OPINION

BLATZ, Chief Justice.

James Jasper’s driver’s license was revoked by the Commissioner of Public Safety because Jasper drove with an alcohol concentration of 0.10 or greater. Jasper petitioned the district court for review of the commissioner’s order of revocation arguing that evidence of his breath-test results was inadmissible because the instrument used to obtain the evidence, the Intoxilyzer 5000, Series 68-01, was not properly approved by the commissioner. The district court sustained the commissioner’s order and the court of appeals affirmed. We affirm.

Jasper was stopped and arrested for driving under the influence of alcohol in Inver Grove Heights, Minnesota, in March 2000. During the course of the stop Jasper consented to a breath test and Officer Joe Gelhaye of the Inver Grove Heights Police Department administered a test using the Intoxilyzer 5000, Series 68-01 (Series 68-01). Jasper’s breath-test results revealed an alcohol concentration of 0.24 and the commissioner revoked Jasper’s driver’s license for a period of 180 days under Minn.Stat. § 169.123, subd. 4(e)(3) (1998). 1

Jasper petitioned the district court for review of the commissioner’s order of revocation as provided for by Minn.Stat. § 169.123, subd. 5c (1998). Jasper argued that the Series 68-01 was not properly approved by the commissioner as required by law and that, consequently, evidence of his breath-test results was inadmissible. Specifically, Jasper argued that approval of the Series 68-01 can be accomplished only through the promulgation of a new rule by the commissioner in accordance with the rulemaking procedures set forth in the Minnesota Administrative Procedure Act (MAPA), Minn.Stat. §§ 14.001-.69 (1998).

The significance of Jasper’s argument is best understood when placed in its statuto *437 ry and historical context. Between 1983 and 1997, law enforcement officials throughout the state used the Intoxilyzer 5000, Series 64 (Series 64), and the Intoxi-lyzer 5000, Series 66 (Series 66), to measure the alcohol concentration of breath samples. In 1984, a statute was enacted providing that the results of a breath test obtained through the use of an “approved” infrared breath-testing instrument are admissible in evidence without antecedent expert testimony establishing that the instrument provides a trustworthy and reliable measure of alcohol concentration. Act of April 23, 1984, eh. 430, § 9, 1984 Minn. Laws 88, 92 (codified as amended at Minn. Stat. § 634.16 (1998)). 2 A party seeking to admit such evidence need only show that the breath test was performed by a person fully trained in the use of the instrument pursuant to training given or approved by the commissioner. Minn.Stat. § 634.16.

In April 1984, the same month during which section 634.16 was enacted, the commissioner promulgated Minn. R. 7502.0420, subp. 2 (2001) (1984 rule). The 1984 rule provides: “The Intoxilyzer 5000 instrument, which uses infrared technology, is approved for use in this state for the purpose of determining the alcohol concentration of a breath sample.”

At some point during the mid-1990s, the Bureau of Criminal Apprehension (BCA) considered replacing the Series 64 and 66 with a new breath-testing instrument. After testing and evaluating two instruments, the BCA decided to replace the Series 64 and 66 with the Series 68-01 and the commissioner subsequently issued three orders regarding the newly selected Series 68-01. Order 101, issued in May 1997, provides: “Pursuant to the provisions of Minnesota Statutes Sections 169.121, 169.123 and 169.128, and Minnesota Rules part 7502.0420, subpart 2, the Intoxilyzer 5000, Series 68, is approved for use in this state for the purpose of determining the alcohol concentration of a breath sample.” In May 1998, the commissioner issued Order 101a, clarifying that Order 101 was meant to apply to the Series 68-01. This was followed by Order 101b in September 1999, stating that Orders 101 and 101a were meant to include all software used by the Series 68-01.

While reserving his argument that the Series 68-01 was not a properly approved infrared breath-testing instrument, Jasper stipulated at the implied consent hearing that Officer Gelhaye was a certified intoxi-lyzer operator trained by the BCA and that the officer administered Jasper’s breath test pursuant to his training. Jasper also stipulated that the March 2000, breath-test results showing that he had an alcohol concentration of 0.24 were accurate and reliable.

At the implied consent hearing, the commissioner called expert witness Gwen Williams, a forensic scientist with the breath-testing section of the BCA laboratory, to testify. Williams had attended *438 courses on the maintenance and repair of the Intoxilyzer 5000 and had testified numerous times as an expert on the subject of breath testing. She testified that the infrared method of analysis used by the Intoxilyzer 5000 was widely accepted within the field of forensic science and that the scientific literature supported the accuracy and reliability of the infrared method. Based on her review of the BCA documents concerning Jasper’s breath test and the stipulations made by Jasper at the hearing, Williams opined that the results showing an alcohol concentration of 0.24 were accurate and reliable.

In addition, Williams testified that, in spite of several differences, 3 the Series 68-01 and the Series 64 and 66 used the same method of infrared analysis and all of the Intoxilyzer 5000 models measured alcohol concentration in “exactly the same” way. She stated further that, although the instruments were the same in terms of their analytical functioning, the changes and modifications she discussed made the Series 68-01 more accurate and reliable than the Series 64 and 66.

The district court sustained the commissioner’s order of revocation and found on two distinct grounds that the commissioner properly approved the Series 68-01. First, the district court concluded in its order that the commissioner properly approved the Series 68-01 by issuing Orders 101, 101a, and 101b. Second, in a memorandum of law incorporated by reference into its order, the district court noted that the “method of infrared analysis employed in each Intoxilyzer machine is identical” and the Series 68-01 was therefore approved by the commissioner’s 1984 rule. Thus, because the method of analysis had not changed, the district court stated that the promulgation of a new rule was unnecessary.

Jasper appealed, and the court of appeals affirmed the district court’s order. The court of appeals held that a new rule approving the Series 68-01 was not required because the record did not reveal any material alterations to the original In-toxilyzer 5000 approved by the 1984 rule. Jasper v. Comm’r of Pub. Safety, No. C1-00-2098, 2001 WL 711337, at *1 (Minn. App. June 26, 2001). The court of appeals stated that, absent such a record, it would defer to the commissioner’s interpretation that the 1984 rule extends to the Series 68-01, including software upgrades. Jasper, 2001 WL 711337, at *3. The court of appeals also concluded that the commissioner established the reliability of Jasper’s breath-test results through Williams’s expert testimony and that Jas *439 per failed to show why the test results were unreliable. Id. at *3. Jasper petitioned this court for review of the court of appeals’ decision and we granted the petition to consider the issue of whether the commissioner properly approved the Series 68-01.

We note at the outset that there is nothing in the record to suggest that the Series 68-01, in general, or the results of Jasper’s breath test, in particular, are untrustworthy or unreliable. Indeed, Jasper stipulated as to the accuracy and reliability of his breath-test results at the implied consent hearing. Furthermore, the commissioner established the trustworthiness and reliability of the results through antecedent expert testimony. Accordingly, even if we were to assume that the Series 68-01 was not properly approved and that section 634.16 does not apply in this case, the record would support the district court’s decision to admit the evidence. 4

Given these factual circumstances, we must decide the preliminary question of whether the approval issue is moot. We do not issue advisory opinions, nor do we decide cases merely to establish precedent. State v. Arens, 686 N.W.2d 131, 132 (Minn.1998); In re Schmidt, 443 N.W.2d 824, 826 (Minn.1989). The mootness doctrine, however, “is a flexible discretionary doctrine, not a mechanical rule that is invoked automatically.” State v. Rud, 359 N.W.2d 673, 576 (Minn.1984). There is a well-established exception to the mootness doctrine, for example, for issues that are capable of repetition yet evading review. See, e.g., State ex rel. Archabal v. County of Hennepin, 505 N.W.2d 294, 297 (Minn.1993); State ex rel. Doe v. Madonna, 295 N.W.2d 356, 361 (Minn.1980). In addition, this court has exercised its discretion to decide issues that are technically moot when the issue is “functionally justi-ciable” and one of public importance and statewide significance that should be decided immediately. State v. Brooks, 604 N.W.2d 345, 347-48 (Minn.2000); Rud, 359 N.W.2d at 576. “A case is functionally justiciable if the record contains the raw material (including effective presentation of both sides of the issues raised) traditionally associated with effective judicial decision making.” Rud, 359 N.W.2d at 576; see also Brooks, 604 N.W.2d at 347-48.

The record in the present case contains detailed information regarding the Series 68-01 and the ways in which it differs from older Intoxilyzer 5000 models. Further, the parties ably briefed and argued the issue of whether the Series 68-01 was properly approved. Thus, we conclude that the approval issue is functionally justiciable. In addition, the Series 68-01 is the only breath-testing instrument currently in use in this state and there has been substantial litigation in the district courts as to whether the instrument was properly approved. 5 Therefore, we conclude that the issue is one of public impor *440 tance and statewide significance that should be decided immediately.

Having decided that the issue is properly before the court, we now address the question of whether the Series 68-01 was properly approved by the commissioner. Our analysis begins with the commissioner’s 1984 rule approving the “Intoxilyzer 5000, which uses infrared technology.” Jasper argues that the 1984 rule does not cover the Series 68-01 because the Series 68-01 differs significantly from the Series 64 and 66. In response, the commissioner argues that any differences between the Series 68-01 and older Intoxilyzer 5000 models are minor and that the 1984 rule is broad enough to include the Series 68-01. The district court found that the method of analysis employed by the Series 68-01 was identical to the method used by the Series 64 and 66 and concluded that the 1984 rule therefore covered the Series 68-01.

When, as here, the language of an administrative rule is clear and capable of understanding, interpretation of the rule presents a question of law reviewed de novo. See St. Otto’s Home v. Dep’t of Human Servs., 437 N.W.2d 35, 39-40 (Minn.1989). As for the district court’s findings of fact, they will not be set aside unless dearly erroneous. Minn. R. Civ. P. 52.01; Astleford Equip. Co., Inc. v. Navistar Int’l Transp. Corp., 632 N.W.2d 182, 193 (Minn.2001). We hold findings of fact as clearly erroneous only when we are “left with a definite and firm conviction that a mistake has been committed.” In re Improvement of Murray County Ditch No. 34, 615 N.W.2d 40, 49 (Minn.2000); N. States Power v. Lyon Food Prods., Inc., 304 Minn. 196, 201, 229 N.W.2d 521, 524 (1975). When findings of fact rest almost entirely on expert testimony, the district court’s evaluation of credibility is of particular significance. In re Knops, 536 N.W.2d 616, 620 (Minn.1995); In re Joelson, 385 N.W.2d 810, 811 (Minn.1986).

Having reviewed the record below, we conclude that the district court’s finding that the Series 68-01 is analytically identical to the Series 64 and 66 is not clearly erroneous. 6 Expert witness Williams repeatedly stated during her testimony that the Series 68-01 and the Series 64 and 66 used the same method of analysis. Jasper did not rebut Williams’s testimony and the district court concluded that she was a credible witness. Based on this record, we are not “left with a definite and firm conviction that a mistake has been committed.” See Improvement of Murray County Ditch No. 34, 615 N.W.2d at 49.

Nonetheless, Jasper argues that our inquiry cannot end here. According to Jasper, the fact that the Series 68-01 and the *441 Series 64 and 66 employ the identical method of analysis does not necessarily lead to the conclusion that the Series 68-01 is covered by the 1984 rule. On the facts of this case, Jasper’s argument is not persuasive. The primary, if not sole, purpose of breath-testing instruments is to measure the alcohol concentration of breath samples. In light of this purpose, the method of analysis used by the instrument to make such measurements is its most essential characteristic. Again, the district court found that the Series 68-01 and the Series 64 and 66 are identical in this regard. Based on the record below, we conclude that the Series 68-01 falls within the scope of the 1984 rule and is therefore a properly approved “infrared breath-testing instrument” as defined in section 169.01, subd. 68. 7

Affirmed.

. Minnesota’s Implied Consent Law was amended and recodified in 2000 and is now found at Minn.Stat. §§ 169A.50-.53 (2000). The 2000 amendments took effect on July 1, 2001. Act of May 15, 2000, ch. 478, art. 2, § 9, 2000 Minn. Laws 1484, 1538. Because this case arises from events that took place before July 1, 2001, all citations are to the pre-amendment version of the Implied Consent Law.

. Section 634.16 states:

In any civil or criminal hearing or trial, the results of an infrared breath-test, when performed by a person who has been fully trained in the use of an infrared breath-testing instrument, as defined in section 169.01, subdivision 68, pursuant to training given or approved by the commissioner of public safety or the commissioner’s acting agent, are admissible in evidence without antecedent expert testimony that an infrared breath-testing instrument provides a trustworthy and reliable measure of the alcohol in the breath.

Under Minn.Stat. § 169.01, subd. 68 (1998), the term " ‘[i]nfrared breath-testing instrument' means a breath-testing instrument that employs infrared technology and has been approved by the commissioner of public safety for determining alcohol concentration.” (Emphasis added.)

. Williams testified as to five differences between the Series 68-01 and the Series 64 and 66. First, two additional filters were added to the Series 68-01 to enhance its ability to discriminate more accurately between ethyl alcohol — the type of alcohol found in alcoholic beverages — and other substances. Second, software upgrades were made that affected the functioning of the printer attached to the Series 68-01 and the appearance of the printouts. Third, the replicate tracking system was changed to prevent the Series 68-01 from accepting a test result if the measurements rendered by the subject and replicate tests varied by 0.006 or more. Williams stated that all models of the Intoxilyzer 5000 performed two measurements — a subject test and a replicate test — on each breath sample. When the results of a subject and replicate test vary by 0.006 or more, the Series 68-01 will reject the results and display a "please blow" message. Earlier models, in contrast, accepted test results with a deviation of 0.006. Fourth, Williams indicated that the Series 68-01 contained different "plumbing,” including a smaller breath sample chamber and a modified exhaust and recirculation system. Finally, the amount of air needed to provide an adequate breath sample for the Series 68-01 was set at 1.1 liters, whereas the Series 64 and 66 were incapable of measuring volume directly and instead used an estimate of volume based on inches of water pressure over time.

. We also note that after we granted Jasper’s petition for review the commissioner promulgated a new rule, Minn. R. 7502.0420, subp. 3 (2001). While not in effect at the time the Series 68-01 was used in the present case, Minn. R. 7502.0420, subp. 3, provides:

The Intoxilyzer 5000 instrument, Series 68, identified by the serial number SN68-01 and followed by four digits, which uses infrared technology, is approved for use in this state for the purpose of determining the alcohol concentration of a breath sample. This approved instrument includes all software updates and changes through Software Version G1408.43 and Slave 75— 0037.

. At Jasper's implied consent hearing, the district court made reference to “numerous cases” in which petitioners argued that the Series 68-01 was not properly approved and noted that district courts have reached opposing conclusions on this issue. Similarly, at *440 oral argument, counsel for the commissioner stated that the reliability of breath-test results has been challenged in approximately 80 cases on the basis that the Series 68-01 was not properly approved.

. While not of precedential value, we find support for our conclusion in the Wisconsin Supreme Court's decision in State v. Busch, 111 Wis.2d 429, 576 N.W.2d 904 (1998). In that case, Busch argued that evidence of his breath-test results was inadmissible on the ground that the instrument used to obtain it was not properly evaluated and approved by the Wisconsin Department of Transportation (WDOT). Id. at 905. The head of the chemical-test section of the WDOT had previously evaluated and approved an earlier model of the instrument, but had not subjected the new model to the same procedure. Id. at 905. At a hearing on Busch’s motion, an expert testified that, while there were some differences between the new and old models, those changes did not affect the analytical functioning of the instrument. Id. at 907. Based on this testimony, the circuit court found that the new and old models were essentially the same machine. Id. at 907-08. On appeal, the Wisconsin Supreme Court held that the circuit court's finding was not an erroneous exercise of the court's discretion. Id. at 911.

. Our holding that the 1984 rule covers the Series 68-01 makes it unnecessary to consider the commissioner’s alternative argument that the Series 68-01 was properly approved by Orders 101, 101a, and 101b. We note, however, that the orders appear to be statements of general applicability and future effect that would otherwise require adherence to the rulemaking procedures outlined in the MAPA. See Minn.Stat. § 14.02, subd. 4 (2000) (defining a "rule” as "every agency statement of general applicability and future effect, including amendments, suspensions, and repeals of rules, adopted to implement or make specific the law enforced or administered by that agency or to govern its organization or procedure”).