Dr. Weir, a moment ago, you said that you weren't receiving any financial awards for testifying in this case; is that correct?
Okay. And would it be fair to say, doctor, that sometimes people say something for money and sometimes people say something because they have a certain political agenda or certain philosophical statement they want to get across. Is that a fair statement?
Okay. And I believe you said a moment ago, sir, when you were describing yourself and the issues that have arisen in forensic DNA typing, you said that you've been very angry at the way these issues have been described in courts over the last several years; is that correct?
And I believe you also said that you were very anxious to get yourself on to the witness stand so your position could be heard by a large audience; is that correct, sir?
And you feel that other people in other cases have been presenting an improper analysis, correct?
And you also felt that the NRC report was presenting improper scientific conclusions, correct?
And haven't you gone to gatherings of scientists or forensic employees and made public statements criticizing the NRC report?
Would you consider yourself as one person who has a certain philosophical or political agenda in this particular issue, sir?
Now, when you calculated these new numbers for the mixtures today on redirect examination from Mr. Clarke, did you take into consideration the extra step that you talked about last week, namely, the FST for substructure?
I haven't completed those numbers, but I have a very good sense of what the numbers will be because the same number of bands in the mixed profile occur in some of the single stain profiles or the same magnitude of the fix will be present.
But did you in fact include that calculation and that factor in the numbers that you put on this board for Mr. Clarke a little while ago?
And, sir, you also testified last week that one of the extra steps that you engage in to be more conservative is that you provide what's called a 99 percent upper confidence limit on a frequency estimate. Do you recall that?
Did you, when you did these calculations for Mr. Clarke and put on the board this afternoon, first take into consideration the 99 percent upper confidence limit for those numbers?
Well, you don't do it first. You do it afterwards. I haven't done it yet, but I know what the effects will be based on the comparisons I've done in similar sized profiles from the mix from the single stains.
Dr. Weir, didn't you tell this jury last week that both those steps, namely, calculating the FST for substructure and providing a 99 percent upper confidence limit will render the frequencies somewhat more common than the regular point estimate would? Isn't that true, sir?
I think I said it goes both ways. Certainly the upper confidence limit is in that direction.
Yes. And that the FST is also in that same direction as making it more common; isn't that correct?
If we use a positive value. We could use a negative value as suggested by the data and go in the other direction.
But if you use the positive value and you use the 99 percent upper confidence limit, you will generate a frequency more common than the numbers you just put on the board for Mr. Clarke in your redirect examination, wouldn't you?
And didn't you also say, sir, that one of the ways that you depart from the other laboratories is that you believe both of those things should be done; isn't that correct?
And, in fact, even now, when you recalculated these numbers after the weekend and you did these new numbers for Mr. Clarke, you again failed to include those two adjustments which would render the frequencies more common than they are on that board; isn't that correct, sir?
Well, sir, had you made the adjustment for substructure that you've characterized before before you put these numbers on the board, wouldn't that have generated frequencies more common than the ones that you put on the board?
The number I would presume is more common. I would--we could calculate it in both directions. We only presented one way, but of course we don't know which way the reality is. So that's true.
But when you said, "We only presented one way," you presented it in a way that if you had done that calculation, it would have made these frequencies even more common than the ones that you put out there; isn't that right?
And, sir, if you had been asked to put in the 99 percent upper confidence limit for these same mixture numbers that you put on the board for Mr. Clarke this afternoon, that too would make the frequencies more common than the ones you put on the board; isn't that correct, sir?
Dr. Weir, it appears from your redirect testimony that there were even more errors in your calculations than it first appeared when we discussed it on cross-examination on Friday; is that correct?
Oh, I see. I understand. No. In fact, there was only one error. I applied all those profiles using the same computer program. There was one mistake. It affected all the profiles that we discussed.
Okay. Now, under the new data that you just provided this afternoon to the ladies and gentlemen of the jury, for instance, on item 29, you expressed a range for two contributors from 1 in 26 all the way up to 1 in 11,306; is that correct?
And on this new data, sir, that you provided on redirect examination, this range reflects one difference and one difference only; namely, it depends on the race or ethnicity of the two contributors, correct, sir?
So the profiles don't change. What you're simply saying is, depending upon the race of those contributors, the frequency of that particular profile could be anywhere from 1 in 26 in the population to 1 in 11,306; isn't that right?
It's almost right. It's based on the four databases I used, the Caucasian and African American and two Hispanics. It could have been, for example, that I had four Hispanic databases and got similar range. I don't know. So there is a difference by self-reported racial designation and also a difference just by the sampling of different people.
Well, doctor, when you say on four different Hispanic databases, you're speculating because you haven't looked at four different Hispanic databases for these particular genetic types, have you?
I think if you remember what I said, I said I don't know. If I had four Hispanic databases, I would get four answers. I don't know what the range would be.
Well, doctor, let me ask you this. Given this range, simply depending on the race of the people who make up the mixture, Mr. Clarke asked you a question, when comparing certain numbers and certain frequencies, how would you characterize that number or characterize that difference. And now I'll ask you the same question, Dr. Weir. How would you characterize the difference between 1 in 26 versus 1 in 11,306?
Mr. Neufeld, Mr. Clarke did not ask me that question. He asked me to compare the results from Friday and the results from today.
Well, I'm going to ask you to compare 1 in 26 to 1 in 11,306, sir. Is that a small difference, a medium difference or a very large difference?
And, sir, again, you started using a gambling analogy under redirect examination. Let me ask you this question, sir. If you were told that the probability let's say at a horse race that you attended at the track for a certain horse coming across the finish lane was 1 in 26 and someone else told you it was 1 in 11,306, is that a meaningful spread that you could evaluate before you decide to bet on that horse?
I think in that circumstance, I wouldn't bet anymore than I would place a specific number on this profile. I think we've been very honest to convey to the Court the frequency estimate of that profile from the knowledge we have available to us. We don't have a precise value. We are laying it all out. The frequency of that profile in a mixture could be somewhere in this range. I think I said these are not enormously rare profiles at the PCR level. I haven't--we haven't hidden anything. That's--that's the range in this particular case.
I had to sleep between midnight and 5:00 A.M. this morning as you know.
I called you last night and told you that as soon as I found out.
I would say that example is a very large difference.
We don't have a precise value. We are laying it all out. The frequency of that profile in a mixture could be somewhere in this range. I think I said these are not enormously rare profiles at the PCR level.