(dissenting). I agree with the majority’s observation that a trial court’s assessment of the various factors bearing upon the best interests of a child must be accorded great weight (see, Matter of Louise E. S. v W. Stephen S., 64 NY2d 946). Appellate Judges, who lack the advantage of observing the demeanor of witnesses, must exercise considerable caution to avoid merely substituting their judgment or preference for that of the trial court. I am compelled to dissent in this case, however, because I am unable to agree with the majority’s conclusion that the trial court carefully weighed the relevant factors in an effort to determine the best interests of the child. The court failed to set forth its findings of fact and analysis of relevant factors bearing upon the best interests of the child in sufficient detail for appellate review. Further, the record does not support the court’s description of the relative interest or bias possessed by each party’s witnesses, and I am constrained to conclude that, in the course of fact-finding, the court accorded undue weight to the testimony of plaintiff’s prior spouse and undue deference to the guardian ad litem’s analysis of the evidence, virtually delegating its fact-finding responsibility to the guardian ad litem. The trial court also improvidently exercised its discretion in failing to direct that a custodial evaluation be performed by a psychologist or psychiatrist. Under the circumstances, the court was deprived of critical probative evidence on the issue of custody.
Numerous factors impact upon a determination of the best interests of a child. Although the number and importance of relevant factors vary with each factual circumstance, such factors generally include (1) the continuity and stability of any existing custodial arrangement, including the relative fitness of the parents, (2) the quality of the home environments that can be provided by each parent, (3) the ability of each parent to provide for the child’s emotional and intellectual development, (4) the financial status and ability of each parent to provide for the child, (5) the individual needs and expressed desires of the child, and (6) the need of the child to live with siblings (see, Matter of Louise E. S. v W. Stephen S., 64 NY2d 946, supra; Eschbach v Eschbach, 56 NY2d 167, 172-173; Fox v Fox, 177 AD2d 209, 210). In the subject case, the trial court *1132mentioned three of those factors, analyzing each factor in a conclusory manner. The court observed that each party was a fit parent and rejected defendant’s contention that plaintiff had abused their child. The court further found that the home environment and structure of daily life provided by plaintiff would better promote the child’s best interests and that the child’s sibling relationship with his brother would be promoted by granting custody to the plaintiff. The court did not mention the factual basis for its conclusions, nor did it indicate the relative weight it accorded to each factor.
The record supports the court’s finding that defendant is a fit parent. Defendant had been the child’s primary custodian from birth. A psychologist who observed defendant and the child, Bradley, testified that they had a positive, loving child-parent relationship; that the relationship between parent and child was appropriate, neither overly solicitous nor overly protective; that he did not observe any signs of maladjustment or abnormal behavior; and that Bradley exhibited the behaviors typical of a child his age. The child’s pediatrician testified that Bradley was healthy and that there was a strong and appropriate bond between mother and child.
The evidence proffered by plaintiff supports a conclusion that plaintiff is a fit parent. Several witnesses testified that plaintiff was a good and loving father to his son of a prior marriage and that he related well with Bradley during visitations. While there is no testimony that would support a finding that plaintiff has physically abused Bradley or that plaintiff has emotional problems that might affect his parenting abilities, I am troubled that the court’s failure to direct that a custodial evaluation be performed deprived the fact-finder of evidence that would have been highly probative on that issue. Most of the plaintiff’s witnesses testified regarding plaintiff’s parenting of his son of a prior marriage; only a few of the witnesses observed plaintiff’s relationship with Bradley during a significant segment of time. Although the Probation Department was directed to perform a custodial evaluation, no evidence was proffered regarding that evaluation. Indeed, the record does not reveal whether such an evaluation was performed. Prior to trial, defendant requested that a custodial evaluation be performed by a psychologist or psychiatrist appointed by the court. That request was denied, without prejudice, apparently because the Probation Department had been directed to perform such an evaluation. At trial, a psychologist was called to testify regarding plaintiff’s relationship with a male child of defendant’s cousin. Allegedly, that *1133child had some difficulty in his relationship with plaintiff, and defendant sought to establish the nature of that difficulty and whether it involved physical or emotional abuse. Because that child’s guardian ad litem did not waive the physician-patient privilege, the psychologist was unable to provide that testimony. That child often was in the company of plaintiff, and such testimony would have been relevant regarding plaintiff’s parental fitness. The psychologist who interviewed and observed Bradley expressed concern regarding Bradley’s behavioral reaction when the psychologist made mention of the father’s name during their sessions. The psychologist acknowledged, however, that he could not render an opinion concerning custody because he did not have the opportunity to communicate with other relevant contacts or to observe the child with other family members. Although the decision whether to direct a psychological evaluation of the participants in a child custody dispute is discretionary (see, Kesseler v Kesseler, 10 NY2d 445, 452; cf., Family Ct Act § 251), that evaluation should have been directed at the time of trial when it became apparent that the probation report was not forthcoming, that no report had been proffered into evidence regarding an evaluation of fitness and home environments by the child’s guardian ad litem, that a genuine factual issue had been raised concerning plaintiff’s emotional relationship with Bradley, and that such an evaluation would, in all likelihood, be the most probative evidence of those factors bearing upon the best interests of the child. In my view, the absence of a custodial evaluation, in the circumstances of this case, rendered the record insufficient to reach a conclusion regarding plaintiff’s fitness (see, Matter of Estrada v Estrada, 154 AD2d 376; Giraldo v Giraldo, 85 AD2d 164). I additionally note that, although it appears that Bradley’s guardian ad litem prepared pretrial reports that may have reflected her interviews with some of the interested parties, those reports were not received into evidence. The guardian’s posttrial report containing a comprehensive evaluation of the credibility of the trial witnesses is no substitute for the guardian’s independent investigation and recommendation of the child’s home environments and the parenting abilities of the parents. The record also reveals that the guardian did not interview plaintiff’s parents, who did have knowledge regarding their son’s relationship with Bradley and his parenting skills.
Plaintiff did not call his parents to testify at trial. The trial court concluded that such failure was not a weakness in the case, because, in its view, "plaintiff’s choice of witnesses *1134showed an emphasis on disinterested witnesses compared to the members of [defendant’s] family” produced by defendant. That comparison clearly is inaccurate. Most of plaintiff’s witnesses were employees of the Buffalo Bills, plaintiff’s former employer and "extended family”, who had very limited observations of his relationship with the son of a prior marriage and none with Bradley; defendant’s cousin, who had been dating plaintiff and who, in her words, was like a sister to plaintiff; the cousin’s mother; and plaintiff’s former wife. Those witnesses were no more disinterested than defendant’s family. Particularly disturbing is the heavy reliance of the guardian and the court upon the testimony of plaintiff’s former wife. The hostility and animosity of that witness toward the defendant, apparently resulting from what the witness perceived to be the defendant’s attempt to interfere with and diminish the witness’ relationship with her son and perhaps other occurrences during their acrimonious relationship, are obvious and color each page of the witness’ testimony. Further, the bias of several of defendant’s family members may well have resulted from the fact that plaintiff instituted a paternity proceeding, questioning whether he was the father of the child and assumed parenting responsibilities only after he received the results of blood grouping tests. The hostility and bias of several of the witnesses for each side merely illustrate further the need for a comprehensive and disinterested custodial evaluation.
With respect to the court’s finding that the child’s sibling relationship with his brother would be promoted by an award of custody to plaintiff, I simply note that the child of plaintiff’s prior marriage is a half-brother; that there is scant evidence of a bond between the half-brothers as siblings; and that there is no evidence that continued primary custody with defendant would disturb or diminish the relationship that has developed between those youths.
I am aware that the parties have already suffered through an acrimonious and lengthy custody trial. In my view, however, neither interests of judicial nor personal economy can supersede the requirement that the best interests of the child be determined upon an adequate record and by a complete and proper analysis of all relevant factors bearing upon those best interests. Thus, I would modify the judgment by vacating the custody award and would remit this matter for a new trial on the issue of custody, with directions that the trial court appoint a psychiatrist or psychologist to perform a custodial evaluation and that a probation report be submitted to that *1135court. (Appeal from Judgment of Supreme Court, Erie County, Fudeman, J.—Custody.) Present—Pine, J. P., Balio, Lawton, Fallon and Davis, JJ.