The UC Davis Legal Studies Research Paper Series recently released an article entitled Revising State Post-Conviction Relief Statutes to Cover Convictions Resting on Subsequently Invalidated Expert Testimony by former Army JAG, and prominent evidence law scholar, Professor of Law Emeritus Edward J. Imwinkelried.  The article posits:

Invalid expert testimony has become a disturbingly recurrent theme in the wrongful conviction studies.

Id. at 3.

To remedy this situation, the article proposes that a statutory remedy ought to be created by which an accused convicted upon invalid expert opinion testimony might obtain a new trial:

It is submitted that an accused ought to be entitled to a new trial when: (1) the accused presents testimony about a new analytic technique developed in subsequent scientific research; (2) that technique yields a different outcome than the expert technique used at the prior trial; and (3) the validation of the new technique is so extensive that it either discredits the prior expert testimony or seriously undermines confidence in its correctness. The Constitution may not mandate a new trial in these circumstances. However, the American criminal justice is intensely concerned about the reliability of the evidence it relies on to justify convictions, and the proposed general standard operationalizes that concern.

Id. at 17.

After proposing his solution, Professor Imwinkelried’s article describes some of the practices which have given rise to the problem of invalid expert opinion testimony in American criminal trials.

First, the article notes that “[i]n a disturbing number of cases, postconviction investigation has revealed that witnesses called as experts have misstated—usually overstated—their credentials.” Id. at 19. By overstating their credentials, the article argues, those experts mislead the finder of fact to accept their opinions with less skepticism than would be directed towards a lay witness, or a less qualified expert.

The article also spends a good amount of time exploring the ways in which experts too often rely upon invalid presumptions or discredited methodology to reach their conclusions. But, one particular expert fallacy receives a bit more of Professor Imwinkelried’s attention:

Statements suggesting or implying greater certainty are not scientifically valid and should not be permitted. In particular, courts should never permit scientifically indefensible claims such as: “zero,” “vanishingly small,” “essentially zero,” negligible,” “minimal,” or “microscopic” error rates; “100 percent certainty” or proof “to a reasonable degree of scientific certainty;” identification “to the exclusion of all other sources;” or a chance of error so remote as to be a “practical” impossibility.

Id. at 25, quoting Report to the President – Forensic Science in Criminal Courts: Ensuring Scientific Validity of Feature -Comparison Methods, President’s Council of Advisors on Sci. and Tech. (2016) at 19. The article explains further:

The basic tenet of metrology is that no measurement is certain. No matter how carefully the analyst conducts the measurement and no matter how well calibrated the measuring instrument is, the analyst can never be certain that he or she has captured the true value of the measurand.

Id. at 26. The scientifically-sound way of presenting a scientific conclusion is to make sure it is articulated as an estimate accompanied by a confidence interval. Id.

The last section of Professor Imwinkelried’s article focuses on why current means of obtaining post-trial relief are insufficient when science is found to be shoddy after the trial is over. The difficulty is particularly prominent in the military justice system where there are no standing trial courts for an accused to seek post-trial relief from. Instead, military defendants must rely upon direct appeal – which may not even be available depending on the severity of the adjudged and approved sentence. Even if direct appeal is available, the standard for obtaining relief tends to be prohibitive. The extraordinary writ process is even more restrictive – not merely by its exceedingly high standard for obtaining relief, but also because the military appellate courts have a very narrow jurisdiction. Accordingly, the article recommends a statutory regime which would allow new trials in deserving cases where improper scientific evidence was permitted to slip pass the gatekeeper.

14 Responses to “Scholarship Saturday: The chronic problem of invalid expert opinion testimony”

  1. SgtDad says:

    Not before time.  The same problem exists, of course, in civil litigation, as well.  A former partner of mine described it thus: “Each side offering its trained seals babbling their own brand of poppycock.”

  2. stewie says:

    All science isn’t equal. The harder stuff isn’t “poppycock.” The softer stuff…

  3. Gloria Delgado-Pritchett says:


    The harder stuff isn’t “poppycock.” The softer stuff…

    I saw whatch you deed there, but jou misspelled “papi.”

  4. DCGoneGalt says:

    There is science and there is social science.
    The word “social” serves the same function as “quasi” does in law school.  In other words, social science isn’t quite science but we want it to seem like it is.

  5. DCGoneGalt says:

    Conclusions based upon scientific data can be abused, such as this post seems to focus on.  However, social science in courts-martial often seems like a Rorschact Test for an “expert” to find whatever they want to find.

  6. Alfonso Decimo says:

    So far, it seems we all share similar bad experiences with pseudo-experts. It seems much more likely the defense, rather than the government, will benefit post-trial when an expert’s basis is later debunked. For this reason, I think prosecutors should be cautious with their use of expert opinion, but the defense should be audacious.

  7. stewie says:

    I don’t think social science is “pseudo”-science. When done right and with the proper caveats, it’s good old run of the mill science.  But it’s not at the certainty level needed for a BRD standard in many, if not most cases.  So you do end up with a GOV expert who finds say a minor or not mental illness and a DEF expert who finds a serious mental illness and who the heck knows where the truth is.

  8. DCGoneGalt says:

    Science deals with the laws of the natural physical universe.  Social science deals with the behavior and decisions of human beings in their social relationships.  For instance, chemistry is a science because there is a right answer and a wrong answer to every question, whether we have yet discovered it or not.  Social science, ex: psychology, is not a science.  It may encompass elements of science, i.e. neuroscience or chemistry for medical treatments, but it is a series of hypotheses and broad concepts with nothing that is definite or repeatable under laboratory conditions because it is dealing with human beings as its subject matter.   
    The idea that “science” can be abused by humans in the investigation/trial process is because human beings introduce their own mistakes and biases into the process.  Social science is more susceptible to abuse, whether deliberate or not, because the instability of human mistakes is baked right into the cake right from the beginning.  It is why DNA is science and experts are left to argue over conclusions based on the evidence (or human mistakes in collecting data) and forensic psychology is a social science and you can have two experts who disagree on the basic concepts at the heart of the field. 

  9. stewie says:

    science is a methodology. It doesn’t deal with anything specifically vice something else. If you are following the scientific method, you are doing science. Social science does follow the scientific method. There is a hypothesis, it gets tested, refined or rejected, which ultimately leads to a working theory which then gets reviewed/challenged.
    The problem with social science is that it’s ability to accurately predict is more limited than some of the harder sciences.
    Because yes humans are more variable than say…rocks, or minerals, or stars.  But that doesn’t make it not science, it just makes it harder to definitively use in such a way as we try in trials with the BRD standard. It’s mighty fine for diagnostic and treatment regimens, etc.
    And I see less about two psychologists disagreeing on basic concepts, and much more about the application of those concepts to an individual human. Then again, a classical physicist thinks a string theorist is just talking gibberish.

  10. DCGoneGalt says:

    String theory, ugh, at least it’s not a holographic universe or multiverse spouting physicist.
    In other news, college campuses seem to be keeping up with the military in the battle to see who can win the Gold Medal in the Batsh!t Crazy Competition at the Sexual Assault Hysteria Olympics:
     USC Student Cleared of Rape, Could Still be Expelled

  11. SgtDad says:

    Roger that.

  12. stewie says:

    What’s wrong with multiverse theory? It explains a lot, it deals with the fact that gravity is weaker than it should be compared to the other fundamental forces, and Brane theory could explain the problems with the Big Bang theory.
    It’s as viable as most of the other big questions we probably will never get answers to at any rate.
    I’m with you on the holographic universe, although holographic theory could resolve the information loss paradox with black holes (with information remaining in 2D form on the event horizon as I understand it). That article does not indicate the school is going through with expelling him, simply that it could expel him because it doesn’t require the same level of proof as a trial court would.
    Not defending it, but a bit premature until they actually expel him.

  13. Vulture says:

    Perhaps we’ve been looking in the wrong place for the interesting questions Stewie.  I’m sure you’ll understand.  One day maybe you’ll turn over a rock and see “Made by god.”

  14. stewie says:

    If I find a rock with God(TM) on it, I’ll let you know…til then, I’m glad folks are using the brains the FSM gave them. ;)