Defendant appeals by right the order changing the physical custody of her son from her sole custody to joint custody with plaintiff. We reverse.
Defendant gave birth to the parties’ only child, Taylor Jordan Winn, on October 9, 1991. Taylor is severely mentally and physically impaired. He suffers from cerebral palsy and epilepsy and is confined to a wheelchair. Although Taylor was five and one-half years old at the time of the custody hearing, he functioned as a one-year-old. Taylor requires constant adult supervision and assistance in all aspects of life and will require some form of adult supervision during his entire life.
The parties separated soon after Taylor’s birth. The June 1993 consent judgment that ended their twenty-two-year marriage reflected the parties’ agreement to joint legal custody, with physical custody to defendant. Plaintiff received reasonable visitation.1 Defendant chose to reside in Traverse City, where she currently cares for Taylor with the help of a part-time nurse. In 1994, defendant adopted a nine-year-old girl, *258with whom Taylor now has a close relationship. Meanwhile, plaintiff established his residence in Charlevoix and eventually retired from his family’s business. Plaintiff also became involved in a long-term relationship with a woman who has two minor children. By all accounts, Taylor recognizes plaintiff’s girlfriend and her children and enjoys their company when in plaintiffs care.
In 1996, the parties obtained extensive training in the use of a ketogenic diet to control Taylor’s epileptic seizures. The diet reduced his daily seizures from twelve to one or two. Taylor has demonstrated significant improvement with the reduction in seizures; he is aware of his surroundings, plays, and expresses himself vocally, but does not speak. With Taylor’s improvement and the changes in plaintiff’s life, plaintiff sought to play a larger role in Taylor’s care. Plaintiff alleged that in response to his request, defendant became more rigid regarding the visitation schedule, denying him scheduled visitation and refusing his requests for alternate and increased visitation. Defendant even insisted that plaintiff, she, and Taylor fly home to Michigan from Florida, rather than allow plaintiff to pick up Taylor for his weekend visitation at a location in Florida instead of the usual location in Traverse City. Plaintiff further claimed that defendant stopped providing him information regarding Taylor’s medical condition, appointments, and schooling.
In March 1997, plaintiff petitioned to change physical custody to a joint custody arrangement under which Taylor would reside with each parent during alternate weeks. After a six-day hearing, the trial court determined that the change was in Taylor’s best interest. The court first evaluated the statutory best *259interest factors,2 finding that factors a, b, c, f, g, and i favored neither party, factors d and e slightly favored defendant, factor h slightly favored plaintiff, and factors j and 1 favored plaintiff.3 Under factor 1, the court found as follows:
This is a special child, with many special needs. Virtually all the witnesses, save the Defendant, testified that Taylor needs to spend as much time with both of his parents (who are also special, dedicated people) as is possible. This is especially true, according to the expert testimony received, *260in the years until he is 11 or 12. But even beyond that, one must consider that Taylor will need lifelong care and both parents must be versed in how to provide that care in minute detail. Not only does each parent have to be comfortable with every aspect of the care of this child, but the child himself must be as comfortable with one parent as he is with the other. It must be pointed out that should something happen to one parent and that parent therefore became unable to care for Taylor, it would fall on the other to assume the full responsibility for ensuring that Taylor’s needs are met. It would also be extremetly (sic) difficult for Taylor to have to make an adjustment that most normal children have difficulty making if one of the parents here did not have his full confidence.
The trial court did not discuss factor k.4
On consideration of these factors, the trial court awarded the parties joint legal and physical custody, with Taylor alternating weeks between his parents’ homes and alternating schools between Charlevoix and Traverse City. The court explained the rationale for its decision to change the established custodial environment:
Considering the complex nature of this case, and considering the extensive testimony and exhibits introduced at the hearing, the court finds that a sufficient change in circumstances has occurred to warrant a change in this custodial arrangement. The court finds that the evidence presented by the Plaintiff meets and exceeds the clear and convincing burden placed upon him by the law.
*261The court finds that it is in the best interests of Taylor Winn that Plaintiff and Defendant share joint legal and joint physical custody of him.
The court has reviewed both Plaintiff’s and Defendant’s proposals for resolution of this matter. The court finds that Plaintiff’s proposal to be the better proposal and it shall be implemented with one additional point.
There has been much conflicting testimony about Taylor’s educational needs, and how to best meet them. Some witnesses felt that two educational settings would be too difficult for Taylor—stressing consistency as the reason. Others felt that as long as the programs were coordinated that two educational settings would be fine. The Defendant has enrolled Taylor in a Montessori program. The Plaintiff feels that the special education program in Charlevoix would be good for the boy. (Plaintiff has no objection to two educational settings.)
The court agrees with the opinion of Ms. Jodi Petry who is a clinical specialist in the pediatric occupational therapy at Mott Children’s Hospital at the University of Michigan. Ms. Petry has examined Taylor on more than one occasion and when asked whether she had an opinion on whether the situation where Taylor lived with one parent one week and the other the next week with two different school programs would that be in his best interests?
Her response was: “I think we really don’t know until it’s tried. I would certainly recommend a trial basis. I think his mother gives him some good things and I think his father gives him some good things. And at this point in time, you need to kind of see how Taylor deals in both settings. You know, he may be at a level in his development that he’s adaptable enough to make that transition. So I would recommend it with a lot of parameters.” Petry Deposition Pg 25, line 23-25; Pg 26, lines 1-9.
The courts (sic) order is that a provision dealing specifically with the allowance of Taylor to attend the Charlevoix educational program when with the Plaintiff, and the allowance that he may attend the Montessori educational program when with the Defendant is to be inserted into the *262Plaintiff’s proposal. The two programs are to be coordinated as much as possible.
Defendant argues that the trial court abused its discretion in changing the established custodial environment to joint custody, with Taylor living wdth each parent during alternate weeks. We agree. This Court reviews the trial court’s findings of fact in a custody matter under the great weight of the evidence standard, the court’s discretionary rulings for a palpable abuse of discretion, and the court’s rulings on issues of law for clear legal error. Fletcher v Fletcher, 447 Mich 871, 876-877; 526 NW2d 889 (1994). Therefore, because the trial court’s custody decision is a discretionary ruling, we review it for an abuse of discretion. Fletcher v Fletcher, 229 Mich App 19, 24; 581 NW2d 11 (1998).
We are mindful that this standard of review is highly deferential to the trial court. Our Supreme Court explained in Fletcher, supra at 880, that the standard of review is “not significantly unlike” the standard set forth in Spalding v Spalding, 355 Mich 382, 384-385; 94 NW2d 810 (1959):
Where, as here, the exercise of discretion turns upon a factual determination made by the trier of the facts, an abuse of discretion involves far more than a difference in judicial opinion between the trial and appellate courts. . . . In order to have an “abuse” in reaching such determination, the result must be so palpably and grossly violative of fact and logic that it evidences not the exercise of will but perversity of will, not the exercise of reason but rather of passion or bias.
In this case, because Taylor’s established custodial environment was with defendant, the trial court could change custody under MCL 722.27(l)(c); MSA *26325.312(7)(l)(c) only if clear and convincing evidence established that the change was in Taylor’s best interest.5 Ireland v Smith, 451 Mich 457, 461, n 3; 547 NW2d 686 (1996). Plaintiff bore the burden of proof in this regard. Heid v AAASulewski (After Remand), 209 Mich App 587, 593; 532 NW2d 205 (1995). Whether plaintiff has met this burden turns on the trial court’s findings regarding the statutory best interest factors.6 MCL 722.23; MSA 25.312(3). The trial court need not, however, give the statutory factors equal weight, and its finding regarding one factor does not necessarily countervail its other findings. McCain v McCain, 229 Mich App 123, 131; 580 NW2d 485 (1998). Further, mathematical equality on the statutory factors does not necessarily amount to an evidentiary standoff that would preclude a change in custody. Heid, supra at 594. The overwhelmingly predominant factor is, as always, the welfare of the child. Id.
Here, the trial court found that statutory best interest factor h slightly favored plaintiff and factors j and 1 favored him. Regarding factor h, the trial court remarked that evaluation of the factor was difficult because Taylor’s physical and mental disabilities *264greatly impair his ability to develop a home, school, and community record. The court also noted that very little information was available about Taylor’s recent home schooling because defendant did not keep records of her efforts to educate him. Plaintiff sought this information, but defendant did not comply with his requests.
Regarding factor j, the trial court found as follows:
The Defendant has shown somewhat less propensity to facilitate the Plaintiffs close and continuing relationship with Taylor over the last couple of years. This is demonstrated by her unilateral decisions concerning education, and also by being less forthcoming with information about the details of some of the boy’s treatment. She is also fairly rigid with the order for parenting time.
Expanding on this theme under factor 1, the trial court recognized that Taylor will need care and supervision throughout his life because of his mental and physical limitations. The court, therefore, determined that Taylor should spend as much time as possible with both parents to facilitate his development, to establish a comfortable relationship with them, and to ensure that plaintiff is capable of caring for Taylor in the event that defendant can no longer provide care.
Taken as a whole, the factors favoring plaintiff concerned Taylor’s need to spend as much time as possible with both parents and plaintiff’s willingness to encourage a relationship between Taylor and defendant. In this regard,7 Dr. Michael Nigro, a neurologist who examined Taylor on three occasions, explained:
*265[T]he formative years is when [Taylor] develops relationships, and he will only get to know each parent by spending time with each parent. And he’ll get to know them on a more personal level as not only the care giver but the person who cares most about him, and that only happens—it basically happens at this time especially in a child who is impaired. He does have limited resources for himself as an individual to take advantage of his experiences. And so if one theoretically wants to limit exposure to one or the other parent, that limited exposure more likely that (sic) is going to result in a limited relationship. So although it may not be the best of worlds, this is the time for him to spend as much time with each parent as he can, because I think he will er\joy the parent and understand who the parent is better today than he will later on.
Other factors, however, favored defendant. The trial court found that factors d and e slightly favored defendant. These factors reflect defendant’s role as Taylor’s primary caregiver since his birth. Defendant has provided Taylor with a stable environment in which he has recently experienced significant improvement in his development. Accordingly, the trial court had to weigh a desire to maintain stability in an environment in which Taylor is thriving against his need for a relationship with both parents. The trial court’s quandary was one that courts face when a child has two parents who have the ability and desire to care for the child, but have divorced.
The parties presented numerous witnesses to advise the trial court in making its decision. Employees of the Charlevoix Public School System, the Traverse City School System, and a Montessori school located in Traverse City testified regarding their ability to coordinate Taylor’s schooling in the event that he alternated weeks in different communities. The witnesses generally agreed that Taylor, unlike other *266children his age, could alternate schools weekly because his educational goals are very basic. The witnesses explained that the focus of Taylor’s education is not on scholastic matters, but rather, on developing communication and motor skills, as well as encouraging social interaction. Accordingly, the logistical obstacle that normally would preclude joint physical custody where parents reside in different communities is not present in this case.
In the end, therefore, the trial court faced the difficult task of attempting to predict how plaintiff’s proposed custody arrangement would affect Taylor. To assist the court, several witnesses testified regarding whether the weekly change in home environment was in Taylor’s best interest. Of these witnesses, three opposed plaintiff’s proposal, two recommended that the trial court adopt it, and one gave a qualified endorsement. On scrutiny, however, the record is devoid of clear and convincing evidence that plaintiff’s proposal was in Taylor’s best interest.
The three witnesses who opposed plaintiff’s proposal testified unequivocally that the joint custody arrangement was not in Taylor’s best interest. Dawn Brown, Taylor’s occupational therapist who provides therapy for him three times weekly, testified that Taylor requires consistency and that he has difficulty adapting to change. She opined that plaintiff’s proposed custodial arrangement would not be in Taylor’s best interest:
I believe that the amount of days that we’re talking, the seven days here, seven days here, is conflicting with what I think is best for Taylor. It’s not 50 percent of the time with John, 50 percent of the time with Jo; it’s just the every other week flip flop that I would be concerned about.
*267Regarding the prospect of subjecting Taylor to two school environments, Brown opined as follows:
My best professional judgment is that it would be very challenging for Taylor to do that. I feel that if you need to try it, try it, and if he fails, he has to go through that. And then he’ll crash and then we’ll have to work on putting the pieces back together.
Brown recommended that the court not attempt to divide Taylor’s education among two schools.
Gail Stahl, Taylor’s nurse, likewise testified that consistency is critical to Taylor’s development. Regarding the proposal that Taylor attend two schools, Stahl opined:
I believe Taylor’s reaction could be—could be positive and it could be negative. I have a tendency to believe that it would be on the most part negative. For a child of that age or of his development, it would seem that consistency would be a vital part of him knowing what to expect. I believe every child has a right to know what to expect what’s going to happen to him. Even if you were to tell Taylor, you’re going to developmental kindergarten this week, and or we’re going to see Mrs. So and So, I don’t know that he would comprehend. He needs to be able to predict that when he goes to school, it’s going to be at one place. He needs to be able to predict that when he goes to Nanna’s house, which is me, that that’s going to be at my home.
Julia Ann Crowter, a clinical psychologist, testified that although a child benefits from equal access to both parents, plaintiff’s proposal was not in Taylor’s best interest:
In terms of my opinion as to what the custody arrangement should be, I don’t have any opinion on that, because I haven’t done a custody evaluation in this case. In terms of *268placement with Mother or Dad, I don’t know Mother or Dad, so I can’t render an opinion.
The only thing I do have information on or an opinion, a professional opinion on, is to the, um, rightness, correctness of having Taylor attend, live in two communities, two different residences and attend two different schools on an every other week basis, and the question is do I think that’s in his best interest, I would have to say no.
Crowter further explained:
[I]n the first place, I think it would be difficult to argue for an arrangement like this given the geographical differences—or distance between Mother and Dad’s home for any child to be able to be educated in, develop in, that much of a disruptive environment every other week setting. Having to get used to two different school settings, two different therapeutic settings perhaps if that was being planned in the Charlevoix area, two different communities and two different homes. I think this opens the door for opportunities for Taylor to be significantly disruptive in terms of his growth and development.
The evidence in favor of plaintiffs proposal, by contrast, consisted primarily of the opinion testimony of two witnesses—Gerald Miks and Dr. Robert Shuman. This testimony does not, however, withstand even the most deferential scrutiny we apply in custody cases. Although Miks, a friend of the court investigator, recommended a joint custody arrangement, he explained that he presumes that joint custody is the appropriate custodial arrangement unless a parent proves otherwise by clear and convincing evidence. Thus, Miks did not objectively assess this case, but rather improperly presumed that joint physical cus*269tody was in Taylor’s best interest.8 Moreover, Miks’ investigation was less than complete. Miks did not send the parties the extensive questionnaire he normally uses to evaluate custody cases. He met with plaintiff two or three times during his investigation, but first saw defendant three days before trial. Miks, in fact, acknowledged the incompleteness of his investigation, stating that he would defer to the experts’ opinions regarding Taylor’s best interest.
Dr. Shuman, Taylor’s neurologist, likewise opined that plaintiff’s proposal was in Taylor’s best interest. Dr. Shuman, however, testified that his ideal custodial arrangement would entail Taylor residing in one location with his parents alternating weeks in that house. Rejecting this arrangement as unfeasible, Dr. Shuman endorsed plaintiff’s proposal. Dr. Shuman reasoned that while Taylor requires a stable environment, his need for contact with both parents during his present stage of development outweighs any need for a single home environment.
Although this Court usually defers to the trial court’s resolution of a dispute among experts, see Fletcher, supra, 447 Mich 890, we cannot ignore the total contradiction between Dr. Shuman’s deposition and trial testimony. Dr. Shuman acknowledged that he opined in his deposition ten days before his trial testimony that plaintiff’s proposal was not in Taylor’s best interest. Dr. Shuman stated that plaintiff phoned him after the deposition to inform him that his deposition testimony would be used to reduce plaintiff’s parent*270ing time, rather than increase it. Dr. Shuman then had a change of heart. In explaining his contradictory testimony, Dr. Shuman stated that he “ha[d] been swayed to believe that one [environment] might be removed from [Taylor].”9 Dr. Shuman, however, could identify no evidentiary support for this belief, other than plaintiffs representations.
The only other witness to opine favorably regarding plaintiff’s proposal was Jodi Petry, an occupational therapist who first evaluated Taylor in 1994. Regarding whether plaintiff’s proposal was in Taylor’s best interest, Petry advocated experimentation:
I think we don’t really know until it’s tried. I would certainly recommend a trial basis. I think that his mother gives him some good things and I think his father gives him some good things. And at this point in time, you need to kind of see how Taylor deals in both settings. You know, he may be at a level in his development that he’s adaptable enough to make that transition. So I would recommend it with a lot of parameters.
[Y]ou still need to recall that he has sensory overload. And if you give him, say, too much vestibular stimulation at *271the same time give him a lot of auditory information and a lot of visual information, it may be too much for his brain to process and he’ll go into sensory overload. And I know Dawn Brown has worked with Mrs. Winn, and I don’t know if she has with Mr. Winn, on recognizing these signs and knowing what to do when that happens. I think both of them need training in that area to make this happen. And I guess the parameters would include that both of them work with Dawn Brown to make sure that that happens in a way that best meets Taylor’s needs. But, you know, I see his dad is doing some more dad kind of things. You know, a little more of the movement stimulation, you know, the roughhousing kinds of things. You know, he needs both.
And I’m comfortable that both parents are learning, and very capable and competent of learning what this child needs. I mean, we all make mistakes as parents, but under the guidance of a therapist, you know, if a mistake was made, then they will learn that this is something we ought to change and how to change it. And I really think that that could happen in this case.
On this record, the trial court abused its discretion in finding that plaintiff met his burden of proving by clear and convincing evidence that the change in physical custody was in Taylor’s best interest. With the exception of Dr. Shuman and Jodi Petry, the witnesses agreed that, while Taylor should spend as much time as possible with both parents, a joint custody arrangement involving a weekly change of physical custody was not in Taylor’s best interest. Dr. Shuman conceded that his recommendation was based in part on plaintiff’s suggestion that defendant or the court sought to further restrict Taylor’s contact with him. Dr. Shuman had strongly opined ten days earlier that the proposed joint custody arrangement was not in Taylor’s best interest. Petry, by comparison, merely *272remained optimistic that Taylor could possibly adapt to the weekly transfer of custody. In light of the uncontroverted evidence that Taylor is developing and receiving excellent care under the current custodial arrangement, the mere possibility that he could weather the weekly transfer of custody does not establish by clear and convincing evidence that the change is in his best interest.
We do not imply that the trial court can never modify a custody order when a child is thriving under the then-existing custodial arrangement. We merely recognize that a child’s development is not something with which courts should experiment and risk disruption. Although ideally a child would develop a close relationship with his loving and caring parents through an equal division of the parenting time, the ideal is difficult to achieve when, as in this case, the child’s parents elect to establish their homes in different communities. This problem is further compounded by the friction that often develops between ex-spouses as they move on with their lives after their divorce. As parents, however, they must cooperate with each other in raising their child, rather than turning to the courts for a rigid custody arrangement, that, while seemingly avoiding conflict and ensuring equal access to the child, may have a devastating effect on the child’s well-being. In this case, we cannot overlook the stress inflicted on a child who is shuffled weekly between parents who live in different communities.
We likewise do not condone defendant’s recent actions that place her interests above Taylor’s need for a relationship with his father. Defendant’s hostility toward plaintiff may well be understandable under *273the circumstances of this case. Both parents, however, must consider the effect of their actions on Taylor and act in his best interest. In this case, the new custodial environment creates a high risk of harming Taylor. In ordering this change in custody the trial court forgot that the paramount consideration in a child custody decision is the child’s best interests, not those of his parents. See Ireland, supra at 466; Held, supra at 595. Accordingly, the trial court abused its discretion in modifying the custody order.
Reversed.