The respondent, Jon D., appeals the Superior Court’s (Dalianis, J.) finding pursuant to RSA chapter 169-C that he sexually abused his daughter, Gina D. Relying on our recent decision in State v. Cressey, 137 N.H. 402, 628 A.2d 696 (1993), he argues that the superior court erroneously based its finding on insufficient evidence because the expert psychological testimony concluding that Gina had been sexually abused was unreliable. In addition, he contends that the superior court improperly reviewed the district court file as part of the de novo trial, and drew an improper inference from his decision not to testify, thus burdening his right against self-incrimination. Because the superior court’s decision was based on erroneously admitted evidence, we reverse and remand for further proceedings.
When Gina D. was two years and ten months old, her mother, Susan D., brought her to Ann Bastille, a child and adolescent therapist. Bastille holds a master’s degree in counseling and psychotherapy and has acquired specialized knowledge in the evaluation and treatment of child sexual abuse. Gina’s mother was concerned about the effects on Gina of her divorce from the respondent, including the custody proceedings. She reported that Gina was engaging in sexualized play and was experiencing nightmares, behavioral difficulties, and mood swings. Gina’s mother also told Bastille that another family member had made allegations of sexual abuse against the respondent. Because of the possibility of sexual abuse, Bastille first began an evaluation to “rule out” the possibility of sexual abuse as the cause of Gina’s reported problems. Bastille had received training with regard to sexual abuse allegations in child custody disputes, and she took care to discount Gina’s mother’s reports and to conduct the major evaluation sessions with Gina without her mother being present.
Bastille eventually came to the conclusion that the respondent had sexually abused Gina. Her conclusion was based largely on verbal and nonverbal disclosures Gina made during the course of Bastille’s evaluation, but was bolstered by Bastille’s observations of Gina’s behavior and emotions. Bastille reported Gina’s disclosures to the division for children and youth services (the division). The division then filed a petition in the Nashua District Court (Howorth, J.), pursuant to RSA chapter 169-C, alleging that the respondent had sexually abused Gina. Following a hearing, the court found that Gina had been abused within the meaning of RSA chapter 169-C and that the respondent was the perpetrator. He appealed to the superior court for a de novo hearing. See RSA 169-C:28 (1990).
*700The superior court heard testimony on behalf of the State from Gina’s mother, Bastille, and Barbara Blue, the child protective services worker who filed the abuse petition. Christopher Garner, Gina’s guardian ad litem in her parents’ divorce action, and Wilfred Derby, a clinical forensic psychologist, testified for the respondent, who did not testify. Like the district court, the superior court concluded that the respondent had sexually abused Gina. The respondent appealed.
The first issue we address is whether the superior court properly admitted and relied on Bastille’s testimony. At the superior court trial, the respondent objected to Bastille’s qualifications to testify as an expert in the field of child sexual abuse. The respondent’s argument on appeal is based on our holding in Cressey, which was not decided until after the respondent filed his notice of appeal. Because the question of the reliability of Bastille’s expert opinion was sufficiently raised below and the parties briefed and argued the issue citing Cressey, we will review the superior court’s ruling in light of our holding in Cressey.
Expert testimony must meet a threshold level of reliability to be admissible under New Hampshire Rule of Evidence 702. Cressey, 137 N.H. at 405, 628 A.2d at 698. Under Rule 702, a psychologist’s expert opinion is not sufficiently reliable to be admitted in a criminal trial as evidence that alleged victims have been sexually abused. Cressey, 137 N.H. at 407, 628 A.2d at 699; State v. Chamberlain, 137 N.H. 414, 417, 628 A.2d 704, 706 (1993); State v. Luce, 137 N.H. 419, 421, 628 A.2d 707, 709 (1993). The present case, however, is a civil proceeding, not a criminal matter, see RSA 169-C:18, I (1990). In addition, the New Hampshire Rules of Evidence do not apply to abuse and neglect proceedings; instead, the court “may admit evidence which it considers relevant and material.” RSA 169-C:12, :18, III (1990). Consequently, the evidentiary holding in Cressey does not directly control the issue of Bastille’s testimony in this case. Nevertheless, we must determine whether the expert opinion testimony by Bastille was relevant and material in light of our conclusion in Cressey that “a psychological evaluation of a potentially abused child does not present the verifiable results and logical conclusions that work to ensure the reliability required in the solemn matter of a criminal trial.” Cressey, 137 N.H. at 407, 628 A.2d at 699.
Because abuse proceedings are not governed by the New Hampshire Rules of Evidence, we look to New Hampshire common law of evidence for guidance in analyzing the materiality and relevancy of the challenged testimony. To be relevant, evidence must *701have probative value that logically supports a resulting conclusion. Goldsmith v. Kingsford, 92 N.H. 442, 448, 32 A.2d 810, 814 (1943). To be material, evidence must be offered to prove a fact in issue. Welch v. Bergeron, 115 N.H. 179, 182, 337 A.2d 341, 344 (1975). Therefore, “testimony must have some tendency to establish a fact of consequence to the determination of the action” in order to be admissible into evidence. Id. Opinion testimony is relevant and material if the witness is qualified to give the opinion offered and if the opinion will aid the trier of fact in the search for the truth. See Peters v. McNally, 123 N.H. 438, 440, 462 A.2d 119, 121 (1983); Rau v. Stores, 97 N.H. 490, 494-95, 92 A.2d 921, 924 (1952).
At the appeal hearing, Bastille testified about her observations, evaluations, and conclusions concerning the possibility that Gina had been sexually abused. Bastille stated that Gina told her, “‘My daddy hurts my back,’” and that “her daddy hurts her with a funny-looking stick.” She testified about a session when Gina talked about her daddy hurting her and Bastille asked her to show on an anatomical drawing where her daddy hurts her with the “funny-looking stick.” The anatomical drawing showed the back of a young girl on one side and the front view on the other side. On the back view of the girl, Gina circled her buttocks area and part of her upper back. After marking the back view, Gina spontaneously turned over the paper to the front view and made marks in the genital area and the area of the breasts. She then began scribbling on the paper very hard and said, “‘He hurts me all over.’”
Bastille’s testimony about Gina’s disclosures and her own observations of Gina’s behavior is merely factual and was not generated by her particular expertise. Her factual testimony is both relevant and material to the question of whether Gina was sexually abused by her father, the respondent. Absent other infirmities that have not been demonstrated on appeal, her factual testimony was properly admitted under RSA 169-C:12.
Bastille’s expert opinion that Gina was sexually abused by her father, however, must assist the court in its search for the truth to be relevant and admissible. Relying on Cressey, the respondent contends that Bastille’s conclusion was unreliable, and therefore inadmissible, because it lacked a sufficient scientific basis. Generally, the basis for an expert opinion is tested through cross-examination of the witness, and objections may affect the weight of the testimony but do not render it inadmissible. Tullgren v. Phil Lamoy Realty Corp., 125 N.H. 604, 609-10, 484 A.2d 1144, 1148 (1984). Expert psy*702chological evidence must be based on factors in addition to the child’s accounts of abuse; otherwise, the testimony is inadmissible as merely an opinion of the child’s credibility. Cressey, 137 N.H. at 407, 628 A.2d at 700; see State v. Huard, 138 N.H. 256, 259, 638 A.2d 787, 789 (1994). In Cressey, however, this court decided that a thorough cross-examination of the psychologist was likely to be ineffective in exposing any infirmity in the interpretive component of an expert opinion that a child had been sexually abused. Cressey, 137 N.H. at 410, 628 A.2d at 701. “The methodology used in the psychological evaluations makes [the psychologist’s] presentation of evidence effectively beyond reproach.” Id.
Bastille testified that her conclusion that Gina had been sexually abused by the respondent was based on the following factors:
“[Gina’s] disclosure and the ongoing emotional disturbance that [Bastille] saw in the office at times, behavioral difficulties when any kind of even nondirective questioning was used, the — the anxiety, the — especially the extensive play and talk around issues of nightmares, which certainly appeared very genuine.”
Gina’s symptoms, according to Bastille, were indicators that Gina had been traumatized. Based on Gina’s disclosures, Bastille concluded that the traumatization was due to sexual abuse. Bastille used a prepared chart showing the signs and symptoms of child sexual abuse in evaluating Gina. She explained that based on her training she was able to identify “victimization themes” in Gina’s nightmares and drawings and in Gina’s explanation of her drawing. Bastille also drew significance from the fact that Gina had included “sexualized parts” in her self-portrait because she did not include those parts on her other drawings of females. Further, she ruled out the possibility that Gina had been “programmed” by Susan to allege that the respondent sexually abused her based on the following “wide range of indicators”:
“Gina disclosed or had indicators of sexual abuse in a wide range of ways in terms of behavioral, what she had done in the diagnostic drawings, her disclosure in terms of having marked that drawing rather than having it be just a verbal disclosure, the — certainly, the richness with regard to her nightmare content and continuing themes of victimization and being hurt — being hurt in the areas of her bum.”
Bastille believed that she had “ruled out” other possible perpetrators of the abuse, other than the respondent, because “to the best *703that I could assess, ‘Daddy’ was the daddy that she visited on the weekend.” During cross examination, Bastille explained that her conclusions were not based on one single factor but that each indicator was “a piece of a total puzzle in doing the evaluation.” Although individual factors could be interpreted differently, together they led Bastille to a conclusion of sexual abuse by the respondent.
Dr. Wilfred Derby, appearing on behalf of the respondent, testified that the indicators listed in the chart relied on by Bastille were not diagnostic of child abuse and that there are no indicators of child sexual abuse other than possibly venereal disease. Instead, he said, the kinds of symptoms Bastille reportedly observed in Gina show emotional problems. He further testified that at this time there is no research which supports relying on the interpretation of a child’s drawings to diagnose sexual abuse. He disputed Bastille’s qualifications to interpret Gina’s dreams and characterized her interpretations as “almost in the realm of fortune telling.” Finally, he noted that Bastille’s sessions with Gina were not taped so that there was no way for anyone else to know what happened and to assess the validity of her observations and methods.
Experts in the field of behavioral science dispute the reliability and validity of diagnoses of sexual abuse based on behavioral characteristics of the child as is indicated by the striking conflict between the expert witnesses in this case. See Cressey, 137 N.H. at 408, 628 A.2d at 700; State v. Michaels, 498 A.2d 489, 499-500 (N.J. Super. Ct. App. Div. 1993); State v. Jones, 863 P.2d 85, 98 (Wash. App. 1993); Myers et al., Expert Testimony in Child Sexual Abuse Litigation, 68 Neb. L. Rev. 1, 69-70 (1989). If Bastille’s evaluation and diagnosis were presented as novel scientific evidence, it would not meet either prong of the test for admissibility articulated in State v. Vandebogart (DNA), 136 N.H. 365, 376, 616 A.2d 483, 490 (1992), as neither the principle nor the techniques of diagnosing sexual abuse used by Bastille have been generally accepted by behavioral scientists.
Bastille based her disputed diagnosis on her evaluation of a variety of factors that depended on her interpretation of Gina’s drawings, nightmares, and behavior. Despite being challenged on cross-examination about her methods and conclusion, Bastille defended her position based on her evaluation of all the pieces of “the puzzle.” An opinion that is impenetrable on cross-examination due to the unverifiable methodology of the expert witness in arriving at the conclusion is not helpful to the court in its search for the truth. See *704Cressey, 137 N.H. at 410, 628 A.2d at 701; State v. Campbell, 127 N.H. 112, 116, 498 A.2d 330, 333 (1985). If the court, as the trier of fact, cannot determine and assess the bases for the expert’s opinion, it also cannot accord the proper weight, if any, to the testimony. In these circumstances, we hold that the opinion testimony given by Ann Bastille that Gina had been sexually abused by the respondent was not sufficiently probative to meet the standard of assisting the court as the trier of fact. Consequently, Bastille’s opinion that Gina was sexually abused by the respondent was not relevant, and the trial court erred as a matter of law in admitting and relying on Bastille’s opinion.
We are not persuaded that the purpose of an abuse proceeding under RSA chapter 169-C, as contrasted with the purpose of a criminal trial on allegations of sexual abuse of a child, justifies using expert opinion testimony of a diagnosis of sexual abuse in an abuse proceeding. Although RSA chapter 169-C “is designed to protect children, not punish parents,” In re Heather D., 121 N.H. 547, 550, 431 A.2d 789, 791 (1981), an abuse proceeding undertakes a solemn decision-making process as does a criminal trial on allegations of sexual abuse. A child’s and his or her family’s life may be seriously and permanently affected by the court’s determination in an abuse proceeding. See In re Tracy M., 137 N.H. 119, 123-24, 624 A.2d 963, 965-66 (1993). The different purposes of an abuse proceeding and a criminal trial are properly served, in part, by their differing standards of proof. In an abuse proceeding, the State must prove the allegations in the petition by a preponderance of the evidence, RSA 169-C:13 (1990), rather than beyond a reasonable doubt, RSA 625:10 (1986). Also, the more relaxed evidentiary standard in abuse proceedings under RSA 169-C:12 allows a broader range of evidence than in a criminal trial, although evidence must meet the relevance and materiality standard. Finally, although abuse proceedings are decided by the court, RSA 169-C:18, :28 (1990), rather than by a jury as is generally true in a criminal trial, opinion evidence nevertheless must be material and relevant. Bastille’s opinion testimony did not meet the necessary standard.
As in Cressey, we caution that the holding in this case does not preclude all opinion testimony from psychologists or therapists in abuse and neglect proceedings. Expert testimony based on psychological evaluation or other behavioral science expertise may be helpful to the court by providing information about a child’s physical and cognitive development and behavior. In particular, such testi*705mony may show the presence of age-inappropriate sexual behavior and knowledge, obsession with sexual abuse, and may explain behaviors that would otherwise suggest that the child lied in making allegations of abuse. Cressey, 137 N.H. at 410, 628 A.2d at 702; Chamberlain, 137 N.H. at 417, 628 A.2d at 706. In addition, we recognize that the behavioral science fields are not static and that research may produce new information and consensus in the scientific community about diagnosing symptoms in sexually abused children.
We will uphold the findings and rulings of the trial court unless they are unsupported by the evidence or are legally erroneous. In re Tracy M., 137 N.H. at 125, 624 A.2d at 966. In assessing the sufficiency of the evidence to support the trial court’s findings, however, we must review all of the evidence presented to the trial court and “determine whether a reasonable person could have found as the trial judge did.” Id. Although Gina’s mother’s testimony is suspect because of her interest in the custody dispute, we cannot say that no reasonable person would have relied on it. Barbara Blue, the child protective services worker, testified that a pediatrician who examined Gina found no physical evidence of sexual abuse although she also testified that a lack of physical evidence is not unusual in sexual abuse cases. Otherwise, Blue’s testimony was dependent on Ann Bastille’s evaluation.
The guardian ad litem’s testimony reported his observations of Gina, her mother, and the respondent. He testified that Gina’s mother made a variety of “wild and exaggerated allegations” about the respondent and his family that the guardian ad litem was not able to confirm or repudiate. He described his visit at the respondent’s house, observing him with Gina for two hours, and noted that Gina was very open and verbal and seemed happy. He stated that he was concerned “about the way [Gina’s mother] has handled the whole divorce from the beginning and the way events seem to coincidentally fit into an abuse-and-neglect case for some benefit at the divorce court level.”
Dr. Derby’s testimony seriously undermined the reliability and validity of Bastille’s interpretations and opinions. He did not examine Gina, however, as Bastille did. Consequently, we cannot say that no reasonable person could have found, as the superior court did, that the evidence at trial supported a finding that Gina had been abused by her father, the respondent.
In making its decision, however, the superior court relied heavily on Bastille’s opinion that Gina had been sexually abused by the respondent. The court entirely discounted Gina’s mother’s testimony *706except for foundational matters and found that Barbara Blue’s testimony did not add anything of significance. The court noted that the guardian ad litem thought that Gina’s mother had suggested the allegations of sexual abuse to Gina, and acknowledged Dr. Derby’s testimony challenging Bastille’s credentials and some of the bases for her conclusions. Finding that “the substance of Ann Bastille’s testimony bears up under scrutiny,” the superior court ruled that Gina had been sexually abused within the meaning of the statute.
As we have explained, Bastille’s opinion' testimony that Gina had been sexually abused by the respondent was not sufficiently reliable to be relevant. Consequently, the court’s reliance on Bastille’s opinion was an error of law. Based on a review of the record and the importance accorded by the superior court to Bastille’s testimony, we do not find that the error was harmless. Cf. In re Kristopher B., 125 N.H. 678, 684, 486 A.2d 277, 282 (1984) (finding erroneously admitted evidence harmless due to sufficiency of other evidence to support the court’s ruling). Therefore, we reverse and remand for further proceedings that are consistent with this opinion. We find no merit in the other issues raised on appeal.
Reversed and remanded.
JOHNSON, J., and Horton, J., dissented; the others concurred.