The opinion of the court was delivered by
Johnson, J.:A law enforcement officer participating in a saturation patrol near a Wichita bar stopped a vehicle driven by William J. Molitor and subsequently conducted a driving under the influence (DUI) investigation. After Molitor failed the horizontal gaze nystagmus (HGN) test but passed the walk-and-tum and one-leg-stand tests, the officer requested a preliminary breath test (PBT), the results of which ultimately led to Molitor s arrest and conviction for DUI.
At a subsequent suppression hearing conducted on appeal to the district court, tire court ruled that the HGN results could be admitted at that hearing to establish the officer s reasonable suspicion of DUI, even though the results were inadmissible at trial. The Court of Appeals affirmed that the HGN test could be used to establish the statutorily required reasonable suspicion of DUI that would permit a request for a PBT. City of Wichita v. Molitor, 46 Kan. App. 2d 958, 959, 268 P.3d 498 (2012). Additionally, the panel held that, even if the HGN test results were excluded, the officer had enough other evidence to form a reasonable suspicion of DUI. We granted review and reverse both the panel and the district court.
Factual and Procedural Overview
On the evening of February 28,2009, Officer Jeremy Diaz, while working with other officers on a traffic and DUI saturation patrol in Wichita, observed Molitor make a right turn at a stop sign without using the turn signal, albeit the officer noted that Molitor had made a complete stop at the sign, had turned appropriately into the correct traffic lane, and had driven straight down the street. The officer effected a vehicle stop based on tire turn signal infraction, and, according to the officer, as Molitor pulled over, his vehicle struck the curb and came to a stop with the tire hallway up the curb. Molitor claimed that he did not drive up on the curb but *253rather bumped into the curb because it was located on the edge of the road. The stop was not videotaped.
Diaz approached the vehicle and observed that Molitor’s eyes were watery and bloodshot and that a strong odor of alcohol was emanating from the vehicle. Diaz asked Molitor if he had been drinking, and Molitor responded that he had consumed two or three beers. Molitor’s speech was not slurred; he had no difficulty producing his driver’s license, insurance information, and vehicle registration; and he did not lose his balance while exiting his vehicle or walking thereafter. The officer continued to smell a strong odor of alcohol as Molitor exited the vehicle.
First, Officer Diaz administered the HGN test, recording that Molitor displayed six out of die six possible clues of intoxication. Next, Molitor scored one out of eight possible clues on the wallc- and-tum test and one out of four possible clues on the one-leg stand test. Both tests require two clues before the results are indicative of unlawful intoxication. Notwithstanding the passing scores on two of the standardized field sobriety tests (SFSTs), Diaz requested diat Molitor submit to a PBT. Molitor agreed to take the test and registered a breath alcohol content (BAC) of .090. After obtaining the PBT result, Diaz asked Molitor to take a trial-quality breath alcohol test, utilizing an Intoxilyzer 8000. This test was conducted about an hour after the initial stop and recorded a BAC of .091.
Molitor was charged and convicted in Wichita Municipal Court of DUI and failing to signal a turn. He appealed to the Sedgwick County District Court, and, prior to trial, moved to suppress the PBT and breath test results. Molitor argued that he had passed the only two “admissible NHTSA [National Highway Traffic Safety Administration] tests.” Therefore, he argued, the evidence did not support that the officer had the requisite reasonable suspicion to request the PBT.
At the suppression hearing, Officer Diaz testified that he had successfully completed training on administering the HGN test. Molitor’s attorney objected, claiming that Kansas caselaw holds that HGN test results are inadmissible in court for any reason. The district court overruled the objection, finding that although an *254HGN test result was inadmissible at trial, it could be used to support “probable cause.” At the conclusion of tire hearing, tire district court judge denied the motion to suppress, finding that under the totality of circumstances, there was reasonable suspicion to request the PBT.
Molitor filed a motion to reconsider, arguing that HGN testing is not admissible in Kansas pursuant to State v. Chastain, 265-Kan. 16, 960 P.2d 756 (1998), and State v. Witte, 251 Kan. 313, 836 P.2d 1110 (1992). The district court denied tire motion to reconsider and held that even though HGN test results were not admissible “in a court of law, it’s admissible for probable cause, it’s admissible for reasonable suspicion.” The district court also concluded that based on “all the circumstances, the driving, the breath, and the officer’s observation of tire defendant in the preliminary tests, that it was proper to request a preliminary breath test.”
Subsequently, Molitor agreed to a bench trial on stipulated facts, with the understanding that he could appeal tire denial of his motion to suppress the PBT and breath test. Based on the stipulated facts, the district court found that Molitor was guilty of DUI and failure to signal a turn. Molitor filed a timely appeal.
On appeal, Molitor argued that the district court abused its discretion by failing to follow binding Kansas Supreme Court precedent holding that evidence of PIGN testing is inadmissible for any purpose. He also asserted that the district court abused its discretion by failing to properly analyze the arresting officer’s opinion testimony pursuant to the provisions of K.S.A. 60-456. As a consequence, Molitor claimed the erroneous admission of the HGN evidence was prejudicial by depriving him of his due process right to a fair and impartial hearing on his motion to suppress.
The Court of Appeals panel first determined that no binding Kansas Supreme Court cases “directly address the issue of whether PIGN evidence may be considered prior to trial as part of the totality of die circumstances in determining if a law enforcement officer had reasonable suspicion to request a PBT.” Molitor, 46 Kan. App. 2d at 963. The panel found that while there was still considerable debate throughout other jurisdictions as to whether HGN test results could be admissible at trial, it was unable to find *255any authority from other jurisdictions holding that HGN test results could not be considered for the purposes of determining probable cause in a DUI case. 46 Kan. App. 2d at 965. The panel concluded that because reasonable suspicion is a less demanding standard than probable cause, “HGN test results may, under appropriate circumstances, be considered as part of the totality of the circumstances in determining whether a law enforcement officer has reasonable suspicion to request a PBT.” 46 Kan. App. 2d at 965.
Interestingly, the panel then essentially rendered its HGN discussion superfluous dictum by proceeding to find that there was “sufficient evidence in the record to support the district court’s conclusion that Officer Diaz had reasonable suspicion” to request the PBT, even without the HGN test results. 46 Kan. App. 2d at 966. Finally, the panel concluded that in light of Diaz’ testimony regarding his successful completion of HGN test training and Mol-itor’s failure to challenge Diaz’ qualifications below, the district court did not abuse its discretion in admitting the officer’s HGN testimony pursuant to K.S.A. 60-456, the provision governing the admission of opinion testimony. 46 Kan. App. 2d at 968.
Molitor filed a timely petition for review, arguing that the Court of Appeals erred in holding that HGN test results are admissible at a suppression hearing. He claims that the test for reliability set forth in Frye v. United States, 293 F. 1013 (D.C. Cir. 1923) (Frye test) had to be met before HGN test results could be considered for any purpose, including a determination of whether reasonable suspicion existed to request a PBT. Molitor also sought review of the Court of Appeals’ determination that the requisite reasonable suspicion existed without considering the HGN test results. This court granted Molitor’s petition for review under K.S.A. 20-3018(b), obtaining jurisdiction under K.S.A. 60-2101(b).
Admissibility of Horizontal Gaze Nystagmus Test Results
The first question we address is whether evidence of the HGN test results was erroneously considered by the district court at the pretrial suppression hearing to make the determination that the *256arresting officer possessed reasonable suspicion to believe that Molitor had been operating a vehicle while under the influence of alcohol, which was a statutory prerequisite for the officer to request the PBT. See K.S.A. 2010 Supp. 8-1012(b) (law enforcement officer may request PBT if officer has reasonable suspicion to believe person guilty of DUI). In his petition for review, Molitor argued that both the district court and Court of Appeals violated the duty for lower courts to follow the precedent of the Kansas Supreme Court.
In a supplemental brief to this court, a subsequently appointed attorney for Molitor took the tack that it was procedural error for the district court to admit the HGN results, because K.S.A. 60-402 makes the same rules of evidence applicable to the pretrial suppression motion hearing as are applicable at the trial. Therefore, under that argument, given that the HGN results are inadmissible at trial, they must also be inadmissible at the pretrial suppression hearing as a matter of statutory procedure.
But we view the issue as more fundamental than construing whether the statutory rules of evidence permitted introduction of the HGN results at a pretrial suppression hearing. For instance, a prior panel of the Court of Appeals found that it was not an abuse of discretion for a district court to admit HGN evidence at a bench trial, notwithstanding its unreliability, because of the presumption that judges, unlike juries, would not be unduly swayed by inadmissible evidence. State v. Ruth, No. 101,209, 2009 WL 3428611, at *3 (Kan. App. 2009) (unpublished opinion). In other words, we are not so much concerned with whether the evidence was procedurally admissible at a particular hearing as we are with the overarching question of whether HGN testing is competent evidence that can be relied upon when determining the existence of reasonable suspicion to believe that a vehicle driver was driving under tire influence, regardless of whether that determination is being made at the scene, at a suppression hearing, or at trial. Cf. State v. Shadden, 290 Kan. 803, 819, 235 P.3d 436 (2010) (“In addition to considering K.S.A. 60-456, a district court must determine whether the Frye test has been met if an opinion is based on sci*257entific methods or procedures and is offered for admission.” [Emphasis added.]).
Accordingly, we will leave for another day any consideration of tire general question as to whether the evidentiary rules at a suppression hearing are more relaxed than those at trial and proceed to consider the Court of Appeals’ holding that HGN testing results may be considered as part of the totality of the circumstances in determining whether a law enforcement officer had the requisite reasonable suspicion to request a PBT. Molitor, 46 Kan. App. 2d at 965.
Standard of Review
The issue before us requires that we review the district court’s legal conclusions, which is a de novo exercise. See Martinez v. Milburn Enterprises, Inc., 290 Kan. 572, 579, 233 P.3d 205 (2010). Moreover, whether tire district court failed to correctly apply the Frye standard for the admissibility of scientific evidence is an abstract question of law subject to de novo review. Kuhn v. Sandoz Pharmaceuticals Corp., 270 Kan. 443, 455-56, 14 P.3d 1170 (2000). More specifically, tire question of whether HGN test results are competent evidence of unlawful intoxication is a question of law. Cf. State v. McClanahan, 212 Kan. 208, 211, 510 P.2d 153 (1973) (competency of evidence is question of law).
Analysis
As a statutory condition precedent to requesting that Molitor submit to a PBT, Officer Diaz had to possess “reasonable suspicion to believe [Molitor] has been operating or attempting to operate a vehicle while under the influence of alcohol or drugs or both alcohol and drugs.” K.S.A. 2010 Supp. 8-1012(b). In other words, to request that a driver submit to a PBT to aid in the establishment of the probable cause necessary to arrest the driver for DUI, an officer must have already acquired a reasonable suspicion that the driver was DUI. We have described reasonable suspicion and its relationship to probable cause as follows:
“ ‘Reasonable suspicion means a particularized and objective basis for suspecting the person stopped is involved in criminal activity. Something more than an unparticularized suspicion or hunch must be articulated. Reasonable suspicion *258can arise from information that is less reliable than that required to show probable cause. Both reasonable suspicion and probable cause are dependent upon the content of information possessed by the detaining authority and the information’s degree of reliability. Quantity and quality are considered in the totality of the circumstances—the whole picture that must be taken into account when evaluating whether there is reasonable suspicion.’ State v. Toothman, 267 Kan. 412, Syl. ¶ 5, 985 P.2d 701 (1999).” (Emphasis added.) State v. Tollman, 286 Kan. 881, 890, 190 P.3d 234 (2008).
The Court of Appeals was persuaded by cases from a number of sister states that have held HGN test results “to be properly considered as a factor in determining probable cause.” 46 Kan. App. 2d at 964. Then, “[bjecause reasonable suspicion is a less demanding standard than probable cause,” the panel made new law in this state, finding “that testimony from a law enforcement officer trained in administering HGN tests may properly be considered as part of the circumstantial evidence used prior to trial in determining whether the totality of circumstances show that a law enforcement officer had reasonable suspicion to request a PBT.” 46 Kan. App. 2d at 965. Given the binding precedent that existed in this state, the panel erred in seeking guidance elsewhere. See State v. Jones, 44 Kan. App. 2d 139, 142, 234 P.3d 31 (2010), rev. denied 292 Kan. 967 (2011) (Court of Appeals duty bound to follow Kansas Supreme Court precedent).
The Court of Appeals recognized the existence of binding Kansas precedent dealing with the use of HGN testing, specifically citing to Shadden, 290 Kan. at 821-22; Chastain, 265 Kan. at 22-23; and Witte, 251 Kan. at 322. The panel observed that in Witte, “the Kansas Supreme Court distinguished [the] HGN test[] from other field sobriety tests because it is based on scientific principles outside the scope of common knowledge.” Molitor, 46 Kan. App. 2d at 962. The panel further acknowledged that, because of that distinction, the Witte court required that the foundational requirements of Frye must be met before HGN test results can be used. But the panel appears to have narrowed the scope of Witte’s foundational requirement to only apply “before HGN test results may be admitted at trial.“ (Emphasis added.) 46 Kan. App. 2d at 962. A closer review of Witte belies the panel’s narrow construction of its holding.
*259When Witte was decided, a person committed the crime of DUI under K.S.A. 8-1567(a)(l) by operating or attempting to operate a vehicle with a blood or breath alcohol concentration of .10 or higher. The opinion noted that NHTSA claimed that “the HGN test is an accurate and effective field sobriety test to determine whether a drivers alcohol concentration is above .10.” 251 Kan. at 315.
Witte then discussed HGN testing and scoring, beginning with a description of nystagmus as “ ‘an involuntary rapid movement of the eyeball, which may be horizontal, vertical, rotatory, or mixed/ ” and defining HGN as “ ‘a jerking of the eyes as they gaze to the side/ ” 251 Kan. at 316. In the field, the trained officer holds an object, such as a pen or a finger, about 12 to 15 inches in front of and level with the driver’s eyes. The driver is to keep his or her head steady and use his or her eyes to follow the object as the officer moves it to the side. The first side movement is all of the way until the driver’s eyes can go no further sideways. Then the movement is repeated from the front to a point that the officer estimates is a 45-degree angle of gaze. The test is performed to both sides, i.e., the left and right eyes are tested separately.
In scoring the test, the officer looks for three possible signs of intoxication for each eye, for a total of six clues. The first intoxication sign—the angle of onset of nystagmus—is premised upon the theory that the more intoxicated a person becomes, the sooner the jerking will occur during the eye’s sideward movement. The NHTSA asserts that the expected angle of onset when the driver’s BAC is .10 is approximately 40 degrees.
The second HGN intoxication sign involves observing how distinct the nystagmus is at maximum deviation, i.e., at the point where the eye is as far to the side as possible. Presumably, this means that the officer must assess the level of jerldness at maximum deviation because the theory is that the jerking will increase in intensity as the level of intoxication increases.
The third sign is smooth pursuit, i.e., the officer assesses the smoothness with which die driver’s eye pursues the object as die officer moves it sideways. This assessment is based upon the sup*260position that the eyes of an intoxicated person often cannot smoothly follow a slowly moving object.
One point is assigned for each clue of intoxication, so diat failing all possible clues earns a score of 6 points. According to NHTSA, a score of 4 or more points indicates a BAC above .10.
Witte’s complaint on appeal was that the district court had erroneously denied his motion in limine to prevent the State from presenting evidence of the HGN results. He claimed that the HGN test is scientific evidence; that the State had failed to establish that such scientific evidence met the Frye test; that the officer had not properly conducted the testing; and that the HGN test is simply not scientifically reliable evidence. The Witte court declared that the questions of whether the HGN test is scientific evidence and whether it meets the Frye admissibility requirements were issues of first impression in Kansas at that time. 251 Kan. at 318.
With respect to the scientific evidence question, Witte considered and rejected the State’s argument, apparently adopted by some other jurisdictions, that the HGN test only involves the officer’s objective personal observation of the driver’s conduct, much the same as the one-leg stand test, and therefore does not require expert interpretation. Witte noted that alcohol’s effect on a person’s sense of balance is common knowledge but that the same could not be said for the principles underlying the HGN test. Accordingly, given that the HGN is based upon scientific principles that exceed common knowledge, Witte held that HGN test results are scientific evidence subject to the Frye foundation requirements. 251 Kan. at 322.
The Witte court then considered the State’s argument that it did not have to establish tire reliability of HGN evidence through expert testimony in this state because other jurisdictions had recognized HGN evidence as being reliable under tire Frye test. After reviewing decisions in other jurisdictions—principally Arizona and Louisiana—that had found HGN testing to be scientifically reliable, the Witte court turned to a discussion of the results of its own research, a significant portion of which called into question the scientific bona fides of the HGN test. Indeed, one cited commentator referred to the HGN as “voodoo science.” 251 Kan. at 326 *261(citing to Pangman, Horizontal Gaze Nystagmus: Voodoo Science, 2 DWI Journal 1, 3-4 [1987]).
One area of concern strikes at tire heart of the HGN theory, i.e., tire angle of nystagmus onset. Although the NHTSA maintains that observing nystagmus at the 45-degree angle correctly foretells a .10 BAC 78% of the time, other researchers dispute that 45 degrees is the appropriate angle of onset. For instance, one authority asserts that 50% to 60% of sober individuals who deviate their eyes more than 40 degrees to the side will exhibit nystagmus that is indistinguishable from alcohol gaze nystagmus. 251 Kan. at 327 (citing Pangman, 2 DWI Journal at 2 [citing Toglia, Electronystagmog-raphy: Technical Aspects and Atlas (1976)]). Accordingly, “[r]esearchers have expressed concern that the 45-degree angle used by the NHTSA will create false positive readings.” 251 Kan. at 328. Some have even criticized the NHTSA study for deliberately screening out persons at high risk for being classified as a false positive and for conducting its tests with mechanical devices that hold the person’s head steady while precisely measuring the angle of lateral deviation of the eye. Of course, in the field, the driver is merely told to hold his or her head steady and the officer estimates the point at which the eye has reached a 45-degree angle.
Another concern addressed was that “many other factors can cause nystagmus,” such as suffering from such innocuous conditions as influenza or eyestrain; or consuming such common commodities as caffeine, nicotine, or aspirin. 251 Kan. at 328. Even “[a]n individual’s circadian rhythms (biorhythms) can affect nys-tagmus readings—the body reacts differently to alcohol at different times of the day.” 251 Kan. at 328.
Perhaps most compelling was the research study done by “[a] prosecution-oriented group in California,” measuring the correlation between a police officer’s estimations of the angle of onset of nystagmus and the actual results of chemical testing of blood samples (as opposed to breath samples). 251 Kan. at 329. Quoting from Pangman, 2 DWI Journal at 3, Witte recited that “ ‘[t]he data in the study revealed that there was virtually no correlation between the actual value of blood alcohol concentration and the predicted value based upon the angle of onset of nystagmus.’ ” 251 Kan. at *262329. After noting that the study group conceded that the HGN should not be used to predict a person’s blood alcohol level, Witte concluded as follows:
“If the Arizona Supreme Court had had this evidence before it, it may not have held that HGN evidence satisfies the Frye admissibility requirements. The reliability of the HGN test is not currently a settled proposition in the scientific community. This court holds that HGN evidence requires a Frye foundation for admissibility. If the Frye foundation is established to this court’s satisfaction, HGN evidence wiE be admitted in other cases without the need to satisfy tire Frye test each time. Before this court rules on whether HGN evidence satisfies the Frye admissibility requirements, a trial court first should have an opportunity to examine, weigh, and decide disputed facts to determine whether the test is sufficiently reliable to be admissible for any purpose in Kansas.” (Emphasis added.) 251 Kan. at 229-30.
Although the case involved the admission of HGN evidence at trial, the Witte opinion did not limit its holding to that scenario. To the contrary, the opinion plainly informed judges and prosecutors that the first tiling that had to happen before any court in this state could admit HGN evidence for any purpose was that the State had to present an appropriate Frye foundation to a trial court which would then be convinced to find that the HGN test was sufficiently reliable to be admissible. Notwithstanding the passage of more than two decades since Witte’s direction, the State has yet to follow the procedure outlined in Witte as being necessary to establish the reliability of the HGN test. Indeed, we are unaware of any proceeding in which the reliability of the HGN has been established under any standard. See Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993) (establishing alternative to Frye test utilized in federal courts); see also K.S.A. 2014 Supp. 60-456(b) (establishing criteria for admitting scientific opinion testimony).
Notwithstanding Witte’s clear directive that a trial court must first “examine, weigh, and decide disputed facts to determine whether the [HGN] test is sufficiently reliable to be admissible for any purpose in Kansas,” State v. Witte, 251 Kan. 313, 330, 836 P.3d 1110 (1992), the Court of Appeals made the curious declaration that “[i]t is important to note that tire issue presented in this case is not whether HGN testing has now reached the level of *263acceptance to satisfy the Frye test.” City of Wichita v. Molitor, 46 Kan. App. 2d 958, 963, 268 P.3d 498 (2012). But, of course, that is precisely the issue in this case, and any case in which the HGN test is involved, until it is proved rehable.
Granted, as suggested above, reasonable suspicion can be established with evidence that is less reliable than that which is required to establish guilt beyond a reasonable doubt or even to establish probable cause. But there is a threshold level of reliability that must be met. One must show that any proffered evidence that is ostensibly based upon scientific principles does, in fact, have some credible correlation to the matter that must be proved. For instance, consider the hypothetical scenario of an officer who testified that tire officer had undergone extensive training in the operation of a Ouija Board; that when a Ouija Board is asked if the driver being tested is DUI, the Board’s arrow will point at “yes or “no”; that random sampling has shown that the Ouija Board correctly identifies when a driver’s intoxication exceeds the legal limit 60% of the time; and that the Board’s arrow pointed at “yes” when asked if Molitor was DUI. Should a court allow the officer to base reasonable suspicion upon the Ouija Board test results? Of course not. And at this point in the state of Kansas, the HGN test has no more credibility than a Ouija Board or a Magic 8 Ball. To change that circumstance, the State needs to prove the legitimacy of the test, as Witte directed.
The panel also declared that it was not deciding the question of whether HGN evidence should be admissible to prove “a specific BAC.” But it is important to keep in mind that the officer must reasonably suspect unlawful activity and it is not unlawful to simply drink and drive. Rather, in order to request a PBT, the officer must have “reasonable suspicion to believe the person has been operating or attempting to operate a vehicle while under the influence of alcohol. . . .” (Emphasis added.) K.S.A. 2010 Supp. 8-1012(b). To be operating a vehicle under the influence of alcohol, pursuant to K.S.A. 2010 Supp. 8-1567(a)(l) or (2), the alcohol concentration in the person’s blood or breath must have reached the level of .08 or more. If, as the above-referenced California study concluded, the HGN is essentially useless in predicting a person’s blood al*264cohol level, then it is difficult to understand how that test can provide reasonable suspicion that a driver was driving under the influence of alcohol, as opposed to driving after two beers.
Accordingly, we hold that the district court and the Court of Appeals erred in allowing the State to rely on the scientifically unproved HGN test results to establish the requisite reasonable suspicion that permitted the officer to request that Molitor submit to a PBT pursuant to K.S.A. 2010 Supp. 8-1012(b). Such a fundamental error cannot be deemed to be harmless, unless the other evidence was sufficient to establish the requisite reasonable suspicion without considering the HGN test results, i.e., unless the panel’s opinion on the use of the HGN test results was merely judicial dictum. See Black’s Law Dictionary 549 (10th ed. 2014) (judicial dictum is opinion “that is not essential to the decision”); see also Law v. Law Company Building Assocs., 295 Kan. 551, 564, 289 P.3d 1066 (2012) (“ ‘Nobody is bound by dictum . . . .’ ”).
Harmless Error
After having determined that the HGN test results were appropriately considered as part of the totality of the circumstances that supported the finding that Officer Diaz had reasonable suspicion to believe that Molitor was operating his vehicle while under the influence of alcohol, the panel then embarked on an analysis of the hypothetical question of whether the evidence would have been sufficient to establish reasonable suspicion if the HGN test results had been excluded. If we were to agree with the panel’s use of the HGN test results as part of the totality of the circumstances affecting the reasonable suspicion analysis, we would forego consideration of the academic question of whether excluding the HGN test would change the result. But given our exclusion of the HGN testing, we must proceed to analyze the other evidence to determine whether the consideration of the HGN test results was harmless error.
Standard of Review
Whether reasonable suspicion exists is a question of law, and appellate courts review this question with a mixed standard of re*265view, determining whether substantial competent evidence supports the district court’s factual findings, while the legal conclusion is reviewed de novo. State v. Thomas, 291 Kan. 676, 688, 246 P.3d 678 (2011).
Analysis
Molitor argues that the district court and the Court of Appeals ignored tire evidence which indicated that he was not impaired by alcohol. The record indicates that the lower courts did mention the exculpatory evidence, but it appears that it was not fully integrated into the totality of the circumstances calculus.
The Court of Appeals listed “the factors supporting reasonable suspicion” as being “striking the curb, very strong odor of alcohol, bloodshot and watery eyes, admission to drinking beer, losing balance during instruction phase of walk-and-turn test, and putting foot down on the one-leg-stand test.” Molitor, 46 Kan. App. 2d at 967. The panel summarily dismissed the exculpatory evidence, as follows:
“We note that there is evidence in tire record that Molitor was able to speak without slurring his words, produced his identification without difficulty, and had only one clue each on tire walk-and-turn test and tire one-leg-stand test. But we do not find that these factors substantially dissipated Officer Diaz’ reasonable suspicion that Molitor had operated a vehicle under the influence of alcohol.” 46 Kan. App. 2d at 967.
After die panel filed its opinion in this case, this court decided State v. Edgar, 296 Kan. 513, 294 P.3d 251 (2013), which involved the question of die role that passing grades on field sobriety tests should play in the analysis of whether the investigating officer possessed the requisite reasonable suspicion to request a PBT. Edgar clarified that “[wjhether a law enforcement officer has the statutorily required reasonable suspicion to request a preliminary breath test is determined by examining the totality of the circumstances existing at the time of the request” and that the driver’s performance on field sobriety tests given before the PBT request is a circumstance that must be included in the totality of circumstances examination. 296 Kan. 513, Syl. ¶ 2.
*266Here, the panel correctly stated that the totality of the circumstances paradigm was applicable, and it appeared to grasp the essence of that test when it declared: “ ‘Quantity and quality are considered in the totality of the circumstances—the whole picture that must be taken into account when evaluating whether there is reasonable suspicion.’ See State v. Toothman, 267 Kan. 412, Syl. ¶ 5, 985 P.2d 701 (1999).” Molitor, 46 Kan. App. 2d at 966. But then, rather than looking at die whole picture to make a single, totality-of-the-circumstances reasonable suspicion determination, the panel broke the analysis into two parts. It first determined that the inculpatory factors it had identified would have justified an officer’s reasonable suspicion, and then it assessed whether the acknowledged exculpatory factors “substantially dissipated” the previously formed reasonable suspicion. 46 Kan. App. 2d at 967. But in exercising the totality of circumstances test for reasonable suspicion, an appellate court should not engage in “ ‘assessing each factor or piece of evidence in isolation. [Citations omitted.]’ ” United States v. Jones, 701 F.3d 1300, 1315 (10th Cir. 2012). The determination that reasonable suspicion existed obtains only after the interaction of all factors is assessed.
Before discussing the interaction of the factors present in this case, we pause to reiterate that, under the applicable version of the statute, Officer Diaz had to reasonably suspect that Molitor was illegally driving his vehicle while under the influence of alcohol before he could request a PBT. K.S.A. 2010 Supp. 8-1012(b). Interestingly, an earlier version of K.S.A. 8-1012 permitted a law enforcement officer to request a PBT based upon “reasonable grounds to believe that the person: (a) Has alcohol in the person’s body; . . . .” K.S.A. 8-1012 (Furse 2001). Although “reasonable grounds” was equated with probable cause (not reasonable suspicion) under the prior statute, the focus of the inquiry was not whether the driver was operating the vehicle under the influence, but rather whether the driver “had alcohol in her body.” Gross v. Kansas Dept. of Revenue, 26 Kan. App. 2d 847, 849, 994 P.2d 666, rev. denied 269 Kan. 932 (2000).
Logically, then, an officer’s subjective observations that the driver smelled of alcohol, or had bloodshot and wateiy eyes, would *267be more compelling evidence where tlie matter to be proved was simply that the driver had alcohol in his or her system rather than where the question is the legality of the alcohol concentration in the drivers body, i.e., whether it had reached the level of .08 or more. Likewise, the nature of the drivers admission to having drunk two or three beers is different for the two inquiries. Under the old statute, it would be compelling evidence that the driver had alcohol in his or her body, whereas, under the current statute, it would be evidence that tends to refute the notion that tire driver was operating the vehicle with an illegal level of alcohol in his or her body, i.e., it is questionable whether two or three beers would raise the alcohol concentration in the breath or blood of a normal size man to .08 or more.
Moreover, an officer s sensory perceptions, such as the strength of the alcohol odor or the condition of the driver’s eyes, are subject to an imprecise personal opinion. Moreover, that subjective assessment might be influenced by the subsequent discovery that the driver failed the PBT. Indeed, the California study on the HGN test discussed in Witte frankly reported that “ ‘the cops fudged the horizontal gaze nystagmus determination to correspond with the already known correct answer determined by the breath test result.’ ” 251 Kan. at 329 (quoting Pangman, 2 DWI Journal at 3).
In contrast, the SFSTs were developed by the NHTSA after both laboratory studies and field studies, from which clues were identified and a scoring criteria developed that would provide an objective assessment as to the probability that the driver’s alcohol concentration was at an unlawful level (.10). See Rubenzer, The Standardized Field Sobriety Tests: A Review of Scientific and Legal Issues, 32 Law & Hum. Behav. 293 (2008). For instance, the arresting officer in the Shadden case testified at trial that if a driver exhibits two clues, he or she fails the SFST, creating a 68% probability that the driver’s concentration of alcohol is .10 or more. State v. Shadden, 290 Kan. 803, 806-07, 235 P.3d 436 (2010). In other words, SFSTs are alleged to result in an objective assessment of the level of alcohol in a driver’s body, rather than just the presence of alcohol in the body.
*268Granted, the officer here testified that Molitor ran into or onto the curb while stopping his vehicle. Obviously, evidence of unsafe driving can suggest intoxication. But that alleged lapse of coordination must be viewed in conjunction with what followed. After stopping the vehicle, Molitor spoke without slurring his words, produced his identifying documents without difficulty, exited and proceeded from his vehicle without losing his balance, and, most importantly, passed the two admissible SFSTs. In other words, under the totality of circumstances, one could not reasonably suspect that Molitor’s balance was impaired by alcohol to the point of being legally under the influence of alcohol.
Moreover, in Pollman, this court set a low bar for the observable indicia of intoxication that can support reasonable suspicion, noting only the smell of alcohol and the driver’s admission to having drunk alcohol, in addition to the acts leading to the criminal obstruction of official duty charges. State v. Pollman, 286 Kan. 881, Syl. ¶ 7, 190 P.3d 234 (2008). But here, tire subjective observations which might suggest to Officer Diaz that Molitor was illegally intoxicated were offset by the objective indications that he was not. Indeed, if Molitor had failed the objectively scored SFSTs, one would suspect that the State would be arguing that the officer’s trained observations were corroborated by the psychomotor testing.
Curiously, the panel padded its description of the intoxication indicia by referring to the one clue on each SFST to which the officer testified. But the officer admitted that Molitor passed the tests, and we have nothing in die record which would tell us what one clue reveals about a person’s alcohol concentration level. Indeed, “[sjeveral studies suggest that cut-off scores are set too low on the psychomotor SFSTs,” and one study “found that over 50% of drivers at .00% BAC failed Walk and Turn.” Rubenzer, 32 Law & Hum. Behav. at 297. The panel should not have deviated from the criteria and scoring of the NHTSA’s standardized testing model to glean reasonable suspicion of DUI from a successful completion of the admissible SFSTs.
In short, we reverse the determinations of both the district court and the Court of Appeals that Officer Diaz possessed the requisite reasonable suspicion that Molitor was operating his vehicle while *269under the influence of alcohol when the officer requested that Molitor submit to a PBT.
Reversed and remanded.
Michael J. Malone, Senior Judge, assigned.* * *