Under article III, section 45 of the Missouri Constitution, the General Assembly enacted H.B. 193 to redistrict Missouri for the election of the United States House of Representatives. Two groups of plaintiffs filed declaratory judgment actions to challenge the constitutional validity of the congressional redistricting map in H.B. 193 (the Map), claiming that it failed to meet the constitutional requirements for compactness. The trial court ruled that the plaintiffs failed to prove that the Map violates the requirement in article III, section 45 that each district be “as compact ... as may be” and entered judgments in favor of the defendants. On appeal, the plaintiffs claim that the trial court’s judgments erroneously interpret the constitutional standard for compactness and that the judgments are against the weight of the evidence.
This Court determines that the trial court did not err in its interpretation of the constitutional compactness standard because the standard does not require absolute precision in compactness and because mandatory and permissible recognized factors can impact the configuration of district boundaries. This Court further holds that the plaintiffs do not prevail on their claim that the trial court’s judgment is against the weight of the evidence. This Court generally refuses to substitute its opinion for that of the trial court on disputed factual issues by re-weighing the evidence in a court-tried case. The parties strenuously disputed whether the challenged districts depart from the compactness principles and if they are “as compact ... as may be,” specifically whether minimal and practical deviations from compactness are supported by factors recognized by this Court. The trial court made credibility assessments and weighed the evidence at trial in reaching its judgments. Because this case involves judgments for the defendants, who have no burden of proof, and because neither party requested findings of fact that would assist in appellate review, this is not a case in which this Court should substitute its judgment for that of the trial court on the evidence regarding disputed factual issues.
Accordingly, this Court affirms the judgments of the trial court.
Procedural History
The Missouri Constitution provides that the General Assembly shall divide the *42state into districts based on the number of representatives to which it is entitled, as determined under the decennial census of the United States. Mo. Const, art. Ill, sec. 45. The Missouri Constitution requires the General Assembly to enact a map with districts that “shall be composed of contiguous territory as compact and as nearly equal in population as may be.” Id. In May 2011, over the governor’s veto, both houses of the General Assembly voted to approve the Map in H.B. 193.
Two groups, the Pearson Plaintiffs and the McClatchey Plaintiffs1 (collectively, “Plaintiffs”), filed actions to challenge the Map under Mo. Const, art. Ill, sec. 45, each asserting claims that the districts were not “as compact ... as may be.” Defendants, Attorney General Chris Roster and Secretary of State Robin Carnahan, as well as intervenors Representative John. J. Diehl and Senator Scott T. Rupp,2 filed motions to dismiss or for judgment on the pleadings. The trial court sustained the motions and dismissed both cases. Plaintiffs appealed to this Court.
This Court consolidated the cases and decided Pearson v. Raster on January 17, 2012. See 359 S.W.3d 35 (Mo. banc 2012) (Pearson I). In Pearson I, this Court affirmed the trial court’s dismissal of Plaintiffs’ petitions on every claim except the claims regarding compactness. Id. at 40. This Court stated:
[I]t is a question of fact, yet to be tried, whether those districts are “as compact and nearly equal in population as may be.” Mo Const, art. Ill, sec. 45 (emphasis added). This Court makes no prejudgment on these issues, or on the compactness of other districts, other than to hold that Plaintiffs have stated a claim as to the compactness of the districts that is subject to proof and defenses in accordance with evidence as in any other lawsuit.
Id. This Court remanded the cases to the trial court for determination of the factual issues. Id.
On remand, the trial court held a trial to determine whether the districts are “as compact ... as may be.” Mo. Const, art. Ill, sec. 45. Plaintiffs and Defendants made various stipulations, and both presented evidence regarding whether the challenged districts are “as compact ... as may be.” After hearing the evidence, the trial court entered judgments in favor of both Defendants on February 3, 2012. It determined that the phrase “as compact ... as may be” means that compactness cannot be achieved with absolute precision and permits districts to be drawn in multiple ways while still meeting the compactness requirement due to other factors. The trial court then found that, “[ujnder the standard and rationale announced by the Supreme Court, and the facts adduced at trial, the Plaintiffs have failed to prove that H.B. 193 is unconstitutional because it is not ‘as compact as may be.’ ” Plaintiffs appeal.
On appeal, the Pearson Plaintiffs and the McClatchey Plaintiffs assert multiple claims of error in the trial court’s judgments. The Pearson Plaintiffs claim that the trial court erred in: (1) applying an improper standard for determining whether the Map is as compact as may be; and *43(2) finding the Map as compact as may be under the facts, in that a visual observation and other maps show that it could be significantly more compact. The McClat-chey Plaintiffs claim that the trial court erred in: (1) improperly interpreting the phrase “as compact ... as may be”; (2) failing to shift the burden to Defendants to justify deviations from compactness; and (3) failing to find that district 5 is not as compact as may be under the circumstances. Plaintiffs’ points on appeal essentially constitute claims that the trial court’s judgments erroneously apply the law and that they are against the weight of the evidence.
Standard of Review
This Court’s decision in White v. Director of Revenue details the applicable standard of review for appeals of court-tried civil cases. 321 S.W.3d 298, 307-08 (Mo. banc 2010). The judgment of the trial court will be affirmed “unless there is no substantial evidence to support it, it is against the weight of the evidence, or it erroneously declares or applies the law.” Id. (citing Murphy v. Carron, 536 S.W.2d 30, 32 (Mo. banc 1976)). The application of this standard of review varies depending on the burden of proof applicable at trial and the error claimed on appeal to challenge the judgment. See In re Adoption of W.B.L., 681 S.W.2d 452, 454 (Mo. banc 1984). The reviewing court cannot review the judgment of a trial court properly under a given standard of review without considering the burden of proof governing the trial court’s determination.
The burden of proof applicable at trial depends on the type of claim presented in the pleadings. This case involves a challenge to the constitutional validity of a statute. For a court to find that a statute is unconstitutional, the plaintiff must overcome a burden of proof that assumes constitutional validity. Mo. Prosecuting Attorneys v. Barton County, 311 S.W.3d 737, 740 (Mo. banc 2010). The statute will not be held unconstitutional unless the plaintiff proves that it “clearly and undoubtedly contravene^] the constitution” and “plainly and palpably affronts fundamental law embodied in the constitution.” Id. at 740-41 (internal quotations omitted); see also St. Louis Cnty. v. Prestige Travel, Inc., 344 S.W.3d 708, 712 (Mo. banc 2011). All doubts are “resolved in favor of the constitutionality of the statute.” Barton County, 311 S.W.3d at 741 (internal quotations omitted).
In addition to the burden of proof, the reviewing court also must apply the proper standard of review for the error claimed on appeal. A claim that there is no substantial evidence to support the judgment or that the judgment is against the weight of the evidence necessarily involves review of the trial court’s factual determinations. See White, 321 S.W.3d at 308. A court will overturn a trial court’s judgment under these fact-based standards of review only when the court has a firm belief that the judgment is wrong. Id. A claim that the judgment erroneously declares or applies the law, on the other hand, involves review of the propriety of the trial court’s construction and application of the law. Id. Implicit in these standards is the recognition that the trial court, in reaching its judgment, is in a better position to determine factual issues than an appellate court reviewing only the record on appeal. See id. at 308-09. In this regard, it is necessary for the reviewing court to treat differently questions of law and questions of fact.
This Court applies de novo review to questions of law decided in court-tried cases. StopAquila.org v. City of Peculiar, 208 S.W.3d 895, 899 (Mo. banc 2006). With respect to such questions, *44“the appellate court reviews the trial court’s determination independently, without deference to that court’s conclusions.” Moore v. Bi-State Dev. Agency, 182 S.W.3d 241, 242 (Mo. banc 2004). “The quintessential power of the judiciary is the power to make final determinations of questions of law,” and courts may not delegate that authority to anyone else. Asbury v. Lombardi, 846 S.W.2d 196, 200 (Mo. banc 1993) (citing Marbury v. Madison, 1 Cranch 137, 5 U.S. 137, 2 L.Ed. 60 (1803)).
In reviewing of questions of fact, the reviewing court will defer to the trial court’s assessment of the evidence if any facts relevant to an issue are contested. White, 321 S.W.3d at 308. A factual issue is contested if disputed in any manner, including by contesting the evidence presented to prove that fact. Id. As enunciated in White, a party can contest the evidence in many ways, such as by putting forth contrary evidence, cross-examining a witness, challenging the credibility of a witness, pointing out inconsistencies in evidence, or arguing the meaning of the evidence. Id. Once contested, “a trial court is free to disbelieve any, all, or none of th[e] evidence,” and “the appellate court’s role is not to re-evaluate testimony through its own perspective.” Id. at 308-09. The trial court receives deference on factual issues “ ‘because it is in a better position not only to judge the credibility of the witnesses and the persons directly, but also their sincerity and character and other trial intangibles which may not be completely revealed by the record.’ ” Id. (quoting Essex Contracting, Inc. v. Jefferson Cnty., 277 S.W.3d 647, 652 (Mo. banc 2009)).
A claim of error on appeal may present a mixed question of law and fact. In such an instance, the reviewing court applies the same principles articulated above except that it is necessary to segregate the parts of the issue that are dependent on factual determinations from those that are dependent on legal determinations. “[W]hen presented with an issue of mixed questions of law and fact, a [reviewing court] will defer to the factual findings made by the trial court so long as they are supported by competent, substantial evidence, but will review de novo the application of the law to those facts.” 5 Am.Jur. 2D Appellate Review § 631 (2012). For example, when the issue is whether Missouri courts have personal jurisdiction over a defendant, a reviewing court defers to the fact-finding court with regard to any facts that are essential to that determination (e.g., the intention to create and sustain contacts with Missouri); however, the ultimate question of whether the exercise of jurisdiction meets the standards of the Missouri long-arm statute and the constitution remains a legal question, which is reviewed independently on appeal. See Longshore v. Norville, 93 S.W.3d 746, 751-54 (Mo.App.2002); see also State v. Werner, 9 S.W.3d 590, 595 (Mo. banc 2000) (applying de novo review to whether a defendant was “in custody” but deferring to the trial court with respect to facts essential to that determination); State v. Brooks, 185 S.W.3d 265, 273 (Mo.App.2006) (“While factual issues on motions to suppress often are mixed questions of law and fact, the trial court’s superior capacity to resolve credibility issues is not dispositive of the ‘in custody’ inquiry.” (citations omitted)). Therefore, it is a matter of deferring to the fact-finder in its assessment of the facts and then applying de novo review in determining how the law applies to those facts.3 White, 321 S.W.3d at 307 *45(“When the facts of the case are contested, this Court defers to the trial court’s assessment of the evidence.”).
Discussion
A. Review of the trial court’s judgment upholding the Map
As stated previously, review of the trial court’s judgment upholding the constitutional validity of the Map requires consideration of the burden of proof at trial and the applicable standard of review. Although the McClatchey Plaintiffs do not challenge that they have the burden to show that the Map clearly and undoubtedly contravenes the constitution, Barton County, 311 S.W.3d at 740-41, they contend that, at a point in the trial, the burden of proof shifts to the Defendants. The Pearson Plaintiffs alternatively contend that review of the trial court’s judgment should be de novo on the ground that this case largely consists of stipulated facts. Neither party is correct as to this Court’s review of the trial court’s judgment.
1. Plaintiffs have the burden of proof at all times
As stated in Pearson I, a redistricting case is “subject to proof and defenses in accordance with evidence as in any other lawsuit.” 359 S.W.3d at 40. Applying general law, “[t]he person challenging the validity of the statute has the burden of proving the act clearly and undoubtedly violates the constitution.” St. Louis Cnty., 344 S.W.3d at 712 (internal quotations and citations omitted). Although prior cases of this Court provide that statutes are “presumed” constitutional, a “presumption” is a term of art that denotes a specific meaning in the law. A presumption is a “standardized practice, under which certain facts are held to call for uniform treatment with respect to their effect as proof of other facts.” McCormick on Evidence, section 337 (6th ed. 2006). The purpose behind stating that statutes are “presumed” constitutional is not to establish a presumption but instead to allocate the burden of proof to the plaintiff for its claim that a statute is unconstitutional — i.e., the plaintiff has the burden to show that a statute “clearly and undoubtedly violates the constitution.” See St. Louis Cnty., 344 S.W.3d at 712 (internal quotations and citations omitted).
In their claim that Defendants fail to meet their burden of proof, the McClat-chey Plaintiffs erroneously state that Missouri law requires the burden of proof to shift to the state after a plaintiff makes an initial showing that a district could be more compact. Once the burden shifts, the state then would have to prove why the district is not substantially more compact. In support of this proposition, the McClatchey Plaintiffs cite the dissent in Preisler v. Kirkpatrick, which sought to shift the burden to the state in a challenge to the compactness of senate districts. See 528 S.W.2d 422, 436 (Mo. banc 1975) (Finch, J., dissenting), overruled in part by Pearson I, 359 S.W.3d at 39. Contrary to the McClatchey Plaintiffs’ assumption, the majority of this Court did not shift the burden in that case, despite the inquiry into the subjective intent of the legislature at that time. See id. at 426; see also Pearson I, 359 S.W.3d at 40 (holding that a subjective test no longer applies and re*46manding for the determination of factual issues “as in any other lawsuit.”).
The McClatchey Plaintiffs also cite as support the United States Supreme Court’s decision in Karcher v. Daggett, 462 U.S. 725, 103 S.Ct. 2653, 77 L.Ed.2d 133 (1983). In relying on Karcher, the McClatchey Plaintiffs fail to recognize the distinction between the standard used by federal courts for challenges to redistricting maps under the United States Constitution and the standard for challenges under the Missouri Constitution. Under art. I, sec. 2 of the United States Constitution, the standard for population equality “requires that the State make a good-faith effort to achieve precise mathematical equality.” Kirkpatrick v. Preisler, 394 U.S. 526, 530-31, 89 S.Ct. 1225, 22 L.Ed.2d 519 (1969). If the plaintiff shows that a deviation from population equality is not part of a good-faith effort of the legislature, federal courts shift the burden of proof, and “the State must bear the burden of proving that eáeh significant variance between districts was necessary to achieve some legitimate goal.” Karcher, 462 U.S. at 730-31, 103 S.Ct. 2653. Unlike the federal population equality standard, the existence of good faith in the legislature or lack thereof is irrelevant under the requirements in the Missouri Constitution. In Pearson I, this Court expressly rejected the good faith standard and held that the applicable standard is the language of the constitution itself, which is an objective standard. 359 S.W.3d at 40.
It is Plaintiffs who seek a declaration that the Map is unconstitutional, and shifting the burden of proof conflicts with their ultimate burden to show that the Map “clearly and undoubtedly” contravenes the constitution. See Johnson v. State, 366 S.W.3d 11, 33 (Mo. banc 2012) (determining that plaintiffs failed to prove their case, because they “failed to prove that it is possible to achieve greater population equality and compactness when considering federal law requirements and other factors.”). Placing the burden of proof on a plaintiff challenging a redistricting map is consistent with the framework used by nearly every state in the nation. The vast majority of states either have expressly rejected shifting the burden of proof4 or have treated redistricting cases the same as other constitutional challenges and maintained the burden of proof on the plaintiff.5 Only four states shift the burden of proof to the defendant after an initial showing by the plaintiff. Of those *47four states, two shift the burden in redistricting challenges based on the state’s constitution without providing any analysis,6 one applies the federal framework without providing any rationale for doing so,7 and the last shifts the burden based on a state statute specifically relating to reapportionment challenges.8
While federal courts apply a burden-shifting framework, they do so under a different constitutional standard, and there is no apparent rationale that supports a departure from the burden of proof applied by this Court in Johnson, at 32-83. Plaintiffs at all times have the burden of proving the Map is unconstitutional. The burden of. persuasion and the burden of production never shift to the defendants. See Anchor Centre Partners, Ltd. v. Mercantile Bank, N.A., 803 S.W.2d 23, 30 (Mo. banc 1991); White, 321 S.W.3d at 305.
2. Whether a district is “as compact ... as may be” is a mixed question of law and fact
Mo. Const, art. Ill, sec. 45 provides the compactness requirement for congressional districts. It states:
When the number of representatives to which the state is entitled in the House of the Congress of the United States under ... each census ... is certified to the governor, the general assembly shall by law divide the state into districts ..., which districts shall be composed of contiguous territory as compact ... as may be.
The determination of whether the compactness requirement is satisfied by the Map involves the determination of a mixed question of law and fact. The meaning of the language in Mo. Const, art. Ill, sec. 45 is a question of law to which de novo review applies. Comm, for Educ. Equal, v. State, 294 S.W.3d 477, 488 (Mo. banc 2009) (“[T]he trial court’s interpretation of the Missouri Constitution [is a] question[ ] of law given de novo review.”). In contrast, determining whether the characteristics of a particular map satisfy the meaning of the “as compact ... as may be” requirement involves questions of fact.9 It is necessary, therefore, to analyze separately the legal determinations from the factual determinations on the issue of whether the challenged districts are as “compact ... as may be.”
B. The trial court did not erroneously declare the meaning of “as compact *48... as may be” under Mo. Const, art. Ill, sec. 45
Mo. Const, art. Ill, sec. 45 requires that the General Assembly draw the House districts according to census figures, making the districts: (1) contiguous territory; (2) as compact as may be; and (3) as nearly equal in population as may be. These requirements are mandatory and objective — each must be satisfied — although the language used in the requirements may allow some flexibility in their compliance. See Johnson, at 23-24. Plaintiffs claim that the trial court erred in defining the standard for the “as compact ... as may be” requirement, which this Court reviews de novo. See StopAquila.org, 208 S.W.3d at 899.
In Pearson I, this Court stated that “the applicable standard for a court in reviewing an article III, section 45 claim is the language of the constitution itself[.]” 359 S.W.3d at 40. This Court assumes that every word in the constitutional provision has effect and meaning. Buechner v. Bond, 650 S.W.2d 611, 613 (Mo. banc 1983). The primary rule is to “give effect to the intent of the voters who adopted the [provision]” by considering the plain and ordinary meaning of the words used. Keller v. Marion Cnty. Ambulance Dist., 820 S.W.2d 301, 302 (Mo. banc 1991); StopAquila.org, 208 S.W.3d at 902.
This Court must consider the phrase “as compact ... as may be” in its entirety in order to ascertain its meaning and also give effect to every word used. See Buechner, 650 S.W.2d at 613. A determination of whether a district fails to satisfy the requirement cannot be accomplished solely by inquiring if it is “compact,” because the modifier “as may be” alters the meaning of that word. See Mo. Const, art. Ill, sec. 45. As with statutes, the construction of constitutional provisions is not to be a hyper-technical process. Donaldson v. Crawford, 230 S.W.3d 340, 342 (Mo. banc 2007).
Accordingly, contrary to the urging of Plaintiffs, the test for whether a district is “as compact ... as may be” is not a two-part test. Instead, the test is a single inquiry as to whether, under the totality of the evidence, the challenged district is “as compact ... as may be.” This test involves a determination of whether there is a departure from the principle of compactness in the challenged district and, if there are minimal and practical deviations, whether the district is nonetheless “as compact ... as may be” under the circumstances. See Pearson I, 359 S.W.3d at 40.
It is necessary to begin with the definition of the word “compact” in determining the meaning of the phrase “as compact ... as may be.” Courts considering the definition of “compact” as used in the context of the reapportionment of districts have recognized two possible definitions. Some courts define “compact” as referring to “the physical shape or size of electoral districts,” while others define it as referring to “closely united territory, a phrase not necessarily limited to physical dimensions.” See Kurtis A. Kempter, Annotation, Application of Constitutional “Compactness Requirement” to Redistricting, 114 A.L.R.5th 311 (2003). A century ago, this Court adopted the latter definition, finding that “compact” for Missouri redistricting purposes means “closely united territory” and, in effect, rejecting the proposition that “compact” refers solely to physical shape or size. See Pearson I, 359 S.W.3d at 38 (quoting State ex rel. Batrett v. Hitchcock, 241 Mo. 433, 146 S.W. 40, 61 (1912)).
Because the word “compact” does not refer solely to physical shape or size, a visual observation, although relevant, is *49not the • decisive factor in determining whether a district departs from the principle of compactness. In fact, scholars have recognized that “compactness” is a vague standard and have developed various statistical measures to be utilized in determining compactness, as shown by two articles that were admitted into evidence. One article states that “multiple measures should be used whenever possible,” and that there is no threshold level that can be shown by statistics.10 Richard G. Niemi, et al., Measuring Compactness and the Role of a Compactness Standard in a Test for Partisan and Racial Gerrymandering, 52 J. Pol. 1155, 1176-77 (1990).
Further, modification of the word “compact” with the phrase “as may be” recognizes that “compactness ... cannot be achieved with absolute precision.” Pearson I, 359 S.W.3d at 39. The existence of multiple districts prevents absolute compactness, as it would be with circular boundaries, because the boundary of one district must fit the boundary of another district, all within state territory lines. The “as may be” standard also recognizes that there are other recognized factors that affect the ability to draw district boundaries with closely united territory. These factors include the impact of the standards for contiguous territory and population equality. See Mo. Const. art. III, sec. 45; see also Pearson I, 359 S.W.3d at 38. By virtue of the Supremacy Clause, U.S. Const, art. VI, cl. 2, the districts also must comply with the United States Constitution and federal laws, such as the Voting Rights Act. See Johnson, at 27-28.
In addition, the phrase “as may be,” modifying the word “compact,” permits consideration of other recognized factors that inherently are included within the constitutional standards governing the reapportionment process, although not expressly articulated as a separate requirement in the constitution. See Johnson, at 27-28. These factors were identified by this Court in Preisler v. Doherty, 365 Mo. 460, 284 S.W.2d 427 (1955), Preisler v. Hearnes, 362 S.W.2d 552 (Mo. banc 1962), and Preisler v. Kirkpatrick, 528 S.W.2d 422. In Doherty, this Court distinguished the process of city officers dividing St. Louis into districts from the process of redistricting the state because “it is obvious that it is much easier for them to make compact districts than for a legislature or commission restricted to county lines (or following town, ward or other district lines)....” 284 S.W.2d at 432. In Hearnes, a case challenging congressional districts, this Court stated that:
[C]ounties are important governmental units, in which the people are accustomed to working together. Therefore, it has always been the policy of this state, in creating districts of more than one county (congressional, judicial or senatorial) to have them composed of entire counties.... We must hold that it was proper for the legislature to follow this policy. In fact, to do otherwise could lead to the most vicious kind of gerrymander. The only departure therefrom in the .1961 Act was in our two largest cities, St. Louis and Kansas City.... Urban conditions may justify this treatment.
362 S.W.2d at 556-57. The Court also stated that “it is not improper to consider the precedents of allocation of counties to existing districts in deciding the composition of new enlarged districts.” Id. at 557.
Later, this Court in Preisler v. Kirkpatrick cited the United States Supreme *50Court in Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964), to recognize that “districting, without any regard for political subdivision or natural or historical boundary lines, may be little more than an open invitation to partisan gerrymandering.” 528 S.W.2d at 425 (internal quotations omitted). This Court also recognized that population density may affect boundary lines, stating that “population density of the state is, of course, uneven and any effort to accomplish both the overriding objective of [population equality] and the preservation of county lines reasonably may be expected to result in the establishment of districts that are not esthetically pleasing models of geometric compactness.” Id. at 426.
As provided in these cases, the language used in the constitutional requirements implicitly permits consideration in the redistricting process of population density; natural boundary lines; the boundaries of political subdivisions, including counties,11 municipalities, and precincts; and the historical boundary lines of prior redistricting maps.12 This Court recently affirmed the continued propriety of recognized, unenumerated factors in Pearson I. See 359 S.W.3d at 40 (recognizing the importance of preservation of “the integrity of the existing lines of our various political subdivisions,” despite not expressly stated as a separate consideration in the constitution).
Interpreting the language “as may be” as allowing for consideration of other recognized factors is consistent with the United States Supreme Court’s requirement for congressional districts to have population equality “as nearly as is practicable” under its interpretation of the Equal Protection Clause in the United States Constitution. See Reynolds, 377 U.S. at 577, 84 S.Ct. 1362. The federal standard permits “minor variations which ‘are based on legitimate considerations incident to the ef-fectuation of a rational state policy.’ ” Swann v. Adams, 385 U.S. 440, 444, 87 S.Ct. 569, 17 L.Ed.2d 501 (1967). As with this Court, the United States Supreme Court recognizes that legitimate considerations include recognition of natural boundary lines, recognition of historical district boundary lines, and respect for boundaries of political subdivisions.13 See id.; Karcher, 462 U.S. at 740, 103 S.Ct. 2653.14 Similar to the United States Su*51preme Court’s interpretation of the “as nearly as is practicable” standard under the Equal Protection Clause, this Court interprets the requirements in the Missouri Constitution to implicitly permit the legislature to comply with federal laws and consider recognized factors yet still comply with the requirements of the Missouri Constitution. The requirement for compactness “as may be” allows for consideration of these recognized factors. See Pearson I, 359 S.W.3d at 39.
This Court’s precedent does not hold that constitutional requirements can be disregarded to consider other factors but instead recognizes that the constitutional requirements themselves incorporate such considerations by use of the standard “as may be.” Plaintiffs recognized that the constitutional requirements incorporate other factors by presenting evidence at trial and making arguments on appeal regarding the boundary lines for historical district maps, counties, and political subdivisions. As part of the standard for the constitutional requirements, federal law and the other recognized factors are in fact of constitutional significance, and this Court recognizes that in its precedent. Each of these factors plays a role in determining the fact-based inquiry of whether a district is “as compact ... as may be.” If a district seems not to be composed of closely united territory because of minimal and practical deviations, the district is still “as compact ... as may be” if those deviations are due to mandatory and permissive factors. “[M]aps could be drawn in multiple ways, all of which might meet the constitutional requirements.” Id. at 39.
Under this construction of the phrase “as compact ... as may be,” the trial court did not erroneously declare or apply the compactness requirement in Mo. Const, art. Ill, sec. 45. The trial court’s judgment, which determined that absolute precision is not required and that other factors may affect compactness, is consistent with the proper construction of the constitutional provision.
C. The trial court did not err in concluding that Plaintiffs failed to prove that the Map clearly and undoubtedly contravenes the constitution
Plaintiffs assert that the trial court’s judgment, which determined that Plaintiffs failed to meet their burden of proof that the Map is unconstitutional, is against the weight of the evidence. This Court will overturn a trial court’s judgment on the ground that it is against the weight of the evidence — with the term “weight” referring to the probative value (not quantity) of the evidence — only if it has a firm belief that the judgment is wrong. White, 321 S.W.3d at 308-09. The weight of the evidence standard from Murphy v. Carron was discussed in State Farm Mut. Auto Ins. Co. v. Allen: “The purpose of the [‘weight of the evidence’] rule should be read in context. The purpose of the rule is to give the findings of fact of the trial court the approximate effect of a jury verdict, especially when weighing and credibility are involved.” 744 S.W.2d 782, 786 (Mo. banc 1988) (footnote omitted). On finding that the record in State Farm Mut. Auto Ins. Co. showed a factual dispute regarding insurance policy coverage, this Court was “not willing to substitute [its] judgment for that of the trial court” by *52weighing the evidence and affirmed the trial court’s judgment Id. at 787.
This Court rarely has reversed a trial judgment as against the weight of the evidence under the Murphy v. Carron standard. In Buckner v. Jordan, for instance, this Court reversed a judgment in favor of a plaintiff as against the weight of the evidence. 952 S.W.2d 710, 712 (Mo. banc 1997). The case involved a determination . of a father’s child support obligation, and the circumstances of the case revealed that the trial court failed to consider an affidavit by the mother that showed her increased income. Id. Because of the complete failure to consider the affidavit by the trial court, this Court weighed the evidence and determined that the trial court erred in its judgment. See id.
This case does not present the rare circumstance when the trial court’s judgment should be reversed as against the weight of the evidence. Significantly, Plaintiffs’ weight of the evidence claim challenges a judgment in favor of Defendants. Defendants have no burden of proof in this case and were not required to present any evidence to prevail. See State Farm Mut. Auto. Ins. Co., 744 S.W.2d at 786 (refusing to weigh evidence in court-tried case in which the trial court’s judgment was for the defendant). Plaintiffs’ claim that the challenged districts were not “as compact ... as may be” required the trial court to weigh evidence and make factual determinations. See Pearson I, 359 S.W.3d at 40 (“[I]t is a question of fact, yet to be tried, whether those districts are ‘as compact and nearly equal in population as may be.’ ”). The parties had the ability to request written findings of fact to identify the precise issues they wanted the trial court to determine for each district, but neither party made a proper request.15 Because the parties did not request findings, this Court does not have the benefit of specific, articulated conclusions as to the determination that the districts are “as compact ... as may be.” Without written findings of fact, this Court views the facts in the light most favorable to the trial court’s judgment. Rule 73.01.
1. The Map, H.B. 193
The Map was admitted into evidence by stipulation of the parties. Plaintiffs challenged the constitutional validity of districts 3, 5, and 6.
*53[[Image here]]
2. Plaintiffs’ burden of proof
As discussed above, Plaintiffs, at all times, bore the burden of proving that the Map clearly and undoubtedly contravened the constitution. Because the standard for determining whether a district is drawn “as compact ... as may be” includes whether any minimal and practical deviations were a result of recognized factors that may affect the district boundaries, Plaintiffs must prove that the boundaries of districts 3, 5, and 6 depart from the principles of compactness and that any deviations were not minimal or practical deviations resulting from applying the recognized factors. Accordingly, Plaintiffs’ showing must account for any minimal and practical deviations occurring as a result of: (1) the interrelationship in standards for the population equality and compactness requirements; (2) the contiguity requirement; (3) federal laws, including the Voting Rights Act; and (4) the recognized factors of population density, natural boundary lines, boundaries of political subdivisions, and historical boundary lines of prior redistricting maps. While the existence of evidence regarding each of these factors satisfies Plaintiffs burden of production, Plaintiffs nonetheless may fail to satisfy their burden of persuasion with the trier of fact that, based on the evidence presented, the challenged districts clearly and undoubtedly contravene the constitution.
3. The trial court did not err in concluding that Plaintiffs failed to prove that district 3 clearly and undoubtedly contravenes the constitution
Plaintiffs claim that the trial court’s judgments finding that they failed to meet their burden of proving that district 3 is not “as compact ... as may be” is against the weight of the evidence. Evidence presented at trial could support the trial court’s determination that the minimal and practical deviations in district 3 are supported by Voting Rights Act con*54siderations. Evidence at trial established that the district boundaries for every district are interrelated and that some districts must be drawn less compactly because of the shape of neighboring districts. Evidence at trial supports a finding that this is the case with districts 1, 2, and 3.
The parties presented evidence that the boundaries for districts 1 and 2 were drawn in a circular manner to comply with the Voting Rights Act by boosting the minority population of district 1 and generally protecting against minority “vote dilution.” The purpose of the Voting Rights Act is to ensure that members of a protected class have the same opportunity as other citizens to participate in the political process and elect representatives of their choice. 42 U.S.C. § 1973(b) (2000). Defendants’ expert, Dr. Hofeller,16 explained that when a district is shaped like a circle, neighboring districts must necessarily be “crescent-shaped.” Because districts 1 and 2 are nearly circles, it can be expected that the surrounding district — district 3— would take on a crescent shape. This accurately describes the areas of district 3 that extend around districts 1 and 2. Given the reasoning for the shapes of districts 1 and 2 and the knowledge that the shape of district 3 must adjust to its adjoining districts, the trial court could conclude that Plaintiffs failed to prove that district 3 was not “as compact ... as may be.”
4. The trial court did not err in concluding that Plaintiffs failed to prove that district 5 clearly and undoubtedly contravenes the constitution
Plaintiffs claim that the trial court’s judgments finding that they failed to meet their burden of proving that district 5 is not “as compact ... as may be” is against the weight of the evidence. In assessing the evidence regarding whether district 5 is sufficiently compact, the trial court considered the stipulations and evidence presented by the parties during the three-day trial. The parties stipulated to the Map and the statistical evidence regarding compactness, population equality, and racial composition of the relevant districts. While these stipulations of fact relieved the parties from proving the matters stipulated, In re Marriage of Hendrix, 183 S.W.3d 582, 591 (Mo. banc 2006), the stipulations do not prove, as a matter of law, that district 5 does not meet the constitutional standard for compactness. Some of the stipulated evidence is favorable to a finding that district 5 is “as compact ... as may be.” Additionally, through evidence presented at trial and the cross-examination of Plaintiffs’ expert, Dr. David C. Kimball, Defendants zealously disputed multiple factual issues relevant to determining whether district 5 was “as compact ... as may be.”
The trial court had evidence before it, the Map, showing that district 5 is not in the shape of a circle or square, which the experts testified are the most compact shapes. If the constitutional standard for compactness were merely that a district be circular or square, the ■visual inspection of district 5 would demonstrate that it is neither.17 The Missouri *55constitutional standard of “as compact ... as may be” requires, instead, that each district be “closely united territory” as may be, rejecting the proposition that “compact” refers solely to physical shape or size. See Barrett, 146 S.W. at 61. The boundaries for district 5 are not so egregiously drawn that it could be found to violate the compactness standard as a matter of law. There exists a question of fact whether it is closely united territory with practical and minimal deviations from compactness resulting from application of recognized factors.
The parties stipulated to the results of eight statistical tests measuring compactness.18 While some of these statistics measure the level of compactness of the Map as a whole, there were also measures of the individual districts. Plaintiffs’ expert, Dr. Kimball, testified that the alternative maps introduced by Plaintiffs scored better on the majority of the compactness measures and, as a result of the statistical measures and his visual examination, he opined that district 5 was not “as compact ... as may be.”19 When cross-examined, however, he testified that the standard for compactness is not a precise standard and that there is no bright line between a compact and non-compact district. He acknowledged that the statistical tests do not produce a percentage threshold for determining whether a district can be more compact. He further testified that it was possible to make Plaintiffs’ alternative maps more compact so, under his comparison standard of compactness, Plaintiffs’ proposed maps of district 5 would not be “as compact ... as may be.” He also testified that he never used any mathematical tests to compare the compactness of congressional districts before this case. Importantly, he testified that he does not know if it is possible to draw the most compact district if controlling for population equality, voting rights, and other factors. He did not consider such factors or anything else, such as the dispersion of population in a district, when forming his opinion of the compactness of district 5.
Defendants presented the testimony of Dr. Hofeller regarding the compactness of district 5. Dr. Hofeller testified as to the principles relevant to determining compactness, the meaning of the phrase “as compact ... as may be” and opined as to whether district 5 meets that standard.20 Dr. Hofeller also testified as to the meaning and application of the statistical measures of compactness to which the parties stipulated. He testified that, while statistical measures are helpful, there is no statistical measure or specific score that conclusively indicates that a map is compact. Instead, he testified that compactness is measured on a continuum. According to Dr. Hofeller, there may be instances in which a district boundary does or does not satisfy the compactness principles, but *56there is no bright line in determining it. Under this continuum, Dr. Hotelier testified that district 5 falls in the area of being compact. This testimony disputes Plaintiffs claim that the districts are not “as compact ... as may be” and was weighed by the trial court in reaching its decision.
Both Drs. Kimball and Hofeller testified that the same eight statistical measures are used to evaluate compactness. These measures were discussed at length in the redistricting articles admitted into evidence at trial. The parties stipulated as to how district 5 scored on each of these statistical measures. District 5 scored well on the measures of compactness that consider area in combination with population. Plaintiffs presented no other evidence regarding population density.
There was also evidence produced at trial disputing whether the boundaries for district 5 were affected by historical district boundaries. This evidence consisted of the prior redistricting maps shown in Appendix A. These maps demonstrated that a portion of Jackson County historically has been carved out of district 5 and appended to other districts (both districts 4 and 6). The current Map only slightly expands that carved-out portion. Plaintiffs’ expert, Dave Rolland, a constitutional attorney, was questioned about these historical maps at length during both direct examination by Plaintiffs and cross-examination by Defendants.
Finally, evidence was presented regarding the effect of political subdivision boundaries on the boundaries for district 5, including both urban and rural areas of Jackson County. The evidence could support a finding by the trial court that, while district 5 divides certain political subdivisions, it maintained the boundary lines of other subdivisions, specifically municipalities north of the Missouri River. For example, the boundary for district 5 generally follows the boundary for Gladstone, which is surrounded by Kansas City. The evidence also showed that district 5 includes a portion of Clay County, including urban portions of Kansas City that continue from Jackson County into Clay County. The trial court could have found that the boundaries of the Map were drawn in consideration of the legitimate factor of keeping a greater portion of Kansas City in district 5.
This record reveals factual disputes regarding whether the deviations in the boundary of district 5 were minimal and practical deviations that could have been drawn to take into account certain recognized factors. See State Farm Mut. Auto. Ins. Co., 744 S.W.2d at 787. The trial court was free to consider the weight and credibility of the evidence on the record. From that assessment, the trial court could have objectively determined that Plaintiffs failed to prove clearly and undoubtedly that district 5 is not as “compact ... as may be.” This Court will not substitute its judgment for that of the trial court by re-evaluating the credibility of that evidence in this case. Id.
5. The trial court’s judgment regarding district 6 is not against the weight of the evidence
Because the trial court did not err in its judgments regarding district 5, and because the boundary in district 5 has a direct correlation to the boundary in district 6, the same analysis applies. Therefore, this Court finds no error in the trial court’s judgment regarding district 6.
Conclusion
Accordingly, this Court finds no error in the trial court’s judgments. The judgments of the trial court are affirmed.
BRECKENRIDGE and FISCHER, JJ., and MITCHELL and LYNCH, Sp.JJ., concur.*57FISCHER, J., concurs in separate opinion filed; LYNCH, Sp.J., concurs in opinion of FISCHER, J.
PRICE, J., dissents in separate opinion filed; STITH, J., and ELLIS, Sp.J„ concur in opinion of PRICE, J. TEITELMAN, C.J., RUSSELL and DRAPER, JJ., not participating.[[Image here]]
[[Image here]]
The Missouri Congressional Map of 1921 placed the northwestern and central-western wards of Kansas City into a district with Platte, Clinton, Clay, and Ray counties. The southwestern wards of Kansas City and eastern and southern Jackson County comprised their own discrete district.
*58[[Image here]]
The Congressional map of 1933 created a separate district for metropolitan Kansas City, and placed the remainder of Jackson County in its own Congressional District.
*59[[Image here]]
The map of 1943 is the same as the 1933 map.
*60[[Image here]]
The map of 1952 continued to give the Kansas City portion of Jackson County its own district (the fifth) but united the remainder of Jackson County with Lafayette, Cass, Johnson, Henry, Davies, Vernon, and Barton counties (the fourth district).
*61[[Image here]]
The map of 1962 expanded the fifth district, and, added Henry and Pettis Counties to the fourth, while taking Barton County away from that district.
*62[[Image here]]
The map of 1972 expanded the fifth district southward, so that approximately the western third of Jackson County’s geographic area was now one district. The map added Barton back into the fourth district, and it also added St. Clair, Hickory, Benton, Morgan, Cooper, Howard, and Saline counties to that district.
*63[[Image here]]
The map of 1982 split Jackson County in half, geographically. The eastern half of Jackson was united with numerous other counties to the south and east.
*64[[Image here]]
The map of 1992 created the downward protrusion. In this map, Jackson County was split between three different Congressional Districts. The bottom south-east corner continued to be untied with the fourth district. The downward protrusion, which is shaped somewhat like the state of Massachusetts, except with the thm hook extending southward instead of eastward, was united with counties in the northwest and the north-central part of Missouri (the sixth district). The remainder of Jackson County comprised district five.
*65[[Image here]]
The map of 2002 extended district five to the southeast corner of Jackson County. Under this map, the downward protrusion continued to be united with the sixth dis-' trict (northwest and north-central Missouri). Only a tiny peninsula of Jackson County remained united with the district four (west central Missouri). In addition, the 2002 map extended district five to include an irregularly shaped portion of Cass County (south of metropolitan Kansas City).
*66[[Image here]]
[[Image here]]
[[Image here]]
The Map in 2012 extends the downward protrusion southward, so that district six contains more of Jackson County. The fifth district picks up the small portion of Jackson County that had continued to be in the fourth district, and it also extends to the east and slightly to the north to include Ray, Lafayette, and Saline counties. A portion of Cass County goes from the fifth district back to the fourth.
. The Pearson Plaintiffs are Kenneth Pearson, Phoebe Ottomeyer, Brian Murphy, Mildred Conner, Timothy Brown, and Joan Bray. The McClatchey Plaintiffs are Stan McClatchey, Donna Turk, Ivan Griffin, Patricia Smith, Molly M. Teichmann, Laura Meeks, and Matt Ullman.
. Intervenors are the respective chairs of the redistricting committees for the House and Senate.
. As discussed infra, parties who desire to know how the trial court sorted out these *45mixed questions of fact and law have the ability to request written findings of fact. Hammons v. Ehney, 924 S.W.2d 843, 849 (Mo. banc 1996). A party who fails to make such a request forfeits that advantage on ap-pelíate review, because the trial court’s findings are considered as having been found in accordance with the judgment. See Rule 73.01(c).
. See, e.g., In re 1983 Legislative Apportionment of House, Senate, and Cong. Dists., 469 A.2d 819, 828 (Me.1983); Ariz. Minority Coal. for Fair Redistricting v. Ariz. Indep. Redistricting Comm’n, 220 Ariz. 587, 208 P.3d 676, 684 n. 7 (2009); Holt v. 2011 Legislative Reapportionment Comm’n,-Pa.-, 38 A.3d 711, 737 (2012); Parella v. Montalbano, 899 A.2d 1226, 1233 (R.I.2006).
. See, e.g., Schneider v. Rockefeller, 31 N.Y.2d 420, 340 N.Y.S.2d 889, 293 N.E.2d 67 (1972); Stephenson v. Bartlett, 357 N.C. 301, 582 S.E.2d 247, 247-48 (2003); In re Constitutionality of House Joint Resolution 1987, 817 So.2d 819, 825 (Fla.2002); Davenport v. Apportionment Comm’n, 65 N.J. 125, 319 A.2d 718, 723 (1974); State ex rel. Gosch v. Lemler, 77 S.D. 23, 84 N.W.2d 418, 419 (1957); Kilbury v. Franklin Cnty. ex rel. Bd. of Cnty. Comm'rs, 151 Wash.2d 552, 90 P.3d 1071, 1076 (2004); In re Reapportionment of the Colo. Gen. Assembly, 828 P.2d 185, 197 (Colo. banc 1992) (“[O]bjectors have failed to come forward with a concrete alternative plan for house districts.... Although the question is close, we conclude that the Final Plan for House District 62 and for the City of Westminster does not violate [the constitution].”); Beaubien v. Ryan, 198 Ill.2d 294, 298, 260 Ill.Dec. 842, 762 N.E.2d 501 (2001). The Supreme Court of Connecticut in Fonfara v. Reapportionment Comm’n, discussed shifting the burden of proof, but expressly did not decide the issue. 222 Conn. 166, 610 A.2d 153, 164-65 (1992).
. See In re Legislative Districting of Gen. Assembly, 193 N.W.2d 784, 791, supplemented by 196 N.W.2d 209 (Iowa 1972), amended sub nom. Matter of Legislative Districting of Gen. Assembly, 199 N.W.2d 614 (Iowa 1972); Jackman v. Bodine, 55 N.J. 371, 262 A.2d 389, 395 (1970).
. See In re Legislative Districting of State, 370 Md. 312, 805 A.2d 292, 325 (2002). Of course, states generally apply the federal burden-shifting framework for redistricting challenges under the federal constitution. See, e.g., Opinion of the Justices, 353 Mass. 790, 230 N.E.2d 801, 804 (1967); State ex rel. Lockert v. Crowell, 656 S.W.2d 836 (Tenn.1983); Egan v. Hammond, 502 P.2d 856, 867 (Alaska 1972).
. In re Reapportionment of Towns of Hartland, Windsor, and W. Windsor, 160 Vt. 9, 624 A.2d 323, 327 (1993).
. Pearson I recognized this distinction when this Court affirmed the dismissal on the pleadings of all of the Plaintiffs’ claims except one, stating there "is a question of fact, yet to be tried, whether those districts are as compact and nearly equal in population as may be. Mo. Const, art. Ill, sec. 45 (emphasis added),” and "[tjhis Court makes no prejudgment on these issues, or on the compactness of other districts, other than to hold that Plaintiffs have stated a claim as to the compactness of the districts that is subject to proof and defenses in accordance with evidence as in any other lawsuit.” Pearson I, 359 S.W.3d at 40.
. That does not mean that these measures are completely irrelevant but rather that they alone do not demonstrate that a map is or is not compact.
. Although this Court's prior opinions, including Pearson I, recognize that county lines may validly be considered in reapportioning congressional and house districts, the Missouri Constitution makes it a mandatory factor in the reapportionment of senate districts. See Mo. Const, art. Ill, secs. 5 and 7.
. Consideration of historical district boundaries allows residents of a district to continue any relationships such residents may have established with their elected representatives and to avoid the detriment to residents of having to reestablish relationships when district boundaries change.
. In Kirkpatrick v. Preisler, the Supreme Court recognized an exception to the consideration of county or municipal boundaries in a challenge to the population equality of congressional districts under the United States Constitution, stating that it is improper to justify deviations based on political subdivision boundaries. 394 U.S. at 533-34, 89 S.Ct. 1225. Since Kirkpatrick v. Preisler, however, the Supreme Court expressly identifies respect for municipal and county boundaries as legitimate considerations in congressional redistricting. See Abrams v. Johnson, 521 U.S. 74, 98, 117 S.Ct. 1925, 138 L.Ed.2d 285 (1997) (citing Karcher, 462 U.S. at 740, 103 S.Ct. 2653); see also Bush v. Vera, 517 U.S. 952, 963, 116 S.Ct. 1941, 135 L.Ed.2d 248 (1996).
.The Supreme Court also has stated that certain factors cannot be considered when seeking to attain population equality under the United States Constitution. In Reynolds v. Sims, the Supreme Court detailed that "history alone [referring to history as the histori*51cal number of districts as a basis for representation, not the location of historical boundary lines as later identified in Swann], nor economic or other sorts of group interests, are permissible factors in attempting to justify variances from population-based representation. ... Considerations of area alone provide an insufficient justification for deviations from the equal-population principle.” 377 U.S. at 580, 84 S.Ct. 1362.
. In any case tried without a jury, "[t]he court may, or if requested by a party shall, include in the opinion findings on the controverted fact issues specified by the party.” Rule 73.01. "[I]t is the parties' duty to specifically request findings of fact and conclusions of law” and to identify “the issues they wish the court to decide.” Hammons, 924 S.W.2d at 849 (emphasis added). "Merely submitting proposed findings to aid the court does not trigger the court’s duty to make findings of fact and law.” Id.
. Much of Dr. Hofeller’s testimony addressed compactness of the Map as a whole. But in Pearson I, this Court particularly noted that the protection of art. Ill, sec. 45 "applies to each Missouri voter, in every congressional district.” Id. at 39. As such, each congressional district must be "as compact ... as may be.” Dr. Hofeller’s testimony regarding compactness of the Map as a whole, therefore, has limited relevance.
. The experts also testified that the impossibility of drawing a redistricting map with all districts being the shape of a circle makes it clear that perfect compactness is not intended by constitutional standards for compactness.
. Although these statistical measures do not establish a threshold for determining when a district is or is not compact, both Plaintiffs’ and Defendants’ experts testified that such measures are relevant to the evaluation of compactness. Therefore, they could have been a factor relied on by the trial court.
. To the extent that Dr. Kimball testified regarding an issue that is solely a question of law, this Court will give no more deference to his opinion than to the trial court's ruling on an issue of law. See Howard v. City of Kansas City, 332 S.W.3d 772, 785 (Mo. banc 2011). This Court could not simply defer to the trial court, an expert, or anyone else as to legal questions without abandoning its constitutional duty to be the final arbiter as to what the law is in Missouri. See id.
.Any testimony of Dr. Hofeller regarding an issue that is solely a question of law is given no deference on the same basis as that set forth in footnote 19.