This is a judicial-disciplinary case. The proceedings commenced with the filing of a complaint with the Advisory Committee on Judicial Conduct (ACJC or Committee) against respondent, Judge Edward J. Seaman, a judge of the Superior Court in Middlesex County. The complaint was made by B.D., a former law clerk of respondent, charging him with judicial misconduct in violation of several canons of the Code of Judicial Conduct and of the Court’s Disciplinary Rules. The gravamen of the complaint was that respondent had abused his authority by mistreating the complainant while she was employed as his law clerk. The mistreatment took the form of various kinds of sexual harassment.
The ACJC issued a presentment in which it found many of the allegations of the complaint to have been established by clear and convincing evidence. The presentment recommended that respondent be publicly censured. Respondent moved for an order dismissing the complaint pursuant to Rule 2:15-13. This Court denied that motion and simultaneously issued an Order to Show Cause why respondent should not be disciplined.
I
This matter first arose when, in August 1989, respondent’s law clerk, B.D., filed a complaint with the Affirmative Action Officer for the Middlesex County Court House. The complaint, denominated an “Affirmative Action Complaint,” alleged that during the course of B.D.’s clerkship, respondent had engaged in a pattern of abusive behavior consisting of sexual harassment of complainant. *73According to the complaint, respondent repeatedly made remarks of a sexual nature to complainant. The complaint also alleged that respondent had repeatedly touched complainant in an inappropriate manner.
That complaint was eventually referred to the ACJC, which interviewed complainant on October 25, 1989. The interview was reduced to writing. As a result of complainant’s interview, the ACJC lodged a formal complaint against respondent charging him with violating the Code of Judicial Conduct. The basis for the charges was the alleged acts of sexual harassment set forth in the Affirmative Action Complaint and further described in the interview. The complaint alleged that by engaging in that course of conduct respondent had violated: Canon 1, “A judge should uphold the integrity and independence of the judiciary”; Canon 2, “A judge should avoid impropriety and the appearance of impropriety in all activities”; Canon 2A, “A judge should respect and comply with the law and should act at all times in a manner that promotes public confidence in the integrity and impartiality of the judiciary”; Canon 3A(3), “A judge should be patient, dignified, and courteous to litigants, jurors, witnesses, lawyers, and others with whom the judge deals in an official capacity ... ”; and Canon 3A(4), “A judge should be impartial, and should not discriminate because of race, color, religion, age, sex, sexual orientation, national origin, marital status, socioeconomic status, or handicap.” Additionally, the complaint alleged that respondent’s actions had violated Rule 2:15-8(a)(6), as “conduct prejudicial to the administration of justice that brings the judicial office into disrepute.”
The ACJC held hearings on March 11 and April 16, 27, and 29, 1992, to investigate the allegations against respondent. Although authorized “to conduct formal hearings with three members in attendance” (R. 2:15-3(b)), no fewer than six of the eight members who participated in the matter were present at any time. At those hearings, the witnesses presented against respondent were complainant; Susan Leib and Robin Pedersen, law clerks of the assignment judge of Middlesex County; and complainant’s mother, C.D. Testifying for the respondent were himself; his wife; *74Judge Breitkopf, the assignment judge; William F. Lamb and Stephen Leary (respectively a prosecutor and a private attorney with whom complainant had interviewed for positions); and Grace Berrue, Nancy Malkiewicz, and Joseph Hixon, members of respondent’s office staff.
Applying a “clear-and-convincing” standard to the evidence adduced, the ACJC, as noted, found that respondent had engaged in a great many of the separate incidents of sexual harassment set forth in the complaint and, by that conduct, had violated Canons 1, 2, 2A, 3A(3), and 3A(4) of the Code of Judicial Conduct, as well as Rule 2:15-8(a)(6). The ACJC recommended that respondent receive a public censure. One member of the ACJC, who concurred in the recommendation of a public censure, found that only three incidents of sexual harassment had been established by clear and convincing evidence.
II
A.
Matters of judicial discipline brought before this Court on the presentment of the ACJC receive a de novo review of the record and are subject to a clear-and-convincing standard of proof. See, e.g., In re Collester, 126 N.J. 468, 476, 599 A.2d 1275 (1992) (applying “clear-and-convincing” standard in assessing evidence in case of judicial discipline). Clear-and-convincing evidence is “that which ‘produce[s] in the mind of the trier of fact a firm belief or conviction as to the truth of the allegations sought to be established,’ evidence ‘so clear, direct and weighty and convincing as to enable [the factfinder] to come to a clear conviction, without hesitancy, of the precise facts in issue.’” In re Boardwalk Regency Casino License Application, 180 N.J.Super. 324, 339, 434 A.2d 1111 (App.Div.1981), modified, 90 N.J. 361, 447 A.2d 1335 (1982) (quoting Aiello v. Knoll Golf Club, 64 N.J.Super. 156, 162, 165 A.2d 531 (App.Div.1960)); see R. Biunno, Current N.J. Rules of Evidence, comment 6 on Evid.R. 1(4) (1993). In our review of a judicial-disciplinary matter, we must engage in an independent *75consideration of the record and determine, as a matter of first impression, the material facts that have been established by clear and convincing evidence. Because the focus of judicial-disciplinary matters is necessarily on the canons of judicial conduct, we inquire into whether the facts as determined demonstrate conduct on the part of the respondent that is incompatible with those canons.
B.
We must state, prefatorily, that the inquiry before the Court is not whether respondent’s behavior constituted sexual harassment as such. Although undoubtedly all forms of behavior that cross the legal threshold of sexual harassment would constitute judicial misconduct, many forms of offensive interpersonal behavior that would violate the Code of Judicial Conduct would not meet the legal definition of sexual harassment. Nevertheless, we cannot overstress that although we must address the ultimate issue of whether judicial conduct violates the canons, the charges of misconduct against respondent equate with sexual harassment. That form of conduct is personally offensive, highly invasive, psychologically hurtful, and often deeply embarrassing to the victim. For that reason, we have chosen to maintain complainant’s anonymity by referring only to her initials, despite the fact that the charges against respondent have become public and complainant’s privacy has been shattered. Charges of judicial misconduct are ordinarily made public when found in a presentment of the ACJC. The ACJC followed conventional practice and used complainant’s full name in its presentment. In the future, we direct that judicial-disciplinary cases involving abuse of the judicial office through sexual harassment, or other activities that humiliate or degrade those with whom a judge comes into contact, should preserve the anonymity of the alleged victim. The purpose behind that practice is to protect the victim’s privacy and encourage reporting of such offenses.
*76C.
Many of the facts forming the general background to the events that are the subject of the complaint are not controverted or materially disputed. The record reveals that complainant first met respondent in March 1988, when she applied for a position as respondent’s law clerk for the 1988-1989 term. Complainant was hired by respondent and began working in September of 1988.
Complainant’s duties included reading and evaluating pre-trial motions, doing whatever legal research was required, and writing recommendations to respondent on the eases that came before him. Complainant worked in a room with about a dozen other clerks. The room was several yards from the office suite in which respondent worked. In the course of her work, however, complainant frequently visited respondent’s chambers and met with him in his office. Complainant testified at the hearing below that she saw respondent “a couple of times throughout the morning, maybe a couple of times throughout the afternoon” in the course of her working day, although as respondent described his schedule, the frequency with which he saw complainant would have been much lower.
Assisting respondent, in addition to complainant, were several other court personnel. Those persons included Grace Berrue, respondent’s secretary; Joséph Hixon, a court aide for respondent; and Nancy Malkiewicz, a court clerk by designation and respondent’s administrative assistant in handling settlement conferences.
Complainant claimed that respondent’s misconduct took place shortly after the commencement of her clerkship in September 1988 and persisted through June 1989. At the hearing below, however, complainant and respondent gave sharply conflicting accounts of the evidence relating to respondent’s alleged abusive conduct.
Complainant testified that respondent, in October 1988, began directing various remarks of a sexual nature at her. Those remarks, according to complainant, continued throughout her *77clerkship. For example, complainant claimed that respondent had a conversation with complainant, sometime in the spring of 1989, in which he expressed the wish that a pen complainant was holding were actually respondent’s penis. Respondent, according to complainant, boasted of his sexual prowess, asked her to repeat a vulgar sexual remark to him, and assured complainant that, were they to have sexual relations on his desk, he would be sure to avoid a crack on the desk that might scratch her. Complainant stated that although she had disregarded those sorts of comments by respondent, he continued to subject her to such remarks.
In addition to her claims of verbal harassment, complainant testified to improper physical contact by respondent.
Two of those episodes, unlike most of the other incidents to which complainant testified, were witnessed by third parties. In the first incident, complainant was speaking with Susan Leib and Robin Pedersen when respondent came into the room and reached under complainant’s mid-calf length skirt, apparently touching complainant’s knee. When complainant recoiled, respondent left the room. In the second incident, complainant was conversing with Pedersen when respondent entered the room, stood behind her, lifted complainant’s skirt, and examined the back of her knees.
Both Leib and Pedersen testified to the first incident. For the second, only Pedersen was present, but she testified in detail as to respondent’s behavior.
Complainant also testified that on other occasions respondent had initiated unwanted sexual contact with complainant. For example, in one of those episodes complainant alleged that respondent grabbed complainant’s hand and attempted to place it on his crotch. Complainant pulled away before her hand made contact with respondent’s body. In another episode, in the fall of 1988, complainant averred that respondent told her that if she wanted a favorable job recommendation from him she would have to sit next to him on his office sofa.
*78Several witnesses for complainant offered testimony about some of the incidents as related to them by B.D. They also testified about her general behavior, attitude, and demeanor during her clerkship while these incidents of sexual harassment were occurring. Complainant herself stated that respondent’s behavior had embarrassed and troubled her. Complainant’s mother, C.D., testified that although complainant seemed quite happy for about the first month of her clerkship, in October 1988 complainant underwent a marked personality change. C.D. noted of complainant that “[s]he became very quiet, stayed in her room a lot. Cried a lot.” That change continued throughout the year. C.D. also testified that complainant had told her of respondent’s salacious remarks about the pen, about his sexual prowess, and about respondent’s attempt to place complainant’s hand on his crotch. Complainant related those incidents to her mother during the fall of 1988, roughly contemporaneous with the events of which she complained.
Complainant’s colleagues, Susan Leib and Robin Pedersen, also testified that complainant had related some of those episodes to them, although not contemporaneously. Both Leib and Pedersen, who witnessed the first skirt incident, suggested that complainant was somewhat bashful and even-naive about sexual matters. Complainant was given the moniker “Sister B.” because, as one witness explained, complainant was “very straight ... [and] rather naive and she was very modest and wouldn’t use swear words, and if people told off-colored [sic] jokes she didn’t understand them, and always kind of saw only the good side to jokes and things like that.”
Leib noted that complainant had told her of respondent’s “pen” remark, and also testified that complainant had told her of respondent’s remark about-joining him on his office couch. Pedersen also testified to being told, by complainant, about the pen and couch remarks. Pedersen, however, also recalled complainant telling her about respondent’s remark about the scratch on his desk and respondent’s request that complainant repeat a “dirty word” to him.
*79According to the testimony, in June 1989, complainant viewed a film on sexual harassment sponsored by the Affirmative Action Office of the state judiciary. As complainant recounted it, the film precipitated the realization that she had been the victim of respondent’s sexual harassment. After speaking about several of the incidents with Susan Leib and Robin Pedersen, complainant requested and received from Judge Breitkopf a transfer from her position with respondent. Judge Breitkopf testified that complainant “made allegations that seemed to [him] to be sufficient that [he] should do something about them,” but Judge Breitkopf s testimony does not further specify those allegations. Complainant thereafter brought the matter to the attention of the Affirmative Action Officer assigned to the Middlesex County Court House. According to her testimony, she later typed out a complaint, labeled “Affirmative Action Complaint,” which she brought to the attention of Judge Breitkopf.
Respondent, for his part, denied all the allegations against him. However, in addition to making general.denials, respondent took issue with several specific allegations of complainant.
With respect to the first “skirt incident” respondent claimed that a paper had fallen to the floor, he had bent down to retrieve it and accidentally startled the complainant. Respondent insisted that he had never touched complainant during the episode. When pressed on why Susan Leib and Robin Pedersen would testify that he had placed his hand under complainant’s dress and had touched her knee, respondent claimed that he did not know why, but that Leib and Pedersen might have been biased against him because complainant had told them that respondent was speaking ill of them to other lawyers.
Although respondent’s staff claimed that they had never seen respondent improperly touch complainant, Joseph Hixon and Nancy Malkiewicz did recall that respondent had put his arm around complainant’s shoulders, and that that was characteristic behavior for respondent. Moreover, Robin Pedersen, when cross-examined on her reaction to the first skirt incident, remarked that “though *80extreme,” the lifting of complainant’s skirt was “not out of character” for respondent. Pedersen further testified that respondent told “dirty jokes to the clerks” and once put his hand on her back. Pedersen commented that although she had not been shocked by respondent’s touching her, she had been surprised that he “went that far.”
With respect to complainant’s overall comportment, respondent and his witnesses depicted complainant as overly sensitive to criticism and responding badly to his legitimate complaints about her work. Respondent’s staff did not find complainant to be particularly bashful. Members of the office staff related that for most of the year complainant had said very complimentary things about respondent and had never spoken about allegations of sexual harassment. By the end, however, complainant’s attitude had changed; she had become “arrogant” toward respondent and “didn’t seem too interested in the job.”
Further, respondent claimed that complainant had not expressed any dissatisfaction with her working environment to him. Evaluating complainant’s work for the fall of 1988, respondent deemed complainant’s work “average.” Respondent noted, however, that the quality of complainant’s work had deteriorated as the year progressed. Respondent testified that he had spoken several times with complainant about deficiencies in her work and had explained to complainant that the timing of her job interviews in the spring of 1989 had been inappropriate. According to respondent, the deterioration in complainant’s work product had been accompanied by a change in complainant’s attitude. Respondent testified that complainant began ignoring him when he was present and speaking ill of him to other court personnel.
Respondent also introduced character testimony from his wife and staff. That testimony noted that respondent was a devoted father and husband; a consistent church-goer; one who enjoyed recreational and sports activities; and a competent, conscientious, and hardworking judge who kept a very regular routine. Al*81though respondent admitted that he told off-color jokes, he insisted that he never made such remarks in complainant’s presence.
Respondent, both on direct examination of his own witnesses and on cross-examination of complainant and some of her witnesses, elicited testimony questioning complainant’s general veracity. That testimony concerned three incidents in which complainant was alleged to have been untruthful.
The first incident involved complainant’s assertion, to Susan Leib, that she had been a “runner-up” for a Rhodes Scholarship. Evidence from the Rhodes Committee contradicted that assertion. The second involved complainant’s stating that she was reluctant to accept a job offer from a law firm because her mother had advised her to have a say in choosing her office and her secretary, and an associate of the firm had advised her that the benefits package was inadequate. Both complainant’s mother and the associate denied rendering such advice to complainant. The third episode involved complainant’s statement, when seeking two days off for bereavement, that her grandmother had died. In fact, complainant’s great aunt was the deceased.
Ill
A.
Our findings of fact depend critically and inescapably on our assessment of the credibility of various witnesses. That behooves us to explain our analysis of the issues of credibility and the reasoning that brings us to our determinations of fact.
In a sexual harassment case litigated under the civil rights laws, the impact of respondent’s conduct on the victim is an essential element of the cause of action. See, e.g., Drinkwater v. Union Carbide Corp., 904 F.2d 853, 859 (3d Cir.1990); Andrews v. City of Philadelphia, 895 F.2d 1469, 1482 (3d Cir.1990); Lehmann v. Toys ‘R' Us, 132 N.J. 587, 626 A.2d 445 (1993). In a judicial disciplinary proceeding, the effect of judicial misconduct on other persons is not an essential element, although it can be a relevant *82factor in assessing the gravity of the misconduct and appropriate discipline. E.g., In re Connor, 124 N.J. 18, 26, 589 A.2d 1347 (1991). Thus, we need not decide, for example, whether the harassment complained of created a “hostile working environment” as such for complainant. However, because the claimed misconduct has taken the form of sexual harassment, there are aspects of sexual harassment cases that are germane to our analysis in this case. In particular, in the adjudication of sexual harassment cases, fact-finding poses special concerns. The weighing of evidence based on the assessment of the credibility of witnesses is an integral part of that process. Hence, the growing body of knowledge about sexual harassment and the law surrounding that subject are highly instructive in guiding our determinations of fact in this case.
One of respondent’s major arguments is that the evidence, consisting primarily of complainant’s own testimony, falls far short of satisfying the. clear-and-convincing standard of proof necessary to support the charges of sexual harassment as the basis of judicial misconduct, and, indeed, the evidence is so wanting that the presentment should have been dismissed. Respondent thus stresses that virtually all of the evidence in support of the complaint is uncorroborated; it consists only of complainant’s assertions concerning the occurrence of the incidents and is not bolstered even indirectly by the testimony of other witnesses. Respondent claims that uncorroborated victim testimony cannot meet the clear-and-convincing standard.
As a matter of principle, and as a practical matter in this case, we reject that contention. True, the majority of incidents recounted by complainant were not witnessed by others and, in general, were not directly corroborated. It does not follow from that, however, that the evidence presented at the hearings below could not meet the clear-and-convincing standard of proof.
First, we observe that uncorroborated testimony of a victim is sufficient to meet the law’s highest standard of evidence — guilt beyond a reasonable doubt. See, e.g., People v. *83Martin, 112 Ill.App.3d 486, 68 Ill.Dec. 151, 445 N.E.2d 795 (1983); In the Interest of Winslow, 46 Ill.App.3d 962, 5 Ill.Dec. 299, 361 N.E.2d 622 (1992); King v. State, 598 N.E.2d 589 (Ind.App.1992); State v. Ryan, 233 Neb. 74, 444 N.W.2d 610 (1989); State v. Middelstadt, 579 P.2d 908 (Utah 1978). For crimes of sexual assault, some state courts have explicitly held that the uncorroborated testimony of a victim witness is sufficient, if credited by the jury, to sustain a finding of guilt beyond a reasonable doubt. See, e.g., State v. Cantrell, 234 Kan. 426, 673 P.2d 1147 (1983). That juries may convict on the uncorroborated testimony of an accomplice, and the burden of proof on virtually all other criminal offenses can be met with the1 testimony of a lone victim or witness, is well established. See Anderson v. Bessemer City, 470 U.S. 564, 575, 105 S.Ct. 1504, 1512, 84 L.Ed.2d 518, 529-30 (1985).
We recognize that the most serious forms of sexual harassment are also the least likely to occur in public and, therefore, the least likely to be witnessed by third parties. See generally Susan Estrich, Sex at Work, 43 Stan.L.Rev. 813 (1991). In sexual harassment cases, the Equal Employment Opportunity Commission (E.E.O.C.) itself can find a cause of action based solely on a reasoned decision to credit the charging party’s testimony. E.E.O.C. Guidelines, at 10-13. Further, as one federal district court has stated, “The court is far from suggesting that a plaintiff claiming sexual harassment must produce corroborating evidence in order to prevail. In the nature of things, sexual harassment is very frequently a secret activity, carried on without the potential embarrassment that onlookers might provide.” Tindall v. Housing Auth., 762 F.Supp. 259, 263 (W.D.Ark.1991).
The clear-and-eonvincing evidence standard has also been applied and satisfied in disciplinary proceedings similar to this one. The Washington Supreme Court, in In re Deming, 108 Wash.2d 82, 736 P.2d 639 (1987), found that the clear-and-eonvincing standard does not require that testimonial evidence be corroborated. In holding that, the court relied on In re McDonough, 296 N.W.2d 648 (Minn.1979), in which the Minnesota Supreme Court rejected *84the contention that the clear-and-convincing standard implies corroboration. In Deming, the judge being disciplined also asserted that the standard of proof could not be met when it was a case of “one person’s word against another.” 736 P.2d at 653. In rejecting that position, the Deming court noted that as long as the trier of fact “can impose discipline with clarity and conviction,” ibid., the clear-and-convincing standard is satisfied, and “no mechanistic corroboration requirement is necessary.” Ibid.
We conclude that uncorroborated evidence may satisfy a burden of proof based on the standard of clear-and-convincing evidence.
B.
We are satisfied from our independent review of the record that the evidence demonstrates clearly and convincingly that respondent engaged in a course of conduct constituting an abuse of his authority as a judge with respect to his supervisory responsibilities over an employee. That conduct consisted of a pattern of sexually harassing behavior that was personally offensive to his employee and inimical to her dignity, privacy, and emotional well-being. We conclude that respondent has thereby violated the Code of Judicial Conduct and the Court’s Disciplinary Rules.
We base that conclusion on the existence of a number of incidents of harassment that collectively and cumulatively indicate and demonstrate a pattern of conduct that consists of repeated acts of sexual harassment and, indeed, imply that other incidents of similar behavior likely occurred whether or not specifically alleged in the complaint or recounted in the testimony.
The incidents that most readily and obviously establish such a pattern of repeated behaviors are those the occurrence of which were directly corroborated by third persons or indirectly corroborated by complainant’s reporting the occurrences of those incidents to other persons. Specifically, Susan Leib and Robin Pedersen both witnessed respondent reaching under complainant’s skirt. Leib testified that as complainant was leaning against a filing *85cabinet in the clerks’ office, respondent “came over to her ... [and] put his hand on her knee or something under her dress.” Pedersen verified that respondent had “reached his hand up under [complainant’s] skirt,” but thought respondent might have touched complainant’s thigh. Pedersen also testified that, later on the same day as the first skirt incident, respondent had “lifted up the back of [complainant’s] skirt and looked at her legs.”
As to the indirectly corroborated incidents — the “pen”, “crotch”, “couch”, “sexual prowess”, “desk”, and “dirty word” incidents — in sexual harassment eases, the victim’s communication of the alleged harassment to others can serve to corroborate or support testimony of those events. Thus, in Spencer v. General Electric Co., 697 F.Supp. 204, 207 (E.D.Va.1988), testimony from a co-worker partially corroborated the plaintiffs complaint of sexual harassment by their supervisor, where the co-worker testified that on one occasion the plaintiff had come to his office, upset and frightened, and had told him that the supervisor had asked her to sleep with him and on several occasions thereafter she had complained to the co-worker of the supervisor’s pressure to have sex with him. In Lehmann v. Toys ‘R’ Us, the victim on several occasions communicated to others the occurrence of incidents of sexual harassment. 132 N.J. at 596, 626 A.2d 445.
We also note that some other state courts apply the “totality-of-the-evidence” standard in disciplinary proceedings. That standard allows uncorroborated victim-witness evidence to be validated by evidence that is corroborated. See, e.g., Deming, supra, 736 P.2d at 657. A corollary of that principle is that corroboration of some evidence can serve indirectly to corroborate or validate other evidence. In this case, we find that the testimony of the two law clerks about the “skirt incidents” bolsters the credibility of complainant’s testimony with respect to both the specific incidents that were the subject of her testimony and the pattern of sexual harassment generally.
Here, complainant did express to her mother, C.D., that respondent had been making lewd remarks to her and that those *86remarks had left complainant deeply disturbed. Similar sentiments were expressed by complainant to Susan Leib and Robin Pedersen. We find that the testimony of C.D. and the two law clerks bolsters the credibility of complainant’s testimony with respect to both the specific incidents that were the subject of the testimony and the pattern of sexual harassment generally.
In reaching that conclusion we have considered and weighed with great care respondent’s attacks on complainant’s credibility on additional grounds. On balance, we do not find that the several factors stressed by respondent impugn complainant’s credibility.
Respondent points to complainant’s statements about the Rhodes Scholarship, the law firm offer, and the death of her grandmother to argue that complainant is untruthful. We find that complainant greatly exaggerated or distorted the truth with respect to those incidents. They demonstrate that complainant is not a scrupulously truthful person and is prone to exaggerate in a self-serving way; they do not demonstrate that she would, or did, lie about something as profoundly destructive of another person as engaging in an unremitting course of sexual harassment. Consequently, we reject respondent’s contention that those incidents compel the sweeping judgment that all complainant’s specific allegations against him are untrue and that a pattern of sexually offensive conduct did not occur.
Respondent further contends that complainant’s testimony was too vague to be believable. The vagueness of her testimony, he urges, greatly undermines the credibility of complainant’s allegations.
Respondent’s argument, in that respect, correctly focuses the analysis on the intrinsic quality of complainant’s testimony. Vagueness or lack of specificity is one aspect of the quality of testimony and its inherent believability. But we note from the outset that our finding of misconduct consisting of a pattern of repeated acts of sexual harassment, on the part of respondent, is directly supported by several specific events that were corroborated, and indirectly and inferentially supported by the likely occur*87rence of other similar events. Nevertheless, the testimony of complainant with respect to such other uncorroborated incidents, though not the subject of specific findings, bear on her credibility.
As we review the record, the vagueness in complainant’s testimony with respect to all of the incidents of sexually offensive behaviors appears to be largely chronological in the sense that the testimony is not precise with respect to the times and dates on which particular incidents occurred. However, complainant’s allegations are not vague with respect to the general periods or the sequence in which incidents occurred.
In adjudicating sexual harassment cases, courts have found that imperfect recall years after a series of events has occurred is not necessarily discrediting. See, e.g., Ashway v. Ferrellgas, 59 F.E.P. 375, 1989 WL 384851 (1989), aff'd in part, rev’d on other grounds, 945 F.2d 408 (1991). In complainant’s case, given the passage of three years between the alleged events and her testimony, complainant’s ability to place events in the time frame of certain months or seasons demonstrates a recall of chronology not nearly as inadequate as respondent claims. The trauma and embarrassment of the events, coupled with a good deal of internalized denial on complainant’s part, may have contributed to the defects in her recollection. We therefore do not find that this form of Vagueness is dispositive of whether complainant’s testimony is worthy of belief.
We also note that complainant’s testimony about many of the alleged incidents is, apart from specific dates and times, quite detailed, e.g., a remark about a crack on respondent’s desk, his comments about his sexual prowess, and the like. In sexual harassment cases, factors adding to the credibility of a complaining witness include the level of detail that is recalled in the recounting of the incident. See, e.g., Christoforou v. Ryder Truck Rental, Inc., 668 F.Supp. 294, 298-99 (S.D.N.Y.1987) (finding plaintiffs testimony to be essentially credible when plaintiff testified to a fairly detailed incident; court did not credit testimony *88about subsequent incidents that were “extremely vague and unspecific”).
In addition, complainant’s testimony was quite consistent. Her testimony did not vary significantly from the versions of the events originally set forth in her Affirmative Action Complaint and later in the interview. Neither of those documents, although marked in evidence, do we consider probative in themselves. Nevertheless, considered only as prior statements, they do not have any impeaching impact. Consistency of testimony, both internally and between witnesses, is an important indicator of truthful testimony. Christoforou Truck Rental, supra, 668 F.Supp. at 299; see Jackson v. Veteran’s Admin., 768 F.2d 1325 (Fed.Cir.1985) (crediting testimony about incident of sexual harassment because of its consistency); Anderson, supra, 470 U.S. at 575, 105 S.Ct. at 1512, 84 L.Ed.2d at 529-30 (in the criminal context, when witness has told a coherent, plausible story that is internally consistent and uncontradicted by extrinsic evidence, that story may be credited).
Respondent also stresses complainant’s motive to he because she was angered and disappointed by her perception that he refused or neglected to give her a favorable job recommendation. However, complainant’s testimony exhibited characteristics inconsistent with the fabrication of a vindictive employee. For example, complainant’s testimony reveals a systematic understatement of the allegations against respondent. Complainant’s testimony concerning certain events — the skirt incidents, the attempted contact with respondent’s crotch and the couch comment — was less inflammatory or extreme than the testimony of the witnesses who had observed those events or had those events recounted to them by complainant. The refusal to conform her testimony to the more serious versions proffered by other witnesses does not suggest making up charges to destroy her former employer.
Respondent further argues that complainant cannot be believed because her conduct throughout was inconsistent with that of a *89person who, as complainant claimed to have been, had been victimized continuously and maliciously over a long period of time.
Respondent points out that complainant never expressed her dissatisfaction with respondent’s alleged behavior until June 1989. Further, not only did complainant not express dissatisfaction, she actually conveyed a sense of good will and respect for respondent for most of her clerkship. Respondent notes, for example, that complainant brought him gifts back from her vacation in New Mexico, asked his advice on a law firm job offer, requested that he swear her into the bar, invited him to a family luncheon that followed her swearing in, and continued to show him small kindnesses around the office, like asking him to a clerks’ pizza party and bringing baked goods to the office to share with respondent.
Complainant acknowledged that she had not reported being harassed by respondent and had shown respondent various kindnesses. However, complainant also stated that the reason for those behaviors toward respondent was her desire to finish her clerkship. As complainant noted, keeping her first job out of law school, which could determine her future opportunities, was very important to her. Complainant also testified that after the fall she began to avoid respondent.' There was testimony from respondent’s staff that, later in the clerkship, complainant spent large amounts of time in the courtrooms of other judges and was rarely around respondent, except when her duties demanded it. Respondent himself noted that complainant began ignoring him and spending a great deal of time watching trials in another judge’s courtroom. Those actions are entirely consistent with the behavior of someone seeking to avoid respondent.
Complainant’s failure to notify others of respondent’s harassing behavior; her continuing to serve as respondent’s clerk even though he was allegedly subjecting her to degrading remarks and unwanted physical contact; her apparently pleasant demeanor toward respondent, including bringing respondent small gifts and offering social invitations — all are aspects of the larger phenomenon of nonreporting. In understanding complainant’s failure to *90report respondent’s misconduct, we note that the E.E.O.C. has found that “the fact that an employee has not promptly reported the sexual advance is not dispositive of the issue of whether or not [the advances] were unwelcome.” Ernest G. Hadley, A Guide to Federal Sector Equal Employment Law & Practice, Ch. VI, B(5) (1990). In sexual harassment litigation, courts and agencies have admitted expert testimony that reticence about complaining is common in cases of sexual harassment. See Robinson v. Jacksonville Shipyards, 760 F.Supp. 1486, 1506-07 (M.D.Fl.1991) (admitting and crediting consultant’s testimony that women may not complain about sexual harassment because of fear, embarrassment, and feelings of futility). The E.E.O.C., in a footnote to Trammel v. Postmaster General, 01871154, 1922/B14 (1988), noted that “[o]ne can be offended by certain conduct while still being reticent to report it for fear of losing one’s job, creating controversy, not being believed, not being supported, or generally making matters worse.” One federal appeals court has described the peculiar problems faced by victims of sexual harassment as a “cruel trilemma ... [i]n which a victim must choose among acquiescence in the harassment, opposition to it, or resignation from her job.” Vinson v. Taylor, 753 F.2d 141 (D.C.Cir.1985) (citing Bundy v. Jackson, 641 F.2d 934, 946 (D.C.Cir.1981) (ruling that in D.C. Circuit plaintiffs would not be required to resist their harassers).
Surveys of victims’ reactions confirm that reasoning. One study has reported that although 53.1% of women surveyed had experienced sexual harassment at work, only 22.5% of the total survey respondents reported ever having discussed the general subject matter with a co-worker. Fewer than one in five women who had experienced sexual harassment at work ever reported it to any authority. Barbara Gutek, Sex and the Workplace 46, 54 (1985). Silence may well signal the shame, humiliation, fear, and dependence of the victim. Susan Estrich, Sex at Work, supra, 43 Stan.L.Rev. at 829. The extent to which a woman may react to insults, propositions, and even physical abuse may have less to do *91with the severity of the harassment than with the woman’s need to keep her job. Id. at 846.
A study directed by Governor Florio, through Executive Order 88, offered many of the same reasons for underreporting. That study found that “[fjear of retaliation, loss of privacy, and unease about confidentiality” along with a “lack of faith in the complaint process” played a significant part in the phenomenon of underreporting in sexual harassment cases. Review Committee on Sexual Harassment, New Jersey Department of Personnel, People Working Together: A Report on Sexual Harassment, section V (1993) (hereafter, “Report on Sexual Harassment”).
Moreover, a survey of similar cases from other jurisdictions reveals that complainant’s apparently-inconsistent behavior is not at all uncommon. See, e.g., In re Seitz, 441 Mich. 590, 495 N.W.2d 559 (1992) (describing disciplinary findings against state court judge, including sexual misconduct, in which court secretary, with whom judge was romantically obsessed, continued to work with judge and support him in his disputes with chief judge until such time as employee left for another position, complaining that judge had made her remaining impossible); In re Buchanan, 100 Wash.2d 396, 669 P.2d 1248 (1983) (describing disciplinary proceedings against state court judge involving sexual misconduct toward women clerks and other court employees, “the employees tolerated this conduct out of fear of reprisals”); see also In re Deming, supra, 108 Wash.2d 82, 736 P.2d 639 (describing disciplinary proceedings against state court judge, involving, inter alia, sexual misconduct, in which court observed that “several witnesses [to judge’s improper touching of a female employee] indicated that they felt unable to respond to Judge Deming’s harassment because they were intimidated by his authority”). A court has found that judicial misconduct occurred even when the victim arguably placed himself at risk. In re Miera, 426 N.W.2d 850 (Minn.1988) (finding sexual misconduct where the victim, a court reporter, complained that state court judge had made sexual advances toward him on an evening when judge stayed at victim’s apart*92ment; victim allowed judge to stay over the very next week, at which time judge again made sexual advances).
Respondent also challenges complainant’s contention that she did not recognize that she was a victim of sexual harassment until seeing a film on the subject. Respondent claims that complainant, as a law student, must have known the definition of sexual harassment long before seeing a film on the subject. Thus, complainant must not have told the truth in her account of how she came to file a complaint against respondent.
That conclusion does not seem necessary to us. Complainant, on cross-examination, explained that she did not recognize, until seeing the affirmative action film, that respondent’s offensive conduct met the legal criteria for sexual harassment. Complainant’s explanation does not suggest that she did not consider herself to have been abused and victimized by respondent’s conduct. Rather, it implies that she did not fully appreciate the special nature of her victimization and its legal implications. Not every law student has studied the elements of a sexual harassment claim. By the time of the affirmative action seminar, June 5,1989, in which she saw the film, complainant’s distress may have grown to the point that it overcame her fear of the harm to her career that such a complaint might bring. The film, besides communicating factual information about what constitutes sexual harassment, very likely may have had the effect of persuading complainant that she need not and ought not tolerate respondent’s conduct. Affirmative action seminars are meant not only to inform but also to embolden and empower employees about the respectful and fair treatment that is due them. Indeed, complainant’s testimony about the effect of the film confinns that purpose: “I believe [the film] gave me the confidence to comply with the sexual harassment laws and to report the sexual harassment that I had experienced.” As a State Commission on Sexual Harassment has found:
[Harassment and other forms of employment discrimination are so pervasive that women, particularly, do not recognize sexual harassment as each episode occurs. Rather, something dramatic such as an unexpected termination or denial of *93promotion will trigger the awareness of a pattern of abuse that will act as a catalyst for filing a complaint. [Report on Sexual Harassment, sec. V.]
In complainant’s ease, we believe the affirmative action film acted as just such a catalyst for the recognition that she had been subjected to a pattern of abusive conduct by respondent.
In considering the failure of complainant to report respondent’s misconduct sooner, we note that complainant did not work directly with other clerks, and was not on particularly close terms with respondent’s secretary, the court clerk, or other judges’ clerks. Therefore, contemporaneous complaints or disclosures to those persons could not be expected and their absence does not demonstrate that complainant fabricated her allegations against respondent. However, the fact that complainant disclosed several incidents of respondent’s harassment to Susan Leib and Robin Pedersen, after she had become friendly with them, lends credibility to complainant’s testimony. The versions of Leib and Pedersen, with regard to those reported incidents, are entirely consistent, and are also largely consistent with complainant’s testimony.
Complainant’s testimony at the hearings below confirms the impressions of Leib, Pedersen, and C.D., complainant’s mother, that complainant was ashamed and upset about respondent’s conduct. Those feelings may well have contributed to complainant’s reluctance to report respondent’s misconduct. Complainant’s testimony, however, adds another dimension to our understanding, of why complainant failed to expose respondent sooner: the vulnerability she felt as a judicial clerk. During cross-examination by respondent’s counsel about her failure to do anything sooner about respondent’s alleged harassment, the following colloquy took place:
Q: Did you tell anybody about the remarks or actions you have just described?
A: No.
Q: Why not?
A: I was confused; I was embarrassed; I was humiliated; I didn’t really know what was going on, and I just wanted to take the first job that I got out of law school and do the best that I could at it.
Q: Did you give any thought to quitting your clerkship?
*94A: Starting right out of law school, I didn’t really want to quit the first job that I ever had.
Plaintiffs reticence is particularly understandable in the context of a judicial clerkship. The judge-clerk relationship is unique. The importance of a judicial clerkship to the career of a young lawyer is enormous. A judicial clerkship can be an auspicious beginning to a legal career. See Trenton H. Norris, The Judicial Clerkship Selection Process: An Applicant’s Perspective on Bad Apples, Sour Grapes, and Fruitful Reform, 81 Calif.L.Rev. 401, 403-04 (1993) (describing importance of clerkship to future careers of new lawyers). Judicial clerkships are marked by both strong dependence and a significant power imbalance between judge and clerk. See Paul R. Baier, The Law Clerks: Profile of an Institution, 26 VandL.Rev. 1125, 1129-31, 1153-61 (1973). The vulnerability of a clerk to a judge is even greater than that in most supervisor-employee relationships. By alienating his or her judge, a clerk risks great professional jeopardy.
Accordingly, we find that complainant’s failure to complain sooner, and her continued kindnesses toward respondent, can be viewed as a need to maintain appearances, and, indeed, served to mask her resentment of respondent’s offensive conduct. That conclusion is strongly supported by the power dynamics inherent in a judicial clerkship.
In determining the issues of credibility, we are further impelled to remark on the adequacy of the written record in the proceedings before us. In making findings of credibility, observing live testimony is doubtless the ideal. That being acknowledged, we are nonetheless convinced that the written record before us has provided an ample basis for an accurate assessment of the credibility of those who testified, particularly of respondent and complainant.
We appreciate that issues of credibility are not like problems in geometry. When articulating the reasons why we have found a solid basis for believing a witness, we do not presume that any single argument will definitively settle the issue or resolve all possible doubts. Arguments for credibility are almost always *95probative, almost never dispositive. The best one can do, when justifying why belief is invested in one witness rather than another, is to offer a plausible explanation.
Although our determination that respondent engaged in a pattern of misconduct in based upon our finding complainant and her witnesses more credible than respondent, we nonetheless emphasize that our finding today is not the result of taking sides in a “swearing contest.” Rather, the episodes that we have found clearly and convincingly to demonstrate a pattern of misconduct— the two “skirt” incidents; the remarks about the pen, couch, respondent’s sexual prowess, “scratch on the desk”, and “dirty word”; and the “crotch” incident — had some direct or indirect corroboration by third parties. Together with the general credibility of complainant’s testimony with respect to other incidents, they provide firm support for our ultimate conclusion that there were repeated acts of a sexually offensive nature by respondent.
C.
Based upon those findings, we conclude that respondent abused his judicial and official authority as a judge with respect to his supervisory responsibilities over an employee. That misconduct is evidenced by a pattern of behavior that was offensive and inimical to the employee. That misconduct violates Canons 1, 2, 2A, 3A(3), 3A(4) of the Code of Judicial Conduct and Rule 2:15-8(a)(6).
The Code of Judicial Conduct comprises seven canons that provide guidance on the manner in which judges are to comport themselves. The canons, although rules to be enforced, also exhibit an aspirational and hortatory character. Because of that, the behaviors encompassed by each canon are not separable into rigid and distinct categories. Nevertheless, the canons are not mere platitudes. They direct each judge in conducting himself or herself in office, and guide the Court in determining when judicial misconduct has occurred.
*96The ultimate and permeating objective of the canons, however, is to maintain the integrity of the judiciary and public confidence in that integrity. Accordingly, the canons evidence concern not only for the reality of judicial integrity, but for the appearance of that reality. See Pressler, Current N.J. Court Rules, comment on Canons 1 and 2 (1993). It is obvious from the canons of the Code of Judicial Conduct that integrity — in both actuality and appearance — can be maintained only if judges demonstrate probity, impartiality, and diligence. In this case, there is no suggestion that respondent showed bias in the exercise of his judicial duties or neglected those duties in any manner. The record, however, gives clear and convincing evidence that respondent showed a lack of uprightness in his dealings with complainant. In that way, respondent compromised the integrity of the judiciary and engaged in manifest improprieties in violation of Canons 1, 2 and 2A. Moreover, although we have noted that respondent maintained his independence when adjudicating cases before him, the attentiveness, consideration and fairness enjoined by Canon 3A(3) and 3A(4) include treating all those with whom a judge deals, including employees, in a dignified and nondiscriminatory fashion. That, clearly, respondent did not do. It is also painfully manifest that respondent’s conduct casts disrepute on the judiciary and was highly prejudicial to the administration of justice. R. 2:15-8(a)(6).
Respondent is guilty of judicial misconduct in violation of the Code of Judicial Conduct and our Disciplinary Rules. What remains to be determined is the appropriate discipline to be imposed.
IV
The single overriding rationale behind our system of judicial discipline is the preservation of public confidence in the integrity and the independence of the judiciary. See In re Coruzzi, 95 N.J. 557, 579, 472 A.2d 546 (1984); In re Spitalnick, 63 N.J. 429, 431, 308 A.2d 1 (1973). As we have noted before, “This Court *97cannot allow the integrity of the judicial process to be compromised in any way by a member of either the Bench or the Bar.” Spitalnick supra, 63 N.J. at 431, 308 A.2d 1. Accordingly, institutional concerns figure prominently in cases involving judicial discipline. As the Supreme Court of Minnesota has observed, the standard of judicial conduct is a high one precisely “so that the integrity and independence of the judiciary may be preserved.” Miera, supra, 426 N.W.2d at 855. Judicial misconduct “brings the judicial office into disrepute and thereby prejudices the administration of justice ... and diminishes public respect for the judiciary.” In re Winton, 350 N.W.2d 337, 340 (Minn.1984). Because public confidence in judges is essential to maintaining the legal system, “misconduct by a judge brings the office into disrepute and thereby prejudices the administration of justice.” Ibid.
Consonant with those institutional concerns, the determination of sanctions in judicial-discipline cases is not so much to punish the offending judge as to “restore and maintain the dignity and honor of the position and to protect the public from future excesses.” Buchanan, supra, 669 P.2d at 1250. We concur with the Supreme Court of Maine in its description of the fundamental purpose behind disciplining a judge:
[T]o instruct the public and all judges, ourselves included, of the importance of the function performed by judges in a free society. We discipline a judge to reassure the public that judicial misconduct is neither permitted nor condoned. We discipline a judge to reassure the public that the judiciary of this state is dedicated to the principle that ours is a government of laws and not of men. [In re Ross, 428 A.2d 858 (Me.1981).]
In keeping with that high purpose, we do not discipline for “mere error[s] in judicial activity or professional activities.” In re Mattera, 34 N.J. 259, 270, 168 A.2d 38 (1961); see also In re Alvino, 100 N.J. 92, 494 A.2d 1014 (1985) (holding that omission of certain administrative duties, when no willfulness was involved and judge sincerely thought matters unimportant, did not warrant judicial discipline). Rather, the disciplinary power is ordinarily reserved for conduct that “is marked with moral turpitude and thus reveals a shortage in integrity and character.” Mattera, supra, 34 N.J. at 70, 168 A.2d 38. Thus, we have held that *98“^Judicial misconduct, for example, involving dishonesty of any kind will ordinarily require removal as the appropriate discipline.” Id. 100 N.J. at 97, 494 A.2d 1014; accord Coruzzi supra, 95 N.J. at 566, 472 A.2d 546 (holding that without exception, a judge who accepts a bribe must be removed from office).
The determination of appropriate discipline requires more than establishing some instance or instances of unethical conduct. See Collester, supra, 126 N.J. at 472, 599 A.2d 1275. We must undertake “a more searching and expansive inquiry ... carefully scrutinizing] the substantive offenses that constitute the core of respondent’s misconduct, the underlying facts, and the surrounding circumstances in determining the nature and extent of discipline.” Ibid. Among the surrounding circumstances to which we give heed are the considerations of public policy and the legitimate interests to which the State of New Jersey has made a governmental commitment. Id. at 473, 599 A.2d 1275.
The commitment of this State and its judiciary to end gender discrimination — and one of its most egregious expressions, sexual harassment — clearly weighs heavily in our determination of the discipline to be imposed on respondent. See Lehmann v. Toys ‘R’ Us, supra, 132 N.J. at 609-610, 626 A.2d 445 (noting, in context of sexual harassment case, that Legislature has declared that “discrimination is ‘a matter of concern to the government of the State, and that such discrimination threatens not only the rights and proper privileges of the inhabitants of the State but menaces the institutions and foundation of a free democratic State’ ”) (citation omitted). The Commissioner of Personnel for the State of New Jersey has described sexual harassment as “a serious problem in our state and our nation. It is behavior that we cannot tolerate.” Report on Sexual Harassment, sec. I.
Other aggravating factors serve to define the gravity of misconduct. Those include the extent to which the misconduct, like dishonesty, or a perversion or corruption of judicial power, or a betrayal of the public trust, demonstrates a lack of integrity and probity, Coruzzi supra, 95 N.J. at 572, 472 A.2d 546; whether the *99misconduct constitutes the impugn exercise of judicial power that evidences lack of independence or impartiality, In re Yaccarino, 101 N.J. 342, 502 A.2d 3 (1988); whether the misconduct involves a misuse of judicial authority that indicates unfitness, ibid.; whether the misconduct,'such as breaking the law, is unbecoming and inappropriate for one holding the position of a judge, Collester, supra, 126 N.J. at 473, 599 A.2d 1275; In re Connor, 124 N.J. 18, 27, 589 A.2d 1347 (1991); whether the misconduct has been repeated, Collester, supra, 126 N.J. at 473, 599 A.2d 1275; and whether the misconduct has been harmful to others, Connor, supra, 124 N.J. at 26-27, 589 A.2d 1347; In re Albano, 75 N.J. 509, 384 A.2d 144 (1978); In re Yengo, 72 N.J. 425, 371 A.2d 41 (1977).
Clearly, respondent has engaged in a most serious form of misconduct. That misconduct involves not only the mistreatment of a person in his employ, but flagrant disregard for the law. Canon 3A(4) directs explicitly that “[a] judge ... should not discriminate because of ... sex.” Sexual harassment of women by men is among the most pervasive, serious, and debilitating forms of gender discrimination.
In assessing aggravating factors, we are also attentive to the harmful impact of the misconduct on respondent’s victim. The record contains testimony that complainant, soon after respondent began his harassing activities, became anxious and depressed. That response is entirely consistent with a growing body of research on the physiological and psychological effects of sexual harassment on its victims. See Report on Sexual Harassment, sec. V (quoting testimony of Myra Terry, President of New Jersey chapter of National Organization of Women, that “[victims of sexual harassment] experience physical symptoms such as problems sleeping, nervousness, headaches, and weight gain or loss ... 90% report suffering psychological stress, 63% physical stress, and 75% find that their work performance is adversely affected”). Respondent has been found to have committed acts of sexual crudity deeply offensive to another person. As the ACJC concluded, respondent “[a]s a judge had the obligation to treat [complain*100ant] with dignity. He did not live up to that obligation.” The harm, as with any act of invidious discrimination in violation of civil rights laws, extends well beyond the victim, but is, as we noted earlier, a threat to democratic society. Lehmann v. Toys ‘R’ Us, supra, 132 N.J. 609-610, 626 A.2d 445.
Additionally, we find especially important the vulnerability of respondent’s victim. See In re Yengo, supra, 72 N.J. at 438, 371 A.2d 41 (finding the vulnerability of victim of judge’s abusive language significant: “She (the victim) was disadvantaged and defenseless ... whereas he was a judge and his conduct must be evaluated as such”) (emphasis in original); see also Albano, supra, 75 N.J. at 514, 384 A.2d 144. As we have noted, judges and their clerks have a relationship unique in our profession. The clerkship has become an informal but fixed institution on which the judiciary critically depends. If women are to gain their rightful representation in the legal profession, they must not be subjected to discrimination, particularly in its most egregious forms. We are well aware that sexual harassment, like other forms of discrimination, debilitates its victims and has, as its ultimate effect, the continued subordination of women.
With respect to mitigating factors, we generally are mindful that a matter represents the first complaint against a judge, of the length and good quality of the judge’s tenure in office, of exemplary personal and professional reputation, of sincere commitment to overcoming the fault, of remorse and attempts at apology or reparations to the victim. Connor, supra, 124 N.J. at 26, 589 A.2d 1347. We have also found relevant consideration of whether a judge found guilty of misconduct will engage in similar misconduct in the future, or whether the inappropriate behavior is susceptible to modification. Ibid.
Some of those factors in mitigation bears on the sanction to be imposed on respondent. We note that he has long served on the bench with distinction. Respondent enjoys an outstanding personal and professional reputation. This complaint is the first charge of misconduct to be brought against respondent.
*101Nevertheless, in assessing mitigating factors, we must note that respondent has not acknowledged his wrongdoing or expressed contrition. Throughout the proceedings, respondent steadfastly maintained his innocence, unequivocally denying that the allegations against him were true. Additionally, respondent tried to cast blame on his victim, depicting her as vindictive and emotionally unstable. Respondent also suggested that two of the witnesses to one of his acts of misconduct, also women clerks, were lying under oath for malicious purposes. We do not penalize respondent for defending himself, but we cannot give respondent the benefit of contrition as a mitigating factor.
Further, although the offensive behavior found to have occurred here was not the second or third time in which respondent was found guilty of such misconduct, it did involve a prolonged and continuing course of behavior. See Corvelli v. Board of Trustees, 130 N.J. 539, 547-48, 617 A.2d 1189 (1992) (holding that in case of police chief who persecuted inferior officer by assigning him to patrol dangerous park' each night, misconduct actually involved multiple, daily decisions and constituted course of conduct rather than isolated incident).
Although we find respondent’s long service on the bench, exemplary reputation, and prior record as factors in mitigation of sanction, those factors are outweighed by the recurrent nature of respondent’s course of misconduct, the severity of the harm inflicted by that misconduct on his victim and on public trust in the judiciary, and the imperative of public policy — the overriding social goal of achieving a society where all citizens are treated with respect regardless of gender — that respondent’s misconduct so obviously violates. Accordingly, we amend the sanction imposed by the ACJC, public censure, and direct that respondent be suspended, without pay, from judicial office for a period of sixty days. During his suspension, respondent is to complete an educational program designed to heighten awareness of what constitutes sexual harassment, and to reinforce what sort of behavior is expected by the sexual harassment policy of the judiciary of this *102State. That program will be approved by the Court, which will also verify respondent’s compliance. Failure to adhere to that directive will precipitate action to impose more severe discipline.
In our system of judicial discipline, suspension stands in order of severity between a censure and permanent removal from judicial office. See Connor, supra, 124 N.J. at 28, 589 A.2d 1347. In Connor the offense was very serious — drunk driving — but we “decline[d] to impose a suspension ... because of his good record as a judge and because his transgressions do not directly affect the performance of his judicial duties.” Ibid. Nevertheless, in Connor the offending judge acknowledged his guilt, made a public apology, and exhibited “genuine self-confrontation and commitment to rehabilitation.” Id. at 26, 589 A.2d 1347. Those .factors, which weighed heavy in the Court’s determination of sanction in Connor are regrettably and manifestly absent in respondent’s case.
In In re Collester, supra, 126 N.J. at 468, 599 A.2d 1275, a judge was convicted of a second drunk driving offense. There we emphasized that the repeated nature of the judge’s misconduct demanded discipline more severe than censure. Id. at 475, 599 A.2d 1275. The Court’s determination, however, was also strongly influenced by other factors, including the judge’s deportment when his misconduct was discovered. Id. at 473-74, 599 A.2d 1275.
In this case, we are confronted with a prolonged course of judicial misbehavior that was especially harmful to its victim. We are convinced that “respondent’s misconduct must seriously shake public confidence in the judiciary, and, unless bolstered by a prompt and appropriate disciplinary response, that confidence is bound further to weaken and erode.” Id. at 476, 599 A.2d 1275. Accordingly, we suspend respondent from judicial office for sixty days. A temporary removal from office will impress upon respondent the magnitude of the offense he has committed, reaffirm public confidence in the integrity of our courts, and provide a *103powerful deterrent to future misconduct, of this type, by respondent or others who hold judicial office.
So ordered.
ORDER
This matter having come before the Court on a presentment of the Advisory Committee on Judicial Conduct, and respondent having been Ordered to show cause why he should not be disciplined, and good cause appearing;
It is ORDERED that JUDGE EDWARD J. SEAMAN is hereby suspended, without pay, from his judicial office for sixty days, commencing July 16, 1993, and ending September 13, 1993; and it is further
ORDERED that during his suspension, respondent is to complete an educational program designed to heighten awareness of what constitutes sexual harassment and to reinforce what sort of behavior is expected by the sexual harassment policy of the judiciary of this State, the program to be approved by the Court, which will also verify respondent’s compliance.
WITNESS, the Honorable Alan B. Handler, Presiding Justice, at Trenton, this 16th day of July, 1993.
WILENTZ, C.J., and CLIFFORD, J., did not participate.
For suspension — Justices HANDLER, POLLOCK, O’HERN, GARIBALDI and STEIN — 5.
Opposed — None.