Dissenting opinion by
ACOBA, J.I respectfully dissent as to the partial majority opinion in Part III, upholding the ID procedure of the Administrative Driver’s License Revocation Office (ADLRO). The ID procedure constituted an unconstitutional limitation on Freitas’s right to a public hearing. The partial opinion in Part III, sanctioning as it does a sign-in procedure at public hearings, will have a deleterious and potentially inhibiting effect on the right to attend similar hearings freely and openly and without needless restriction, but more troubling, it diverts focus in any particular case from measures actually aimed at preventing disruptions and ensuring safety. I would hold that Freitas’s hearing should have been open without the restrictions imposed by the ADLRO procedure and order that future hearings be so conducted subject only to security measures previously identified by the Department of Public Safety (DPS) that are appropriate.
I.
First, in my view, the essential supplemental findings of the ADLRO hearing officer, including the reference to the budget capabilities of the ADLRO, see infra page 54, 116 P.3d page 696, are not supported by substantial evidence. In the absence of substantial evidence, the findings were clearly erroneous and the conclusions from which they were derived, wrong. Second, assuming arguendo there was substantial evidence, the record gives rise to a definite and firm conviction that a mistake was made. Third, the hearing officer did not apply the test adopted in Freitas v. Admin. Dir. of the Courts, State of Hawai’i, 104 Hawai'i 483, 92 P.3d 993 (2004) [hereinafter, Freitas L], correctly and, thus, committed reversible error in her legal conclusion. On the grounds set forth herein, a person exercising reasonable caution would not conclude that the evidence submitted was of sufficient quality so as to support the conclusion that the ADLRO ID procedure prevents disruption of hearings, and is the least restrictive manner of implementing security. Finally, the cases relied upon by the majority to uphold the ID procedure concern the sixth amendment right to a public trial, *48and, hence, are distinguishable and inapplicable.
II.
This court previously held inter alia, that “because ADLRO hearings are quasi-judicial administrative hearings, due process requires that the hearings be public, and ... Freitas was entitled to a hearing on his objections to the ADLRO sign-in and identification procedure limiting public access to his hearing.” Freitas I, 104 Hawai'i at 483-84, 92 P.3d at 993-94. Thus, on June 16, 2004, this court remanded this case temporarily to the ADL-RO to afford Freitas a hearing on his objections to the ID procedure that limited public access to his hearing. Id. at 484, 92 P.3d at 994. The hearing was held on July 14, 2004 at the ADLRO offices. Following remand, the ADLRO submitted supplemental conclusions of law and an order on the public hearing issue.
On behalf of the ADLRO, the deputy attorney general called two witnesses, Lloyd Shimabuku, security consultant to several Waikiki hotels and special agent with the state Department of the Attorney General, and Ronald Sakata, Chief Adjudicator of the ADLRO. Respondent-Appellee Administrative Director of the Courts, State of Hawai'i (Director) also submitted two articles: “A Situationist Perspective on the Psychology of Evil: Understanding How Good People are Transformed Into Perpetrators,” by Phillip G. Zimbardo and “Identity and Anonymity: Some Conceptual Distinctions and Issues for Research,” by Gary T. Marx.
Freitas’s counsel called four witnesses, Dr. Reneau Charlene Ufford Kennedy, psychologist, Mr. R. Patrick McPherson, attorney, Ms. Lois Perrin, Director of the American Civil Liberties Union, Hawai'i, and Mr. Michael Nakamura, retired Chief of the Honolulu Police Department.
The Record on Appeal also contains a written security assessment prepared by the Department of Public Safety (DPS) for the ADLRO entitled, “Security Assessment, The Judiciary, Administrative Driver’s License Revocation Office, 3875 South King Street” (the Security Assessment).
III.
Pertinent here, Freitas contends in his supplemental brief that (1) “the hearing officer ignored all evidence contrary to her preconceived determination to uphold the ADL-RO sign-in procedure”1 and (2) the hearing officer’s findings of fact are clearly erroneous and her conclusions of law are contrary to established law. To these contentions Appel-lee essentially responds that the supplemental findings and conclusions are not contrary to the evidence or the law.2
Any restriction on the right to a public hearing must comport with the three-part test adopted in Freitas I:
[T]hat the regulation serve an important governmental interest; that this interest be unrelated to the content of the information to be disclosed in the proceeding; and that there be no less restrictive way to meet that goal.
104 Hawai'i at 489, 92 P.3d at 999 (quoting Brown & Williamson Tobacco Corp. v. Fed. Trade Comm’n, 710 F.2d 1165, 1179 (6th Cir.1983) (citing United States v. O’Brien, 391 U.S. 367, 377, 88 S.Ct. 1673, 20 L.Ed.2d 672 (1968))) (emphasis omitted). Because Freitas asserts a constitutional violation, in applying the three-part test, we are free to *49exercise our own “independent constitutional judgment based on the facts of the case.” Ka Pa‘akai O Ka ‘Aina v. Land Use Comm’n, 94 Hawai'i 31, 41, 7 P.3d 1068, 1078 (2000) (internal quotation marks, brackets, and citations omitted).
IV.
The hearing officer made twenty-five supplemental findings and four supplemental conclusions. On appeal we review findings of fact under the clearly erroneous standard. Child Support Enforcement Agency v. Roe, 96 Hawai'i 1, 11, 25 P.3d 60, 70 (2001) (quoting Gump v. Wal-Mart Stores, Inc., 93 Hawai'i 417, 420, 5 P.3d 407, 410 (2000)). Findings of fact are “clearly erroneous when the record lacks substantial evidence to support the finding.” Id. (quoting In re Water Use Permit Applications, 94 Hawai'i 97, 119, 9 P.3d 409, 431 (2000)). “‘Substantial evidence’ is credible evidence which is of sufficient quality and probative value to enable a person of reasonable caution to support a conclusion.” Id. (brackets, internal quotation marks, and citation omitted); see Leslie v. Tavares, 91 Hawai'i 394, 399, 984 P.2d 1220, 1225 (1999). Findings of fact are also clearly erroneous when “despite substantial evidence to support the finding ..., the appellate court is left with the definite and firm conviction in reviewing the entire evidence that a mistake has been made.” Child Support Enforcement Agency, 96 Hawai'i at 11, 25 P.3d at 70. See Lanai Co. v. Land Use Comm’n, 105 Hawai'i 296, 314, 97 P.3d 372, 390 (2004) (observing a “definite and firm conviction” that the Land Use Commission “made a ‘mistake’ ” in its enforcement of an order). “Hawai'i appellate courts review conclusions of law de novo, under the right/wrong standard.” Leslie, 91 Hawai'i at 399, 984 P.2d at 1225 (citations omitted).
V.
Preliminary, it must be noted that in a letter regarding ADLRO security renovations, the ADLRO detailed prior security incidents over the “history of the program.”3 The letter also states that ADLRO is concerned that “persons whose licenses have been or are in jeopardy of being revoked, in particular when we do start having drug related cases, will become unmanageable and/or violent.”
The Security Assessment was specifically prepared by the DPS for the ADLRO. It precisely addresses the matters of security at ADLRO meetings. In this document the DPS conducted an examination of the ADL-RO premises at the ADLRO’s request.4 The objective was to determine measures necessary to protect the health and safety of the employees and community:
In an effort to meet Public Safety’s goals of ensuring the health and safety of employees and community, the department conducted a Building Security Assessment for the Judiciary’s Administrative Driver’s License Revocation Office at their request. The recommendations contained in this report are for their use in determining *50what level of security is necessary for their operations.
(Emphases added.) Significantly, the Security Assessment does not identify the anonymity of members of the public which the sign-in procedure is designed to counteract as a threat to security. Indeed, a perusal of the Security Assessment reveals that DPS did not recommend the current sign-in procedure at all. The Security Assessment was prepared by the DPS after interviewing ADLRO employees and surveying the ADLRO building, hearing rooms, and office and, hence, is substantial evidence of ADLRO security problems and remedies. Yet the hearing officer did not reference the Security Assessment in her supplemental findings.
VI.
The pertinent supplemental findings are as follows:
3. The ADLRO instituted this ID procedure as a security measure to prevent unknown members of the general public from entering the inner-office area.
[[Image here]]
5. The ID procedure provides a reasonable means of identifying and apprehending those persons who might engage in unlawful or inappropriate behavior at an administrative hearing or within the inner-office area. (Emphasis added.)
6. The ID procedure provides a deterrent for those persons seeking entry past the front desk/reception counter, including those persons who wish to attend hearings ....
7. This deterrent effect arises out of the fact that persons who know that their identity has been recorded will generally be less likely to engage in unlawful or inappropriate behavior for the simple reason that they know they can be held accountable.
8. A person who remains anonymous ... is more likely to engage in inappropriate behavior ....
9. Although the ID procedure is not a perfect security measure, it is a fundamental first-step in the ADLRO’s security measures.
10.Mr. Sakata, as Chief Adjudicator of the ADLRO, instituted this ID procedure based upon his experience and common sense understanding of human behavior.
[[Image here]]
[[Image here]]
12. Articles ... provide further support for this finding, because these articles support the principle that anonymity makes people more likely to engage in aggressive, evil, destructive, or unlawful behavior....
13. This Hearing Officer finds that these two articles support the view that the ID procedure, by directly stripping a person of his or her anonymity, lessens the likelihood that the identified person will engage in unlawful, harmful, or otherwise inappropriate behavior at the administrative hearing and within the inner office area. ...
14. Mr. Partington also elicited testimony from former police chief Michael Na-kamura that the ID procedure would have ■little benefit to security. This Hearing Officer finds that this testimony was not particularly persuasive in light of the testimony of not only the ADLRO Chief Adjudicator Ronald Sakata, but the testimony of security expert Lloyd Shimabuku, and since Mr. Nakamura conceded that the ID procedure could have some deterrent effect. ...
15. With respect to attorney R. Patrick McPherson’s testimony ... in which McPherson acknowledges that no state court, trial or appellate level, requires one to show identification and sign-in in order to attend a court proceeding, this Hearing Officer finds this testimony unrelated to the ADLRO’s unique circumstances in which, unlike the court buildings, the area to which counsel, respondent, and/or other members of the public are requesting access, includes undifferentiated access to the hearing room as well as other areas of the ADLRO office....
16.... [T]he ADLRO does not have separate public and non-public access area. This distinguishing factor is critical and material in determining whether the ADL-RO’s ID procedure is warranted.
*5117. This Hearing Officer finds that other security measures—including a metal detector, x-ray machine and conveyor belt, a hand metal-detecting wand, and someone to operate these devices, or posting sheriffs or security guards (armed or unarmed) in or near the hearing room—would be expensive and beyond the budget capabilities of the ADLRO. ...
18. In addition, metal detectors, x-ray machines and hand metal-detecting wands would do nothing to stop a person intent on accosting ADLRO staff or hearing attendees by hand, arm, leg or foot, nor would such devices prevent someone from causing a vocal or verbal disturbance to the administrative proceeding. The ID procedure, on the other hand, could possibly deter such inappropriate behavior.
[[Image here]]
20. In addition, this Hearing Officer finds that even if such additional security measures were in place—e.g., a metal detector, x-ray conveyor belt, and hand wand, or a security guard—the ID procedure would provide an additional security benefit in the form of deterrence.... There is no other less intrusive way to achieve this particular form of deterrence—based upon depriving a person of his or her anonymity—other than to have the ADLRO’s ID procedure in effect. Although security cameras ... do remove some level of anonymity, they still leave a person the chance of remaining unidentified. ...
[[Image here]]
22.This Hearing Officer finds that there is no less intrusive way to provide the unique deterrent effect created by the ADLRO ID procedure than to maintain the ID procedure. No other security measure could fully substitute for the special and unique deterrent effect brought about by requiring the showing of a picture ID and sign-in, as it is the most effective ... way of eliminating one’s anonymity. ...
23. This hearing officer finds that although the ID procedure is not perfect— e.g., people can sometimes obtain fake ID’s, and some people will engage in bad behavior regardless of being pre-identi-fied—it remains a useful and reasonable security measure for the ADLRO. And the procedure is a very easy and simple process for a prospective attendee to meet. A driver’s license, state I.D. or other acceptable picture identification is all that is necessary.... Indeed, a driver’s license, state I.D., or other acceptable picture identification, is something people need for all sorts of everyday activities, including for example: check cashing, banking, and air travel.
24. This hearing officer finds that although the ID procedure may deprive a person of his or her anonymity—indeed that is precisely why the ID procedure has an effective deterrent effect—that is not an especially significant intrusion because a person attending the hearing would have their face seen by hearing participants in any event. Furthermore, the ADLRO as a matter of policy does not distribute the sign-in list to anyone, except in the event someone on that list engages in unlawful activity or creates a disturbance.
(Emphases added.) On the grounds set forth herein, the relevant supplemental findings are not supported by substantial evidence. The said findings are considered in seriatim.
1. Finding no. 3 states that “[t]he ADL-RO instituted this ID procedure as a security measure .... ” The record shows that the ID procedure was introduced by Sakata, based on his “common sense, experience” that the ID procedure would have a deterrent effect.5 However, as mentioned, the ID procedure is not identified or recommended by the Security Assessment, supra. Additionally, at the time the ID procedure was instituted, Sakata did not even know of the *52social science articles that were subsequently introduced at the remand hearing to justify the procedure, and as he conceded, the articles had nothing to do with his decision to institute the procedure.6 Consequently, the articles not only lack a credible basis for the purpose of the hearing, see discussion infra; they are completely irrelevant to the ADL-RO’s decision to implement the ID procedure.
2. Finding no. 5 states that “[t]he ID procedure provides a reasonable means of identifying and apprehending those persons who might engage in unlawful or inappropriate behavior....” (Emphasis added.) Finding no. 23 repeats that the ID procedure is useful, reasonable, easy and simple.
However, the reasonableness of the procedure is a question of law, and insofar as it relates to a fact, is not supported by substantial evidence. Appellee’s own security expert, Shimabuku, agreed that without training personnel to recognize “fake” photo ID’s, the ID procedure is “relatively useless.”7 Sakata testified that ADLRO staff is not trained to recognize false ID’s, and Chief Nakamura testified that such ID’s are “relatively easy” to obtain in Hawai'i.
3. Findings nos. 6, 7, 8, 14, and 18 state that the ID procedure acts as a deterrent. The assertion is presumably supported by the two articles and the testimony of Sakata and Shimabuku. However, this determination is not supported by the articles submitted by Appellees. In light of the ADLRO letter, the testimony of Chief Nakamura, and the Security Assessment, the testimony of Sakata and Shimabuku as discussed, did not constitute substantial evidence to support these findings.
a. Although the hearing officer relies heavily upon the two articles, neither article can be accepted as credible evidence. In his article, Marx states unequivocally that he focuses on concepts, not actual behavior.8 The author also admits, with respect to Section B,9 that he is simply reporting justifications for concealment and revelation, not endorsing these justifications.10 As Marx *53himself will not endorse the validity of the “claimed empirical consequences,” little credence can be attributed to any findings of fact resting upon this article.
Zimbardo’s article similarly does not constitute credible evidence relevant to security measures at ADLRO hearings. This article does not concern security measures at public hearings but rather, “generic forms of institutional evil, such as poverty, prejudice or destruction of the environment by corporate greed.” It therefore is not meaningfully relevant to the question of public hearing security measures.11 Further, the article is seemingly driven by an overt political view, and cannot be accorded the status of unbiased scientific or social-scientific reports for the purpose of this case.12 Findings 12 and 13, based on these articles, are similarly unsupported by the record.
b. Appellee’s contention that the anonymity of members of the public poses a threat to ADLRO hearings is also unsupported by the ADLRO’s letter, swpra. In that letter listing prior and anticipated incidents, the ADLRO identified threats to the physical safety of ADLRO employees. Any threat posed because of the anonymity of members of the public is absent.
c. The testimony of Chief Nakamura likewise does not support a finding of deterrence. While this court usually disinclines review of a hearing officer’s findings based on oral testimony, finding no. 14 appears to mischaraeterize the testimony of Chief Naka-mura. In that finding, the hearing officer stated that “Mr. Nakamura conceded that the ID procedure could have some deterrent effect....” (Emphasis added.) To the contrary, Chief Nakamura’s testimony was that the procedure was “close to” “useless.” The relevant testimony reads as follows:
[Deputy Attorney General]: Chief Naka-mura, you were saying that signing requirements, sign in and identification showing requirement would have little impact upon security, are you saying that this requirements has no impact on security and is absolutely useless?
Nakamura: It’s not absolutely useless, close to it.
[[Image here]]
[Deputy Attorney General]: But it could have some impact?
Nakamura: If I had to rate that with every option available, it would probably be at the bottom of the list.
(Emphases added.) Chief Nakamura’s testimony does not support a finding that he conceded the ID procedure has a deterrent effect. Although the hearing officer decided, in finding no. 14, that Chief Nakamura’s “testimony was not particularly persuasive,” it would appear evident that the testimony of the former police chief of Honolulu should have been accorded substantial weight in the areas of public disruption, violent acts, and security in public places based on his training, expertise, and experience and in the absence of any finding that his testimony should be disregarded in this respect.
d.As mentioned previously, the hearing officer did not reference the Security Assessment in her supplemental findings. But, the assessment is substantial evidence of ADL-*54RO security problems and remedies. Again, a perusal of the Security Assessment reveals that DPS made specific recommendations, none of which included the current sign-in procedure.
In sum, the documentary evidence and oral testimony in the record do not support a substantial connection between disruption and deterrence of threats and the current ID procedure. Although Appellee argues the ID procedure may facilitate locating an individual after the fact, ADLRO staff are not trained to recognize fake ID’s, see supra note 7; hence, there is no evidence that the ID procedure advances even this limited goal.
4. Finding no. 9 states that the ID procedure is “a fundamental first-step in the ADL-RO’s security measures.” There is no substantial evidence in the record to support a finding that the ID procedure is either “fundamental” or a “first-step” in appropriate security measures. On the other hand, the Security Assessment precisely enumerates the measures necessary to insure security and thus is substantial evidence of the “fundamental” and necessary “steps” required.
5. Findings nos. 15 and 16 discuss the unique circumstances of the ADLRO building. But these findings, unlike the Security Assessment, are not tailored to the specific safety requirements of the facility; hence, these findings are not supported by substantial evidence.
6. Finding no. 17 indicates that other security measures, “including a metal detector, x-ray machine and conveyor belt, a hand metal-detecting wand, and someone to operate these devices or posting sheriffs or security guards (armed or unarmed) in or near the hearing room—would be expensive and beyond the budget capabilities of the ADL-RO....” Contrary to the partial majority opinion in Part III, majority opinion at 39-40, 116 P.3d at 681-82, other than Sakata’s bare testimony, nothing was submitted in the record to support this finding. Moreover, the relevant inquiry on remand was the application of the Freitas I test, supra.13 Additionally, the record indicates that the ADLRO has, on occasion, requested and been afforded deputy sheriffs to provide security, as occurred at this remand hearing.
7. Finding no. 20 states that “[tjhere is no other less intrusive way to achieve this particular form of deterrence—based upon depriving a person of his or her anonymity—other than to have the ADLRO’s procedure in effect. Although security cameras ... do remove some level of anonymity, they still leave a person the chance of remaining unidentified.... ” (Emphases added.) Finding no. 22 essentially restates finding 20.
Findings 20 and 22 must be rejected for two reasons. First, they state a conclusion of law that the ID procedure is the least intrusive means. Even if construed as a fact, these findings are controverted by the recommendations in the Security Assessment, which are substantial evidence of the means for ensuring security at ADLRO hearings.
Second, the findings erroneously limit potential security measures to those that require the public to sign-in and produce a picture ID. The governmental interest at stake is the security at agency hearing. Limiting this interest to security that is based on deprivation of anonymity leads to the syllogistic conclusion that only deprivation of anonymity can secure against the threat of anonymity.
It should also be noted that in footnote 6 to finding 22, the hearing officer states that “IDs are required for entry to circuit court chambers.” The relevant inquiry concerns the right of public access to a public hearing. Court chambers are not the equivalent of public hearing rooms. Therefore, procedures for court chambers entry do not constitute relevant or substantial evidence and the reference thus is clearly erroneous.
8. Finding no. 24 states that the intrusion posed by the ID procedure is insignificant because “a person attending the hearing would have [his] face seen by hearing partici*55pants in any event. Whether an intrusion is insignificant is a question of law and should not be couched as a finding. In addition, it would appear plain that being required to sign one’s name on a roster maintained by a state agency and to produce a picture ID is not equivalent to merely having one’s face seen by participants at an ADLRO hearing. Relatedly, the assertion that the sign-in list is not distributed does not accurately reflect the record. According to Sakata’s testimony, the sign-in list remains on the office counter all day and its subsequent custody is apparently entirely subject to Sakata’s discretion.
Thus, the record lacks substantial evidence to support findings that the current ID procedure (1) advances the governmental interest of safety at the hearings, (2) deters security threats at ADLRO hearings, and (3) is the least restrictive means of achieving security at ADLRO hearings.
VII.
Aside from the erroneous findings, the hearing officer incorrectly applied the Frei-tas I test. Thus, her conclusions were wrong. The four conclusions state:
1. The ID procedure serves an important governmental interest: namely, enhancing security and avoiding disruptions at ADLRO administrative hearings.
2. This interest ... is unrelated to the content of the information to be disclosed in the administrative proceeding.
3. There is no less restrictive way to fully serve this important governmental interest in enhancing security and avoiding disruptions at ADLRO administrative hearings and in-office area, other than to continue with the ID procedure. Although other measures can add to security as well, there is no other less intrusive means of achieving the unique deterrent effect that arises out of depriving a person of his or her anonymity. The ADLRO ID procedure is the least intrusive means of achieving this unique deterrent effect.
4.The ADLRO ID procedure is therefore fully warranted, and does not imper-missibly interfere with a respondent’s right to a public hearing.
(Emphases added.) Inasmuch as the test in Freitas I answers a constitutional question of law, the hearing officer’s application of the test must be reviewed de novo, under the right/wrong standard. See Bank of Hawaii v. Kunimoto, 91 Hawai'i 372, 387, 984 P.2d 1198, 1213 (1999).
VIII.
Logically, the first step in the Freitas I analysis is to identify the ADLRO’s “important governmental interest.” 104 Hawai'i at 489, 92 P.3d at 999. In its letter, see supra, the ADLRO noted its concern that “persons whose licenses have been or are in jeopardy of being revoked ... will become unmanageable and/or violent.” The hearing officer accepted this concern in conclusion no. 1 by identifying “enhancing security and avoiding disruptions at ADLRO administrative hearings” as the “important governmental interest” to be served by the ID procedure.
This would appear to satisfy the first element of the Freitas I scrutiny, see Regal Cinemas, Inc. v. Mayfield Heights, 137 Ohio App.3d 61, 738 N.E.2d 42, 48 (2000) (determining that “security at venues that attract a large number of people in a congested area at the same time” is “an important government interest”); Klein v. Leis, 99 Ohio St.3d 537, 795 N.E.2d 633, 640 (2003) (concluding that “[ejnsuring public safety is an important government interest”); In re Rules Adoption, 120 N.J. 137, 576 A.2d 274, 281 (1990) (holding that “institutional security” at a prison is an “important government interest”), and Freitas does not dispute that the ADLRO’s concern constitutes a valid interest.
IX.
Under Freitas I, the next inquiry applicable to this case14 is determining whether the ID procedure constitutes the least “restric*56tive way to” “enhanc[e] security and avoid[ ] interruptions at ADLRO administrative hearings.” 104 Hawai'i at 489, 92 P.3d at 999. As there is no substantial evidence that the ID procedure advances the governmental interest at stake, concluding that this procedure is the least restrictive means of achieving that interest is also untenable. For the reasons set forth previously, the findings do not support a conclusion based upon substantial evidence that the ID procedure in any way enhances security or prevents disruptions at ADLRO hearings.15 Instead, the record indicates that because ADLRO staff are not trained to recognize false ID’s, the current ID procedure provides no deterrent. Even if the staff were trained, substantial evidence did not exist to support a finding that the ID procedure would in fact reduce security threats.
At this point, the hearing officer’s analysis, as exhibited in conclusion no. 3, blurred the first and third prongs of the Freitas I test. Essentially, the hearing officer confused the “important governmental interest” with the least “restrictive way to meet” that interest. To reiterate, in conclusion no. 3, the hearing officer determined
[t]here is no less restrictive way to fully serve this important governmental interest in enhancing security and avoiding disruptions at ADLRO administrative hearings and in-offiee area, other than to continue with the ID procedure. Although other measures can add to security as well, there is no other less intrusive means of achieving the unique deterrent effect that arises out of depriving a person of his or her anonymity. The ADLRO ID procedure is the least intrusive means of achieving this unique deterrent effect.
(Emphases added.) This analysis is flawed for two reasons. First, the “unique deterrent effect that arises out of depriving a person of his or her anonymity” is not the “important governmental interest” that was asserted by the ADLRO and identified in conclusion no. 1. In asking whether there is no less restrictive means to meet the goal, the hearing officer wrongly redefined the governmental interest as whether “[t]here is no other less intrusive means of achieving the unique deterrent effect that arises out of depriving a person of his or her anonymity.” (Emphasis added.) The governmental interest at stake is the security of the agency hearing, not the most efficacious way of depriving a person of his or her anonymity.
Second, this “unique deterrent effect” constitutes the means of achieving the ADLRO’s interest in enhanced security and in minimizing disruptions at hearings, not the interest itself. The conclusion that “[t]he ADLRO ID procedure is the least intrusive means of achieving this unique deterrent effect” of “depriving a person of his or her anonymity” is factually true. Requiring persons to present proper and valid identification no doubt strips them of their anonymity. But this is not the issue to be decided.
The proper inquiry is whether the ID procedure, i.e., depriving persons of their anonymity, is the least “restrictive way to” “en-hanc[e] security and avoid[] disruptions at ADLRO administrative hearings.” The governmental interest at stake is the security at agency hearings. As indicated before, equating this interest to the deprivation of anonymity leads to the syllogistic conclusion that only deprivation of anonymity can secure against the threat of anonymity and results in the hearing officer’s wrong conclusion.
Based on the reasons enumerated before, the ID procedure was not shown to be the least restrictive means of meeting the governmental interest in “enhancing security and avoiding disruptions.” On the other hand, credible evidence in the form of the Security Assessment set forth security measures previously calculated to the specific situation of the ADLRO. Appellees introduced no credible evidence balancing the al*57ternatives set forth in the assessment as required under the Freitas I test. The hearing officer thus erred in her application of the Freitas I test.
X.
To support its holding that “the ADLRO’s identification and sign-in procedure does not impermissibly infringe upon Freitas’s constitutional right to a public hearingQ” majority opinion at 37, 116 P.3d at 679 (emphasis added), the majority relies upon cases addressing the sixth amendment right to a public trial, see majority opinion at 37-39, 116 P.3d at 679-81. In our prior opinion remanding the case to the ADLRO, however, we distinguished between the sixth amendment right to a public trial and the right to a public hearing asserted by Freitas. See Freitas I, 104 Hawai'i at 486 n. 7, 92 P.3d at 996 n. 7 (distinguishing State v. Ortiz, 91 Hawai'i 181, 981 P.2d 1127 (1999), because it involved “a criminal proceeding subject to the right to a public trial afforded by the [sjixth [ajmendment and [a]rt. VII § 14 of the Hawaii State Constitution and this case is an administrative proceeding”). Inasmuch as this ease concerns a quasi-judicial administrative proceeding before the ADLRO, Freitas I, 104 Hawai'i at 489, 92 P.3d at 999, and “due process requires that [such] hearings be public,” id., the defendant’s Sixth Amendment right to a public trial in a criminal prosecution is not implicated. To intimate otherwise, as the partial majority opinion in Part III does, see majority opinion at 39, note 2, 116 P.3d at 681, note 2, would obscure the “automatic reversal” rule under the sixth amendment applied in criminal cases, and the balancing test we had adopted in Freitas I to be applied where the due process clause pertains.16
Nonetheless, the majority cites to United States v. DeLuca, 137 F.3d 24 (1st Cir.1998), a sixth amendment right to a public trial case, for the proposition that this court should be “hesitant to displace the ADLRO hearing officer’s judgment call in these circumstances.” Majority opinion at 38, 116 P.3d at 680 (quoting DeLuca, 137 F.3d at 34) (brackets omitted). In DeLuca, the First Circuit afforded the trial court “substantial deference” in its “assessment that the screening procedures were warranted,” observing that such “difficult judgments are matters of courtroom governance which require a sensitive appraisal of the climate surrounding a trial and a prediction as to the potential security or publicity problems that may arise during the proceedings.” 137 F.3d at 34 (internal quotation marks and citation omitted).
Assuming, arguendo, the applicability of DeLuca, it should be emphasized that the screening procedure used in that ease “was reasonably designed to respond,” id. at 35, to the concerns specific to the defendants who “either were directly associated with prior efforts to obstruct fair fact[-]finding through untruthful trial testimony, or were found to possess the present means as well as ample inducement ... to sponsor similar efforts in the case,” id. In contrast here, the ID procedure was not shown to be “reasonably designed to respond” to a specific security threat at ADLRO hearings. As stated previously, the Security Assessment prepared by DPS for the ADLRO hearings did not recommend the sign-in procedure and there was no evidence to support the conclusion that the ID procedure would prevent disruptions at the hearings. Thus, the hearing officer is not entitled to the same level of “judgment call” deference afforded the trial judge in DeLuca.
The majority cites to a second right to a public trial case, Williams v. Indiana, 690 N.E.2d 162 (Ind.1997), and asserts that our prior remand for a hearing was “[cjonsonant *58with the Williams rationale[J” Majority opinion at 38-39, 116 P.3d at 680-81. Williams requires a court to “provide the reasons for its decision to authorize the procedures” and to create a record “clearly substantiating] the need for these additional precautions.” 690 N.E.2d at 170. As previously discussed, however, the hearing officer’s findings do not “substantiate the need for,” id. at 170, the ID procedure. Moreover, Williams requires a weighing of “the prospective benefits to the order and security of the courtroom with the burdens to the defendant, the press, and the public.” Id. The hearing officer apparently found in finding no. 17 that the “budget capabilities of the ADLRO” outweighed the burdens of the ADLRO’s sign-in and identification procedure. I cannot accept, as the majority does, this “fiscal feasibility]” justification, majority opinion at 39-40, 116 P.3d at 681-82, for the implementation ofl an ID procedure that, according to the record, including testimony by Chief Nakamura, is unlikely to yield worthwhile security benefits at the ADLRO hearings.
The majority also cites to United States v. Brazel, 102 F.3d 1120 (11th Cir.1997). But like DeLuca, the sign-in procedure in Brazel was upheld based on the trial judge’s “own observations for more than a week ... that individuals had been coming into the courtroom and fixing stares on the witnesses and possibly government counsel.” Id. at 1156. Thus, the court itself had observed a threat that jurors or witnesses might be improperly influenced. Id. at 1155. No such evidence of a similar threat was apparent at the ADLRO hearings and, therefore, I cannot agree with the majority’s assessment that the sign-in procedure is “reasonably tailored to meet the security needs of ADLRO hearings.” Majority opinion at 39, 116 P.3d at 681.
Moreover, the defendants in Brazel “objected that the identification procedure could have a chilling effect on the public, because some people might fear that if they identified themselves (by name, address, and birth date), a computer check might be run and they might be suspected of being a part of the drug conspiracy.” Id. at 1156. Thus, it was logical that an identification requirement would dissuade those with criminal histories, the very ones likely to be improperly influencing the jurors and witnesses, from entering the courtroom and interfering with the court proceeding. The sign-in procedure utilized at the ADLRO hearings is not supported by similar logic, but stems from a sweeping conclusion that depriving a person of his or her anonymity will minimize disruptions at the ADLRO hearings.
Whereas the sign-in procedure in Brazel was justified by the overt instances of intimidation observed by the judge herself and designed to exclude the sources of the intimidation, the sign-in procedure here is not similarly justified. Rather, it is based upon an amorphous threat to security at the ADLRO hearings and may exclude not just the sources of a supposed disruption, but individuals who, as stated in our prior opinion in this case, are entitled access to quasi-judicial proceedings in order to ensure that “the liberty and property of the citizen shall be protected by the rudimentary requirements of fair play” and to maintain “public confidence in the value and soundness of this important governmental process.” Freitas I, 104 Hawai'i at 489, 92 P.3d at 999.
Finally, it should be noted that DeLuca and Brazel involved case-specific approaches aimed at threats unique to the immediate proceeding before the trial judges. Therefore, these cases cannot serve as authority for the ADLRO’s permanent across-the-board sign-in procedure.
XI.
Under the evidence produced at the remand hearing, Freitas’s revocation hearing should have been free of the identification and sign-in procedure. I would order that future ADLRO hearings be open to the public without the requirement of ID and sign-in restrictions and that recommendations of the DPS as are appropriate be implemented.
. Freitas maintains that the hearing officer ignored all contrary evidence because Sakata, the hearing officer's supervisor, implemented the security measure now under scrutiny. Finding no. 10, supra, confirms that Sakata implemented the ID procedure.
. The Director argues that all of the hearing officer's findings and conclusions are valid and that this court must not review credibility of witnesses or weight of evidence, citing State v. Jenkins, 93 Hawai'i 87, 101, 997 P.2d 13, 27 (2000). Assuming its relevance, Jenkins does not represent a complete abrogation of an appellate court’s right to review findings of fact based on witness testimony.
[Vlerdicts based on conflicting evidence will not be set aside where there is substantial evidence to support the trier of fact’s findings. We have defined "substantial evidence” as "credible evidence which is of sufficient quality and probative value to enable a [person] of reasonable caution to support a conclusion.”
Id. at 101-02, 997 P.2d at 26-27 (internal citations omitted).
. The letter lists the following incidents:
— One arrestee lunging at hearing officer during hearing!;]
— One life time revocation individual writing, calling and visiting ADLRO with a series of threatening letters to ADLRO staff and family!;]
— ITjhree bomb threats, two resulting in police investigation, one made on the record at hearing!;]
— [B]ullet holes shot through street front windows on two occasions (undetermined if a direct result of ADLRO activities)!;]
— [N]umerous instances of persons at front desk or on the phone with irate and aggressive behavior shown!;]
— [S]everal instances of obviously intoxicated persons at front desk and/or attending hearings!;]
— One arrestee entering the ADLRO office with a plastic grocery bag filled with tools (pick axe, hammer and other unidentifiable objects) on one of several visits to ADLRO. ADLRO requested that sheriff(s) be on site at ADLRO during the scheduled hearing.
. For example, the Security Assessment identified the hearing rooms as "critical areas."
The critical areas are the Hearing Rooms where Referees meet with clients and their attorneys. The Referees are unprotected should clients become angry or violent. The closed doors present a serious problem with ensuring the Referees and attorneys health and safety.
. ADLRO Chief Adjudicator Ronald Sakata testified:
[Deputy Attorney General]: So, in your opinion, basically removing a person’s anonymity acts as a natural deterrent to wrongful or inappropriate conduct?
Sakata: I believe so, yes.
[Deputy Attorney General]: And what do you base this view on?
Sakata: Well, common sense, experience.
(Emphasis added.)
. The relevant testimony reads:
[Counsel for Freitas (Counsel)]: Mr. Sakata, when did you first see these psychology articles?
Sakata: Within the last couple of weeks.
[Counsel]: They had nothing to do with your adoption or your policy, did they?
Sakata: No.
(Emphases added.)
. Director’s security expert Shimabuku testified as follows:
[Counsel]: [D]oes it matter if the staff of this office or anywhere the sign in procedure is trained to look for fake ID’s? Would that be an important consideration?
Shimabuku: Yes, I would think so.
[Counsel]: And if somebody comes in with a fake ID, then the identification and sign in requirement is rather meaningless, isn't it?
Shimabuku: I think they would, if they were trained to identify fake ID’s, then something could be done. The person could be prevented from coming in for the hearing.
(Emphasis added.)
Sakata testified that the staff had no training in recognizing fake ID's:
[Counsel]: Now what training is given to your staff to identify false or fake identification?
Sakata: No formal training.
[Counsel]: Any informal training?
Sakata: Only the common sense thing about looking at a photo ID and matching it up with the face. ...
(Emphases added.)
Freitas's security expert, retired Chief Naka-mura testified as follows:
[Counsel]: [H]ow hard is it to get a fake identification here in Hawaii?
Nakamura: It's relatively easy based on current technology with computers.
. The relevant statement in Marx’s article states:
In this article I layout some of the conceptual landscape and some research issues. This emphasis is on the cultural level ... more than on describing actual behavior.
(Emphasis added.)
. Finding 12, infra, quotes exclusively from Section B.
. Section B, "Socially Sanctioned Contexts of Concealment and Revelation" is composed of Part 1, Rationales in Support of (Full or Partial) Anonymity and Part 2, Rationales in Support of Identifiability. Footnote 4, in the Introduction to Section B, states as follows:
I make these observations as a social observer and not as a moralist or empiricist (in the sense of subjecting claims to some kind of empirical standard). I argue neither that these justifications are necessarily good, nor that the claimed empirical consequences (and no unintended or other consequences) necessarily follow. To have a pony in those races requires analysis beyond the scope of this paper. Here, I simply take claimed justifications at face value and report them.
*53(Emphases added.)
. Relevant statements in Zimbardo’s article include:
This behaviorally-focused definition [of evil] makes an agent of agency responsible.... It excludes ... the broader, generic forms of institutional evil, such as poverty, prejudice, or destruction of the environment by agents of corporate greed. But it does include corporate responsibility for marketing and selling products with known disease-causing, death-dealing properties, such as cigarette manufacturers, or other drug dealers. It also extends ... to encompass those in distal positions of authority whose orders or plans are carried out by functionaries. This is true of military commanders and national leaders, such as Hitler, Stalin, Mao, Pol Pot, Idi Amin ....
(Emphasis added.)
. For example, the article states that "[t]he 'war on terrorism' can never be won solely by current administration plans to find and destroy terrorists ... "[military commanders ... such as Hitler, Stalin, Mao, Pol Pot, Idi Amin, and others who history has identified as tyrants for their complicity in the deaths of untold millions of innocent people. History will also have to decide the evil status of President Bush’s role in declaring a pre-emptive, aggressive war against Iraq in March, 2002, with dubious justification, that resulted in widespread death, injury destruction and enduring chaos.” (Emphasis added.)
. The hearing officer dismissed ffiese other measures as "expensive and beyond the budget capabilities of the ADLRO." Such fiscal concerns, however, are an irrelevant consideration in the Freitas I test, where the issue is whether the government's regulation is the least restrictive means of achieving its asserted goal.
. Although Freitas argues that "the record utterly fails to support [the] conclusions [of law,]” he does not argue that the hearing officer erred in applying the second Freitas factor concerning the content of the information to be disclosed in the proceeding. Therefore, the hearing officer’s assessment of that factor need not be addressed.
. Taken as a whole, the findings do anything but address the crucial issue. They state that the ID procedure "is a reasonable means” and "provides a deterrent,” (finding 5) "lessen[s] the likelihood” and "may discourage,” (finding 13) "could potentially deter,” (finding 18) and would "provide an additional measure of deterrence,” (finding 21). (Emphases added.) But they do not establish that the ID procedure is the least restrictive "reasonable means” or "deterrent,” or that such a policy is the least restrictive way to "discourage” and “deter" disruptive behavior at hearings.
. In the event that the sixth amendment right to a public trial was denied, then such denial would be "considered a 'structural defect affecting the framework within which the trial proceeds, rather than simply an error in the trial process itself,’ ” State v. Ortiz, 91 Hawai'i at 193, 981 P.2d at 1139 (quoting Arizona v. Fulminante, 499 U.S. 279, 310, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991)) and the case would be "subject to 'automatic reversal.'" Id. (quoting Neder v. United States, 527 U.S. 1,7, 119 S.Ct 1827, 1833, 144 L.Ed.2d 35 (1999)). Thus, insofar as this matter is not a criminal prosecution, the Sixth Amendment and Article I, Section 14 of the Hawai'i Constitution do not apply and Freitas is not guaranteed a Sixth Amendment public trial in this particular administrative hearing.