J. The issue raised in this appeal is whether in a trial pursuant to G. L. c. 123A, § 9, the Commonwealth may satisfy the requirement that it produce expert testimony to prove continuing sexual dangerousness by relying solely on the community access board’s (CAB)1 unanimous opinion. That opinion *124was introduced by a board member who is also a qualified examiner as defined under G. L. c. 123A, § 1. In the circumstances of this case, where the board member testified solely as the CAB’s representative, and the qualified examiners who examined the petitioner in connection with the c. 123A, § 9, proceeding determined that the petitioner is no longer a sexually dangerous person, we conclude that the Commonwealth did not satisfy the requirement.
Background. In 1992, the petitioner pleaded guilty to two counts of indecent assault and battery on a child under fourteen. The victim was his then three year old daughter. He was sentenced to a ten-year term of incarceration at MCI, Concord, followed by five years of probation. 2
While in prison, the petitioner began and then discontinued sex offender treatment several times, eventually completing the treatment programs for phases I through III.3 In the spring of 1998, he was transferred to the Massachusetts Treatment Center (MTC) in order to participate in phase IV of the treatment program, but was terminated from that program in October, 1998, due to lack of participation. He then entered a phase III group for those who had been terminated from phase IV, and his participation improved markedly.
In October, 2000, the petitioner was returned to MCI, Norfolk, *125because he was approaching his sentence completion date. Prior to his release, however, the Commonwealth filed a petition pursuant to G. L. c. 123A, § 14, seeking the petitioner’s commitment as a sexually dangerous person. In November, 2002, following a jury-waived trial, a Superior Court judge found the petitioner to be a sexually dangerous person, and issued an order of civil commitment to the MTC for a period of from one day to life.4
On January 10, 2003, the petitioner filed a petition in the Superior Court for Bristol County, seeking his discharge pursuant to G. L. c. 123A, § 9; the case was immediately transferred to the unified session of the Superior Court for Suffolk County. Both the Commonwealth and the petitioner demanded trial by jury.5 As required by the statute, the judge appointed two qualiexaminers,6 Frederick Kelso, Ph.D., and Jeffrey Miner, Ph.D., both designated forensic psychologists, to examine the petitioner and submit reports to the judge.
Kelso and Miner each concluded that the petitioner was no longer a sexually dangerous person and opined that the petitioner was at “low risk” to reoffend. At the same time, the CAB conducted an evaluation of the petitioner and prepared its report. The CAB unanimously concluded that the petitioner remained a sexually dangerous person and presented a “high risk” of reoffending if released.
At a jury trial beginning on June 13, 2006, the Commonwealth *126called three witnesses: Gregg Belle, Ph.D., a forensic psychologist and qualified examiner who was the CAB’s representative and the author of the CAB report introduced at trial; Sarah Kelly, a master’s level forensic psychologist and one of four MTC therapists assigned to the petitioner’s treatment team; and Frederick Bohning, a correction officer who had observed two instances of the petitioner staring at other residents’ children in the MTC visiting room on May 29, 2000.
Belle testified that the CAB had interviewed the petitioner and reviewed his criminal background, personal history, and treatment progress.7 The CAB’s conclusion that the petitioner remained a sexually dangerous person was based on the petitioner’s diagnosis of pedophilia8 and his refusal to participate in certain recommended treatment,9 and evidence that he suffered from cognitive distortions.10
Belle explained that the CAB was particularly concerned that the petitioner still exhibited deviant sexual arousal to children.11 The CAB expressed “no confidence” that the petitioner would not reoffend, given that he had demonstrated sexual arousal to children while at the MTC. See infra at 127. Belle noted that the petitioner’s candid statement to the CAB that he was in the early stages of treatment, and his lack of an adequate relapse prevention plan, further supported the CAB’s conclusion that *127the petitioner was at “high risk” of reoffending. Belle also testified that the petitioner’s entire treatment team was unanimous in its opinion that the petitioner remained sexually dangerous.
Belle further testified that he had been a forensic psychologist for six years and had himself been appointed a qualified examiner six months earlier. He had no experience directly treating sex offenders; his experience as a forensic psychologist involved evaluating prison inmates, including sex offenders, for other purposes. Since his appointment, he had performed one evaluation as a qualified examiner, and had never testified as such.
Kelly, a therapist on the petitioner’s unit, testified about the degree of the petitioner’s involvement in treatment.12 She noted that the petitioner had made little progress toward his treatment goals, was often passive and needed to be prompted to participate, and sometimes responded with nonrelevant or inappropriate remarks during group sessions. She also testified that the petitioner had only an inadequate, two-page release plan. On cross-examination, Kelly testified that she was unaware that the petitioner had completed a transition planning program in 2001 and received a certificate for that program. The plan and the certificate were in the petitioner’s record and were introduced at trial through Kelso’s testimony.
The Commonwealth’s final witness, Bohning, testified that on May 29, 2000, he observed the petitioner staring at a visitor’s two year old child, and later at another visitor’s thirteen year old boy and fifteen or sixteen year old girl, in the MTC visiting room. Bohning subsequently issued a disciplinary report describing the incidents; this report was admitted in evidence at trial.
At the close of the Commonwealth’s case, the petitioner orally moved for a directed verdict. The judge reserved ruling on the motion, and the petitioner proceeded to present his case. He called both of the appointed qualified examiners, Kelso and Miner. Kelso testified that he had nineteen years of experience in clinical and forensic psychology. He had been appointed a qualified examiner in 1994, and had performed evaluations and *128testified in courts throughout the Commonwealth “thousands” of times. Miner testified that he became a designated forensic psychologist in 1989, and had worked as a full-time forensic evaluator for the courts since 1997. He had been appointed a qualified examiner in December, 2005.
Both Kelso and Miner agreed that the petitioner suffered from pedophilia, but testified that he was no longer a sexually dangerous person. Each stated that he believed the petitioner did not suffer from a mental abnormality or defect as defined in G. L. c. 123A, § 1. Each noted that the petitioner had not made much progress in treatment during the first year of his commitment to the MTC, but that treatment records from the last year and one-half to two years showed some progress.
Like the CAB, the appointed qualified examiners used an “empirically guided” approach to assessing the petitioner’s risk. Unlike the CAB, they each also performed a statistical risk assessment as part of their evaluations. Both Kelso and Miner scored the petitioner as a “1,” or “low” risk, on the “Static-99,” an actuarial tool routinely employed to evaluate the risk of reoffense among those previously convicted of being sex offenders.13 They testified that the Static-99 evaluates only static factors such as an offender’s previous criminal history, and only states the statistical rates of reoffense for someone with the petitioner’s characteristics, rather than predicting individual risk. In deriving the petitioner’s Static-99 score, each noted that the petitioner had committed only one sex offense, had no history of sex offenses early in life, had offended against his own child but not against anyone outside his family, had been married for more than two years, and had no male victims.
Each examiner explained that incest-only offenders, such as *129the petitioner, are statistically the least likely of all sex offenders to reoffend and that those with only female victims are less likely to reoffend than those with male victims. Both Kelso and Miner testified that, because of his single incest offense and a number of “protective” factors, they believed the petitioner was at lower risk than indicated by his Static-99 score. Miner classified the petitioner as at “very low” risk.
Both qualified examiners discussed the petitioner’s governing offense with him; they explained that the petitioner now felt shame and remorse about his actions, understood the harm he had caused his daughter, and understood his previous thinking “very clearly” as a cognitive distortion involving issues of “jealousy and power and control and entitlement.” Miner testified that, despite the petitioner’s cognitive limitations and lower intellectual ability,14 he was “impressed” that the petitioner had made “significant progress” in some areas of treatment, “less progress” in others, and that, as the concepts presented in treatment became increasingly complex, the petitioner struggled initially, faltered, but then “perservered to the next level.”
While Kelso and Miner considered the incidents in the visiting room to be significant, both stated that they had discussed the incidents with the petitioner and considered them in their risk assessment.15 They noted that the petitioner had not spoken with the children, tried to communicate with them, or taken any physical action towards them. They stated that the petitioner now understood his fantasies about the sixteen year old girl to be “deviant” and that there had been no record of incidents in the visiting room since that time.
Each qualified examiner also considered dynamic factors routinely used to evaluate a particular individual’s risk of re-*130offending. Both Kelso and Miner considered the petitioner’s positive relationship with his mother and brother, and his plans to live with his mother and work part-time at a transmission shop with his brother, as protective factors. They considered his admitted deviant arousal as a risk factor. They also noted the fact that he would be on supervised probation for five years, with mandated sex offender treatment, as a protective factor.
Further, we note that Kelso testified that in ten out of fourteen probable cause hearings, and ninety out of ninety-five G. L. c. 123A, § 9, hearings in which he had performed evaluations, he had opined that the offender was sexually dangerous. Miner testified that he had performed six c. 123A, § 9, evaluations as a qualified examiner, and had found the offender to be sexually dangerous in five of the six instances.
Following the qualified examiners’ testimony, the judge held a hearing on the petitioner’s renewed, written motion for a directed verdict.16 The focus of the hearing was the absence of testimony by a qualified examiner that the petitioner was currently sexually dangerous. The judge concluded that the Commonwealth’s evidence was insufficient and allowed the petitioner’s motion. She stated that she based her decision “upon what’s before the Court, and I also take into consideration the posture of the case, the testimony I’ve heard from Dr. Kelso and Dr. Miner, as well as all of what the Commonwealth presented.”17
The judge then dismissed the jury and ordered the petitioner *131released, but allowed the Commonwealth’s motion to stay the release pending its motion to reconsider. After the judge denied the motion to reconsider, she allowed the Commonwealth’s motion for a further stay of release pending the outcome of this appeal.
Discussion. The Commonwealth contends that the judge erred by directing a verdict in favor of the petitioner. “The question here boils down to whether, after viewing the evidence (and all permissible inferences) in the light most favorable to the Commonwealth, any rational trier of fact could have found, beyond a reasonable doubt, the essential elements of sexual dangerousness as defined by G. L. c. 123 A, § 1.” Commonwealth v. Boyer, 61 Mass. App. Ct. 582, 589 (2004).
Expert testimony is required to prove sexual dangerousness. See Commonwealth v. Bruno, 432 Mass. 489, 510-511 (2000); Commonwealth v. Dube, 59 Mass. App. Ct. 476, 482-483, 489 (2003). In determining whether a petitioner remains a sexually dangerous person, the Commonwealth must use the same definition of sexual dangerousness, and must meet the same standard of proof, proof beyond a reasonable doubt, as in an initial commitment proceeding pursuant to G. L. c. 123A, § 14. See Wyatt, petitioner, 428 Mass. 347, 354 (1998); Dutil, petitioner, 437 Mass. 9, 11-12, 16 (2002).
The critical issue here is whether the Commonwealth presented the necessary expert testimony. The Commonwealth argues that Belle’s testimony was sufficient to satisfy its burden of providing expert testimony. The Commonwealth emphasizes that, in addition to being a member of the CAB, Belle is a qualified examiner and that, at the time of trial, one of the Department of Correction (DOC) representatives on the CAB, Ira Silverman, was also a psychologist and a qualified examiner.18 See note 7, supra. The Commonwealth further contends that the CAB report, *132approved in this case by three psychologists — the two statutorily required psychologists under G. L. c. 123A, §§ 6A and 9 (see note 1, supra), and Silverman — and presented by Belle, contains sufficient expert opinion from which a jury could reasonably infer that the petitioner remains a sexually dangerous person. We disagree.
The Commonwealth’s argument ignores the fact that Belle testified only as the CAB’s representative, and not as an independent expert. While Belle’s qualifications and curriculum vitae were presented to the jury, he did no more than speak to the CAB’s report. Moreover, because Belle was called as the CAB’s representative, the petitioner did not have an opportunity to cross-examine Belle as an expert, nor did he have any motivation to fully explore Belle’s level of expertise. Accordingly, the Commonwealth cannot reasonably rely on Belle’s testimony to meet its burden of producing expert testimony.
Furthermore, contrary to the Commonwealth’s assertion, the CAB report is not the equivalent of an evaluative report and diagnosis submitted by an expert. Although a CAB report offers the CAB’s opinion regarding a petitioner’s sexual dangerousness, the CAB’s conclusion is reached by a vote of three DOC representatives and two DOC-appointed psychologists, who may or may not be qualified examiners. G. L. c. 123A, §§ 6A, 9. See note 1, supra. The report itself consists largely of reviews of earlier CAB reports, summaries of annual treatment reports and treatment team notes, intake notes, criminal history, and DOC records.
According to Belle’s testimony, the CAB reviews and reports on four to five residents per week, in order to comply with its statutory mandate to annually “review” each individual committed to the MTC, and relies heavily on the treatment team’s annual treatment review report in reaching its conclusions. Here, Belle testified that the CAB’s group interview of the petitioner lasted forty-five minutes to an hour. The CAB does not review the petitioner’s medical or psychiatric records, nor does it make use of the actuarial tools commonly employed by qualified examiners and other forensic experts.
That the CAB report by itself cannot be considered a substitute for expert testimony is underscored by contrasting the report *133with those prepared by the appointed qualified examiners. In contrast to the CAB members’ review, the designated qualified examiners devoted significant amounts of time to an in-depth review of all of the petitioner’s records and conducted individual interviews with the petitioner.19 Kelso testified that he typically devoted several hours to reading a petitioner’s records at the MTC, obtained additional information and the petitioner’s criminal history from the legal office, and then interviewed the individual for one to two hours, as he did here, prior to developing his report. Miner testified that he performed a comprehensive review of the petitioner’s records and then interviewed the petitioner for three and one-half hours. See generally Commonwealth v. Connors, 447 Mass. 313, 317-318 & n.7 (2006) (contrasting the “imbalance” between expert testimony based on interviews and expert testimony based on reviews of records).
The Commonwealth’s next argument, that the Legislature intended the CAB to assume the role of an expert in G. L. c. 123A, § 9, trials, is similarly unavailing. This assertion is based on the provisions in G. L. c. 123A, §§ 6A and 9, which state that the CAB report “shall be admissible” in all c. 123A, § 9, proceedings. Because a civil commitment or continuing commitment affects a petitioner’s fundamental liberty interest, we apply strict scrutiny in construing the statute and in determining whether the proceedings satisfy due process requirements. See Commonwealth v. Gillis, 448 Mass. 354, 357-358, 364 (2007). See also Commonwealth v. Nieves, 446 Mass. 583, 596-597 (2006); Commonwealth v. Sargent, 449 Mass. 576, 579-580 (2007); Page v. Commonwealth, 13 Mass. App. Ct. 384, 385-389 (1982).
The Commonwealth’s interpretation of the CAB’s role is not supported by the plain language of the statute; the statute explicitly specifies that a variety of documents are admissible at a G. L. c. 123A, § 9, trial. “[Statutory language is the principal source of insight into legislative purpose.” Commonwealth v. Poissant, 443 Mass. 558, 562 (2005), quoting from Local 589, Amalgamated Transit Union v. Massachusetts Bay Transp. *134Authy., 392 Mass. 407, 415 (1984). In the same sentence in which it authorizes admission of the CAB report, c. 123A, § 9, also authorizes the admission of “juvenile and adult court and probation records;” “psychiatric and psychological records;” and “all relevant materials prepared in connection with” any application for participation in a community access program under G. L. c. 123A, § 6A. In addition, c. 123A, § 9, makes admissible “any other evidence that tends to indicate” that the petitioner “is a sexually dangerous person.” Because the CAB report is included among a number of admissible documents, there is no basis for concluding that the Legislature intended to elevate the report to the level of expert testimony. See Commonwealth v. Nieves, 446 Mass, at 590 (“Where the statute is clear, we interpret it as written”).
We further observe that the statute limits the CAB’s role to two specific functions, “consider[ing] a person’s placement within a community access program and conducting] an annual review of a person’s sexual dangerousness.” G. L. c. 123A, § 1. See note 1, supra. The CAB’s relatively minor statutory role is illuminated by comparing it to that of the qualified examiners, who are “central to the statutory scheme” throughout the commitment and discharge proceedings. See Commonwealth v. Bradway, 62 Mass. App. Ct. 280, 283-285 (2004) (noting statutory “specificity” of qualified examiners’ role and their qualifications, and “statute’s reliance upon the involvement and input of qualified examiners”). Whereas the statute requires the designated qualified examiners to perform a detailed evaluation and diagnosis of every petitioner who seeks discharge under c. 123A, § 9, and to submit their reports to the court, no similar requirements are imposed on the CAB.20 Although the CAB’s annual report “may” be “updated” when an individual files a c. 123A, § 9, petition, the statute neither requires such an update nor requires that the CAB report be submitted to the court.21 G. L. c. 123A, § 9. Furthermore, a petitioner may refuse to be interviewed by the *135CAB in conjunction with a c. 123A, § 9, proceeding, and, if so, the CAB report is written by relying on the treatment record. In contrast, a petitioner must meet with the qualified examiners in order to pursue his petition for release. Ibid. Sheridan, petitioner, 412 Mass. 599, 603-604 (1992).
We reject the Commonwealth’s contention that it need only offer “some” expert evidence of sexual dangerousness. A review of a petition for discharge requires an expert to explain the “clinical basis for linking past sexual misconduct with present behavior to produce a diagnosis of a currently sexually dangerous person.” Commonwealth v. Boyer, 61 Mass. App. Ct. at 589, quoting from Poulin, petitioner, 22 Mass. App. Ct. 988, 999 (1986). The CAB report, even presented by a CAB representative who is also a qualified examiner, cannot substitute for this expert evaluation. Accordingly, we conclude that there was no error in the judge’s decision to allow the petitioner’s motion for a directed verdict.
Judgment affirmed.
The CAB is “a board consisting of five members appointed by the commissioner of correction, whose function shall be to consider a person’s placement within a community access program and conduct an annual review of a person’s sexual dangerousness.” G. L. c. 123A, § 1. “Membership shall include three department of correction employees and two persons who are *124not department of correction employees, but who may be independent contractors or consultants. The non-employee members shall consist of psychiatrists or psychologists licensed by the [C] ommonwealth. ’’ G. L. c. 123A, § 6A. Although not statutorily required, the CAB also conducts an evaluation and issues an updated report if the annual review is more than six months old, when a petitioner files a G. L. c. 123A, § 9, petition for discharge. Unless otherwise noted, G. L. c. 123A, §§ 1, 6A, and 9, in this opinion are as they appear in St. 1993, c. 489, §§ 1,4, and 7, respectively.
The incidents occurred over a six-month period between September, 1990, and April, 1991. The petitioner admitted that after he separated from his wife, he sexually assaulted his then two and one-half to three year old daughter ten to twenty times, during her visits every other weekend. During treatment, the petitioner stated that the assaults escalated from exposing himself to rubbing his penis between his daughter’s “butt cheeks and vagina,” ejaculating on her chest, and oral and vaginal penetration.
These designations were used by the Commonwealth’s previously-contracted treatment provider. The current provider, Forensic Health Services, Inc., employs a two-level system under which the previous phases I through in are designated “pre-treatment.”
We recently affirmed the order of civil commitment in an unpublished memorandum and order issued pursuant to our rule 1:28, as amended, 46 Mass. App. Ct. 1001 (1998). See Commonwealth v. Johnstone, 71 Mass. App. Ct. 1108 (2008).
An individual committed under G. L. c. 123A, § 14, may petition for discharge once every twelve months by filing a petition under G. L. c. 123A, §9.
A qualified examiner is “a physician who is licensed pursuant to section two of chapter one hundred and twelve who is either certified in psychiatry by the American Board of Psychiatry and Neurology or eligible to be so certified, or a psychologist who is licensed pursuant to sections one hundred and eighteen to one hundred and twenty-nine, inclusive, of chapter one hundred and twelve; provided, however, that the examiner has had two years of experience with diagnosis or treatment of sexually aggressive offenders and is designated by the commissioner of correction. A ‘qualified examiner’ need not be an employee of the department of correction or of any facility or institution of the department.” G. L. c. 123A, § 1.
The other members of the CAB were Frederic Krell, Ph.D., a psychologist; Ira Silverman, Ph.D., a forensic psychologist who was a Department of Correction representative and a qualified examiner; Joseph Murphy, deputy superintendent of the MTC; and John Novero, director of civil commitments of the MTC.
Belle stated that the definition of pedophilia in the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders (4th ed. rev. 2000) is sexual arousal to prepubescent children under thirteen years of age, where the victim is more than five years younger than the offender.
Specifically, the CAB noted that the petitioner had consistently rejected its recommendation that he undergo a penile plethysmograph examination, which physically measures sexual arousal to various stimulae.
Belle described the petitioner’s previous belief that his daughter enjoyed his assaults as one example of the petitioner’s cognitive distortions.
“Belle testified in detail about the petitioner’s statement to the CAB that he had masturbated “15 to 20” times to fantasies about raping the fifteen to sixteen year old girl he had been observed staring at in the visiting room in May, 2000. In these fantasies, the girl’s father watched the petitioner commit the rape, while she begged the petitioner to stop. See infra at 127.
Kelly testified that she had a Master of Arts degree in forensic psychology and had worked as a therapist for one year, but was not a licensed therapist; the MTC position was her first experience providing therapy.
The Static-99 assessment includes ten factors, each of which is scored from a “0” to either a “1” or a “2”. Over-all scores range from 0 to 12, with a score of 0 or 1 designated as “low” risk, 2 or 3 as “medium-low” risk, 4 or 5 as “medium-high” risk, and 6 or higher as “high” risk.
Kelso and Miner scored the petitioner a “1” on the factor involving previous violent nonsexual offenses. They noted the petitioner’s conviction for assault and battery on both his wife and daughter, for an incident in April, 1989, when he hit his wife and shook his then year old daughter, that resulted in his wife’s decision to seek a divorce. Both Kelso and Miner testified that each had reviewed the petitioner’s records, had discussed these incidents with the petitioner, and had taken them into account in his evaluation.
The treatment team and everyone who has interviewed the petitioner agree that he has cognitive limitations. He was reported to have an IQ of 69, to have been in special classes or resource rooms in school, to have had to repeat three grades in elementary school, and to have dropped out of high school, but later completed his GED. His reading comprehension has been measured at a mid-ninth grade level. He finds writing very difficult, and in addition to being placed in the specialized unit for those with disabilities, he is also enrolled in a “non-readers” workbook class.
Although the petitioner stated that the two older children were together with one family and the toddler was with another, the petitioner admitted only to staring at, and fantasizing about, the teenage girl. See note 11, supra.
The record indicates that the petitioner had planned to call two additional experts, forensic psychologists Joseph Plaud and Leonard Bard, who would have testified that the petitioner was no longer sexually dangerous. The judge had approved and authorized payment for these experts, one of whom was present in court during the last day of trial. They did not testify only because the petitioner’s motion for a directed verdict was allowed before they were called.
At the hearing, the parties and the judge discussed the absence of controlling law on this question. Despite the Commonwealth’s contention otherwise, the judge’s decision does not explicitly state whether, in her view, G. L. c. 123 A, § 9, requires that one of the qualified examiners opine that a petitioner is currently sexually dangerous in order for the Commonwealth to establish a prima facie case. The Supreme Judicial Court has not decided “whether the statute implicitly requires that at least one of the qualified examiners support the position that a defendant is sexually dangerous in order for the Commonwealth to proceed to trial.” Commonwealth v. Poissant, 443 Mass. 558, 560 n.6 (2005). Although the petitioner urges us to reach this issue, we need *131not do so, given our conclusion that the Commonwealth did not sustain its burden of providing expert testimony.
We note with some concern that Silverman, now chair of the CAB, previously testified as an independent expert at the petitioner’s probable cause hearing, as well as at the petitioner’s civil commitment trial. Because Silver-man’s opinion in the proceeding at issue here appeared only through the CAB report and Belle’s testimony about the CAB’s conclusion, the petitioner had no opportunity to cross-examine Silverman for any potential bias.
The statute mandates that the “examiners shall have access to all records of the person being examined.” G. L. c. 123A, § 9. See note 6, supra.
The statute mandates that, in every G. L. c. 123A, § 9, petition, the qualified examiners “shall conduct examinations, including personal interviews,” of the petitioner, and “file with the court written reports of their examinations and diagnoses, and their recommendations for the disposition” of the petitioner’s case.
Section 8 of the former version of G. L. c. 123A, St. 1985, c. 752, § 1, *135however, mandated that the restrictive integration review board (RIRB), the CAB’s predecessor, provide an updated report whenever an individual filed a petition for release under G. L. c. 123A, § 9, as in effect prior to St. 1993, c. 489, § 7. This requirement was eliminated when the RIRB was replaced with the CAB. But see note 1, supra.