LEONARD SAXE, Sworn
THE COURT: If you would tell us, please, your full name.
THE WITNESS: Leonard Saxe.
THE COURT: And spell that for the reporter, please.
THE WITNESS: S-A-X-E.
THE COURT: And your date of birth please.
THE WITNESS: June 12th, 1947.
THE COURT: And we have Dr. Saxe's direct testimony, which will be Exhibit 11 when it's appropriately marked, and you may cross examine, Mr. Good.
MR. GOOD: Thank you, Your Honor.
(Testimony, as above, received
and marked Exhibit No. 11.)
CROSS-EXAMINATION BY MR. GOOD
Q Good morning, Dr. Saxe. Dr. Saxe, is it correct to say that, with respect to your work concerning polygraph, that you've been an evaluator of the scientific evidence about the polygraph?
A Yes.
Q And in that role as scientific evaluator of the polygraph, you have relied upon your expertise in reviewing experimentation with the polygraph, is that true?
A In part, yes.
Q And by "experimentation," you're talking about experimentation using the scientific method.
A Yes.
Q And as an evaluator, you evaluate the scientific methodology, or lack thereof, that is in the experiment.
A In part, yes.
Q And it's correct to say that you have expertise in the scientific methodology of experimentation.
A Again in part, yes.
Q I don't mean that's all the things you're an expert on, but that's one of the things you're an expert about.
A Yes.
Q And you have utilized that expertise in examining reports, published reports, about the scientific experimentation with the polygraph.
A Yes.
Q Now, you are not yourself, sir, a psychophysiologist, that is not your specialty, is that right?
A No, I'm not, I'm not a psychophysiologist.
Q So that you have applied to -- well, is it correct to say that your expertise about research methodology is not specific to psychophysiology?
A It's not specific to psychophysiology.
Q That's correct.
A That's correct.
Q And you have applied, have you not, general principles of the methodology of experimentation to polygraph?
A I've applied the general principles of psychology, psychological tests; a part of that are criteria having to do with methodology.
Q And you're speaking again of the methodology of experimentation.
A Depending on how you define "experimentation," yes.
Q Experimentation, just so the Court understands what we mean by "experimentation," this is the use, in the polygraph context, of the polygraph instrument to measure the physiological responses of a subject who is hooked up to a machine. That could be an experiment. Would you follow that?
A That could be an experiment, yes.
Q And part of the experiment.
A Part of the experiment.
Q And then to manipulate the conditions under which the person is, while hooked up to the machine, subjected to various stimuli.
A That would also be --
Q Part of the experiment.
A -- part of an experiment.
Q And so your role has been to evaluate various experiments that have been performed using the polygraph.
A The distinction that I would draw is that much of my expertise is evaluating real-world evidence, psychological tests, psychological interventions. And so if you include experiments to be data that are available in the natural world, what might be called natural experiments, that is also part of my expertise.
Q So the literature refers to that sort of experiment as "field studies," isn't that right?
A Yes.
Q So the literature really breaks this type of experimentation into two categories: laboratory experimentation and field studies. Is that accurate?
A Well, in the research it's sometimes called "analogue studies" and "field studies," there are also theoretical studies, studies that test theoretical principles. It can either be laboratory or field.
Q And your role has been to examine the methodology in each of those categories of experimentation and testing, is that right?
A It depends on how you define "methodology." If you define methodology to include the scientific principles that are part of the methodology, then yes.
Q Now, with respect -- you used the word I think "analogies," is that the word you used?
A Analogue.
Q Analogue. That would be equivalent with laboratory studies, is that correct?
A Well, some of them are conducted in field studies, it might be a study conducted in a prison where you have created a laboratory in a prison. But it's an attempt to be an analogue of the real-world situation. It's not quite a psychologist's laboratory, not quite the real world.
Q And at any rate, when the experimentation is of the laboratory or analogue sort, is it correct to say that, with respect to the polygraph, an experiment of that sort involving the polygraph, that that, you would have designed an experiment using proper methodology, to hold a number of things constant, that is, to take certain things, parts of this experiment, and if you're going to compare them to something else, you would need all of those things to be the same. Is that right?
A In that one part of an experimental design.
Q So that, for example, in the literature there's been some concern about the comparability of a population of college students undertaking the test, as opposed to a population of people who might be criminal suspects.
A Yes.
Q And there has been some criticism of the utility, the scientific utility, in evaluating polygraph when used for forensic purposes of these studies for that reason.
A I would say that's one of the reasons.
Q Well, that's a reason.
A That is "a reason," yes.
Q And another reason might be that, in that type of situation, one has only a simulated crime as part of the experiment, not a real crime. That would be another limitation of the utility of the experiment?
A Yes, depending on what theoretical construct the study was attempting to assess.
Q And just so we're clear about the polygraph --
MR. GOOD: Excuse me, Your Honor, one moment.
THE COURT: We'll take a brief recess.
(Recess.)
THE COURT: All right, I'll just remind you you're still under oath, and you can resume, Mr. Good.
MR. GOOD: Thank you, Your Honor.
CROSS-EXAMINATION BY MR. GOOD, Resumed
Q Dr. Saxe, the polygraph instrument, there would be no polygraph examination without a polygraph instrument, is that true?
A Yes.
Q And you were, is it fair to say, the principal author of the Office of Technology Assessment of the Polygraph?
A The Office of Technology Assessment Report on "The Scientific Validity of Polygraph," yes, I was the principal author.
Q And in that report, there's a description of the instrument. And just to save time, the report states that there's, that there is a standard understanding of what the instrument measures, is that true?
A That there's a typical device, circa 1983, which is used in a so-called polygraph exam.
Q And that's when one refers to "polygraph exam," one is referring to an examination which makes use of that instrument.
A In this context, yes, but in another context, not necessarily.
Q But in the context of, when a word is used, "polygraph," when you use it, it refers to such an instrument.
A Well --
Q The standard instrument. Now, the standard instrument -- your answer is yes?
A Well, there's a typical, there was as of 1983, a standard machine that was used as part of a so-called polygraph examination.
Q And that machine measures several physiological responses, is that correct?
A That's correct.
Q And those are relative blood pressure?
A Yes.
Q Is one of them.
A Yes.
Q And that's done by fixing a, what's commonly referred to as a "blood pressure cuff" to the subject of the examination?
A Sphygmomanometer, yes.
Q Sphygmomanometer. We'd better spell that for the --
THE COURT: Sphygmomanometer.
MR. GOOD: Right. S-p-h-y-g-m-o-m-a-n-as in Nancy-o-m-e-t-e-r.
Q That's the blood pressure cuff?
A Yes.
Q And that measures relative blood pressure of the subject.
A Yes.
Q And it also measures the electrodermal response?
A Yes.
Q Which is the sweat?
A The measure is based on sweat.
Q Okay. And the third measurement is respiration.
A Yes.
Q So when you refer to a polygraph examination in your work, the instrument involved is that instrument.
A Typically.
Q Well, would it be a misnomer to refer to an instrument as a polygraph if it doesn't measure those three things?
A There are some studies that have measured all three of those. There are some polygraphs that measure additional physiological indices. There are computerized polygraphs.
Q But the standard understanding as of 1983 and since, is that it has to have at least those three, correct?
A Yes.
Q It might have others on top of that, other bells and whistles, but it's got to have electrodermal response, it's got to have respiration, and it's got to have relative blood pressure, otherwise it's not a polygraph, is that right?
A Again, that is the typical field polygraph tool.
Q Now, sir, returning to experimentation, and the methodology of experimentation, when one does experimentation, when people have done experimentation with the polygraph, they use an instrument that has those three measurement characteristics as one element of the experiment, correct, one part of the experiment.
A There are experiments that have only used one or two channels. It depends on the experiment. If the experiment is designed to replicate the situation in the field, then it would use those three channels. But experiments, because experiments are designed to test ideas, test constructs, don't necessarily have to have all three of those channels, and measure them in the same way that is done on a field polygraph examination.
Q And it's true then, that if you want to use the data from an experiment to evaluate the field use of the instrument which you said has those three channels at least, one would have to have in the experiment, to hold that same thing equal, you'd have to have the same instrument to work with, wouldn't you?
A Not necessarily. Again, if the purpose of the experiment is to test one of the underlying principles having to do with the polygraph, it may be possible to use either the cardiovascular or the electrodermal response without the respiration.
Q Well, would it be proper to refer to an instrument that doesn't contain all three channels as a "polygraph"?
A Well, as long as it has more than two, or it has two, it's measuring multiple measures, it's coming out, it comes out on a graph, it is a polygraph.
THE COURT: Excuse me, Mr. Good, are you going somewhere with this?
MR. GOOD: Yes, I am, Your Honor.
Q Now, Dr. Saxe, you've agreed that the field instrument is a three-channel instrument, and it has those three specific measures, not any three, but those three, correct?
A Yes.
Q Now, sir, when that --
A Although if I can just, to add to that, typically there are two channels of respiration measured, rather than just one.
Q Okay. And to maximize the degree to which an experiment involving the testing either the theory or the empirical data about the polygraph examination used in the field, one would use an instrument that has those three channels, isn't that true?
A It depends on the theory that you're testing.
Q Well, if you are testing the theory that the polygraph with those three channels is or is not vulnerable to countermeasures, you want to do an experiment to determine whether or not countermeasures will be effective in interfering with the accuracy of the device, one would use the same device as the device used in the field, would you not?
A Not necessarily. I mean you could have what would be a conservative test of a countermeasure hypothesis by just looking at the electrodermal response. If we could demonstrate that the electrodermal response, with minimal training, can be altered, that would be powerful data, that would be very useful data, even though the situation is different in that case than with a field situation.
Q Now, but it's correct to say then, that you could test the effectiveness of the countermeasure against each of the measurements of the device, and you could that separately, is that right?
A You could do it separately, yes.
Q But if you want to determine whether or not the device, measuring all these things simultaneously, is vulnerable to a countermeasure, one would need to use a device that measures all three responses simultaneously, would you not?
A Again, I don't see the necessity. I could see, for example, just demonstrating - that's been done by Drs. Honts and Raskin - that with minimal training you can defeat a polygraph test, you demonstrate that on one or two channels, you have raised serious theoretical as well as empirical question about the procedure.
Q Well, what I'm asking you, sir, is, what -- if I wanted to determine whether or not countermeasures are effective against the device that measures all three simultaneously, one would have to do the experiment with all three operating simultaneously, isn't that right?
A Sounds logical, yes, as you stated it, yes.
Q And it's also true that when one reports the results of the experiment using your example - that if I wanted to do an experiment and determine the effectiveness of the countermeasure as against, say, the respiration channel - that I would report that the countermeasure is effective against the respiration channel. Isn't that right?
A In general, yes.
Q And you would not report that the countermeasure is effective against a channel that you hadn't tested, isn't that true?
A It depends on what other controls are present in the study.
Q Well, if you conducted the study without, for example, the blood pressure, and you've done a countermeasure study, you could not report as a result of the experiment that the countermeasure is effective against that channel, because you hadn't tested that in the experiment, isn't that true?
A You wouldn't say specifically that you have tested that channel if you hadn't tested it, that's correct.
Q And, now, sir, you have indicated that one countermeasure that's effective against the polygraph is whether or not, if you instill in a person who's subjected to a polygraph a disbelief in the effectiveness of the polygraph, that he or she, that the results would be rendered inaccurate by conditioning the person to believe that the test doesn't work, isn't that true?
A In rough terms, yes.
Q And you've made that statement in your affidavit before the Court in paragraph ten, isn't that right?
A Yes.
Q And in paragraph ten on page seven, at the end of the, the paragraph that ends at the top of page seven --
A Yes.
Q -- you've cited an article which you then attached to your affidavit, am I right?
A That's correct.
Q And, in other words, the reference says "Saxe, 1991a."
A Right.
Q And that, in turn, is attachment three to your affidavit.
A I believe so.
Q And at page 411 --
A Right.
Q -- of the article, this is the part of the article, it's in the second column --
MR. GOOD: Does the Court have a copy of this, Your Honor?
THE COURT: I have it right in front of me.
MR. GOOD: Good.
Q The second column on page eleven has a reference to "Table 1" that's on the next page.
THE COURT: When you say "page eleven," you mean page 411.
MR. GOOD: 411, I'm sorry.
Q There's a paragraph toward the bottom of page 411 that refers to "page one," which appears on the next page, 412, am I right?
A That's correct.
Q And in the paragraph prior to that, that is, the paragraph prior to the reference to "page one," you mention a study, an experiment by Saxe, meaning yourself.
A Yes.
Q Schmitz and Zaichkowsky, is that -- can you pronounce that for the Court?
A Zaichkowsky.
Q Zaichkowsky. And you give the date "1987." You say that you tested the hypothesis and the hypothesis, the way in which you tested that hypothesis is that persons were subjected -- subjects of this experiment that you described, "... had the option of taking some money from a desk drawer, with the promise that they could retain the money if they could pass a polygraph test. Subjects were told that a new type of computer polygraph was being tested. In one condition, the test was described accurately in terms of the authors' doubts about its validity; in another condition, the standard polygrapher's injunction about the test's infallibility was given. Subjects also had the test demonstrated, except that the demonstration failed in the condition in which they were told of the problems with the polygraph." Have I read that correctly?
A I believe so.
Q And you report that, "the results were clear-cut. When subjects believed in the efficacy of the test, detection rates were obtained" --
THE COURT: Slow down.
MR. GOOD: I'm sorry.
Q "... detection rates were obtained that were similar to those found by other analogue researchers." By that you mean when the subject was told that the test works to detect falsehood or deception.
A What I meant there was that we found results that were similar to the results that other researchers who create these analogue conditions find.
Q And that, the last sentence is, "When subjects did not believe in the efficacy of the procedure, however, none of the guilty subjects were detected and errors were again made in misidentifying innocent subjects as guilty." And those results are reflected --
A In table one.
Q In table one on the next page.
Now, sir, you, in writing this article, attachment three, rely upon this as evidence that the polygraph instrument, including the three channels that you've agreed are the polygraph instrument, is vulnerable to this type of differential instruction, or to the person, isn't that true?
A That's true.
Q And it's also true, is it not, that the reference to this experiment appears on page 415 in the list of "References." In the second column it says, "Polygraph Tests as Placebos. Unpublished transcript."
A It says "Unpublished manuscript."
Q I'm sorry, "Unpublished manuscript." Is that right?
A That's correct.
Q And so that when one reads this article on page 411, the reference to the "Saxe, Schmitz, and Zaichkowsky" article, is the same article referred to --
A Yes.
Q -- in the "Reference." Okay. And that article was submitted by you for publication to what journal, sir?
A I'm actually not sure that I formally submitted it for publication.
Q Well, isn't it true that you submitted for publication this article to the "Journal of Applied Social Psychology," sir?
A No, I don't believe so.
MR. GOOD: May I approach the witness, Your Honor?
THE COURT: Yes.
MR. JALELIAN: Your Honor, I'd like the record --
(Document handed to counsel.)
Q Dr. Saxe, is that the article to which we've been referring, the manuscript?
A This is the manuscript, yes.
Q And it's true, is it not, that the table that appears in the article in attachment three, appears as table two in the manuscript?
A I believe so, yes.
MR. GOOD: Your Honor, may that be marked as an exhibit?
THE COURT: Any objection?
MR. JALELIAN: No.
THE COURT: It may be marked Exhibit 12. And if I may see it.
(Article, as above, received
and marked Exhibit No. 12.)
THE COURT: Do you need it, Mr. Good?
MR. GOOD: No, Your Honor, thank you.
Q Now, do you have that before you?
A No, I gave it to the Court.
MR. GOOD: Your Honor, may I just give the witness another copy?
THE COURT: Yes.
(Document handed to the witness.)
Q Now, sir, turning to page three, to page two rather, this article was written in 1987, is that right?
A I believe so.
Q And the OTA report was written, or published rather in 1983, right?
A Also written in 1983.
Q It was written and published in 1983, okay.
Q Now, on page two of this manuscript, there's a paragraph with the heading "Technique." Do you see that?
A Yes.
Q And the second sentence in that paragraph describes the polygraph instrument as, "a physiological" --
THE COURT: You don't have to read it. I have it in front of me and the witness has it called to his attention. But what's the question?
Q The question is this: Dr. Saxe, in that paragraph you state that the polygraph instrument measures respiration, heart rate, and skin conductivity.
A That's correct.
Q And, but you've already testified that the instrument which is understood as the polygraph in the field and in the OTA study, measures respiration, skin conductivity, and blood pressure, isn't that true?
A Well, I think the term that we used was "relative blood pressure." Here, it's three indicators of arousal.
THE COURT: Well, let's cut to the chase, shall we? Did you mean to say "heart rate," or did you mean to say something else? Because you certainly did say "heart rate."
THE WITNESS: The language is not accurate. I meant to say that it measures indicators of arousal and that one of them is related to heart rate.
THE COURT: Excuse me, are you telling me that the polygraph measures heart rate?
THE WITNESS: No, I'm not. I'm saying that it measures something related to heart rate, called relative blood pressure.
THE COURT: Blood pressure is related to heart rate, is that what you're telling me?
THE WITNESS: No, I'm not saying that, Your Honor.
THE COURT: Then why don't we just say that you made a mistake, and where you said "heart rate," you should have said blood pressure.
THE WITNESS: Well, the reason I'm not saying that is because it doesn't actually measure blood pressure either. But this is poorly written. I did not intend to say that the polygraph measures heart rate.
Q And it would be incorrect to say that the polygraph, to refer to an instrument that measures -- strike that.
Now, this is, again, is the study that you rely upon as evidence for the Court that instilling disbelief in the accuracy of the test in the subject, functions as a countermeasure, isn't that right?
A It's one of the studies.
Q It's the study that you've cited on page 411 and 412, isn't that right?
A Yes, it is.
Q And you describe on page six of the manuscript the "instrumentation" that you used in this experiment to establish, rather in order to be able to report as you did in this manuscript, that the polygraph is vulnerable to this countermeasure, isn't that true?
A Yes.
Q And the instrumentation that you describe on page six in the middle paragraph, is actually referred to by you as a "polygraph device." Is that right?
A Yes.
Q But, in fact, this device that you used in this experiment, measures heart rate, is that right?
A That's correct.
Q And heart rate, by "heart rate" we mean beats per minute.
A That's correct.
Q That's not relative blood pressure, is it, sir?
A No, they're not exactly the same.
Q Well, beats per minute is one thing, and relative blood pressure is quite another, is it not?
A It is, yes.
Q And so to refer to this device that you used in this experiment as a "polygraph instrument," is inaccurate, isn't that true?
A It is not the same instrument as an instrument used in a field test of a polygraph exam conducted in the context of a criminal investigation.
Q And you have submitted this as authority to the Court, as scientific evidence for the Court to rely upon, as evidence that the countermeasure of a person being told the test doesn't work, will be effective against the field device, isn't that right?
A The basis for that --
Q No, I'm just asking whether or not it's true that you have relied upon this evidence using a non-field device, a device that doesn't have all the characteristics of the field device, as evidence with respect to the field device, isn't that true?
A Yes, I have.
Q And you have inserted this table, the same table that's in your article, in your manuscript, in this article that's part of attachment three.
A I have, yes.
Q And you have not disclosed, have you, sir, in the article that you submitted to the American Psychologist, that the device that you used to determine whether or not this countermeasure was effective, was not the field device, have you, sir?
A I referred readers, and some people have asked for copies of the original manuscript, I referred readers for details obviously to the original manuscript.
Q Well, what I'm asking you, sir, is, you have referred on paragraph, rather page 412, rather 411, to this experiment as one involving a polygraph, isn't that true?
A That is true.
THE COURT: Mr. Good, I do believe that, as far as I'm concerned, you've made the point. Now, if there's an additional point you want to make, I'm all ears.
Q Well, now with respect to methodology, sir, you've reviewed the literature about the polygraph, have you not?
A I have.
Q Many times.
A Yes.
Q And you've made criticisms of the literature about, or some of the literature about the polygraph, on the basis that people have drawn inferences from their experiments that are not warranted by science, isn't that true?
A That's true.
Q And is it correct to say, sir, that to assert, as you have in this article, attachment three to your affidavit, that the field instrument is vulnerable to a subject not believing, or being told not to believe in the test, when the test involved a field instrument, that this experiment doesn't support that as a matter of science, does it, sir?
A Well, I would say that it does. And I would say that because we were able to replicate for guilty subjects in the deceptive condition the finding that other people who do, who view different devices have found, that, in fact, it indicates that our "polygraph," put it in quotation marks if you want, is comparable to the kinds of measures, the kinds of sets done in other settings.
If you look at table one, you'll see that in the positive-information condition, we were actually able to detect four of the guilty subjects. And the interesting thing is that we were not able to detect subjects in the negative condition.
So perhaps this is not equivalent to the polygraph that was done in the laboratory as a biofeedback device that is used in the field, but we were still able to replicate the findings of polygraphers and people who used field test machines.
Q Well, you wouldn't say, would you, sir, that the use of, the same instrument that you used in this manuscript, or in this experiment described in the manuscript, that using the same criteria that you've used to evaluate the literature about the polygraph, that it would be sound science to rely upon the performance of this instrument - that is, this instrument that only measures one of the three channels - as evidence about the accuracy of the field test, would you, sir?
A "Sound science" means having a sound idea. And all I've suggested here, and suggested in my affidavit, is that there is this placebo explanation that I have offered in the literature, supported by a number of studies, not only this study, it is one of the possible explanations which might be offered for the conditions under which polygraph tests make errors.
Q And it's your testimony that you don't know whether this was submitted for publication in the "Journal of Applied Social Psychology"?
A I am fairly confident that it was not, although it's been a long time.
Q And given the assertions in this manuscript about, the assertion that the data gathered during this experiment shows that the polygraph field instrument is vulnerable to this placebo effect, it's your testimony that this manuscript provides scientific evidence on that point, is that right?
A Well, tell me the statement that you're referring to.
MR. GOOD: If I may just have a moment, Your Honor.
Q Beginning on page five, you describe the hypothesis being tested under the experiment described in the manuscript, is that correct?
A Yes, it is.
THE COURT: I think it's in the plural.
MR. GOOD: "Hypotheses," correct. I'm sorry, Your Honor.
Q And it states that, "It is hypothesized that polygraph tests assess deception accurately only when the subject believes that the procedure is efficacious."
A Yes.
Q And it then goes on to refer in that paragraph to "polygraph tests." And this is in the overall context of an article in the preceding section, the section preceding "Hypotheses" describes the research concerning polygraph tests, including the OTA, the Office of Technology Assessment, the field instrument, the studies about field instrument, isn't that true?
A I would have to re-read the manuscript. What I was talking about was --
Q Sir, I just asked you, isn't it true that the manuscript refers in the section preceding the part about "Hypothesis," with the research about the field test, the field instrument? Isn't that true?
A I don't see the word "field" in the paragraph.
Q Well, that's my word. The research is about, it mentions, for example, in the description of the research, your own the Office of Technology Assessment study.
A Yes.
Q And we've already gone over the fact that that study refers to the field instrument with the three channels at a minimum.
A I have said that in the field, when it is used, those are typically the three channels, yes.
Q And the OTA report describes the instrument that is the subject matter of the study as "the three-channel instrument," isn't that right?
A That is correct.
Q And so a reader of this manuscript would understand that the experiment that you performed is relied upon as evidence of the vulnerability of the field instrument to this placebo effect, isn't that true?
A Well, what I think I was saying and was trying to do, was to critique the underlying idea, being the notion --
Q I'm not asking you what you were trying to say. What I'm asking you is, that you tendered this experiment with a non-field instrument as evidence concerning the performance of the field instrument.
A That is correct.
Q That is correct. And what you didn't do when you offered this manuscript that resulted in the publication of the article which appears as attachment three, is disclose to the reader of the text of that article that you used in this experiment an instrument which is not a polygraph.
MR. JALELIAN: Objection. Asked and answered.
THE COURT: We're really going over material we've gone over a lot. But if there's a purpose, you just go right ahead.
MR. GOOD: Very well, Your Honor.
Q Now, it's accurate to say, is it not, that the data in the laboratory and the field concerning the false-negative rate --
A Yes.
Q -- is that the rate is somewhere, approximately ten percent, isn't that true?
A I don't think it's appropriate to average the rates either from laboratory or field studies. There are too many conditions that have to be attached to them.
Q Okay. But there are field studies conducted -- let me ask you this. There are field studies conducted with the RCMP as the examiners, where the false-negative rate is zero, isn't that true?
A The false-negative -- are you going from false-positive to false-negative?
Q False-negative.
A I thought it was ninety-two percent, but it could be zero.
Q Zero. And there is, one of those studies - I think you mentioned ninety-two percent - in one of those studies of the RCMP was done by Professor Iacono.
A That's correct.
Q And he had one false negative result, if you recall?
A I believe so.
Q And so his rate was two percent.
A You can't do it that way. You can't just talk about the false negatives without talking about the false positives, because if you call everybody "guilty," you can be a hundred percent accurate. But you'd also be making lots of errors on the other side.
THE COURT: Excuse me, I have to get something. I'll be right back.
(Brief pause in the proceedings.)
THE COURT: All right, you may resume.
Q In this instance, in the Patrick and Iacono study that we've been discussing, that's a study involving actual polygraph examinations in actual criminal investigations, isn't that right?
A That's correct.
Q And for purposes of this, this court, you understand, is considering a situation in which the defendant has undergone a polygraph examination, and there were no indications, according to the examination, or rather, it is probable, according to the examination, that her answers are truthful. Do you understand that?
A I understand that.
Q And the concern there would be whether or not that was erroneous.
A I understand that.
Q And the way in which one describes the rate of error in that context is the false-negative rate, isn't that true?
A That's correct.
Q And the false-negative rate reported by Patrick and Iacono for persons in this field study who took the examination and passed, was two percent, isn't that true?
A That's true.
Q And that study involved -- is it true to say that Professor Iacono was someone who, prior to the conduct of this study with the RCMP, was very concerned to do it, to get improvement in the level of certainty about the ground truth? Isn't that true? That prior to that he may have been critical of field studies because of the lesser degree of evidence about ground truth?
A He's been critical both before and after.
Q And he reported in this article that we've been referring to, that he refers to this study as a one in which there is a high degree of verification of ground truth with respect to these results reported in connection with the false-negative rate.
A But he also says, or, what he says is that the high rate of verification is the problem, that these guilty subjects were verified because of a confession, and that makes them very different kinds of cases than typical cases.
THE COURT: Why does it make them different?
THE WITNESS: Well, because if a subject confesses after questioning in a polygraph examination, and the polygraph examiner says, "Okay, I read the charts as the person is deceptive," those are very different from cases, like the present matter, where you don't have a confession.
THE COURT: I thought the confession was only for the purpose of establishing that, when they read the polygraph as indicating deception, and the person said the person was innocent, the confession just established that, in fact, they were deceiving.
THE WITNESS: One of the confusing pieces of the literature is to figure out how independent those are. But even if they are independent, what you have to realize is that the cases where you have a confession may be different kinds of cases, different kinds of individuals, than in your normal case or other cases, where a subject continues to maintain her innocence. We don't have any evidence, we don't have any way to know in those cases whether the polygraph is correct or incorrect.
THE COURT: Well, if the question is, "What is the truth with respect to the issue, is the person telling the truth when he proclaims his innocence," and if the person is told, "The polygraph interpretation indicates that you were deceptive when you said you were innocent." And if he then says, "Well, as a matter of fact I am guilty," I don't see why that's not valid data.
THE WITNESS: It's because, if those are the only cases that you look at, if the only cases that you put in that box are the cases of a guilty person who has confessed, then we don't know what the validity of polygraph is for the guilty person who hasn't confessed. Maybe a guilty person who confesses is the kind of person who is so guilt-ridden, who is so anxious about what they have done, so anxious and stressed by what they have done, that they feel that they have to confess. Whereas another subject, who either doesn't believe that they did what they're accused of doing, or doesn't feel anything about that, they're not included in any of these validation studies.
THE COURT: The question is, does the test properly tell us when somebody is lying.
THE WITNESS: It doesn't - it can't.
THE COURT: And the way it does that in this sample that we're talking about, is to ask people, "Did you do it?" Now, when the person says "no," and the machine indicates in the interpretation of the operator, that the person is being deceptive --
THE WITNESS: Right. Those aren't included in these studies unless there is some kind of, what Raskin has called, "incontrovertible extra evidence," incontrovertible physical evidence that they were the ones who did the crime.
But in this study, where the criterion was confession, the only guilty subjects that you get are the subjects that confessed.
THE COURT: Yes, but those aren't necessarily the only people who lie.
THE WITNESS: That's exactly my point. That's exactly why this test has not been validated. That is the underlying fallacy. And Iacono, the author of this study, has written extensively about that, and why it's a misinterpretation to use these data to say that you don't get false-negative conclusions.
THE COURT: Thank you.
MR. GOOD: Is Your Honor done?
THE COURT: Go ahead.
Q Now, Dr. Saxe, it's true that, with respect to the evaluation of whether or not we have a false-negative result in the Iacono study, is we're now talking about a person who passed the test, but is nevertheless determined to be guilty. Isn't that right?
A Please restate your question.
Q A false negative circumstance --
A Right.
Q -- is one in which a person has passed the test, and found to be truthful by the examiner, but is nevertheless guilty.
A That's correct.
Q And the way in which one determines guilt, using the word "guilty" in those circumstances, according to the report of the Iacono study, is not limited to the confession, isn't that true? It's also true that they rely on investigative information gathered prior to the test, as well as information, other information that would corroborate the confession.
A I believe - and at least the way -- I believe that in virtually all cases, it was verified only by confession. I think it was close to a hundred percent of the cases.
Q You're saying exclusively by confession?
A I believe. I would have to take a look at the study.
MR. GOOD: Well, Your Honor has the article. And rather than take the time --
Q There are, in the Honts studies with the RCMP, that cites additional evidence to establish ground truth, is that right, that is not just a confession, is that right?
A That is correct.
Q And the purpose there is to meet criticism that you don't know whether the person, this is the ground truth criticism, that we need to determine what the ground truth is, isn't that right? And that these studies were an attempt to meet that criticism by getting evidence, not just the confession, but also the additional evidence. For example, if someone confessed and told them where the bones were buried, and they sure enough found the bones where the person said, that's corroboration of a confession.
A Yes, that's, again, that's the fallacy of the logic, that's the problem. That if an innocent person is found deceptive, that's a false negative, the person is going to continue to protest their guilt, than if they were --
THE COURT: You test their guilt or --
A Well, in other words, if an innocent person is found deceptive, in other words, the lie detector finds them --
Q False positive.
A False positive, excuse me. Finds them guilty, they're going to continue to protest --
THE COURT: Well, let's be a little more accurate. The machine finds, that is the machine as interpreted, doesn't find them guilty, it finds them deceptive.
THE WITNESS: Deceptive, excuse me, right.
A That there is somebody else out there who did it, if that is an incorrect judgment. And so other evidence may, another witness may come forward, evidence may come forward to clear. In the case of guilty person who is incorrectly diagnosed as non-deceptive, there is nobody else out there, so you're much less likely to find out that you made an error.
Q Just so we're clear about this, Dr. Saxe, in the column of these studies that deal with the false-negative rate, we're talking about instances in which people passed the test, correct?
A Yes.
Q And we're talking about each instance in which a subject took the test, and is determined by the examiner to be truthful to the relevant questions, correct?
A Correct.
Q And the methodology of the experiment is then, or the methodology of the study, is then to look at evidence other than the polygraph to determine ground truth.
A Usually, it's the opposite way around.
Q Is it true that, during the course of the study, that one then looks, the authors of these studies then looked at the investigative files about the case to determine whether or not the person was, in fact, guilty, correct?
A Not correct as you described it.
Q Is it true that the files were ignored, the investigative files?
MR. JALELIAN: Objection. I'm just unclear as to what study we're talking about.
MR. GOOD: We're talking about Honts, the RCMP studies, okay?
Q What was done was to look at the police file, other than the polygraph file. Whatever information had been gathered by the RCMP about the guilt or innocence of the subject of the test was looked at by the authors of the study, isn't that true?
A Yes.
Q So that you're now going to investigate, independent of the polygraph test, to determine the ground truth with respect to each instance in which a person was determined by the examiner, the polygraph examiner, to be truthful, isn't that true? And the proponents of this study did that as to each instance in which a person was determined to be truthful.
A I don't think they did it in each instance in which a person was determined to be truthful. They did it, it was the opposite way around, that they looked for evidence, for ground truth evidence. And then from among those studies where they have the ability to make the judgment about ground truth, they went and looked at the polygraph.
Q Well, isn't it true that, as to the people who were found truthful, that as to the category of people who were not part of the false-negative, that is, the people who were declared truthful and determined by the study to be, according to the ground truth investigation of truthful, that there was an investigation to determine whether or not in fact someone else was guilty of the crime. Isn't that true, that -- so what you just said is incorrect, that is, we didn't just look at whether or not there's ground truth evidence to corroborate the fact that a person wrongly passed the test. In instances in which people passed the test, there was independent corroboration in the form of a confession usually of somebody else, that that somebody else committed the crime, rather than the test subject. Isn't that true? That's what the --
THE COURT: Wait a minute, what's the answer to the question?
A We're going in circles here.
In the case of a guilty person, you can corroborate a guilty person's testimony by somebody else who's confessed.
Q No, I'm asking you about an innocent person, a person who passed the test. And the question is whether they're innocent, okay?
A Okay.
Q Now, a person who has passed the test, for all the people who passed the test, we're going to determine whether or not they're guilty or innocent. Iacono found in his RCMP study that, as to one person who passed the test, there was evidence of guilt. Okay? As to everyone else who passed the test, there was independent evidence that someone else was guilty, not the subject. Isn't that true?
A As I understand --
THE COURT: Isn't it true that there was independent exonerative evidence? That's the question.
MR. GOOD: Exactly.
A In some cases, yes.
Q Isn't it true that the study states that in each instance in which a person was found to be truthful, and was not part of the set of people who make up the false negatives, which is one, okay, that there was independent evidence of the guilt of someone other than the test subject, exonerative evidence, isn't that true, for each instance?
THE COURT: Well, I thought we had asked that, and I thought the witness said, "In some cases."
MR. GOOD: What I'm saying is, if he looks -- well, what I'm asking him for, Your Honor, is --
Q If you looked at these studies, wouldn't you find in the reports of these studies that there was independent exonerative evidence for each person who passed, that they were, in fact, innocent?
A I'm reading from a paper that Iacono has written. He says, "In a study by Patrick and Iacono, sixty-five percent of the innocent suspects were confirmed."
Q What I'm asking --
THE COURT: All right, excuse me. On that happy note, we will take the luncheon break. We will resume at 2:30.
(Lunch Recess.)
Created and maintained by Cara Lundquist