, insisting his testimony was 'informative' without providing numbers. Judge Ito intervened to tell Harmon to rephrase multiple times.

strategic

Rockne Harmon•Dr. Terence Speed

Harmon pressed Speed on his 1-in-50 to 1-in-200 error rate figures from direct, pointing out they came from bone marrow proficiency tests rather than forensic DNA labs, and that Speed had made no attempt to apply them to this specific case.

revealing

Dr. Terence Speed•Rockne Harmon

Speed volunteered unprompted that he does not believe Weir's population frequency estimates, despite never having criticized them on direct examination — a disclosure that drew a sustained objection for being beyond scope.

strategic

Credibility Attacks (3)

⚔ Dr. Terence Speed

Internal inconsistency / no statistical foundation

Harmon established that Speed raised error possibilities (1-in-50 to 1-in-200) drawn from bone marrow programs rather than forensic DNA labs, made no attempt to quantify errors in this specific case, provided no statistical evidence any errors actually occurred, yet argued error rates play a 'more important role' than Weir's frequency estimates — leaving his testimony as advocacy without numbers.

⚔ Dr. Terence Speed

Prior inconsistent position / scope limitation

Harmon used the NRC's own language (which Speed endorsed) — that error and frequency 'cannot and should not be combined' — to argue that Speed's statement that error 'plays a more important role' violated the NRC caution he claimed to endorse.

⚔ Professor Bruce Weir

Prior testimony / internal contradiction

Speed used Weir's own prior testimony to argue Weir held a 'paradoxical' position — acknowledging errors are possible while rejecting the existence of an error rate — though this backfired somewhat as Harmon used the same testimony to show Weir explicitly rejected incorporating error rates into frequency estimates.

Witness Demeanor

Speed is articulate and resistant, frequently attempting to elaborate past yes/no answers

Speed requests questions be repeated multiple times, possibly buying time to formulate careful responses

Speed becomes somewhat defensive by the end, retreating to 'I believe my testimony has been informative'

Objections

18 objections (6 sustained, 7 overruled)

Proceeding 7241 • 162 utterances • Defense witness

Criminal Trial

Department 103