Exploiting the credibility of science

As a result of science’s great achievements in giving us reliable and accurate knowledge, it has tremendous credibility. Unfortunately, that increase in credibility has not been matched by a corresponding increase in the public’s awareness of how science acquires that knowledge and the limitations on it. This makes it possible to dazzle people with ‘science’.

John Oliver looks at one area where this has had very negative consequences and that is in the use of forensic evidence in criminal cases where overzealous prosecutors and less than careful forensic technicians can combine to result in gross miscarriages of justice, where juries and judges are swayed far too much by what they think is the certainty of the evidence presented to them and innocent people suffer.

I am always a little puzzled by those who are somewhat cavalier about the fact that innocent people end up in prison and have even been executed. They shrug it off, not seeming to realize that for every innocent person in prison, an actual guilty person is roaming free, able to commit more yet more crimes. You would think that that fact alone would make people want to be as sure as possible that the right person is found guilty.


  1. jrkrideau says

    The John Oliver piece is not available on this side of the lake so I probably am repeating points he made though I manage to find a close-captioned” summary and a blurry video of the CSI skit.

    Generally speaking, forensic science is an oxymoron. Most of the so-called forensic sciences have no theoretical basis or justification and there has been little to no research into the reliability or validity of the techniques. My impression is that the little “research” that has been done is usually of poor quality.

    The US National Research Council report “Strengthening Forensic Science in the United States: A Path Forward” (2009) https://www.nap.edu/catalog/12589/strengthening-forensic-science-in-the-united-states-a-path-forward is a fascinating read. The tone of stunned disbelief is amazing. The committee are astounded by the almost complete separation of “forensic sciences” from mainstream science and some of the claims made for the various “sciences”.

    If you have the time I’d recommend grabbing a copy and dipping into a couple of chapters. The polygraph is a good read.

    As far as I have been able to tell, the polygraph has no theoretical or empirical support and probably should be lumped in with homeopathy as pre-scientific quackery.

    The “REPORT TO THE PRESIDENT Forensic Science in Criminal Courts: Ensuring Scientific Validity of Feature -- Comparison Methods” Sept 2016 looks like an interesting update on this 2009 report but I have not read it.

    I think that there are several reasons why “who are somewhat cavalier about the fact that innocent people end up in prison and have even been executed”.

    I think the first one is that they still believe the trial verdict at some level, they may not have the scientific literacy to understand the issues and, in the USA, most of the wrongfully convicted are minorities.

    At the professional Justice level, it may be an unwillingness to admit the emperor has no clothes because once it is accepted that much or even most of the forensic evidence used in criminal trials has been pure crap it opens criminal justice systems around the world general discredit. How many prosecutors and judges want to think that they have sent shiploads of innocent people to incarceration or death?

    From a straight logistics and culpability point of view what does this do the the legal system? Who knows how many criminal convictions are tainted and open to appeal?

    Here is an instance in Boston area where one “rogue” lab chemist caused more than 21,000 low-level drug cases to be dropped. And here the science itself was not questioned. https://www.nbcnews.com/news/us-news/epic-drug-lab-scandal-results-more-20-000-convictions-dropped-n747891

  2. says

    Even “properly done science” doesn’t mean done correctly. A Taiwanese man, Chen Long-Qi, was falsely convicted of rape in 2012 via a commonly used DNA test. It was only after a legal defence group (and Chen going on the run to avoid prison) that people began to pay attention, that a full DNA test was done, proving his innocence. How many other false convictions have there been that didn’t require falsified evidence or coerced confessions, where “proper procedure” was followed?


  3. jrkrideau says

    # 3 Intransitive
    I’d argue that the example is a perfect example of (currently) valid scientific techniques applied incompetently. From a quick glance at the article, it looks like a formalistic application where the technician did not understand the basic science.

    How many other false convictions have there been that didn’t require falsified evidence or coerced confessions, where “proper procedure” was followed?

    Probably less than a googol (i.e. < 10^100) if we are being optimistic.

    Forensic pseudo-sciences combine very well with false memory (See work by Elizabeth Loftus or perhaps some of the newer work by Julia Shaw (https://blogs.scientificamerican.com/mind-guest-blog/what-experts-wish-you-knew-about-false-memories/)

    There is strong evidence that confessions are often false due to police misconceptions and faulty interviewing training. I don't believe that one can call the confessions " coerced" as in "with rubber hoses" but the results are the same or worse as the interogee may end up believing they had committed a crime. (Shaw' mentione'd above has some interesting lay articles on this probably listed here http://www.drjuliashaw.com/in-the-media.html)

    Eye-witness accounts often stink and many, possibly most, identity parades are a crap-shoot with loaded dice.

    And none of this covers sheer chance mistakes, typos in data bases, and so on.

    All of this happens when the police and other players in the system have the best of intentions. Of course, we must remember what road is paved with good intentions.

Leave a Reply

Your email address will not be published. Required fields are marked *