OPINION
ANN CRAWFORD McCLURE, Chief Justice
Appellants (collectively the Loeras) appeal a take nothing judgment following a jury trial. The case arose out of a vehicular accident between the Loeras’ pickup truck and a tractor-trailer driven by Joe Fuentes and owned by Nabors Well Services, Ltd. (collectively referred to as Na-bors). Nabors defended the case in part based upon a seat belt defense: they contend the Loeras’ failure to wear seat belts was the cause of their injuries. The jury awarded substantial damages and found Nabors and its employee partially culpable, but based upon findings concerning the non-use of seat belts, the trial court entered a take nothing judgment.
When we first heard this case, we reversed the take nothing judgment and remanded the.case for a new trial because the controlling law at the time would not allow for the admitted evidence supporting Nabors’ seat belt defense.1 After we had decided this case, the Texas Supreme Court in Nabors Well Services, Ltd. v. Romero, 456 S.W.3d 553 (Tex.2015) concluded that intervening legislative and societal changes required a change in the law. Id. at 566-67. It held that relevant evidence of the use or nonuse of seat belts (or other pre-occurrence injury-causing conduct) is admissible if such conduct caused or was a cause of the damages sought. Id. at 563. The court then vacated our judgment in this case, and remanded it for consideration in light of Romero. Nabors Well Services, Ltd. v. Loera, 457 S.W.3d 435 (Tex.2015).
We now face additional questions not addressed in our original opinion, including how the seat belt defense was submitted to the jury, whether the experts that Nabors used to advance the seat belt defense pass muster under TexR.Evid. 702, and whether the Loeras were operating as a single business enterprise. For the reasons noted below, we affirm the judgment as to Armando Loera individually and as to Mo-rayrna Loera, but reverse and remand as to the claims made on behalf of Josephina Loera.
[765]*765FACTUAL SUMMARY
On a clear sunny day in October 2006, Morayma Loera was driving a 2004 Dodge Ram pickup truck westbound on Highway 67 towards Marfa. Her mother, Josephi-na, was in the front passenger seat and her father, Armando, was in the rear passenger seat behind Morayma. Josephina was a double amputee. Her left leg had been amputated above the knee and her right leg had been amputated below the knee due to complications from diabetes. Because of an earlier stroke, she had lost most of the movement on the left side of her body. Some evidence supports the contention that while she was initially buckled in, she had unfastened her seat belt with her right hand because it was uncomfortable. Other evidence supports the claim that she had her seat belt buckled. There was no dispute, however, that Armando and Morayma were not secured in their seat belts. Morayma had in fact fastened her seat belt behind her to defeat the vehicle’s seat belt alarm.
Near the intersection of Highway 67 and a small county road, Morayma came up on two tractor trailer rigs owned by Nabors, also traveling westbound. The lead truck was driven by Joe Fuentes. Both rigs had slowed down and moved to the right in preparation for an upcoming left hand turn. Nabors contended the trucks had moved over just enough to set up for the left hand turn, both with their turn signals on. Conversely, the Loeras contended the trucks had moved all the way onto the paved shoulder, as if to let them pass. Morayma indeed attempted to pass and as she went around the rear truck, the lead rig turned left blocking her path. She hit her brakes and swerved, but impacted the tractor just at the point of its rear wheels. Prior to the accident, Morayma was traveling approximately 70 miles per hour. Two accident reconstruction experts testified in the case; one estimated her speed at the point of impact at between 30 to 35 miles per hour, and the other expert believed her impact speed was between 40 to 50 miles per hour. All three of the Loeras suffered injuries from the collision.
Josefina and Armando filed suit against Nabors alleging various claims of negligence and vicarious liability. Morayma was joined as a third party defendant and she then asserted her own claims against Nabors.
The Charge and Verdict
The jury was asked five liability questions. In question one, the jury found that Morayma and her parents were engaged in a joint business enterprise at the time of the accident, which would thus impute Mo-rayma’s negligence to Josefina and Armando. In questions two and three, the jury found Joe Fuentes to be 50% negligent in causing or contributing to cause the “occurrence or injury,” Nabors 10% negligent, and Morayma Loera 40% negligent.
The jury was also asked two questions regarding the Loeras’ failure to wear seat belts and answered as follows:
Question No. k;
Was the non-use of a seat belt by any of the persons named below negligent and a proximate cause of the injuries, if any?
Answer ’Yes’ or ’No’:
a) Morayma Loera: Yes
b) Josefina Loera: Yes
c) Armando Loera: Yes.
Question No. 5:
If you answered Yes’ to Question 4 for any of those named below, then answer the following question. Otherwise, do not answer the following question.
Assign percentages of responsibility only to those you found caused or contributed to cause the injury due to non-[766]*766use of a seat belt. The percentages you find, if any, are separate percentages for each individual below. The percentage of responsibility attributable to any one person named below is not necessarily measured by the number of acts or omissions found.
For each person you found negligently caused or contributed to cause the injury due to non-use of a seatbelt, find the percentage of responsibility, if any, attributable to each for such non-use:
a) Morayma Loera: 100%
a) Josefina Loera: 100%
b) Armando Loera: 100%
In a series of damage questions, the jury found that the Loeras suffered approximately $450,000 in damages as a result of the collision. But based upon the answers to questions four and five, the trial court entered judgment in favor of the defendants and ordered that the Loeras take nothing.
The Appeal
The Loeras raise four issues on appeal. Issue One contends that the jury’s answer to the seat belt questions was worded in such a way as to make the answers immaterial, and consequently the trial court erred in entering the take nothing judgment based on them. Issue Two challenges the sufficiency of the evidence to support the jury’s answer to question one (the joint venture issue) such that Moray-ma’s negligence cannot be imputed to Josefina and Armando. Issue Three claims that the seat belt defense was not available under Texas law. Finally, Issue Four, asserts the trial court erred in allowing two expert, Drs. Funk and Smith, to testify in support of the seat belt defense.
On original hearing of this case, we addressed only Issue Three. With the legal basis for the seat belt defense now firmly resolved by Nabors Drilling v. Ramirez, we turn to the remaining three issues.
CHARGE ERROR
The Loeras first contend that in question five, the jury was asked no more than who was responsible for each of them not buckling their seat belt. Isolating only the last sentence of the question, it reads:
For each person you found negligently caused or contributed to cause the injury due to non-use of a seatbelt, find the percentage of responsibility, if any, attributable to each for such non-use.
The Loeras contend that it really asked the jury to find the percentage of responsibility “attributable to each [plaintiff] for such non-use [of seat belts].” The correct inquiry, they contend, would ask the percentage of responsibility for their injuries attributable to not using their seat belts. Because the percentage of responsibility question, as asked, yielded nothing material to the case, the Loeras urge that the trial court abused its discretion in using the answers to formulate the judgment. See C. & R. Transp., Inc. v. Campbell, 406 S.W.2d 191, 194 (Tex.1966)(a jury finding may be disregarded when the issue is immaterial). They raised this issue in their post-trial filings, which were overruled by the trial court.
Nabors responds that the Loeras waived any error in the wording of the charge by not raising that complaint in the charge conference. The Loeras objected to the submission of questions four and five, but only on the grounds that the seat belt defense was not a legally viable defense, and not because the question was improperly worded. A party is generally obligated to timely object to the defective form of a jury issue or instruction. See Tex.R.App. P. 33.1; Tex.R.Civ.P. 274; In re B.L.D., 113 S.W.3d 340, 349 (Tex.2003). That obligation serves an important pur[767]*767pose; a timely and specific objection allows the trial court to wordsmith the charge and thereby avoid the wasted judicial resources of unnecessary appeals and retrials. Id. at 350; In re C.O.S., 988 S.W.2d 760, 765 (Tex.1999). Here for instance, the defect the Loeras identify in question five could have been easily fixed by the modification, or perhaps even the deletion of the last prepositional phrase of the question.
The Loeras respond that they are not asserting error in the formulation of the question; they never wanted it submitted at all. The error they claim is in the entry of the judgment based on the answer to a meaningless jury question. Nabors carried the burden of obtaining findings to support its defense. See Ramos v. Frito-Lay, Inc., 784 S.W.2d 667, 668 (Tex.1990)(noting general obligation of party to obtain findings supporting claim). If an entire defensive theory is omitted from the charge, the Loeras are correct that they had no obligation to object. Tex. R.CxvP. 279. Rather, any error in considering the jury’s answer is preserved by post-trial motions such as they filed. See J & C Drilling Co. v. Salaiz, 866 S.W.2d 632, 640 (Tex.App.—San Antonio 1993, no pet.)(disjunctive liability that finding that “Defendant 1 or Defendant 2” was negligent did not establish that any particular Defendant was liable and post-trial motion preserved error).
We agree that as the issue is presented to us, the Loeras have preserved the error they claim. Neither party contends that the wording of the charge was confusing or ambiguous. Instead, the Loeras contend that the question unambiguously asks only about allocating responsibility for the decision to use or not use a seat belt. Conversely, Nabors contends that the question asks about allocating responsibility for the injuries caused by the non-use of the seat belts. Given only those two options, we agree with Nabors.
The question asked for a percentage of responsibility “attributable to each for such nonuse.” [Emphasis added.] The use of the term “such” modifies “non-use” and necessarily refers the jurors back to use of the same term elsewhere in the charge. American Heritage Dictionary, 1215 (1985)(defining “such” as “being the person or thing implied or previously mentioned”); Webster Third New Int’l Dictionary 2283 (2002)(defining “such” to include “having a quality already or just specified”); Black’s Law Dictionary 1661 (10th ed.2014)(defining “such” as “[t]hat or those; having just been mentioned”); Coker v. Texas Alcoholic Bev. Commn., 524 S.W.2d 570, 574 (Tex.Civ.App.—Dallas 1975, writ ref d n.r.e.)(use of word “such” in statute meant to refer back to the preceding language).
The instruction just above the actual question refers to non-use in the context of “those you found caused or contributed to cause the injury” from the non-use. The question itself focuses the jury on those “you found negligently caused or contributed to cause the injury due to non-sue of a seatbelt....” In question four, which was the predicate question to question five, the jury had just decided whether the “non-use of a seat belt” by Moray-ma, Josefina, Armando was negligent and the proximate cause of their injuries. As our supreme court has stated, context is everything:
When construing statutes, or anything else, one cannot divorce text from context. The meaning of words read in isolation is frequently contrary to the meaning of words read contextually in light of what surrounds them. Given the enormous power of context to transform . the meaning of language, courts should resist rulings anchored in hyper-[768]*768technical readings of isolated words or phrases. The import of language, plain or not, must be drawn from the surrounding context, particularly when construing everyday words and phrases that are inordinately context-sensitive. [Emphasis added.]
In re Off. of the Atty. Gen. of Texas, 456 S.W.3d 153, 156 (Tex.2015). The Loeras singular focus on only the last prepositional phrase of question five falls into that same trap.
We think the reference to “such non-use” had the jury focus on the injury causing potential of the seat belt non-use, and not merely the mechanical act of buckling, unbuckling, or failing to buckle the belt itself. That construction seemed to be recognized by the lawyers at trial. In closing argument, both counsel for the Lo-eras and Nabors argued that question five asked the jury to find the degree to which the non-use of seat belts caused the injuries sustained. Charge error is one instance in which we review the closing arguments to understand whether the charge’s wording results in reversible error. See Timberwalk Apartments, Partners, Inc. v. Cain, 972 S.W.2d 749, 755-56 (Tex.1998)(in conducting harm analysis as to charge error, court reviews the entire record, including closing arguments, and finding reversible error, in part, based on closing arguments). We think it equally important to consider how the charge was argued in determining the contextual meaning of the jury question asked.2 We accordingly overrule Issue One.
THE EXPERT TESTIMONY
In Issue Four, the Loeras challenge the admissibility of the expert testimony used to support the seat belt defense. Nabors advanced its theory through James Funk, PhD, a biomechanical expert, and Dr. Harry Smith, a licensed physician and engineer. We first summarize the opinions that they expressed and the basis for those opinions.
Opinions On How The Injuries Occurred
Dr. Funk focused on the Loeras’ body kinematics in the crash; that is, how their bodies moved during the impact. His analysis utilizes Newton’s First Law of Motion: an object in motion stays in motion in the same direction and at the same speed unless an external force is applied to it. The Loeras were inside a pickup moving in a particular direction and speed and would maintain that same direction and speed until some outside force acted on them. When the pickup abruptly came to a standstill from the impact with the Na-bors truck, their bodies would continue in the same direction and speed unless arrested by. a seat belt, or impacted by something in the interior of the cab.
For Morayma Loera, that something in the cab was the airbag and steering wheel. She suffered a head injury, bruising to her chest, and a left elbow fracture. The pickup had a second generation airbag that was intended to inflate, but then give way as the accident unfolded. It was designed to work in conjunction with the force arresting seat belt system, so as to allow the occupants to “ride down” their inertia. [769]*769Without the seat belt, Morayma moved forward at her immediate pre-impact speed, hit the airbag which gave way, and then impacted her chest on the steering wheel. Dr. Funk identified indentions on the steering wheel consistent with that scenario. She also suffered a coronoid fracture of her elbow which Dr. Smith opined was from her arms being pushed away from the steering wheel by the airbag’s deployment. Dr. Funk was less sure about what impacted her elbow. She further sustained a blow to the back of her head that she testified was from Armando’s head coming forward from the back seat and striking her. Dr. Funk verified this injury mechanism, noting a specific study on occupant-to-occupant impacts for unbelted passengers.
Similarly, Armando’s injuries were consistent with flying forward and impacting the back of Morayma’s seat with his knees, and the back of Morayma’s head with the front of his. He sustained a fracture of the acetabulum, which is the part of the pelvis forming the hip socket. Both experts explained that when Armando’s knee moved forward and struck the back of Morayma’s seat, the force of the impact was transferred through the knee to the femur and back to the acetabulum. Dr. Funk identified indentions in the seat back consistent with that impact. Armando sustained a head injury, consistent with having struck the back of Morayma’s head as he moved forward.
Josephina was the most seriously injured occupant.3 She suffered a fracture of her right femur which required an amputation of an additional portion of that leg just above the knee. Dr, Funk testified that fracture was caused by her knee impacting the glove box. He believes her left leg also struck some part of the dash, fracturing the neck of her femur at the hip’s ball and socket joint. She had a right wrist fracture, which showed axial loading consistent with her hand being extended to brace herself during the impact. She suffered a de-gloving injury to her right arm, which included compound fractures to her ulna and several broken wrist bones. Dr. Funk testified the bone fractures were from impact with vehicle’s window frame. Dr. Smith theorized this injury occurred when her arm got caught up in the unfastened seat belt. She suffered a lung contusion and rib fracture, which Dr. Funk testified came from impact with the airbag.
She also suffered a cut to her liver when it detached from a tendon, and she suffered extensive internal bleeding when a benign mass was severed from her uterus. One or both of these turned out to be life threatening injuries, and required an immediate laporatomy to control the bleeding. Dr. Funk expressed no opinions on these particular injuries. Dr. Smith testified at trial these injuries occurred because the bodies’ organs, like the body itself, follow Newton’s First Law, and they too will move forward until their energy is arrested. The liver thus tore from its tendon and the benign mass tore from the uterus in the impact.
Opinions On How Seat Belts Would Have Prevented Serious Injuries
For each of the Loeras, both Drs. Funk and Smith testified that wearing seat belts would have avoided or at least reduced the risk for major injuries. Dr. Funk based this opinion in part on the design charac[770]*770teristics of the Loera’s vehicle, as confirmed through crash testing with test dummies. The 2004 Dodge pickup showed significant crush damage to the front end, but very little intrusion into the cab area. The vehicle is designed this way so that most of the force of the impact is absorbed by the vehicle structure, and not the occupants. Crash dummy testing has shown that wearing seat belts will provide “good injury protection” in frontal collisions even at speeds from 30 to 38 miles per hour. The combination of the second generation airbag and force arresting seat belt allow the test dummy to ride down the impact as the front end dissipates the force of impact by giving way (crushing).
Dr. Funk also generally relied on a number of published studies on prior injury accidents. The studies are based on data from the National Automotive Sampling System/Crashworthiness Data System (NASS-CDS) which is a sampling of nationwide police reports in tow-away crashes. The NASS-CDS database includes information on different kinds of accidents (i.e. front end, side impact, rollover), the type and severity of injuries that each occupant experienced, the use of seat belts, and the force of impact. Dr. Funk here looked to the studies which measured the incidence and degree of injury from front end impacts at comparable speeds and concluded that belted occupants had a greatly diminished risk of serious injury at a frontal impact of 30 to 40 miles per hour.4
Dr. Funk then attempted to tie the specific injuries for each of the Loeras (as disclosed in the medical records) to his theory. For Morayma, her chest injuries would have been “reduced in severity or eliminated” because her chest would not have impacted the steering wheel. Neither Dr. Funk nor Dr. Smith claimed Mo-rayma would have avoided her elbow fracture. Rather, they contended it was not a serious injury. Similarly, Dr. Funk testified that Armando would not have suffered his head injury (nor Morayma hers) when he flew forward and struck her from behind. His knees would also not have impacted the back of her seat, thus sparing him the hip fracture.
Finally, Dr. Funk testified that Josephi-na would have minimized or avoided “many” of her injuries had she been belted. Her stump, chest, or elbow would not have struck the interior structure of the vehicle, thus avoiding those injuries. He declined to express any opinion regarding her internal bleeding from the detached liver and benign mass. None of the studies he discussed addressed those particular internal injuries. Dr. Smith testified that a seat belt’s retractor spool which is designed to let out a small segment of the belt would have assisted in reducing the risk of these injuries but he cited no studies or other data supporting that claim.
The Basis Of The Challenge
The Loeras filed a pre-trial motion challenging Dr. Funk and Smith’s opinions based on their qualifications along with the [771]*771relevance and foundational reliability of the opinions, and reurged the same objections at trial. In a reply brief to this court, they concede that both experts are qualified.5 Rather, they presently focus on only three arguments: (1) both experts assumed the seat belts in the Loeras’ vehicle were functioning, but never tested that assumption; (2) both experts failed to exclude objects in the compartment as a possible source of injury; and (3) none of the studies upon which they relied tested the effectiveness of seat belts on a double amputee such as Josefina. We limit our discussion to those three arguments.
Standards For Admission Of Expert Testimony
We review a trial court’s rulings on the admissibility of evidence for an abuse of discretion, including evidentiary rulings on expert testimony. Helena Chem. Co. v. Wilkins, 47 S.W.3d 486, 499 (Tex.2001); Broders v. Heise, 924 S.W.2d 148, 153 (Tex.1996). “The test for abuse of discretion is whether the trial court acted without reference to any guiding rules or principles.” E.I. du Pont de Nemours & Co., Inc. v. Robinson, 923 S.W.2d 549, 558 (Tex.1995).
With regard to experts, those guiding rules and principles are found in Tex. R.Evid. 702 which provides:
A witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if the expert’s scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue.
Id.6 Rule 702 requires at least three predicates: the witness must be qualified; the opinion must be relevant; and the opinion must be based on a reliable foundation. See Wilkins, 47 S.W.3d at 499; Robinson, 923 S.W.2d at 556. Only relevance and reliability are at issue here.
Expert opinion testimony is relevant when it is “sufficiently tied to the facts of the case [so] that it will aid the jury in resolving a factual dispute.” Robinson, 923 S.W.2d at 556 [citations omitted]. The requirement incorporates traditional relevancy analysis under Tex.R.Evid. 401 and 402. Robinson, 923 S.W.2d at 556. Simply put, irrelevant evidence is of no assistance to the jury. Id. Conclusory or speculative opinion testimony is not relevant because it does not tend to make the existence of a material fact more probable or less probable. See Tex.R. Evid. 401; Coastal Transport Co., Inc. v. Crown Cent. Petroleum Corp., 136 S.W.3d 227, 232 (Tex.2004).
Rule 702 also requires an expert’s testimony to be reliable. Robinson identifies six factors useful in determining reliability: (1) the extent to which the theory has been or can be tested; (2) the extent to which the technique relies upon the expert’s subjective interpretation; (3) whether the theory has been subjected to peer review and/or publication; (4) the technique’s potential rate of error; (5) whether the underlying theory or technique has been generally accepted by the relevant scientific community; and (6) the non-judicial uses which have been made of the theory or [772]*772technique. Robinson, 923 S.W.2d at 557. These factors are non-exclusive as Rule 702 requires a flexible inquiry. Id. (the factors “will differ with each particular case.”). For instance, some of these factors do not always fit well into automobile collision cases. Ford Motor Co. v. Ledesma, 242 S.W.3d 32, 39 (Tex.2007).
Subsequent case law dictates that reliability requires more than simply satisfying the Robinson factors. Gharda USA Inc. v. Control Solutions, Inc., 464 S.W.3d 338, 348-49 (Tex.2015); Volkswagen of Am., Inc. v. Ramirez, 159 S.W.3d 897, 904-05 (Tex.2004). Expert testimony might also be unreliable if “there is simply too great an analytical gap” between the data on which the expert relies and the opinion offered. Gammill v. Jack Williams Chevrolet, Inc., 972 S.W.2d 713, 726 (Tex.1998), quoting Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997). “Whether an analytical gap exists is largely determined by comparing the facts the expert relied on, the facts in the record, and the expert’s ultimate opinion.” Gharda, 464 S.W.3d at 348. Analytical gaps arise when experts improperly apply otherwise sound principles and methodologies, the expert’s opinion is based on incorrectly assumed facts, or the expert’s opinion is based on tests or data that do not support the conclusions reached. Gharda, 464 S.W.3d at 348 [citation omitted]. A court is “not required ... to ignore fatal gaps in an expert’s analysis or assertions that are simply incorrect.” Volkswagen, 159 S.W.3d at 912; Cooper Tire & Rubber Co. v. Mendez, 204 S.W.3d 797, 800-01 (Tex.2006). Nor should courts accept the mere ipse dixit of an expert (or, it is because I say it is). City of San Antonio v. Pollock, 284 S.W.3d 809, 816 (Tex.2009)(collecting cases). But, however these issues may play out, it is not the court’s role to decide if the expert’s opinions are correct, only that they are reliably formed. Exxon Pipeline Co. v. Zwahr, 88 S.W.3d 623, 629 (Tex.2002).
Application Of The Standards To Drs. Funk And Smith
Assumptions of Seat Belt Effectiveness
Both experts assumed that the seat belts in the 2004 Dodge pickup truck were functional and would have worked as intended had they been worn. They were largely unable to verify that 'fact. By the time they inspected the vehicle it had been partially cannibalized for parts. The seats were missing. Consequently, only the driver’s seat belt was still in the truck.
The Loeras contend the failure to verify that these particular seat belts worked is fatal to the admissibility of the opinions. We disagree. First, there was some evidence that would show the seat belts were operational. Morayma testified that each of the three seat belts latched into place; her belt was in fact latched behind her. She testified that she was unaware of any defect with the seat belts. The truck had recently passed a state inspection and one of the required items to pass that inspection is a functional front seat belt. Tex. Transp.Code Ann. § 548.051(a)(ll) (West Supp.2015)(“A motor vehicle ... registered in this state, must have the following items inspected at an inspection station or by an inspector: ,.. front seat belts in vehicles on which seat belt anchorages were part of the manufacturer’s original equipment.”). The vehicle was only two years old at the time of the collision. Dr. Funk was able to verify that the seat belt mechanism and retractor worked for the driver’s seat belt which was still available at the time of his inspection. He also knew that Morayma’s seat belt which was engaged behind her loaded properly in the collision in that it kept her seat back from collapsing when Armando hit it from the [773]*773back. Dr. Smith testified that Josefina’s unlatched seat belt loaded properly while her arm was entwined in it.
As the Loeras point out, testing is an important aspect of validating an expert’s theory. In Whirlpool Corp. v. Camacho, 298 S.W.3d 631, 642 (Tex.2009) the expert failed to test several predicates of his causation opinion that a component of a clothes dryer caused a house fire. Consequently, his stated conclusion was no evidence of a product defect. Like Camacho where the key components of the dryer were no longer available for testing, at least two of the seat belts here were unavailable for inspection and testing. But unlike Camacho, there has been some testing at least of the design and performance of the second generation airbags and force arresting seat belt systems with test dummies. General crash tests with similar seat belt systems of course do not prove that the particular seat belts in this particular vehicle would work as designed. But short of recreating the accident by using the very belts at issue with crash dummies, one would never precisely know how these particular belts would function. At some level, all experts make assumptions. In automobile crash cases, they might assume some mechanical system—the brakes, accelerator, engine, etc.—were functioning as they were designed. Those assumptions can of course be challenged, particularly where there is some evidence to suggest they were not operating as designed. But lacking any evidence that these seat belts had some defect (and given the only evidence to the contrary) we cannot say the trial court abused its discretion in overruling this particular objection.
The Loeras also rely on Wal-Mart Stores, Inc. v. Merrell, 313 S.W.3d 837, 840 (Tex.2010) which held that an expert’s theory that a house fire was caused by a halogen lamp was little more than speculation. The expert had failed to exclude other causes, and reasoned that because halogen lamps can cause fires, the lamp in that case actually did cause the fire at issue. Id. The Loeras contend that by the same token, just because seat belts generally might work does not prove these particular seat belts would work. We see a difference, however, between an assumption that all halogen lamps will cause a fire because some do, versus an assumption that the Loeras’ seat belts likely worked because most do (as shown by the data from real world crash studies). Viewed as a statistical probability, the fact that a few lamps malfunction does not make it more likely that all lamps malfunction. But the fact that most seat belts work as designed makes it more likely that any particular seat belt will likely work. This distinction is particularly true here when the only evidence before us suggests these particular seat belts did in fact work.
Excluding Other Variables— Compartment Items
The Loeras also contend that neither expert accounted for the possibility of injury from flying cargo inside the pickup cabin. Just as an unrestrained person can strike and injure another occupant, unrestrained cargo can similarly cause injury. Both experts were generally aware that there was loose cargo inside the cab, but neither specifically detailed how it reacted to the force of impact, other than any cargo would also follow Newton’s First law—it would move in the same direction as did the Loeras.
The Loeras claim that Nabors has entirely failed to respond to this argument in their briefing. We believe that they did respond, at least by contending that each of the major injuries resulted from impacts with the interior of the vehicle, or impact [774]*774with each other, that were established in the record. By evidencing how the Loeras impacted the vehicle, or each other, Na-bors necessarily disproved the alternative hypothesis that they were hurt from flying debris. Morayma, for instance, testified that her father’s head hit the back of her head. Her father agreed. This would necessarily exclude flying cargo as the source of both her and her father’s head injuries.
The other injuries were to the front of the Loeras’ bodies striking something in front of them. Cargo items therefore would only be relevant if they were in front of the Loeras. The experts discounted cargo items in that position based on the post-accident photos. Pictures depicted a blanket, television, clock, and pictures in the back seat beside Armando’s seat, and nothing in the floorboard in front of him. The testimony discussed only a small ice chest, purse, and blanket in Josephina’s floorboard. Prom the photos, Dr. Funk identified mostly fragile items that were all still intact. The Loeras’ own accident re-constructionist testified that post-accident photos showed a few small things in the front of the truck, and nothing that would cause significant injury to an occupant. The trial court did not abuse its discretion in excluding the experts’ opinions on this basis.
Application Of Existing Seat Belt Studies To A Double Amputee
Dr. Smith testified that Josefina likely would not have suffered her more serious injuries had she worn a seat belt. Dr. Funk expressed the same opinion excepting only the internal injuries. Together, they opine that a 73-year-old person with extensive pre-existing conditions (double amputations and diabetes) would not suffer any serious injuries in a frontal impact at 30 to 40 miles per hour, so long as she was wearing a seat belt. Contrary to Nabors’ suggestion in some of its earlier briefing in this case, we believe that this proposition would require expert testimony.
Generally, “[e]xpert testimony is required when an issue involves matters beyond jurors’ common understanding.” [Citations omitted.] Mack Trucks, Inc. v. Tamez, 206 S.W.3d 572, 583 (Tex.2006). That common understanding must be sufficient to allow a jury to make a finding with reasonable probability. Id. In the context of product defects, for instance, the supreme court has often required expert testimony to support a jury finding that a product defect caused the plaintiff harm. Gharda, 464 S.W.3d at 348 (whether drum of chemical was contaminated and thus caused a fire); BIC Pen Corp. v. Carter, 346 S.W.3d 533, 542 (Tex.2011)(whether child resistant lighter would have prevented accident); Tamez, 206 S.W.3d at 582-83 (whether defective fuel system caused the release of diesel fuel and then ignited); Nissan Motor Co. v. Armstrong, 145 S.W.3d 131, 137-38 (Tex.2004)(whether product defect caused a motor vehicle to accelerate unintentionally). Expert testimony may also be needed to establish the extent to which medical expenses and treatments are causally related to an accident. Guevara v. Ferrer, 247 S.W.3d 662, 669-70 (Tex.2007). We agree that experts are not always required. JLG Trucking, LLC v. Garza, 466 S.W.3d 157, 163-64 (Tex.2015)(expert testimony to show evidence of injuries from one auto accident contributed to the injuries being claimed by the plaintiff in another accident). But the proposition that an already frail person in a thirty to fifty mile an hour frontal collision would suffer no serious injuries if wealing a seatbelt, when the same person suffers catastrophic life threatening inju[775]*775ries when not belted, is outside the common knowledge of a jury.
Dr. Funk’s rationale for the no-injury proposition is based on the studies and testing that he cited. And indeed, empirical studies can provide a valid basis for establishing general propositions of causation, so long as the studies are otherwise shown to be reliable. See Merrell Dow Pharm., Inc. v. Havner, 953 S.W.2d 706, 713 (Tex.1997)(use of epidemiological studies to establish general causation in drug defect case); Bostic v. Georgia-Pacific Corp., 439 S.W.3d 332, 351 (Tex.2014)(same for toxic tort causation). But even if a study showing general causation is accepted, the party must show that their circumstances are similar to the group analyzed in the study. Bostic, 439 S.W.3d at 347, 358; Pollock, 284 S.W.3d at 815 (expert’s reliance on studies showing association between exposure to benzene at 31 ppm with leukemia was no evidence that plaintiffs exposure, which was far less, caused injury); Matt Dietz Co. v. Torres, 198 S.W.3d 798, 804 (Tex.App.—San Antonio 2006, pet. denied)(plaintifPs dose and exposure to particular chemical not shown to be similar to those in studies being offered). Similarly, Dr. Funk attempted to establish through studies on large numbers of vehicular accidents and crash tests that occupants wearing seat belts generally suffer far fewer serious injuries than those that are unbelted. Josephina Loera challenges the application of these studies to her situation because none of these published studies dealt with double amputees. Dr. Funk agreed that he has not relied on any study that looked at only a population of double amputees or even that his studies included double amputees. If the challenge here was focused on some seemingly irrelevant distinction—i.e. whether the occupants were left handed, or people with red hair—we might dismiss it out of hand. But Dr. Funk admitted that having legs does play some role in the functioning of a seat belt, at least to the extent it provides the occupant with leverage during the accident.7 Any person who has sat in a chair with their ieet on .the ground understands the aide they provide in maintaining proper posture..
To be sure, Dr. Funk ■ testified at the Robinson hearing that the general principles of seat belt effectiveness applied to a double amputee as they would a person with their-feet planted on the floorboard. He makes this. conclusion based on the premise of the lap belt lying across the sturdy bones of the iliac crest in the spine from which it can arrest any forward body movement. We find no evidence in the record, however, that the lap belt engaged across Josephina’s pelvis, and not for instance, across her abdomen. Dr. Funk also testified that he had once tested a seat belt in a crash dummy whose legs were removed. He was never asked the results of that test, but it suggests that the effectiveness of a seat , belt on an amputee is at least a testable hypothesis. Dr. Smith expressed, the opinion had Josephina been properly wearing a seat belt, she would not have injured her abdomen, right arm, or her legs, but he offered only his theory that the force arresting characteristics of [776]*776the seat belt reduced the inertia of the liver and the benign cyst sufficiently to prevent any tearing. He offered no published peer reviewed study or other proof to support that theory.
Essentially we are left with Dr. Funk’s and Dr. Smith’s theory that a double amputee will fare as well as a person with legs in an accident provided they wear a seat belt. While that may ultimately turn out to be a correct hypothesis, on this record, it is supported by nothing more than the experts’ claim that it is true. And because an expert’s ipse dixit is not a valid basis for an expert opinion, we are unable to accept the admissibility of that evidence, and neither should have the trial court.
Was The Error Harmful?
Josephina Loera carries the burden to show that the erroneous admission of the evidence probably caused the rendition of an improper judgment. Tex. R.App.P. 44.1(a)(1). As part of that inquiry, we must review the entire record, “considering the state of the evidence, the strength and weakness of the ease, and the verdict.” Reliance Steel & Aluminum Co. v. Sevcik, 267 S.W.3d 867, 871 (Tex.2008)(internal quotation marks and citation omitted). She must show the excluded evidence was “crucial to a key issue” and if so, that error was likely harmful. Id. at 873. “By contrast, admission or exclusion is likely harmless if the evidence was cumulative, or if the rest of the evidence was so one-sided that the error likely made no difference.” Id.
We conclude that the error in the admission of Drs. Funk’s and Smith’s opinions that Josephina would not have sustained any serious injury had she been wearing a seat belt was harmful error. The testimony went to the very question upon which the take nothing judgment was rendered. It was the only testimony from which the jury could have concluded that she would not have suffered any injuries had she been wearing a seat belt. We have already concluded that the proposition that a frail person in a front-on collision at thirty to fifty miles per hour would suffer no injuries requires expert opinion. Without this testimony, we find no other evidence to support the jury’s answer to question five. Accordingly, we sustain Issue Four, but only as to the claims of Josephina Loera, and remand that claim for a new trial.
SINGLE BUSINESS ENTERPRISE
The Loeras also challenge the sufficiency of the evidence in the record to support the jury’s finding that they were operating in a single business enterprise (thus imputing Morayma’s negligence to Josephina and Armando). Because the jury’s answer to question five was an independent basis to have entered a take nothing judgment against Morayma and Armando, and is unaffected by the single business enterprise issue, we find no need to address the issue as to them. Because we have reversed the judgment against Josefina based on the admission of the expert testimony and remanded it for a new trial on the merits, we also decline to review the record from this trial on the single business enterprise question. Accordingly, we decline to review Issue Two.8
[777]*777REMAND IN THE INTEREST OF JUSTICE
Alternatively, the Loaras ask us to remand the case in the interest of justice. Tex.R.App.P. 43.3. They claim their trial strategy, and in particular their approach to the charge, was based on the existing law at the time. In both Romero and in its decision remanding this case, the supreme court noted its preference for a single question which encompasses both the injury causing and occurrence causing conduct. Romero, 456 S.W.3d at 563; Loera, 457 S.W.3d at 437.9 The Loeras contend this is a change in the law which merits remand of the entire case back for new trial.
When the supreme court changes the law, it sometimes exercises its discretion to remand a case for new trial, thus not penalizing a party’s good faith reliance on existing precedents. See Boyles v. Kerr, 855 S.W.2d 593, 603 (Tex.1993)(“Remand is particularly appropriate where the losing party may have presented his or her case in reliance on controlling precedent that was subsequently overruled.”); Murray v. San Jacinto Agency, Inc., 800 S.W.2d 826, 830 (Tex.1990)(remanding in the interest of justice when the court modified prior precedent regarding when statute of limitations began to run); L.M.B. Corp. v. Gurecky, 501 S.W.2d 300, 303 (Tex.1973)(remanding case when party did not seek affirmative jury finding based on existing law at the time). While the Texas Supreme Court has the power to remand in the interest of justice under Tex, R.App.P. 60.2 (governing dispositions before Texas Supreme Court), that same authority is not co-extensive for the courts of appeals. See Davis v. Bryan & Bryan, Inc., 730 S.W.2d 643, 644 (Tex.1987)(court of appeals can only reverse when there is error in the trial court’s judgment and absent such error, a court of appeals cannot reverse the trial court’s judgment in the interests of justice); Estate of Clifton v. Southern Pac. Transp. Co., 709 S.W.2d 636, 639 (Tex.1986)(stating that a court of appeals cannot reverse an errorless judgment); Chrismon v. Brown, 246 S.W.3d 102, 116 (Tex.App.—Houston [14th Dist.] 2007, no pet.) (discussing distinction between power to remand in interest of justice between Texas Supreme Court and courts of appeals). We can reverse the trial court’s judgment in the interest of justice only when the trial court’s judgment is in error. With regard to the Armando Loera and Morayma Loera, we find no error in the judgment as to them. Accordingly, we overrule their alternate request that we remand their claims in the interest of justice.
CONCLUSION
We affirm the judgment as to Morayma and Armando and overrule Issues One and [778]*778Four. We decline to reach Issue Two. Issue Three was resolved by Nabors Well Services, Ltd. v. Romero, 456 S.W.3d 553 (Tex.2015) and we therefore overrule that issue was well. We sustain Issue Four as to Josephina and reverse the judgment as to her and remand for a new trial.