Challenging the Science: The Kelly Reliability Hearing

Unlawful Is Not the Same as Unreliable

These two arguments get confused constantly, so start here.

A motion to suppress is about the government’s conduct. Did the police have the legal authority to stop you, arrest you, or take your blood? If they did not, the evidence is excluded no matter how accurate it turned out to be. A flawless laboratory result from an unlawful arrest is still suppressible.

A reliability challenge is about the evidence itself. Even where every step was lawful, the State still has to show that the science behind its number is sound enough for a jury to hear. A blood draw taken under a valid warrant, by a qualified person, from a lawful arrest, can still be excluded if the analysis behind the result does not survive scrutiny.

Different questions, different standards, different remedies. A case can raise both, and many do. The lawfulness argument is covered on How a Motion to Suppress Works. This page is about the other one.

The Three Kelly Prongs

Texas used to evaluate scientific evidence under the Frye general acceptance standard, which asked whether the theory and technique were generally accepted as reliable in the scientific community. In 1992 the Texas Court of Criminal Appeals decided Kelly v. State, concluded that Frye was no longer part of Texas law, and replaced it.

Under Kelly, evidence derived from a scientific theory must satisfy three criteria in any particular case to be considered reliable:

• The underlying scientific theory must be valid.
• The technique applying the theory must be valid.
• The technique must have been properly applied on the occasion in question.

The court also listed factors bearing on reliability, and they are not exclusive: how far the relevant scientific community accepts the theory and technique as valid, the qualifications of the experts testifying, the existence of literature supporting or rejecting the theory and technique, the potential rate of error of the technique, the availability of other experts to test and evaluate it, how clearly the theory and technique can be explained to the court, and the experience and skill of the person who actually applied the technique on this occasion.

Note that last one. The person who ran the test is a reliability factor, not a side issue.

One more point that closes off the State’s most common answer. Prosecutors often respond that gas chromatography or infrared breath testing is not novel science, so no reliability showing is required. Texas rejected that. The admissibility criteria apply to all scientific evidence offered under Rule 702, because there is no value in having a different standard of admissibility for novel scientific evidence than for established evidence.

Who Bears the Burden

The party offering the evidence has to establish that it is reliable. In a DWI case that is the State, because the State is the one putting the number in front of the jury.

The standard is higher than the ordinary one. Because lay people have difficulty evaluating the reliability of scientific testimony, the burden of persuasion is enhanced: the proponent must persuade the trial court by clear and convincing evidence that the evidence is reliable and therefore relevant. Not by a preponderance. Clear and convincing.

There is a catch that matters more than anything else on this page. You have to ask for the hearing. The court does not screen the State’s science on its own initiative. Scientific evidence is subject to a Rule 702 analysis before it comes in, but the analysis happens because the defense requests it. A case where nobody asked is a case where the State’s number goes to the jury with no gatekeeping at all.

The Hearing Itself

The hearing is held outside the presence of the jury. The judge acts as gatekeeper, hears the testimony, and decides whether the jury gets to hear the evidence at all.

In practice the State calls the analyst, the technical supervisor, or the expert whose opinion it wants admitted, and the defense cross-examines on the three prongs. The defense can call its own expert. The judge can rule on some prongs and defer others.

A real case shows both how it works and how it goes wrong. In a retrograde extrapolation case, the defense asked the trial court for a Kelly gatekeeper hearing outside the jury’s presence on the State’s expert testimony. Because the defense asked, the State was required to prove reliability by clear and convincing evidence. The court found the first two prongs satisfied but did not rule on whether the technique had been properly applied, and offered to continue the gatekeeping hearing on that question. Defense counsel instead chose to let the experts testify in front of the jury with a running objection. The experts then told the jury the driver was above 0.08 at the time of driving.

On appeal, that failure to obtain an adverse ruling on the third prong meant the error was waived. The appellate court never reached the merits.

The lesson is not subtle. The reliability challenge is won or lost on the record, and a running objection is not a ruling.

The Third Prong Is Where DWI Cases Are Won

Attacking the underlying science is almost always the wrong fight. Texas courts have long since accepted the basic theory behind breath and blood alcohol testing, and a defense that opens by arguing gas chromatography does not work will lose the room.

The realistic posture is the third prong: this technique, on this sample, on this occasion, by this analyst.

Consider a reported case. A DWI defendant did not question the underlying scientific theory of gas chromatography at all. His argument was narrower and much harder to answer: the technique was invalid and improperly applied because gas chromatography had not been shown reliable for the sample actually tested. The alcohol analysis run roughly three and a half months after collection described the blood as normal but with a greenish-brown color. The trial court held a hearing outside the jury’s presence and let the evidence in, and the conviction was affirmed on appeal.

Read that as the shape of the argument rather than as a win. The State’s analyst answered that nothing in his training or experience indicated blood color affects alcohol concentration, that he had seen green blood fairly often, and that blood turns green when it is not refrigerated, which tends to lower the ethanol reading rather than raise it. On retesting, the reading in that case dropped by 0.009.

What the case demonstrates is where the leverage lives. Nobody had to disprove chromatography. The question was whether a method validated on properly stored blood was validly applied to a sample that had visibly changed in storage, and answering it required the analyst to account for the specific condition of the specific specimen.

What Feeds a Reliability Challenge

A third-prong challenge is only as good as the underlying facts about how the test was actually done. Those facts live on the forensic pages:

• The Blood Test covers the laboratory methodology and where it breaks down.
• In-Vitro Fermentation and Sample Integrity covers what happens to a specimen between the needle and the instrument.
• Gas Chromatography and the Chromatogram covers whether the peak the analyst called ethanol was only ethanol.
• Intoxilyzer 9000 Problems covers the breath instrument and its records.

The full evidence section is indexed on Tests and Evidence.

Motions in Limine

A reliability challenge asks the court to exclude a body of evidence. A motion in limine is the narrower tool: it asks the court to order that specific testimony not be mentioned in front of the jury until the court has ruled on whether it is admissible.

The difference is about damage control. Some evidence is harmful the moment a juror hears it, even if the judge later strikes it. A motion in limine gets the fight resolved before that happens, and in a DWI case it is typically filed to cover the roadside test testimony, statements taken without warnings, the Romberg balance test, portable breath test results, and any mention of a breath or blood number during opening statement before the result has actually been admitted.

The Emerson Limit on HGN

The horizontal gaze nystagmus test involves the science of how alcohol affects eye movement, which makes it scientific evidence subject to Rule 702. Two years after Kelly, the Texas Court of Criminal Appeals addressed it directly in Emerson v. State.

Emerson settled the first two prongs against the defense. The theory underlying HGN is sufficiently reliable, and the technique as designed and promoted by NHTSA is reliable. The third prong, whether the officer administered the test to protocol on this occasion, is open in every case, and it is worked through clue by clue on How to Challenge Field Sobriety Test Results.

What belongs on this page is the limine consequence, because it is the part the State crosses more often than it should. A witness qualified on the administration and technique of HGN may testify about how a defendant performed, but may not correlate that performance to a precise blood alcohol concentration. HGN is qualitative evidence that alcohol was consumed. It is not a number, and it cannot be converted into one.

The violation usually arrives indirectly. Officers are trained using field tests verified against breath instrument results, and testimony about that training can imply a relationship between clue count and a quantitative alcohol result without ever stating one. Keeping that implication out is a standard limine request. See The HGN Eye Test.

Portable Breath Test Results

The small handheld device an officer may use at the roadside is not the evidential breath instrument, and its result does not belong in a DWI trial.

Texas breath testing regulations require that instruments used for evidential purposes be approved or certified by brand or model by the scientific director of the Department of Public Safety, that reference sample devices used with evidential testing be approved, and that testing techniques, methods, and programs used for evidential purposes be approved and certified. A portable device does not meet that predicate. On top of that, the arresting officer cannot testify as an expert on the scientific reliability of the handheld unit or even explain how it distinguishes alcohol from other substances, which means the result also fails Rule 702 and Kelly.

The hearing on that request is grounded in Articles 28.01(6) and 38.22(6) of the Code of Criminal Procedure.

How the evidential instrument differs from the roadside device is covered on The Breath Test.

Getting the Material the Challenge Requires

No reliability challenge is possible without the underlying data. A third-prong argument is a document argument: the chromatograms, the calibration and maintenance records, the standard operating procedures, the analyst’s training file, the storage and refrigeration history, and the raw instrument output rather than the one-page report.

That material comes through discovery, and getting all of it is its own piece of work. See Discovery.

What It Takes to Run One

A reliability challenge is a scientific cross-examination conducted under a legal standard. It requires counsel who can read the underlying data rather than the summary, and it usually requires a defense expert to evaluate the State’s methodology and, where the challenge goes to the jury instead, to testify.

Two things are worth being direct about. First, credentials matter here in a way they do not in most of a DWI case, because the argument is made in the analyst’s own vocabulary. Deandra Grant Law has two ACS-CHAL Forensic Lawyer-Scientists on staff. See What to Look For in a DWI Lawyer.

Second, expert costs are billed separately from the legal fee. An independent analyst, a retest, or a testifying expert is an additional expense, and it should be discussed at the outset rather than discovered later. See What DWI Fees Cover.