Introduction
Charles Williams (“Williams”) appeals the trial court’s judgment, entered after a bench trial, convicting him on one count of possessing a controlled substance. Williams, a student at a public high school, was searched, by school officials and forced to remove items in his pockets after he arrived 30-minutes late to school. Among the items in Williams’s pocket was a substance with a cocaine base. Williams contends that the trial court should have suppressed all evidence relating to his possession of a controlled substance because the evidence was the fruit of an unreasonable search under the Fourth Amendment. Recognizing that we review the constitutionality of school searches under a more relaxed standard than otherwise is required under the Fourth Amendment, we nevertheless hold that the search of Williams violated his right to be free of an unreasonable search for two reasons. First, the school official conducting the search lacked any reasonable, individualized suspicion that Williams possessed contraband on his person. Second, the school’s policy to conduct a hand search of all late-árriving students was, ünder all the circumstances, unreasonable. Because the school official’s search, of Williams violated the Fourth Amendment, the trial court should have suppressed the resulting evidence. We reverse and remand for a new trial.
Factual and Procedural History
Williams was a student at Vashon High School (“Vashon”) when he was arrested there for possession of a substance with a cocaine base. The State charged Williams as an adult with one count of possessing a controlled substance—a class C felony un*692der Section 195.202.1 Before trial, Williams moved to suppress “any and all articles seized and intended to be used against [Williams], and any testimony regarding such evidence.” Williams waived his right to a jury trial. The trial court held a bench trial, which adjudicated both the suppression motion and Williams’s guilt.
At trial, the State’s main witness was DeAndre Duncan (“Duncan”), a school-safety employee at Vashon. Duncan’s duties were to “patrol the school and protect and check anyone that comes in.” Duncan knew Williams from prior interactions at the school. When asked about those interactions, Duncan stated, “Just the normal routine just get to class. Nothing negative.”
Duncan testified that, on the day of arrest, Williams arrived at school at least 30-minutes late with a “group of young men.” Duncan instructed the group, as they came through the front door, to remove their shoes and all items from their pockets. The students were then required to go through a metal detector. Williams passed through undetected. After the metal detector, Duncan “immediately” hand searched the students. When Duncan attempted to pat down Williams’s back pocket, Williams told him to stop. Suspecting that Williams had something “that didn’t belong” in school, Duncan brought Williams to the security office.
In the security office, Duncan demanded that Williams remove whatever was in his pocket. Williams removed a white “rocky substance” in a plastic wrapping. According to Duncan, Williams characterized this substance as “dope.” Duncan handcuffed Williams and called Vashon’s resource officer, DeAndre Davis (“Officer Davis”). Officer Davis worked for the St. Louis Metropolitan Police Department as a detective, but he was assigned to Vashon as the resource officer. Officer Davis arrived at the security office, seized the drugs,2 and took Williams into custody. Williams was moved to Officer Davis’s office, which was also at the school. Officer Davis started the booking process in his office before conveying Williams to the police station for formal booking. At some point during this process, Williams was read-his Miranda3 rights and stated that he understood those rights. According to Officer Davis’s testimony at trial, Williams said that he “found” the contraband and “never should have picked [it] up.”4
A central issue at trial was the constitutionality of the hand search made by Duncan when Williams first entered the school, and consequently whether evidence concerning Williams’s drug possession should have been suppressed as the fruit of an unreasonable search under the Fourth Amendment. Duncan testified that he searched Williams according to school policy. According to Duncan, all students were required to pass through a metal detector to enter the school. But for all students who were late to school by 30 minutes or more, school policy required an additional “hand check.” Duncan described this procedure vaguely:
A hand check entails removing all items from your pocket and even maybe shoes or—and to pull your pockets out and for *693me to check the front area of the body down the sleeves to pat the pockets and to—and even to the ankles, all the way down to the ankles.
The purpose for this policy, according to Duncan, was to protect the students, teachers, and staff. Duncan further stated that the reason for this hand-check policy addressed “[s]uspicion on late arrival.” When asked what was suspicious about late arrival, Duncan replied, “Because of the neighborhood we have a lot of drug activities and violence down in that neighborhood, so the school [Vashon] and the St. Louis Public Schools asked us to have a more detailed check on later arrival for students at the time.” Duncan stressed that his hand search of Williams when he entered the school merely followed the school’s policy.
After hearing the testimony,5 the trial court overruled the suppression motion. The trial court found that the school’s policy of suspicionless searches for all late-arriving students was justified because it was uniformly applied and done for safety purposes. The trial court also found that the search’s scope was reasonable because it was limited to the emptying of pockets and a pat-down—no strip search occurred. The trial court found the search to be reasonable.
The trial court then found Williams guilty. The trial court suspended imposition of the sentence, placing Williams on probation. Williams’s probation was later revoked and he was sentenced to three years in prison. Williams now appeals.
Points on Appeal
Williams raises three points on appeal. Point One argues that the trial court erred in overruling his motion to suppress evidence of the drugs found on his person because the hand search of Williams violated his Fourth Amendment rights. Point Two contends that the trial court erred by admitting, as evidence, Williams’s statements in the security office acknowledging the presence of “dope.” Point Three assigns error to the admission of Williams’s statements that he “found” the drugs and should not have picked them up.
Standard of Review
“When reviewing the trial court’s overruling of a motion to suppress, this Court considers the evidence presented at both the suppression hearing and at trial to determine whether sufficient evidence exists in the record to support the trial court’s ruling.” State v. Pike, 162 S.W.3d 464, 472 (Mo. banc 2005). We will reverse only if the ruling was clearly erroneous. State v. Loggins, 445 S.W.3d 105, 109 (Mo. App. E.D. 2014). We defer to the trial court’s factual findings and credibility determinations. Id. But, an alleged Fourth Amendment violation is a question of law. State v. Selvy, 462 S.W.3d 756, 764 (Mo. App. E.D. 2015). As such, the reasonableness of a public-school search under the Fourth Amendment is reviewed de novo. Doe ex rel. Doe v. Little Rock Sch. Dist., 380 F.3d 349, 352 (8th Cir. 2004).
Discussion
We begin our discussion with an overview of the Fourth Amendment’s application to searches of students by public-school officials. After analyzing Duncan’s search of Williams under two frameworks, we conclude that the search was unconstitutional. Finally, we determine that the evidence deriving from the unconstitutional search should have been suppressed as the fruit of a poisonous tree.
*694I. The Fourth Amendment in Public Schoóls
The Fourth Amendment to the United States Constitution provides that “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated....” U.S. Const, amend. IV. The Fourth Amendment, by virtue of the Fourteenth Amendment, applies to searches by public-school officials, as they are considered state actors. New Jersey v. T.L.O., 469 U.S. 325, 334, 105 S.Ct. 733, 83 L.Ed.2d 720 (1985) (citing Elkins v. United States, 364 U.S. 206, 213, 80 S.Ct. 1437, 4 L.Ed.2d 1669 (1960)).
“[T]he touchstone of the constitutionality of a governmental search” is “reasonableness.” Bd. of Educ. of Indep. Sch. Dist. No. 92 of Pottawatomie Cnty. v. Earls, 536 U.S. 822, 828, 122 S.Ct. 2559, 153 L.Ed.2d 735 (2002). Fourth Amendment rights in public schools are different from those rights possessed elsewhere, as the “reasonableness” inquiry cannot disregard the school’s tutelary responsibility for children. Vernonia Sch. Dist. 47J v. Acton, 515 U.S. 646, 656, 115 S.Ct, 2386, 132 L.Ed.2d 564 (1995) (“Vernonia”). Further, securing order in a public-school environment sometimes requires greater controls over students than those over adults. Earls, 536 U.S. at 831, 122 S.Ct. 2559. As such, the United States Supreme Court has dispensed with the need for public-school officials to obtain a warrant based on probable cause before a search at school, because the warrant requirement would “unduly interfere with the maintenance of the swift and informal disciplinary procedures [that are] needed.” Id. at 828-29, 122 S.Ct. 2559.
In the end, whether a public-school official’s search of a student is unconstitutional depends on the reasonableness of the search under all circumstances. T.L.O., 469 U.S. at 341, 105 S.Ct. 733; Vernonia, 515 U.S. at 652, 115 S.Ct. 2386; Earls, 536 U.S. at 828, 122 S.Ct. 2559. The United States Supreme Court has employed two different frameworks in analyzing the reasonableness of searches conducted by public-school officials.
The first framework applies when the search is based on some level of individualized suspicion that a student has violated the law or school rules. See T.L.O., 469 U.S. at 345-46, 105 S.Ct. 733 (search of a student’s purse after a teacher caught the student smoking in violation of school rules); Safford Unified Sch. Dist. No. 1 v. Redding, 557 U.S. 364, 373-74, 129 S.Ct. 2633, 174 L.Ed.2d 354 (2009) (strip seárch of student based on suspicion that she was distributing contraband). Under this first framework, the reasonableness of a search “involves a twofold inquiry”: (1) whether the action was “justified at its inception,” and (2) whether the search as actually conducted was “reasonably related in scope to the circumstances which justified the interference in the first place.” T.L.O., 469 U.S. at 341, 105 S.Ct. 733. The Supreme Court has required only reasonable suspicion—rather than probable cause—to justify a search, or a “moderate chance” of finding evidence of the wrongdoing suspected.6 Redding, 557 U.S. at 371, 129 S.Ct. 2633.
*695The second framework recognizes that individualized suspicion is not an irreducible requirement of the Fourth Amendment’s reasonableness inquiry. Vernonia, 515 U.S. at 653, 115 S.Ct. 2386. Thus, even absent individualized suspicion, a search might pass constitutional muster under a broad balancing test. The United States Supreme Court has applied the second framework twice, both in the context of suspicionless drug testing—under a specific school policy—of public-school students participating in extracurricular activities. See Vernonia, 515 U.S. at 654-664, 115 S.Ct. 2386; Earls, 536 U.S. at 830-838, 122 S.Ct. 2559, Under the Vernonia/Earls framework, a court must weigh the intrusion on the student’s Fourth Amendment rights against the promotion of legitimate governmental interests. Earls, 536 U.S. at 830, 122 S.Ct. 2559. Three factors guide this analysis: (1) “the nature of the privacy interest allegedly compromised,” (2) “the character of the intrusion imposed,” and (3) “the nature and immediacy of the government’s concerns and the efficacy of the [p]olicy in meeting them.” Earls, 536 U.S. at 830-838, 122 S.Ct. 2559; see also Vernonia, 515 U.S. at 654-664, 115 S.Ct. 2386. While we are aware of no Missouri case applying' the Vernonia/Earls framework, the Eighth Circuit has applied it to suspi-cionless searches conducted according to a school policy. See Doe, 380 F.3d at 354-57 (searching thé students’ belongings, without individualized suspicion, in randomly selected classrooms under school policy); see also. Hough v. Shakopee Pub. Sch., 608 F.Supp.2d 1087, 1109 (D. Minn. 2009) (school policy of daily, suspicionless searches of students).
The parties here do not agree on the proper framework to apply. Williams argues that individualized suspicion is required under T.L.O. and that Duncan did not have reasonable suspicion to force Williams to . empty his pockets and to pat him- down. Conversely, the State emphasizes the language of Vernonia and Earls in asserting that no individualized suspicion was necessary because Duncan conducted the search according to the school’s policy. Even assuming arguendo that reasonable suspicion was required, the State asserts that Williamses tardiness provided the reasonable suspicion to justify, the search. The parties raise plausible arguments under each of the frameworks developed hy the Supreme Court in T.L.O. and Vernonia/Earls. Accordingly, we see legitimate reasons for analyzing this case under both frameworks, and we will do so in turn.
II. Application of . T.L.O.—Suspicion-Based Searches
Under the two-pronged T.L.O. framework, the search of Williams was reasonable if it was (1) “justified at its inception” and, if so, (2) whether the search conducted was reasonably related in scope to the circumstances that justified the interference in the first place, 469 U.S. at 341, 105 S.Ct. 733. Ordinarily, a search is justified at inception when “there are reasonable grounds for suspecting that the search wjll turn up evidence that the student has violated or is violating either the law or the rules of the school.” Id. at 341— 42, 105 S.Ct. 733. This test amounts to reasonable suspicion, allowing a search if “a moderate chance of finding evidence of wrongdoing” exists. Redding, 557 U.S. at 371, 129 S.Ct. 2633. A search is permissible in scope when “the measures adopted are reasonably related to the objectives of the search and not excessively intrusive in light of the age and sex of the student and the nature of the infraction.” T.L.O., 469 *696U.S. at 342, 105 S.Ct. 733. There must exist some nexus between the item searched for and the suspected violation. Id. at 345, 105 S.Ct. 733.
The State argues that Williams’s tardiness alone provides the requisite suspicion to justify the pat-down search. The State’s conclusory argument stems from Justice Breyer’s concurrence in Earls, which noted that lower courts have used tardiness as a court-approved factor in justifying suspicion-based searches. Earls, 536 U.S. at 842, 122 S.Ct. 2559 (Breyer, J., concurring).7 However, Justice Breyer’s statement does not suggest that tardiness functions as a litmus test or per se rule establishing reasonable suspicion.8 Tardiness, at best, is merely a factor in the reasonable-suspicion analysis. We found no Missouri case applying T.L.O.’s reasonable-suspicion test to a student search, let alone considering a student’s tardiness in that context. Other jurisdictions support our skepticism that a student’s tardiness or truancy at school, alone, establishes reasonable suspicion to justify this search.
For example, an Indiana appellate court considered the search of D.M., a student at an alternative school who entered his class nearly one hour late. D.M. v. State, 902 N.E.2d 276, 277 (Ind. Ct. App. 2009). D.M.’s teacher overheard him tell another student that he had been shopping that morning and that he “had a stack.” Id. The teacher searched D.M.’s jacket, finding stolen credit cards and a set of keys for a stolen car. Id. The State defended the search’s reasonableness because of D.M.’s limited expectation of privacy as a student at school, his late arrival to class, his comments to the other student about the “stack,” and the fact that other students (not the defendant) had recently possessed weapons and drugs at school. Id. at 279. The court held that this evidence lacked justification for the search at its inception because no reasonable grounds existed for suspecting that D.M. possessed contraband. Id. (applying T.L.O.). While the court was “reluctant to interfere with a school’s disciplinary policies, the standard enunciated in T.L.O. commands that students’ legitimate privacy rights must, nonetheless, be balanced against the need of school officials to deal effectively with the threat of drugs and violence.” Id.; see also D.I.R. v. State, 683 N.E.2d 251, 253 (Ind. Ct. App. 1997) (holding no justification existed for a search at its inception when an official reached into the student’s pocket “merely because she arrived late for class”).
The Supreme Court of California also rejected tardiness or truancy as a sole factor justifying a search of a public-school student. In re William G., 40 Cal.3d 550, 221 Cal.Rptr. 118, 709 P.2d 1287, 1297 (1985). In William G., an assistant principal encountered three students who were late to class. Id., 221 Cal.Rptr. 118, 709 P.2d at *6971289. Upon questiqning, William made furtive gestures in attempting to hide his calculator case, which had an odd-looking bulge. Id The assistant principal found contraband upon a search of the calculator case. Id. In applying T.L.O., the court held that the assistant principal “articulated no facts to support a reasonable suspicion that William was engaged in a proscribed activity justifying a search.” Id., 221 Cal.Rptr. 118, 709 P.2d at 1297, In so holding, the court stressed that the record reflected a “complete lack of any prior knowledge or information” relating to William’s possession, use, or sale of contraband. Id. The student’s tardiness and truancy provided no reasonable basis for conducting a search of any kind. Id.
Applying T.L.O. and William G., a New Mexico court arrived at a similar conclusion on similar facts. See State v. Pablo R., 139 N.M. 744, 137 P.3d 1198, 1202 (N.M. Ct. App. 2006). A school official encountered a high-school student out of class without permission. Id The school official testified about a few similar encounters in the past with Child. Id. at 1200, On those occasions, the school official simply told Child to return to class. Id. This time, however, Child acted “a little nervous” and fidgety, so the school official suspected that he might have a weapon or marijuana on him. Id. In holding that a subsequent search of the Child’s person was not justified at its inception, the court stressed that the school official did not suspect Child of engaging in criminal activity, did not smell marijuana, had no other history of wrongdoing with Child, and had no knowledge or information concerning any wrongdoing on this occasion, other than his being out of class. Id. at 1202. The court found “no logical connection between the search of Child for contraband and the suspected violation of being out of class without a pass.” Id. Because the school officials had “no idea what Child might have had in his possession upon searching him, or why the search might have revealed evidence of a violation of the law or school rules,” the court found no “reasonable suspicion to justify the search of Child at its inception.” Id. at 1203.
The Supreme Judicial Court of Massachusetts also agrees. See Com. v. Damian D., 434 Mass. 725, 752 N.E.2d 679, 683 (2001). A school official ordered Damian to empty his pockets, remove his shoes, and submit to a pat-down search. Id. at 681. The only basis for the search was Damian’s truant behavior—missing class on one day, and then missing his subsequent disciplinary hearing. Id. at 682. The court stressed that Damian was not searched for evidence of his truancy; instead, he was searched for contraband. Id. at 683. “It was pure speculation to conclude that, because Damian was out of class for a period of time during the day, he was likely to have contraband.” Id. Possibility and speculation, like hunches and unparticularized suspicion, the court reasoned, do not constitute reasonable grounds for the search of a student. Id. Thus, the court held that the search was not justified at its inception. Id.
Our research has revealed only one case approaching the State’s argument that tardiness or truancy, alone, justifies searching a student’s person. See Coronado v. State, 835 S.W.2d 636 (Tex. Crim. App. 1992). In Coronado, based on a tip from another student, an assistant principal suspected that appellant had been selling drugs at the school. Id. at 637. The assistant principal confronted appellant, searched him, and found no drugs that day. Id. The assistant principal asked him, “Do you sell drugs?” and appellant replied, “Not on campus.” Id. A few weeks later, the assistant principal learned that appellant was leaving school and was in the parking lot on a pay phone. Id. After requiring appel*698lant to come back inside the school, the assistant principal determined that appellant’s excuse for his truancy was false. Id. Appellant then was subjected to a series of searches, which first included a pat-down search “out of a concern for safety.” Appellant was subsequently asked to remove his shoes and socks, to pull down his pants, to open his locker, and to unlock his car, where drugs were finally found. Id. at 637-39. Applying T.L.O., the court found that the assistant, principal had reasonable grounds to suspect that appellant was violating school rules by “skipping.” Id. at 641. Thus, the Court reasoned that school officials had reasonable grounds to investigate why appellant was attempting to leave school. Id. While the Court concluded that school officials were “justified in ‘patting down’ appellant for safety reasons,” it reversed appellant’s conviction because the subsequent searches were not reasonably, related (T.L.O. prong 2) to the .circumstances that caused the suspicion of appellant skipping school. Id.
Here, Williams was 30-minutes late to school. Williams was subjected to a hand search only because he was late to school. We agree with the other jurisdictions outlined above that a student’s tardiness, standing alone, does not establish the requisite reasonable suspicion that allows school officials to search a student’s person. See William G., 221 Cal.Rptr. 118, 709 P.2d at 1297 (tardiness or truancy provides no reasonable basis for conducting a search of any kind); Pablo. R., 137 P.3d at 1202 (“no logical connection between the search of Child for contraband and the suspected violation of being out of class without a pass”); Damian D., 752 N.E.2d at 683 (“pure speculation to conclude that, because Damian was out of class for a period, of time during the day, he was likely to have contraband”). Here, Duncan testified that the pat-down search was conducted because of “suspicion on late arrival.” When specifically asked why he patted down Williams, Duncan said only “suspicion.” Duncan did not justify in the record the basis for any suspicion, other than pointing to a school policy that allowed him to pat down students who arrived 30-or-móre-minutes late. Tardiness, alone, does not justify a search of Williams’s person.
Further, we are not dissuaded from our holding by Coronado, in which the facts are appreciably distinguishable. In Coronado, the assistant principal had specific, individualized information that appellant had been selling drugs at school, and the appellant (weeks before the search) had admitted that he sold drugs outside of school. Thus, the assistant principal had individualized evidence of appellant’s recent involvement in the drug trade. No individualized evidence connecting Williams to drug use or possession exists here. Further, the court in Coronado allowed the initial pat-down search “for safety reasons,” suggesting (albeit, not explicitly in the court’s opinion) that appellant, who admitted selling drugs outside of school, might have been armed with a weapon. No genuine safety concern existed here because Williams had already passed undetected through a metal detector, which would have alerted Duncan to guns, knives, or other metallic weapons.
While Duncan testified that the neighborhood surrounding the school had “a lot of drug activities and violence,” the record contains no evidence connecting this surrounding drug activity and violence to Williams, or even to the school generally. The State presented no evidence that tardy students were more likely to bring contraband into the school. Nor was there evidence that the school officials suspected those students who arrived late to school were ,in any way involved in the surrounding violence and drug activity. The record *699is devoid of any evidence to suggest that the school had any reasonable basis for suspecting that Williams was involved in any illicit activity. To the contrary, Duncan testified that he. knew Williams personally and when asked about his prior interactions with Williams, Duncan stated, “Just the normal routine just get to class. Nothing negative.” In this way, this case resembles Pablo R. In both cases, the school officers knew that the students searched had no history of wrongdoing (other than their tardiness and truancy). As in Pablo R., the State here failed to elicit any specific, articulable facts to support a reasonable suspicion that Williams was carrying a weapon9 or drugs, or was engaging in any prohibited activity that would justify searching his person.
Given the record before us, Duncan reasonably could not have concluded, before the search, that there was “a moderate chance of finding evidence of wrongdoing” on Williams. Redding, 557 U.S. at 371, 129 S.Ct. 2633. Thus, no reasonable suspicion existed, and the search was not justified at its inception. The search fails the first prong of T.L.O.
III. Application of Vernonia and Earls—Suspicionless Searches
Although the search of Williams is not justified based on reasonable suspicion under the T.L.O. test, we now consider whether the school search was reasonable under the Vernonia/Earls framework: Does a neutral school policy authorizing suspicionless searches of all students who arrive more than 30-minutes late align with the constitutional safeguards against unlawful search and seizure?
As discussed above, this analysis requires us to weigh Williams’s privacy rights against the school’s interest in safety and order. We again find no Missouri case applying the Vernonia/Earls framework in schools. However, reviewing Ver-nonia, Earls, and Doe (an Eighth Circuit decision) illustrates the contours of this balancing framework. All three cases applied the following guiding factors: (1) “the nature of the privacy interest allegedly compromised,” (2) “the character of the intrusion imposed,” and (3) “the nature and immediacy of the government’s concerns and the efficacy of the [p]olicy in meeting them.” Earls, 536 U.S. at 830-38, 122 S.Ct. 2559. We will discuss Vernonia. Earls, and Doe in turn before applying the framework to Duncan’s search of Williams.
A. Vernonia Sch, Dist, 47J v, Acton
In Vernonia, a 6-3 .Supreme Court considered the school district’s “Student Athlete Drug Policy.” 515 U.S. at 650, 115 S.Ct. 2386. The policy authorized random urinalysis drug testing of students who ■participated in the school district’s athletics programs. Id Testing was conducted on all athletes at the beginning of their season and on a randomly drawn subset (10%) each week during the season. Id. Students gave a urine sample in the bathroom; males stood at a urinal with a school official behind them, and females used a private stall with a school official outside listening for “normal sounds of urination.” Id. The Supreme Court held that the Fourth Amendment imposes no irreducible requirement of individualized suspicion. Id. at 653, 115 S.Ct. 2386. Instead, the ultimate inquiry is reasonableness, which the Supreme Court analyzed by looking to the factors expressed above. Id. at 652, 115 S.Ct. 2386.
*700The first factor was the nature of the privacy interest upon which the search intrudes. Id. at 654, 115 S.Ct. 2386. The Court recognized that the nature of a school’s power over students is custodial and tutelary, which permits a degree of supervision and control that government officials cannot not exercise over free adults. Id. at 655, 115 S.Ct. 2886. Thus, students within the public-school environment have a lesser expectation of privacy than do members of the general population. Id. at 655-66, 115 S.Ct. 2386. Legitimate privacy expectations are further lessened for these student athletes because the school’s locker rooms had no individual dressing rooms, no private showers, and not even all toilet stalls had doors. W. at 657, 115 S.Ct. 2386. Finally, the Court found that students’ privacy interests in that case were further reduced because all tested students had voluntarily gone out for the team, thereby voluntarily subjecting themselves to a higher degree of regulation than the general student population. Id. Students who voluntarily participated in school athletics had reason to expect intrusions—through, for example, preseason physical exams, minimum grade point averages, and rules of conduct—upon their normal rights and privileges, including their privacy. Id. Thus, the Court found that the students had a diminished expectation of privacy. Id.
The second factor to be scrutinized was the character of the intrusion at issue. Id. at 658,115 S.Ct. 2386. The Court identified two privacy-invasive aspects of the way the urine samples were collected. Id. First was the manner of collection. Id. Male students produced samples at a urinal, while fully clothed and only observed by a male monitor from behind; female students produced samples in an enclosed stall, with only a female monitor standing outside listening only for sounds of tampering. Id. Under such conditions, the Court considered the intrusion “negligible,” comparing these circumstances to nearly identical conditions encountered in public restrooms. Id. The second aspect of intrusion on the students’ privacy was the information the testing disclosed about the student. Id. The Court found it important that the tests checked only for the presence of drugs, and did not probe into a student’s medical condition, for example if the student was epileptic, pregnant, or diabetic. Id. Further, the test results were seen by a limited class of school personnel, and the results were not turned over to law enforcement or even used for school disciplinary proceedings. Id. A positive test result merely affected participation in sports. Id. Thus, the Court concluded that the invasion of privacy was “not significant.” Id. at 660,115 S.Ct. 2386.
The third factor considered was the nature and immediacy of the governmental concern, and the efficacy of the means for meeting that concern. Id. Regarding the nature of the concern, the Court acknowledged that deterring drug use by our nation’s schoolchildren was important, particularly because high-school years are the time when the physical, psychological, and addictive effects of drugs are most severe. Id. at 661, 115 S.Ct. 2386. Further, the Court recognized that the effects of drug-infested schools hurt not just the users but also the entire school community. Id at 662, 115 S.Ct. 2386. In this case, the searches were narrowly confined to student athletes, where drug use posed an elevated risk of physical harm resulting from the potential impairment of judgment, slow reactions, and a lessening of the perception of pain. Id.
Regarding the immediacy of the concern, the Court deferred to the district court’s factual findings, which showed that a large segment of the student body, particularly the student-athlete segment, “was in a state of rebellion.” Id. at 663,115 S.Ct. *7012386. Disciplinary actions had reached “epidemic proportions,” largely fueled by “alcohol and drug abuse as well as by the student’s misperceptions about the drug culture.” Id. Thus, the Court found that the district court’s factual findings demonstrated the immediacy of the school district’s concern when it established the drug-testing policy. Id.
Finally, regarding the policy’s efficacy in addressing the problem, the Court found it “self-evident” that the drug problem facing the school district-fueled largely by the “role model” effect of athletes’ drug use— was effectively addressed by ensuring that those athletes did not use drugs. Id After balancing the factors, the Supreme Court held that the school district’s drug-testing policy was reasonable and thus constitutional. Id. at 664-65, 115 S.Ct. 2386.
B. Bd. of Educ. of Indep. Sch. Dist. No. 92 of Pottawatomie County v. Earls
Seven years later, in a 5-4 decision, a sharply divided Supreme Court held another student drug-testing policy to be reasonable under the Fourth Amendment. Earls, 536 U.S. at 825, 122 S.Ct. 2559. The policy at issue was much like the one in Vernonia, except that it applied to all students who voluntarily participated in any extracurricular activity, rather than just athletes. Id. The Court applied the same factors in balancing the reasonableness of the policy.
First, in addressing the nature of the privacy interest compromised by drug testing, the Court again recognized that a public-school student’s expectation of privacy in school is less than that of an adult in public. Id. at 830-31,122 S.Ct. 2559. The students reasoned that they had a higher expectation of privacy than the student athletes had in Vernonia because they were not athletes subject to preseason physicals and communal undress. Id. at 831, 122 S.Ct. 2559. Rejecting this argument, the Court found that this consideration, while a factor, was not “essential” to its holding in Vernonia. Id. The Court stressed that the students still voluntarily subjected themselves to extracurricular activities, like in Vernonia, which imposed a greater degree of regulation and, in turn, further diminished the expectation of privacy for the participating students compared to the general student body. Id. at 831-32, 122 S.Ct. 2559. Thus, the students affected by the drug-testing policy had a limited expectation of privacy. Id. at 832, 122 S.Ct. 2559.
Second, Court found the privacy intrusion imposed by this drug-testing policy to be indistinguishable from the intrusion in Vernonia. Id. The urine-collection procedure was “virtually identical” to the procedure in Vernonia. Id. Also like Vernonia, the test results were kept in confidential files, shared to school officials only on a need-to-know basis, and were not disclosed to law enforcement. Id. at 833, 122 S.Ct. 2559. A positive result only affected participation in the extracurricular activities. Id. The Court concluded that the invasion of privacy was “not significant” because of the “minimally intrusive nature of the sample collection and the limited uses” to which the test results were put. Id. at 834, 122 S.Ct. 2559.
Third, the Court found that the nature and immediacy of the school’s concern, and the efficacy of the search, were sufficient to justify the search. Id. Regarding the nature of the school’s concern, the Court cited the importance of preventing drug use by students. Id. Drug abuse, the Court recognized, is especially important because the “evil is being visited not just upon individuals at large, but upon children for whom [the school] has undertaken a special responsibility of care and direction.” Id. The nationwide drug epidemic makes *702the war against drugs a pressing concern in. every school. Id.
Regarding immediacy, while the facts in Earls did not describe the presence of a drug problem of “epidemic” proportions as in Vernonia, the Court noted that the school district presented specific-evidence of drug use by its students, -Id. at 834-35,-122 S.Ct. 2559, Cautioning that the Fourth Amendment does not require the existence of a particularized or pervasive drug problem, the Court nevertheless noted that the school district presented sufficient evidence of student drug use to “shore up” the need for its drug-testing policy. Id. at 835, 122 S.Ct. 2559. The need to deter childhood drug use, combined with the specific evidence of drug use at the school, provided the necessary immediacy for "the drug-testing policy. Id. at 836, 122 S.Ct. 2559.
Finally, regarding efficacy, the Court held that testing students who participated in extracurricular activities was a “reasonably effective means of addressing” the school district’s legitimate concerns about preventing, deterring, and detecting drug use. Id. at 837, 122 S.Ct. 2559, On balance, the school district’s policy, like in Verno-nia, was reasonable and thus constitutional. Id. at 838, 122 S.Ct. 2559.
Four members of the Earls Court dissented, however, concluding this policy was unreasonable. Id. at 842, 122 S.Ct. 2559 (Ginsburg, J., dissenting). In particular, the dissent focused on the third factor and contrasted the “alarming situation” presented in Vernonia with the school district’s problems in Earls. The dissent characterized the situation in Earls as similar to every other school and presented no major or immediate problem at the time. Id. at 849, 122 S.Ct. 2559. Without a demonstrated drug-abuse problem, the dissent found that the efficacy of the school district’s solution to the problem was “greatly diminished.” Id. at 850, 122 S.Ct. 2559; Moreover, the dissent stressed that the Vernonia policy was limited to athletes, whereas the policy at issue “indiscriminately subjected” all students who participated in extracurricular activities. Id. at 851, 122 S.Ct. 2559. In Vernonia, much of the justification rested on the student-athlete safety, where the risk of immediate physical harm was particularly high. Id. Such a risk of physical harm for members of the choir, for example, was lower. Id. at 851-52, 122 S.Ct. 2559. On balance, the dissent concluded that the drug-testing policy was unreasonable and thus unconstitutional. Id. at 854, 122 S.Ct. 2559,
C. Doe ex rel. Doe v. Little Rock Sch. Dist.
If Earls can be characterized as a case narrowly meeting the constitutional requirements of reasonableness for a suspi-cionless school search, the Eighth Circuit, in Doe v. Little Rock Sch. Dist., addresses a school search that failed to meet the minimum requirements. of reasonableness. See 380 F.3d 349 (8th Cir. 2004). Jane Doe was a secondary-school student. Id. at 351. One day, school officials ordered all students in Doe’s classroom to remove everything from their pockets, to leave all belongings in the room, and to exit the room. Id. School officials searched the items left behind and found marijuana in Doe’s purse. Id. The school district had a policy of regularly searching randomly selected classrooms in this manner. Id. Following Vernonia and Earls, the Eighth Circuit applied the balancing test to determine the search’s reasonableness. Id. at 353-56.
Regarding the first factor—the nature of the student’s privacy interests—the court recognized that public-school students have a legitimate, though limited, expectation of privacy in their personal belongings at school. Id. at 353. Students’ belongings retain some degree of privacy *703because, at a minimum, students may legitimately need school supplies, keys, money, necessaries of personal hygiene, and non-disruptive yet highly personal items like photographs, letters, and diaries. Id. (quoting T.L.O., 469 U.S. at 339, 105 S.Ct. 733). The Eighth Circuit distinguished Doe from Vernonia and Earls, as those cases involved sub-populations of the student body. Id. In Vernonia and Earls, the students voluntarily participated in extracurricular activities that imposed separate systems of rules and enhanced regulation, and they thereby waived certain privacy expectations for the privilege of participating in the activity. Id. at 354-55. By contrast, in Doe, the Eighth Circuit explained that because the search policy applied to the entire student body, the school district could not plausibly claim that the students had voluntary traded some privacy interests for a benefit or privilege. Id. Thus, the Doe court concluded that these public-school students retained some legitimate expectation of privacy in their persons and belongings that was higher than the students’ privacy interests were in Vernonia and Earls. Id. at 354.
The second factor, the degree of intrusion, was pivotal to the Doe court’s analysis. Unlike the “negligible” intrusions present in Vernonia and Earls, the Doe Court found that the classroom searches invaded students’ privacy interests in a major way: “A search of a child’s person or of a closed purse or other bag carried on her person, no less than a similar search carried out on an adult, is undoubtedly a severe violation of subjective expectations of privacy.” Id. (quoting T.L.O. 469 U.S. at 337-38, 105 S.Ct. 733). The court recognized that students often carry personal items in their pockets and bags, and, whether or not carrying contraband, students surely feel uncomfortable or embarrassed when school officials rifle through their belongings. Id. at 354-55. The court found that the search policy “wholly obliterated” the students’ privacy interests in their bags. Id. at 355. The court emphasized the distinction between these physical intrusions and searches by drug-sniffing dogs or metal detectors (both commonly used at schools). Id. Searches using metal detectors and dogs are minimally intrusive, and provide an effective means for producing the individualized suspicion necessary for more intrusive searches.10 Id. By contrast, the search method employed in the Little Rock School District was highly intrusive, as school officials indiscriminately rifled through students’ bags. Id.
Another vital factor, noted by the court in its intrusiveness analysis was the purpose for which school officials used the fruit of the searches. In Verno-nia and Earls, the drug-test results were not disclosed to law enforcement; the most serious discipline was exclusion from extracurricular activities. Id. The limited consequences in those cases, the Eighth Circuit stressed, “contributed significantly to a conclusion that the’ invasion of the students’ privacy was insignificant.” Id. In “sharp contrast,” the fruit of the searches in Doe was turned over to law-enforcement officials, used in criminal proceedings, and, for Doe, resulted in a misdemeanor conviction. Id. Thus, the Court found that the character of this intrusion 'was qualitatively more severe than in Vernonia and Earls. Id. Rather than acting to promote student welfare, the school officials, in large part, played a law-enforcement role.11 Id.
*704In considering the third factor—nature, immediacy, and efficacy—the Eighth Circuit held that the Fourth Amendment’s reasonableness- inquiry is a sliding scale: the more compelling the governmental interest, the more the government is justified in conducting serious intrusions on legitimate expectations of privacy. Id. at 355. Thus, the nature and immediacy of the governmental concern must be “important enough to justify the particular search at hand.” Id. (emphasis in original) (quoting Vernonia, 515 U.S. at 661, 115 S.Ct. 2386). The court then concluded that the school district failed to identify a particularized need sufficient to justify the substantial intrusions upon the students’ privacy interests. Id. at 356. The court recognized that all schools have a strong interest in minimizing the harm done by the existence of drugs and guns. Id. But, the school district expressed only “generalized concerns” about weapons and drugs in its schools. Id. These generalized concerns, according to the court, distinguished the search at issue from Vernonia and Earls, as the school districts in both of those cases presented evidence of specific drug problems in their schools before the drug-testing policies were implemented.12 Id.
On balance, the Eighth Circuit found the random search of students’ belongings to be unreasonable:
While the line separating reasonable and unreasonable school searches is sometimes indistinct, we think it plain enough that the [school district's search practice crosses it. In light of the government’s legitimate interest in maintaining discipline and safety in the public schools, the privacy that students in those schools are reasonably entitled to expect is limited. The [school districtj’s search practice, however, effectively reduces these expectations to nothing, and the record contains no evidence of unique circumstances that would justify significant intrusions. The mere assertion that there are substantial problems associated with drugs and weapons in its schools does not give the [school district] carte blanche to inflict highly intrusive, random searches upon its general student body. [Id. at 356-57.]
D. The Search of Williams
In the case before us, Duncan searched Williams according to an established school policy. All students first pass through a metal detector. The policy, as Duncan described it, also required the hand search of students who arrive more than 30 minutes late to school. A hand search involved late-arriving students removing items from their pockets,13 and a school official patting them down. Duncan described the procedure for a pat down vaguely, but it appears that school officials would pat down the front of the student’s chest (for a male student, at least), then down the arms to the end of the sleeve, *705then down the legs to the ankles. Pockets were patted too. Duncan testified that the policy’s purpose was to protect the students, teachers, and staff. Guided by the Vernonia, Earls, and Doe, we will consider the three guiding factors as they apply here.
1.Nature of the Privacy Expectation
Williams’s privacy expectation was the same as the students’ in Doe. While students do not “shed their constitutional rights ... at the schoolhouse gate,” those rights are also less than an adult in public. Vernonia, 515 U.S. at 655-56, 115 S.Ct. 2386 (internal citation omitted). Nevertheless, students retain an expectation of privacy in their bodies and in their personal belongings at school. Doe, 380 F.3d at 353. Like in Doe, Williams’s privacy expectation was not diminished below that of the general student population by voluntary participating in extracurricular activities, as in Vernonia and Earls, Doe, 380 F.3d at 353-54. Williams had a limited, but legitimate, expectation of privacy in his body and in the belongings carried in his pockets.
2.Character of the Intrusion
We next consider the nature of the intrusion. How does the intrusion compare to that of the students’ legitimate privacy interests in Vernonia, Earls, and even Doe? In Vernonia and Earls, the Supreme Court held that collecting the urine sample caused a “negligible” intrusion because it was no more invasive than using any public restroom. Here, the search policy required students to empty their pockets and subject themselves to touching by school officials. Searching a student’s person, or a bag carried on his or her person, severely violates subjective expectations of privacy. T.L.O., 469 U.S. at 337-38, 105 S.Ct. 733. In one sense, the school search policy at Vashon is less intrusive than in Doe, because the school officials there rifled through all of the students’ belongings, in addition to the contents of the students’ pockets. On the other hand, school officials in Doe did not touch students. The physical touching of the students at Vashon exacerbates the intrusive nature of the .search.
We also find it critical to our analysis that school officials immediately turned over the results and fruit of their search to law enforcement. The pat-down search of Williams by Duncan led to the discovery of a cocaine-based substance, which immediately resulted in a call to Officer Davis, who arrested and booked Williams for a felony. Unequivocally, the law-enforcement involvement here is qualitatively more intrusive than in Vernonia and Earls. See Doe, 380 F.3d at 355. Even in Doe, the student was charged only with a misdemeanor, while Williams was charged with a felony carrying a potential sentence of years in prison. Rather than a search conducted to promote students’ welfare and safety, the policy here operated, in large part, with a law-enforcement purpose. See id. As a result, the intrusiveness of the search under a constitutional analysis is very, very high.
3.Nature and Immediacy of the Governmental Interest, and the Efficacy of the Means for Meeting That Interest
The final factor to be considered is the nature and immediacy of the government’s concern, and the efficacy of the policy in meeting those concerns. The record indicates that the school’s interest in its search policy was to “protect the students and the teachers and staff.” We note that Williams had already passed through a metal detector before he was hand searched and he does not challenge the school’s use of the metal detector. Thus, in justifying the *706hand search, the State can only rely on the school’s interest in finding drugs or other non-metallic contraband.
We recognize the legitimate and important nature of the school district’s interest in keeping drugs out of its schools. We understand the vital importance of the governmental concern in preventing all school drug use, Earls, 536 U.S. at 834, 122 S.Ct. 2559, the significance of which is magnified by the feet that drug use affects all children and staff at the school, Vernonia, 515 U.S. at 662, 115 S.Ct. 2386. We find this concern particularly pressing given Duncan’s testimony that the neighborhood surrounding Vashon experienced a lot of violence and drug activities. Vashon must, ensure that vices in the surrounding neighborhood do not spill into the classroom. But given the facts of this case, the school’s concern is not sufficiently particularized to override Williams’s right to be free from unreasonable search and seizure.
Importantly, the record lacks specific evidence of the immediacy of the* concern about drug use and violence in this school. All students were already required to pass through a metal detector, which would alert school officials to the most imminently harmful weapons—guns, knives—and would produce reasonable suspicion for a more intrusive search. See Doe, 380 F.3d at 355. The record contains no evidence that the drug problems in the surrounding neighborhood spilled into the school. The record contains no evidence linking the tardiness of students to participation in the drug activity in the neighborhood. There was no evidence that students loitering outside the school or, tardy coming to school were participants, as either buyers or sellers of drugs, in drug trafficking or any other unlawful activity occurring in the vicinity of the school.
Nor does our record contain any evidence illuminating the thought process of school or district officials in implementing this policy. In Vernonia, for example, specific evidence established a serious drug problem in which the student body (in particular, athletes) was in a “state of rebellion” fueled by alcohol and drug use. 515 U.S. at 662-63, 115 S.Ct. 2386. Surely, specific evidence is not a per se requirement in establishing a search’s reasonableness, .but such evidence “shores up” the need for the policy. See Earls, 536 U.S. at 835, 122 S.Ct. 2559. The immediacy here is no , more than in Doe: a generalized concern about keeping drugs out of school. Doe, 380 F.3d at 356.
Moreover, the record lacks any evidence suggesting that Vashon’s search policy was effective in combatting drug use in the school. In Earls and Vernonia, the policy was found to be very effective: the goal was to reduce drug use by the relevant students, and drug testing is very effective in identifying drug use. Vernonia, 515 U.S. at 663-64, 115 S.Ct. 2386; Earls, 536 U.S. at 837-38, 122 S.Ct. 2569. In- one sense, the policy here can be shown to be effective. Common sense dictates that it will be increasingly difficult for students to carry drugs into a school building when they are subjected to a pat-down search of pockets upon entry. This policy might deter some students from bringing contraband into school, provided they are at least 30 minutes late. But in another sense, the policy has no effect on drug-possessing students who arrive to school in a timely maimer. And, again, the record contains absolutely no evidence linking tardiness to contraband possession at Vashon. Absent any connection in our record between tardiness and drug or weapon possession, we find that the efficacy of Vashon’s policy markedly diminished. Cf. Earls, 536 U.S. at 850, 122 S.Ct. 2559 (Ginsburg, J., dissenting) (finding that the sparse evidence of a drug problem at the school “greatly diminished” *707the efficacy of the school’s policy). Stated differently, because we have no evidence of any particularized need the school was trying to address, other than a generalized interest in school safety, it is difficult to determine if the means employed—searching tardy students—were effective.
4. Conclusion
Vashon’s policy of searching tardy students did not meet the reasonableness standards articulated in Vernonia and Earls. Williams’s reasonable expectation of privacy exceeds that of the students in Vernonia and Earls because Vashon’s policy applied to the general student body, and was not limited to students who voluntarily participated in an extracurricular activity. Further, the character of the intrusion here was decisively greater than in Vernonia and Earls, and slightly greater than in Doe, because Williams was hand searched, immediately handed over to law enforcement and charged with a felony. While we acknowledge and support the school’s legitimate interest in preventing drugs in schools, the concern expressed in this case is no different than the generalized interest in keeping drugs out of school found in Doe. While a generalized interest might justify a minimally invasive search, given the record before ■ us, we conclude that Vashon’s generalized interest was not sufficiently compelling to justify the significant intrusion on Williams’s reasonable and legitimate expectation of privacy.14 See Doe, 380 F.3d at 356-57. We therefore hold that the search was unreasonable and .thus unconstitutional under the Fourth Amendment.
IV. Exclusion of Evidence
Point One assigns error to the admission of the drugs (State’s Exhibit 2)' and any testimony in support thereof. Points Two and Three both assign error to the admission of Williams’s, subsequent statements as the fruit of a poisonous tree. Point Two addresses Williams’s statement to Duncan, in the security office, acknowledging that the white, rocky substance was “dope.” Point Three addresses Williams’s statement made to Officer Davis that Williams “found” the contraband and “never should have picked [it] up.”15
*708Generally, evidence discovered and later found to be derivative of a Fourth Amendment violation must be excluded as fruit of the poisonous tree. State v. Miller, 894 S.W.2d 649, 654 (Mo. banc 1995). In determining whether evidence is the fruit of a poisonous tree, we ask whether, granting establishment of the primary illegality, the evidence “has been come at by exploitation of that illegality or instead by means sufficiently distinguishable to be purged of the primary taint.” State v. Lingar, 726 S.W.2d 728, 737 (Mo. banc 1987), abrogated on other grounds by State v. Taylor, 238 S.W.3d 145, 150 (Mo. banc 2007) (quoting Wong Sun v. United States, 371 U.S. 471, 488, 83 S.Ct. 407, 9 L.Ed.2d 441 (1963)).
Indeed, the State makes no argument that the statements were acquired by means sufficiently distinguishable from the unconstitutional search, such that they were purged from the taint of the unconstitutional search. The State’s only argument is that no poisonous tree existed— the search was constitutional. But we have rejected that argument. Following the initial unconstitutional search by Duncan, Williams stopped Duncan from patting his back pocket. Williams’s refusal to allow a search of his back packet prompted further investigation in the security office, which immediately led to Duncan discovering the drugs (Point One) and Williams’s comment that the drugs were “dope” (Point Two). Officer Davis soon arrested and booked Williams. At some point in that process, Williams told Officer Davis that he found the drugs and never should have picked them up (Point Three).
All three pieces of evidence derived directly and immediately from, and were still tainted by, the unconstitutional search. The evidence, therefore, must be suppressed as the fruit of the poisonous tree. Points One, Two, and Three are granted.
Conclusion
The judgment of the trial court is reversed, and the case is remanded for a new trial.
James M. Dowd, P.J., concurs. Gary M. Gaertner, Jr., J., concurs in result.. All statutory references are to RSMo (Cum. Supp. 2011).
. Williams does not dispute that the substance contained a cocaine base.
. Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966).
.Our record is unclear about where Williams made these statements. The transcript suggests that the statements occurred in Officer Davis’s school office, but they might have occurred at the police station.
. Williams did not testify.
. We note that Duncan was an employee of the school district and not a police officer. Much litigation has spawned over school resource officers who are employees of a police department, and whether a reasonable-suspicion standard is appropriate when police officers, rather than school officials, conduct the search. See, e.g., Shade v. City of Farmington, 309 F.3d 1054, 1061 (8th Cir. 2002). Because Duncan was a school employee, and Officer Davis (a police officer) was not contacted *695until after the complained-of search, we do not address this issue.
. Justice Breyer cited a law-review article (Martin H. Belsky, Random vs, Suspicion-Based Drug Testing in the Public Schools—A Surprising Civil Liberties Dilemma, 27 Okla. City U. L. Rev. 1, 20-21 (2002)), which in turn cited an unpublished district-court opinion (Bartram ex rel. Bartram v. Pennsbury Sch. Dist., CIV. A. 98-6159, 1999 U.S. Dist. LEXIS 7916 at *25, 1999 WL 391480, at *7 (E.D. Pa. May 24, 1999)).
. In context, Justice Breyer was actually skeptical of using tardiness as a factor, the use of which he thought was too subjective of a criterion that risked unfairly targeting members of popular groups or leaving "those whose behavior is slightly abnormal stigmatized in the minds of others." Id at 841-42, 122 S.Ct. 2559. Thus, Justice Breyer’s opinion used tardiness (and other subjective factors that lower courts had used) to advocate against requiring individualized suspicion— and, thus, he concurred in upholding a suspi-cionless search in Earls. Id.
. In fact, because Williams had already passed through a-metal detector without setting it off, Duncan could not reasonably suspect that Williams was carrying a gun or another metallic weapon (such as a knife).
. In other words, a minimally invasive, sus-picionless dog or metai-detector search (its reasonableness analyzed under Vemo-nia/Earls) can produce reasonable suspicion to perform a more invasive search of the person (analyzed under T.L.O.).
. Judge Beam, in a concurring opinion, was especially concerned with turning over evi*704dence to law enforcement, concluding that law-enforcement involvement had a "profound impact” on the "bottom-line reasonableness inquiry” Id. at 359. In other words, Judge Beam reasoned, the school reduced the extent to which it was acting under its "custodial and tutelary responsibilities” and, instead, brought its actions much closer to "the investigatory role for which the probable-cause standard is intended.” Id. (quoting Earls, 536 U.S. at 830, 122 S.Ct. 2559, which was quoting Vernonia, 515 U.S. at 656, 115 S.Ct. 2386).
. We reiterate the Earls Court’s rejection of the contention that specific evidence of problem at the school was required to justify a search. Earls, 536 U.S. at 835, 122 S.Ct. 2559. Nevertheless, the more invasive the search, the more justification a school needs on the other side of the scale, and specific evidence “shores up” that justification. Id.
. Duncan said that a hand search entails "maybe” removing shoes.
. We recognize the concern of the concurring opinion in this case—in fact, we do not in any way "dissuade public schools from crafting constitutionally appropriate pat-down policies that protect students and staff in light of the serious drug and violence problems facing school communities today.” We do not suggest that school policies allowing suspicionless searches are per se unconstitutional; indeed, since the touchstone of our analysis is the reasonableness of the policy, we necessarily require a fact-based analysis of the individual circumstances in each case. We are attempting to balance the limited, but legitimate, privacy interests of public-school children with the paramount interest of schools in keeping those children safe. It is incumbent upon the government—the school or the prosecution using the resulting evidence—to justify the reasonableness of a sus-picionless search policy. Again, the more invasive the search is on a student’s privacy interests, the more the State must justify that intrusion. Of critical importance, and lacking in the facts before us, is any nexus between the search policy here (searching late-arriving students) and achieving the school’s stated purpose of protecting students, teachers, and staff. The mere assertion of generalized concerns about drugs and weapons in its schools, which is undoubtedly a concern for every school, does not give a school carte blanche to inflict highly intrusive searches on the general student body without further justification, See Doe, 380 F.3d at 356-57.
. With regard to Points Two and Three, the State urges us to review for plain error because Williams did not specifically object to the testimony about Williams's statements. But Williams’s counsel objected "pursuant to [the] suppression motion" before all pieces of ■ relevant evidence (in Points One through Three) were admitted. The motion sought suppression of all articles seized and any testimo*708ny regarding such evidence. Williams’s statements at issue in Points Two and Three were certainly regarding the seized evidence. The trial court noted counsel's objection and explicitly stated that it would be continuing. See State v. Pike, 162 S.W.3d 464, 472 (Mo. banc 2005) (after the trial court overruled the defendant’s motion to suppress, the objection was preserved when counsel requested at trial that his motion to suppress be granted and received a continuing objection). In fact, the trial court noted the continuing nature of the objection three separate times throughout trial. The claims in Points Two and Three were preserved.