Jury service and spotting the liars

[Capsid]

I'm doing my public duty as a juror for 2 weeks. It's an interesting experience but one I feel could be streamlined if there was a better way of identifying when someone is lying when in the witness box. The evidence given to the jusry is mainly anecdotal. The jury is instructed to acquit the defendant if they are not sure about the evidence and with many conflicting witness accounts the result would be one of not guilty.

How reliable then is reading body language to detect if someone is lying? I first thought of Derren Brown who is very adept at reading people and his website on channel 4 (http://www.channel4.com/entertainment/tv/microsites/M/mindcontrol/fom/index.html) (UK TV company) lists several books on the subject. Of course this has been attempted by the use of polygraphs (which I understand are not reliable/don't work). Is reading body language equally unreliable? I think my deliberations with fellow jurors would have been facilitated by some training in this.

[Rob Lister]

I'm doing my public duty as a juror for 2 weeks. It's an interesting experience but one I feel could be streamlined if there was a better way of identifying when someone is lying when in the witness box. The evidence given to the jusry is mainly anecdotal. The jury is instructed to acquit the defendant if they are not sure about the evidence and with many conflicting witness accounts the result would be one of not guilty.

How reliable then is reading body language to detect if someone is lying? I first thought of Derren Brown who is very adept at reading people and his website on channel 4 (http://www.channel4.com/entertainment/tv/microsites/M/mindcontrol/fom/index.html) (UK TV company) lists several books on the subject. Of course this has been attempted by the use of polygraphs (which I understand are not reliable/don't work). Is reading body language equally unreliable? I think my deliberations with fellow jurors would have been facilitated by some training in this.

I think you should play it by ear and not try to use any techniques that claim to detect lies.

Consider, if the burden of proof is BaRD, and your technique has a reliability factor less than that, is justice served?

Just remain the skeptic (giving the accused the benefit of the doubt) through-out the trial.

[Capsid]

I think you should play it by ear and not try to use any techniques that claim to detect lies.

Consider, if the burden of proof is BaRD, and your technique has a reliability factor less than that, is justice served?

Just remain the skeptic (giving the accused the benefit of the doubt) through-out the trial.I'm not intending to use any techniques, it's too late for that now anyway. My query was really about if the trials process could be improved in the future if such techniques were used and what is the reliability of them?

[TKOne]

I read an article a few years ago (I think in New Scientist, but am not certain) about testing how well people could detect lying. Various people, such as lawyers and police officers, claimed to be able to tell from reading someone's body language. The testing showed that none of the people did well, with one exception. The exception was people who worked in espionage--CIA, MI5, etc.; those people performed well (I think ~80%).

There seem to be two ways to interpret that. (1) You can tell if someone is lying, via body language, but it takes a great deal of professional training to acquire the skill. (2) Some people are naturally talented at determining if someone is lying, and those people tend to be employed in espionage.

[Capsid]

So if the skill can be taught then I propose that juries comprise trained people in body language reading/detecting lies. This might reduce the number of court cases with defendants pleading guilty as well as speed up the deliberation process.

[timhau]

I'd like to see evidence that detecting when a complete stranger is lying is a skill that can be taught. It certainly can't be honed to a degree where it would be 100% reliable, so using trained lie detecting specialists could well a way to enable juries to make mistakes with confidence.

One of the most underwhelming documentaries I ever saw dealt with the ability of 'body language experts' to tell when a person is lying. The experts first went on and on about the tell-tale signs that expose a liar, and then they showed their art in practice. As a demonstration, they had chosen Bill Clinton's famous "I did not have sexual relations with that woman, Ms. Lewinsky", which they exposed as a lie -- in a program filmed in the 2000s. Whoo-pe-doo-doo-doo.

[Dan O.]

Any well known technique for reading body language could also be used to train a witness to lie using that language. The proper way to resolve conflicting statements is to probe for additional evidence. It's too bad that juries aren't empowered to do that.

One technique that I've found useful as a juror in a trial is to assume that anyone that is speaking while not under oath is not necessarily telling the truth.

[TKOne]

timhau, the article that I read was about complete strangers.

The article also described a lie-detecting device that actually works, albeit in a limited context. Suppose that you have visited some site, and are later shown the site. It turns out that a certain part of your brain gets activated--a part that does not get activated if you have never see the site before. So, if a criminal suspect claims that he has never visited, say, a certain home, and you show him pictures of that home, you can tell whether or not he is lying, but monitoring certain parts of his brain.

This is claimed to be guaranteed (though I do not know if it has been tested on highly-drugged people). Then you can compare the device results with what lawyers/policemen/spies/etc. claim to be able to tell via body language.

[timhau]

I remember an article in which some of the body language based lie-detection stuff was taught to law enforcement officials. Their lie detection skills were tested before and after the course, and the way I remember it was that while the course gave them more faith in their skills in lie detection, their performance actually deteriorated -- it enabled them to go wrong with confidence.

However, take all that with several grains of salt, since I can't find the reference (it's amazing how much bs comes up on google if you try 'lie detection' and 'body language').

[timhau]

timhau, the article that I read was about complete strangers.

OK. I'm not saying that lie-detection isn't a skill or that some people aren't better in it than others. What I do doubt is whether it is a skill that is easily taught to just about anybody.

The article also described a lie-detecting device that actually works, albeit in a limited context. Suppose that you have visited some site, and are later shown the site. It turns out that a certain part of your brain gets activated--a part that does not get activated if you have never see the site before. So, if a criminal suspect claims that he has never visited, say, a certain home, and you show him pictures of that home, you can tell whether or not he is lying, but monitoring certain parts of his brain.

That's interesting; monitoring brain activation could well be useful in this context. However, since it's activation that is measured, one question springs to mind: can this method distinguish between situations where the suspect has been in the pictured home earlier and ones where he hasn't been in that particular home, but it strongly reminds him of a familiar place?

[Wolfman]

I would have several concerns with such a methodology.

1) Juries are supposed to be chosen at random from society at large; but your proposal would essentially require that juries would have to be made up of 'professionals' who had been trained in such methods. This would create a significantly smaller pool of potential jurors, thereby significantly decreasing the odds of being able to put together an impartial jury. Consider a trial in a smaller community, where only 50 or so people had undergone the necessary training to be able to "read" whether people are lying or not. The odds that most of those 50 people know the people involved in the case, and/or have read/heard prejudicial information prior to the trial, are quite high.

Finding a jury that is as unbiased as possible, and will consider the evidence fairly, is a crucial component of the entire judicial system; your proposal would significantly undermine that aspect.

2) Attempting to train a jury who had not previously been fully trained and tested in such methods would be an exercise in disaster; there is no way that you would be able to give more than a very brief, cursory explanation of the methods involved (or are you going to incarcerate every jury for an additional 2 weeks before the trial begins in order to train and test them?). The result would be not only people who didn't really understand the methodology, but who in addition would make wrong conclusions and thereby render poor verdicts, based on faulty impressions rather than an evaluation of the actual physical evidence.

3) These methods have proven notoriously inaccurate when used on people from other cultures. Used on strangers from your own culture, there may be a certain degree of accuracy; but used on strangers from another culture, where physical cues and mannerisms may have an entirely different meaning, is an exercise in futility. So you would have to train people not only in how to use these techniques, but in addition you'd have to train them in discerning such cues in a wide variety of different cultures.

4) I do not believe that even the most practiced and competent 'reader' would be able to boast complete accuracy or certainty in their readings. Derren Brown, while quite brilliant, still makes mistakes...and he's dedicated years and years to this. How competent would an average juror be? I do not believe a high enough level of competence could ever be reached or maintained by a jury of 12 people that this could prove to be a viable method.

[Capsid]

I would have several concerns with such a methodology.

1) Juries are supposed to be chosen at random from society at large; but your proposal would essentially require that juries would have to be made up of 'professionals' who had been trained in such methods. This would create a significantly smaller pool of potential jurors, thereby significantly decreasing the odds of being able to put together an impartial jury. Consider a trial in a smaller community, where only 50 or so people had undergone the necessary training to be able to "read" whether people are lying or not. The odds that most of those 50 people know the people involved in the case, and/or have read/heard prejudicial information prior to the trial, are quite high.

Finding a jury that is as unbiased as possible, and will consider the evidence fairly, is a crucial component of the entire judicial system; your proposal would significantly undermine that aspect.

2) Attempting to train a jury who had not previously been fully trained and tested in such methods would be an exercise in disaster; there is no way that you would be able to give more than a very brief, cursory explanation of the methods involved (or are you going to incarcerate every jury for an additional 2 weeks before the trial begins in order to train and test them?). The result would be not only people who didn't really understand the methodology, but who in addition would make wrong conclusions and thereby render poor verdicts, based on faulty impressions rather than an evaluation of the actual physical evidence.

3) These methods have proven notoriously inaccurate when used on people from other cultures. Used on strangers from your own culture, there may be a certain degree of accuracy; but used on strangers from another culture, where physical cues and mannerisms may have an entirely different meaning, is an exercise in futility. So you would have to train people not only in how to use these techniques, but in addition you'd have to train them in discerning such cues in a wide variety of different cultures.

4) I do not believe that even the most practiced and competent 'reader' would be able to boast complete accuracy or certainty in their readings. Derren Brown, while quite brilliant, still makes mistakes...and he's dedicated years and years to this. How competent would an average juror be? I do not believe a high enough level of competence could ever be reached or maintained by a jury of 12 people that this could prove to be a viable method.
Yes my proposal is full of holes I agree. I'm being quite radical and throwing the idea out for discussion. I appreciate your input.

1) the system might work in countries with a large population?

2) I'm proposing a professional group of jurors trained in reading body language, that would be their job. Could that work? After all they would be experienced in the jury system and expedite the process.

3) Good point. Is body language so different between cultures?

4) Mistakes may happen but using a group of professionals (not average jurors) may help to reduce that.

[Wolfman]

2) I'm proposing a professional group of jurors trained in reading body language, that would be their job. Could that work? After all they would be experienced in the jury system and expedite the process.
It seems to me that what you are describing, essentially, is the role of the judge. A person generally has a choice between trial by jury, or trial by judge. The judge is a professional, well versed in various points of the law; a jury is non-professional, drawn from one's peers.

If you were to propose such training for judges (and if you could eliminate other concerns regarding the accuracy of such methods), you might arguably have a case.

But in regards to a jury, one of the most basic principles in choosing a jury is that they be as unbiased as possible...that they do not know or have any connection with the person being tried, that they have not heard any news about the case that might be prejudicial to their deliberations, etc.

If you had "professional jurists", they would essentially have to be paid regular salary as full-time professionals (as it would be impossible for them to hold other jobs when they are constantly being called away for jury duty). It would be economically unfeasible to pay full time salaries to an excessively large population of jurists, so you would have to keep the number of professional jurists to a relative minimum. Which all means that the odds of being able to find an unbiased jury is very significantly decreased. In order to even begin to implement what you suggest, you'd have to change the entire system and definition of what a jury is.

I'd suggest, if you want to pursue this as a theoretical model, that you focus on judges, rather than juries, for such training.

[CptColumbo]

Some states allow the jury to ask a witness questions, you might want to make sure you're in one of them before you try it. It might help clarify a few points.

[balrog666]

So if the skill can be taught then I propose that juries comprise trained people in body language reading/detecting lies. This might reduce the number of court cases with defendants pleading guilty as well as speed up the deliberation process.

Aside from being utter crap, it's a stupid idea. Would you want your life hinging on some bozo who's only 80% correct? Of even 99%?

To take it just one step further into outright lunacy, just train the judge, let him interview the defendant, and then there's no need for a trial in the first place! Hell, give that judge a 9mm with one bullet and he can save us from all those expensive appeals too. Gee, does that sound good to you?

Oops, that's already be said.

[Dan O.]

The whole purpose of a trial was to read body language. Where do you think the word testimony comes from? In the old testament there are several references to putting ones hand under another's thigh to ascertain that they are telling the truth.

[Puppycow]

I applaud the idea that we need to find a way to make the trial process more accurate, but I don't put much stock in looking for body language cues. witnesses are going to be very nervous even if they are truthful, wheras some people are very smooth liars. A nervous but honest person might appear more suspicious than a smooth liar.

A recent study (http://www.sciencedaily.com/releases/2007/06/070628161330.htm) suggests that juries get the verdict wrong about 1 in 8 times, which is uncomfortably high. If an innocent defendant has a 1 in 8 chance of being convicted, to me that means we need creative new ideas to improve the process and help people analyze facts better. For example, there are studies showing that eyewitness testimony is much less reliable than most juries think it is.

If we had more sceptical people on juries, I think it would help.

[TKOne]

OK. I'm not saying that lie-detection isn't a skill or that some people aren't better in it than others. What I do doubt is whether it is a skill that is easily taught to just about anybody.

I agree.

That's interesting; monitoring brain activation could well be useful in this context. However, since it's activation that is measured, one question springs to mind: can this method distinguish between situations where the suspect has been in the pictured home earlier and ones where he hasn't been in that particular home, but it strongly reminds him of a familiar place?

Good question. I don't know. (I've just tried googling to find the article, but didn't, sorry. It was some years ago.)

[Wolfman]

A recent study (http://www.sciencedaily.com/releases/2007/06/070628161330.htm) suggests that juries get the verdict wrong about 1 in 8 times, which is uncomfortably high.
It is an interesting argument; they take cases where a jury finds a verdict, and ask the judge presiding over that case what his verdict would have been. If the judge's verdict would have been different than the jury's verdict, then it is assumed that the jury's verdict was 'wrong'. According to this methodology, 1 in 8 jury verdicts were 'wrong'.

But one crucial problem with this argument; I'd be willing to bet that in by far the majority of instances, if you were to put together a panel of eight randomly chosen judges from across the U.S., and have them look at verdicts of other judges, you'd find at least one in eight of those judges disagreeing with the verdicts made by other judges.

There is a faulty assumption, in my opinion, that because judges are "professionals", that therefore they are "better"; but this ignores the fact that judges are largely political appointees. There are tons of judges out there who are incredibly incompetent (just read the stories about judges who masturbate under their cloak during a trial, or who sleep through crucial testimony), but who maintain their positions because of political connections. And many verdicts are rendered based not on a pragmatic evaluation of the evidence, but on political agendas of the politicians who appoint those judges.

One of the main strengths of a jury is that it is largely immune to political pressure, or corruption; something that is most definitely not true of judges.

No system is perfect. Which is why defendants are given the choice for themselves, whether they have trial by jury, or by judge.

But the "conclusion" that one out of eight jury verdicts were "wrong" just because the judge in the case would have rendered a different verdict? Sorry, but the way this research was done, I'd consider the conclusions as meaning less than nothing.

[Capsid]

It seems to me that what you are describing, essentially, is the role of the judge. A person generally has a choice between trial by jury, or trial by judge. The judge is a professional, well versed in various points of the law; a jury is non-professional, drawn from one's peers.

If you were to propose such training for judges (and if you could eliminate other concerns regarding the accuracy of such methods), you might arguably have a case.

But in regards to a jury, one of the most basic principles in choosing a jury is that they be as unbiased as possible...that they do not know or have any connection with the person being tried, that they have not heard any news about the case that might be prejudicial to their deliberations, etc.

If you had "professional jurists", they would essentially have to be paid regular salary as full-time professionals (as it would be impossible for them to hold other jobs when they are constantly being called away for jury duty). It would be economically unfeasible to pay full time salaries to an excessively large population of jurists, so you would have to keep the number of professional jurists to a relative minimum. Which all means that the odds of being able to find an unbiased jury is very significantly decreased. In order to even begin to implement what you suggest, you'd have to change the entire system and definition of what a jury is.

I'd suggest, if you want to pursue this as a theoretical model, that you focus on judges, rather than juries, for such training.Good points again. I suppose I am proposing that the whole system is changed. Can a jury really be free of prejudice or bias? I don't think so, in the time I have spent on a case it was clear that some of my fellow jurors had decided on their decision not on the evidence (such as it was) but on their perception of whether the defendant was the type to commit a crime regardless of whether he had committed the specific crime he was being tried for or not. I'm just exploring an idea as to whether their might be a better way. Probably not.

[Gravy]

Another obvious problem with this proposal is that all lawyers, defense or prosecution, would train their clients to exhibit "honest" body language.

[Wolfman]

Another obvious problem with this proposal is that all lawyers, defense or prosecution, would train their clients to exhibit "honest" body language.
Oh, Gravy...lawyers are only interested in the truth, surely they would not intentionally train their clients to intentionally deceive a judge or jury just so that they could win the case...

[Capsid]

Another obvious problem with this proposal is that all lawyers, defense or prosecution, would train their clients to exhibit "honest" body language.
They do it already, the questions are asked in a way that requires little emotion, most often it's a simple yes or no. Further, the witnesses are not allowed to speculate.

[Mangafranga]

It is an interesting argument; they take cases where a jury finds a verdict, and ask the judge presiding over that case what his verdict would have been. If the judge's verdict would have been different than the jury's verdict, then it is assumed that the jury's verdict was 'wrong'. According to this methodology, 1 in 8 jury verdicts were 'wrong'.Emphasis in above and below is mine. This isn't an accurate rendition of the methodology. See below for the article's rendition of the methodology.
To obtain a numerical estimate of jury accuracy, some assumptions were made, as is the case for virtually any statistical analysis of social groups or programs. A key assumption of Spencer's study is that, on average, the judge's verdict is at least as likely to be correct as the jury's verdict.

Without assumptions, a 77 percent agreement rate could reflect 100 percent accuracy by the judge and 77 percent accuracy by the jury, or 100 percent accuracy by the jury and 77 percent accuracy by the judge, or 88 percent accuracy by both, or even 50 percent accuracy by both if they often agreed on the incorrect verdict.

With the assumption of the Spencer analysis that judges are at least as accurate as jurors after completion of all testimony, we can get an estimate of jury accuracy that is likely to be higher than the actual accuracy. Thus, the 77 percent agreement rate means that juries are accurate up to 87 percent of the time or less, or reach an incorrect verdict in at least one out of eight cases.

[timhau]

I wonder if the juries were, in general, more likely to acquit or convict than the judge.

[Wolfman]

Emphasis in above and below is mine. This isn't an accurate rendition of the methodology. See below for the article's rendition of the methodology.
That's fine; but to determine the validity of these results, I'd want to see the same study done with results from judges only, and see how often judges agreed on the verdicts of other judges. I'd be very willing to bet that the rate of agreement/disagreement between different judges would be equal to or greater than the rate of agreement/disagreement between judges and juries.

I'm also concerned that although the article states that this was based on "271 cases from four areas", it does not state how many judges were involved. I highly doubt that it was 271 different judges; rather, that it was a limited pool of judges who submitted their conclusions in a number of different cases.

If that is true, then if only one or two of those judges was incompetent, that would entirely destroy the validity of any results from this study.

There are 271 unique juries involved in this study...but how many unique judges? 10? 15? 20? How many incompetent judges would it take to render any conclusion from such a study as completely useless?

Before I could even begin to consider these conclusions as valid, I'd want to see two further steps taken. First, that exactly the same study be done with judges examining the verdicts of other judges, to see how often they agree on the verdicts; and two, that where the verdicts of judges and juries disagree, they be examined by a panel of three other judges to see whether or not they agree with the decision of the first judge.

The article talks about using these statistics to protect people from wrongful convictions based on faulty verdicts rendered by juries; yet they ignore the easiest way of determining if this is actually a valid concern/conclusion.

It is simple. We've got tons of examples of wrongful convictions, where a person was convicted, then subsequently found to be innocent. So take all those cases, and divide them into verdicts that were rendered by judges, and verdicts that were rendered by juries. If it is found that a significantly higher number of wrong verdicts are coming from jury trials, that would indicate a very strong statistical argument that juries are more likely to render wrong verdicts.

My proposal would be far easier to undertake, would use information that is easy to get, and could be used to render conclusions far more authoritative and far less speculative than anything in this study. Personally, I think this is a massive waste of money to generate results of dubious value at best.

[Mangafranga]

I'm also concerned that although the article states that this was based on "271 cases from four areas", it does not state how many judges were involved. I highly doubt that it was 271 different judges; rather, that it was a limited pool of judges who submitted their conclusions in a number of different cases.

If that is true, then if only one or two of those judges was incompetent, that would entirely destroy the validity of any results from this study.

There are 271 unique juries involved in this study...but how many unique judges? 10? 15? 20? How many incompetent judges would it take to render any conclusion from such a study as completely useless?I agree that the actual conclusion they came up with (that the wrong verdict was given in 1 out of 8 cases) could in fact be damaged by such a scenario (and indeed that such a scenario is possible does damage the conclusion). However I do not agree that "any conclusion" would be completely useless. The reason I checked the article to begin with was because I assumed that they wouldn't restrict themselves to a conclusion about juries. I assumed their conclusion would be more like this (from the article)-Without assumptions, a 77 percent agreement rate could reflect 100 percent accuracy by the judge and 77 percent accuracy by the jury, or 100 percent accuracy by the jury and 77 percent accuracy by the judge, or 88 percent accuracy by both, or even 50 percent accuracy by both if they often agreed on the incorrect verdict.i.e. that there is a loss of accuracy somewhere, and it is either in juries or in judges, and it is of this range of degree. And of course if some judges are known to be dodgey, then the lack of accuracy might not equate to actual bad outcomes, as if a lawyer knows the judge to be dodgey, they would then opt for a jury trial. It is simple. We've got tons of examples of wrongful convictions, where a person was convicted, then subsequently found to be innocent. So take all those cases, and divide them into verdicts that were rendered by judges, and verdicts that were rendered by juries. If it is found that a significantly higher number of wrong verdicts are coming from jury trials, that would indicate a very strong statistical argument that juries are more likely to render wrong verdicts.But they didn't draw the conclusion that "juries are more likely to render wrong verdicts." They concluded that, assuming juries are not more accurate than judges, the wrong verdict was given by juries in at least 1 out of every 8 cases. So the conclusion was not about seeing who was more accurate, it was about seeing how accurate juries were, given some assumptions about the relative accuracy of judges and juries.

[Puppycow]

Wolfman: I think the study is at least good enough to suggest that more research should be done. To disparage "the conclusions as meaning less than nothing" is to say that science itself is meaningless. It is like saying that we shouldn't even question the current system, but instead have unquestioning faith in it. The article was published in a peer-reviewed scientific journal. The 1 in 8 figure is only suggestive, not written in stone. Further studies with different methodologies would help shed more light and further increase the confidence level.

One possibility would be to have two independent juries and two independent judges judge each case (this study might be costly, but it would be worth it). It's horrible to imagine that so many innocent people might be convicted, so either way, research should be applied to find the true answer. I think and hope that in reality it is probably less than 1/8, but I think it is well worth doing the research, if only to increase our peace of mind and confidence in the system.

I wonder if the juries were, in general, more likely to acquit or convict than the judge.While the probability of an innocent defendant being convicted is apparantly higher, the number seems to be less. (Maybe this is because most defendants are guilty, so the higher rate of wrongful conviction is still less in absolute numbers than the number of guilty defendants acquited.) Still, one in ten convictions is estimated to be a "wrongful conviction."Abstract
Average accuracy of jury verdicts for a set of cases can be studied empirically and
systematically even when the correct verdict cannot be known. The key is to obtain a
second rating of the verdict, for example the judge’s, as in the recent study of criminal
cases in the U.S. by the National Center for State Courts (NCSC). That study, like the
famous Kalven-Zeisel study, showed only modest judge-jury agreement. Simple estimates
of jury accuracy can be developed from the judge-jury agreement rate; the judge’s verdict
is not taken as the gold standard. Although the estimates of accuracy are subject to error,
under plausible conditions they tend to overestimate the average accuracy of jury verdicts.
The jury verdict was estimated to be accurate in no more than 87% of the NCSC cases
(which, however, should not be regarded as a representative sample with respect to jury
accuracy). More refined estimates, including false conviction and false acquittal rates, are
developed with models using stronger assumptions. For example, the conditional
probability that the jury incorrectly convicts given that the defendant truly was not guilty (a
“type I error”) was estimated at 0.25, with an estimated standard error (s.e.) of 0.07, the
conditional probability that a jury incorrectly acquits given that the defendant truly was
guilty (“type II error”) was estimated at 0.14 (s.e. 0.03), and the difference was estimated at
0.12 (s.e. 0.08). The estimated number of defendants in the NCSC cases who truly are not
guilty but are convicted does seem to be smaller than the number who truly are guilty but
are acquitted. The conditional probability of a wrongful conviction, given that the
defendant was convicted, is estimated at 0.10 (s.e. 0.03).

[Gravy]

They do it already, the questions are asked in a way that requires little emotion, most often it's a simple yes or no. Further, the witnesses are not allowed to speculate.Yes, and more than that, in a major trial witnesses are often coached on their demeanor, body language, eye contact, clothes, etc. Many people who appear to be composed and confident of their innocence are in fact guilty skulking dogs, and others who are innocent can give the impression that they aren't.

[Wolfman]

To disparage "the conclusions as meaning less than nothing" is to say that science itself is meaningless.
Perhaps I was a little harsh; but there is too great a tendency to confuse "statistics" with "science". "Statistics" can be used in pretty much any way you want, to prove anything you want, by scientist and woo alike.

Without adequate info about how many judges were involved, and a comparative study of judges reactions to other judges rulings, I'm not sure that these figures or conclusions have any real substantiation at all; I'm quite sure that someone else could take EXACTLY the same stats and reach a completely different conclusion.

Statistics are too often used to prove a pre-determined conclusion, rather than as a method of reaching a conclusion. And that, most assuredly, is not science.

[TragicMonkey]

When I was a child, my mother thought she had a foolproof way to detect if I were lying. She'd give me this ridiculous glare and if I laughed (always, because it was hilarious) she concluded that I must be lying. Ironically, she was invariably wrong because I was such a good liar that she never suspected actual lies. She only applied this test to implausible-but-true things, and therefore managed a 100% wrong rate. She'd have been more accurate flipping coins.

The lesson is, you may observe someone's body language and reactions but you don't know what it means. My boss thought I was upset about something said at a meeting because I was fidgeting. I wasn't even paying attention--there was a little piece of plastic in the chair that was sticking me. And I know a guy who everyone thinks is always hitting on them, because he keeps winking roguishly at them. It's a tic. He's been in fights because of it.

The best way to judge if someone's lying is to assess their story, not their movements. Because even honest people can have a rash or an upset stomach.

[gnome]

I don't think there's any way to make eyewitness testimony more reliable. it's focused a lot on TV and in movies because it's very personal and lends itself to character acting. But it's usually not the most important aspect of a trial--instead it's the evidence that backs up the stories that matter. If you're having to decide by body language alone who's telling the truth, there's probably a significant reasonable doubt.

[richardm]

An interesting article (http://www.simonyi.ox.ac.uk/dawkins/WorldOfDawkins-archive/Dawkins/Work/Articles/1997-11-16trialbyjury.shtml) by Richard Dawkins regarding juries, which seems to touch on similar ground to the study's supposition that judges are at least more consistent than juries. It begins with typical bluntness:


Trial by jury must be one of the most conspicuously bad good ideas anyone ever had.

Would you bet on two independent juries reaching the same verdict in the Louise Woodward case? Could you imagine even one other jury reaching the same verdict in the O.J.Simpson case? Two judges, on the other hand, seem to me rather likely to score well on the concordance test.

[Beerina]

When I was a child, my mother thought she had a foolproof way to detect if I were lying. She'd give me this ridiculous glare and if I laughed (always, because it was hilarious) she concluded that I must be lying. Ironically, she was invariably wrong because I was such a good liar that she never suspected actual lies. She only applied this test to implausible-but-true things, and therefore managed a 100% wrong rate. She'd have been more accurate flipping coins.


Nah, it works well. Sit both kids down on the sofa, and ask, "Who did such-and-such?", and if one of them smiles, bingo.

[gnome]

An interesting article (http://www.simonyi.ox.ac.uk/dawkins/WorldOfDawkins-archive/Dawkins/Work/Articles/1997-11-16trialbyjury.shtml) by Richard Dawkins regarding juries, which seems to touch on similar ground to the study's supposition that judges are at least more consistent than juries. It begins with typical bluntness:

It's tricky, because having juries is a safeguard against a bribed, politically connected, or otherwise influenced judge.

[Thanz]

An interesting article (http://www.simonyi.ox.ac.uk/dawkins/WorldOfDawkins-archive/Dawkins/Work/Articles/1997-11-16trialbyjury.shtml) by Richard Dawkins regarding juries, which seems to touch on similar ground to the study's supposition that judges are at least more consistent than juries. It begins with typical bluntness:
Dawkins article is wrongheaded. He assumes that the principal of juries to find twelve truly independent assessments of the evidence, and that simply isn't the case. If it were, there would be no deliberations of juries at all. They would simply cast their votes in isolation of each other and be done with it. He says in his article "Perhaps there is something to be said for the collective wisdom that emerges when a group of twelve people thrash out a topic together, round a table." and then dismisses it saying "But this still leaves the principle of independent data unsatisfied."

Well, the collective wisdom is the point of the jury system. Not the principle of independant data.

If all they wanted was independent data, they could have trial by twelve gull herring chicks.

[brodski]

Dawkins article is wrongheaded. He assumes that the principal of juries to find twelve truly independent assessments of the evidence, and that simply isn't the case. If it were, there would be no deliberations of juries at all. They would simply cast their votes in isolation of each other and be done with it. He says in his article "Perhaps there is something to be said for the collective wisdom that emerges when a group of twelve people thrash out a topic together, round a table." and then dismisses it saying "But this still leaves the principle of independent data unsatisfied."

Well, the collective wisdom is the point of the jury system. Not the principle of independant data.

If all they wanted was independent data, they could have trial by twelve gull herring chicks.

It goes further than that, trial by jury is not designed just so that (so far as is reasonably practicable) justice is done, they are also designed so that it is seen to be done. They are a truly democratic institution, in that it is ordinary people, chosen at random from society who get to determine if someone has transgressed the rules of society- it is not a system easily manipulated by totalitarian impulses of states.

Any system of trial which replaces juries must weigh up its (presumably) superior truth finding abilities against the lack of corruption, abuse and the overall openness and sense of participation in justice which eth jury system helps to foster.

[LashL]

(removes lawyer hat)

On a personal level, I am not a big fan of juries. Frankly, and I know this will sound harsh, I think that for the most part, jurors are cabbages. Truth be told, if I was ever personally accused of a crime, I would want to have my trial adjudicated by a judge alone, unless I was actually guilty, in which case I would opt for a jury.