TWA appeals the decision of the Labor and Industrial Relations Commission finding that James Thatcher sustained a 34 1/3 percent disability for binaural occupational hearing loss at the 180-week level.
Facts
James Thatcher filed a worker’s compensation claim against TWA for disability associated with occupational hearing loss. Thatcher was a welder for TWA for twenty-six years. Thatcher claimed that he was exposed to high levels of noise associated with rivet guns, high-speed grinders and air drills. On March 15,1995, Thatcher was transferred to a different area at the TWA base, resulting in less noise exposure. This transfer came after he told his employer that he was developing hearing problems.
TWA agreed that Thatcher was covered by Missouri Workers’ Compensation Act and that he had sustained an occupational disease in the form of hearing loss arising out of and in the course of his employment. The dispute between the parties concerns the extent of Thatcher’s disability and its calculation.
The Hon. R. Carl Mueller, Jr., Administrative Law Judge, entered Findings of Fact and Conclusions of Law and awarded Thatcher a five percent body as a whole disability for tinnitus,1 reimbursement of medical expenses and future medical treatment in the form of digital hearing aids and necessary replacement of the aids. TWA did not request review of these findings and awards. Further, Mueller found that Thatcher sustained a disability totaling 34 1/3 percent for bin-aural occupational hearing loss at the 180-week level. TWA appealed this portion of the award to the Labor and Industrial Relations Commission, which affirmed and adopted the previous Findings of Fact and Conclusions of Law entered. TWA raises two points on appeal.
In its first point it argues that the Commission erroneously applied the law and more specifically misinterpreted the phrase “lowest measured losses” in § 287.197.4, RSMo 2000. As a result of this misinterpretation, TWA claims, the Commission utilized incorrect hearing loss measurements in calculating the percentage of workers’ compensation disability under the statutory formula. If TWA is correct, then the Commission calculated a percentage of disability for hearing loss that did not follow the formula set forth by the statute.
In its second point TWA claims that the Commission erred in applying retroactively the 1998 amendments to § 287.190 which increased the number of weeks of compensation for total hearing loss from 168 weeks to 180 weeks. TWA asserts the statutory amendments were substantive and therefore not subject to retroactive application.
Testing Hearing Levels
The testing of hearing loss is extremely complicated. An understanding of basic *536terms is necessary to understand the issues in this case.2 The term decibel is used to express the intensity of sound. The hearing threshold level is the softest level that a person can hear a pure tone generated by an audiometer at a given frequency. Measurement is by decibel hearing threshold level (dbHTL). Frequency is the number of sound waves or vibrations in a specified time period. The unit of measurement is cycles per second (Hz). Thus, if a given individual first detected an audiometer tone at 15 decibels for 500 Hz, his hearing threshold level for that frequency would be 15 dB. If the same individual first detected the tone at 25 decibels for 1500 Hz his hearing threshold level for that frequency would be 25 dB. The decibel scale is a logarithmic scale in which 0 dB approximates the threshold of hearing in the middle frequencies for young adults.3
A threshold shift is a change in hearing sensitivity. Such changes may be temporary, temporary threshold shift (TTS) or permanent, permanent threshold shift (PTS). A loud noise may cause a temporary worsening in hearing sensitivity (TTS) that may persist for 14 hours or longer. Determination of a permanent loss of hearing sensitivity (PTS) thus requires testing on separate dates. Not all temporary or permanent threshold shifts result from noise exposure. When a PTS can be attributed to noise exposure it is referred to as a noise induced permanent threshold shift (NIPTS). A hearing loss is the amount of impairment, in decibels, measured as a set of hearing threshold levels at specified frequencies.
An audiometer is an instrument for measuring individual hearing acuity or ability. It produces a chart, graph or table (audiogram) showing an individual’s hearing threshold levels (HTL), the ability to hear, as a function of frequency (Hz). An audiometer is calibrated according to certain uniform standards. ANSI (American National Standards Institute), formerly known as ASA, is a voluntary membership organization that develops consensus standards nationally for a wide variety of devices and procedures. ISO is the international standards organization that performs a similar function.
Hearing Loss and Workers’ Compensation
Litigation seeking workers compensation coverage for gradual hearing loss as an occupational disease increased in the late 1940’s. Some states initially refused to recognize the compensability of hearing loss due to protracted noise exposure. States that did recognize the theory of compensability struggled with how to treat hearing loss as a scheduled loss under workers’ compensation statutes as com-pensable when no actual loss of earnings could be shown.4 Missouri first fell into the camp denying compensation under the then existing laws. Marie v. Standard Steel Works, 311 S.W.2d 368, 371 (Mo.App.1958). While the case was pending before *537the Supreme Court, an Advisory Committee to the General Assembly considered legislation. The Supreme Court eventually reversed Marie, 319 S.W.2d 871 (Mo. banc 1959), holding that hearing loss was compensable as an occupational disease generally under Missouri’s existing workers compensation statutes. There was no issue in Marie of how the Industrial Commission, without specific statutory guidance, evaluated and determined that Mr. Marie’s hearing loss was work related or how the Commission determined his disability rating. Despite the reversal by the Supreme Court, the General Assembly went forward based on the Committee’s recommendations and enacted legislation specifically recognizing work induced hearing as compensable and setting forth a detailed method for determining the various issues presented by such claims. Larson, supra note 4, at § 52.05[5]. These enactments are now numbered as §§ 287.067 and 287.197.
The Missouri Hearing Loss Statutes
The Missouri Act, once considered the most advanced statute on the subject, contains the following features: (1) a six-month waiting period after separation from the noisy work before evaluation of a loss to determine whether the degree of permanent hearing loss is permanent; (2) a determination that hearing loss be limited to the frequencies of 500,1000 and 2000 cycles per second;5 (3) the percent of hearing loss should be calculated as the average, in decibels, of the thresholds of hearing for the three frequencies; (4) if the hearing losses average less than 15 decibels in the three frequencies the hearing loss will not constitute a compensable hearing disability and if the losses average 82 decibels or more then the compensable hearing loss is deemed total or 100 percent; (5) deducting one-half decibel from the total average decibel loss for each year of the employee’s age over forty at the time of last exposure in recognition that age itself is a contributing factor to hearing loss; (6) a method of computing bin-aural percentage of loss based on the individual hearing losses in each ear to arrive at a final binaural hearing impairment.
The relevant portions of the Missouri occupational hearing loss law for purposes of the arguments here are § 287.197.2 and § 287.197.4:
The percent of hearing loss ... shall be calculated as the average, in decibels, of the thresholds of hearing for the frequencies of five hundred, one thousand and two thousand cycles per second, Pure tone air conduction audiometric instruments, approved by nationally recognized authorities in this field, shall be used for measuring hearing loss.... (§ 287.197.2)
In measuring hearing impairment, the lowest measured losses in each of the three frequencies shall be added together and divided by three to determine the average decibel loss. (§ 287.197.4) (emphasis added).
Pursuant to the authority of § 287.650, the Industrial Commission in 1959 issued an administrative rule, 8 CSR 50 5.060. Although subsequently amended in 1967, certain important provisions remain unchanged. Subsection (11) of the rule expands upon the statutory directive to use recognized audiometric instruments:
The reference zero levels of the audiometer used to measure hearing levels *538must be explicitly identified either as ASA-1951 ... or as ISO....
The specification of what standards for audiometers would apply was necessary for more than identification of a recognized authority. Section 287.197.2 specified that losses of 15 decibels or less were not com-pensable and losses of 82 decibels or more were considered total losses. But how were 15 and 82 to be measured? That answer was in the ASA-1951 and ISO standards which each establish reference zero levels for audiometers. Subsection (15) of the administrative rule then provides:
For every decibel that the corrected6 average exceeds fifteen (15) decibels based on the ASA-1951 reference levels or twenty-six (26) decibels on the ISO reference levels, an allowance of one and one-half percent (1½%) shall be made up to the maximum of one hundred percent (100%) which is reached at eighty-two (82) decibels on the ASA-1951 reference levels and at ninety-three (93) decibels based on the ISO reference levels.
Thus, Rule 8 CSR 50-5.060 implicitly recognizes that at that time there were two organizations that established reference (calibration) standards for audiometers. Important for our later consideration is that the two standards, although using different numerical values, establish equivalent decibel levels. In other words 15 decibels under the ASA-1951 standard is the equivalent of 26 decibels under the ISO standard. To further complicate matters, although the regulation remains unchanged, ANSI (the successor to ASA) adopted a new standard in 1969. The ANSI-1969 standard uses a reference level of 25 decibels rather than the reference level of 15 dB under the ASA-1951 standard. Similarly, a reference level for maximum impairment of hearing of 92 dB is used by the ANSI-1969 as opposed to the reference level of 82 dB under ASA-1951. Apparently there is also a 1989 ANSI standard. Although the record does not reflect the differences between the 1969 and 1989 standards, it appears that the 1989 ANSI Standard established a reference level of 26 dB rather than the 25 dB used in ANSI-1969 and the 15 dB used in ASA-1951. This distinction may be important because the employee’s expert (Dr. Korun-da) used the ANSI-1989 standard and the employer’s expert (Dr. Thedlinger) used the ANSI-1969 standard. Our difficulty, as hopefully will be understood in the subsequent discussion, is whether there is an equivalency between the various standards and, if so, what is the nature of that relationship.
Procedural History
For purposes of understanding and reviewing the Administrative Law Judge’s findings of fact and conclusions of law, and the subsequent affirmation by the Labor and Industrial Relations Commission, it is necessary to set out the audiometric measurements utilized by both expert physicians. Dr. Korunda’s records reflect:
Thresholds re ANSI 1989 Test 500 Hz 1000 Hz 2000 Hz
Right Ear A 15 35 70
B 25 25 70
*539[[Image here]]
The numbers in bold indicate the hearing levels that Dr. Thedlinger added together for division by three to obtain an average hearing loss in each ear.7
Ultimately, after correction for age and a conversion of the monaural loss to bin-aural measurement, Dr. Korunda arrived at 23.08 percent of impairment and Dr. Thedlinger arrived at 21.83 percent of bin-aural impairment. The Administrative Law Judge and the Commission eventually concluded that both doctors had incorrectly calculated Mr. Thatcher’s hearing loss for several reasons and then calculated the percentage of binaural impairment at 34.33 percent.
The Commission’s Findings of Fact and Conclusions of Law
The Commission found that both doctors had made two errors in their calculations. It noted that both doctors in calculating the average at each frequency had used the lowest decibel measurement at each frequency (described as “best at each frequency”) in the charts above. The Commission found this to be an erroneous interpretation of the term “the lowest measured losses” in § 287.197.4. Because the audiometer measures the point at which the tone is first detected the lower decibel measurement would reflect, where multiple tests were done, as by Dr. Korun-da, the best hearing at that frequency. The Commission held that the statute re*540quired use of the highest decibel measurement at each frequency, or in other words the worst hearing measured in the multi-pie tests. The contrast of interpretations is demonstrated below:
Calculation of average hearing levels
Dr. Korunda best each frequency: 500Hz lOQOHz 2000Hz
Right ear 15 25 70
Left ear 10 25 80
Commission worst each frequency: 500Hz 1000Hz 2000Hz
Right ear 25 35 80
Left ear 15 35 80
When the averaging of the scores in each frequency is calculated, Dr. Korunda arrives at an average loss of 36.66 dB in the right ear and 38.33 dB in the left ear. The Commission calculated an average loss of 46.67 dB in the right ear and 43.33 in the left ear.
The Commission further found that both doctors had erred in following the formula set out in 8 CSR 50-5.060 and particularly in following subsection (15) by calculating the hearing loss after correction for age from the average hearing levels shown above on the amount of deviation from 26 dB rather 15 dB. In other words the Commission believed that the doctors should have subtracted 15 dB from the average hearing level rather than 26 dB. This, the Commission concluded, resulted from doctors’ use of the regulation (8 CSR 50-5.060) which it found conflicted with the statute. The Commission based this conclusion on the language of § 287.197.2 which makes hearing loss compensable only above 15 dB. The Commission, therefore, corrected Dr. Korunda’s calculations in two respects: (1) as shown above by using worst hearing levels for each frequency (to determine average hearing level), and (2) by also reducing that product by 15 dB rather than 26 dB to arrive at the average hearing loss in each ear.
Points on Appeal
In its first point on appeal, TWA argues that the Commission erred in calculating the extent of Thatcher’s compensable hearing loss under § 287.197. TWA argues that the Commission misinterpreted the phrase “lowest measured loss” to mean greatest hearing loss, and the correct interpretation is the common ordinary usage of the word “lowest,” which in this case would mean the measurement which showed that Thatcher had sustained the least amount of hearing loss. As part of this point TWA argues that the Commission erred in using 15 dB as the reference point rather than 26 dB as used by both doctors in their calculations.
In its second point, TWA argues that the Commission erred in applying 1998 amendments to § 287.190 which increased the number of weeks of compensation for total deafness in both ears from 168 weeks to 180 weeks. TWA claims that the 1998 amendments were substantive in nature and should not have been applied retroactively where Thatcher’s date of injury preceded the effective date of the new enactment.
Standard of Review
The parties disagree as to the correct standard of review governing this case. *541TWA claims that where issues of law exist in a workers’ compensation case, the Court of Appeals reviews the case independently. Holaus v. William J. Zickell Co., 958 S.W.2d 72, 74 (Mo.App.1997). Further, TWA cites White v. Dallas & Mavis Forwarding Co., 857 S.W.2d 278, 280 (Mo.App.1993) and states that decisions involving the interpretation or application of law are not binding on an appellate court and fall within the appellate court’s province of review and correction.
Conversely, Thatcher argues that in a workers’ compensation claim the court may only review questions of law and modify, reverse, remand or set aside an award only where: (1) the Commission acted without or in excess of its powers; (2) the award was procured by fraud; (3) the facts found by the Commission do not support the award; or (4) there was not sufficient evidence in the record to warrant the making of the award. Strange v. SCI Bus. Prods., 17 S.W.3d 171, 172 (Mo.App.2000). Thatcher relies upon a case from this court holding that “the determination of a specific amount of percentage awarded to a claimant is a finding of fact within the province of the Commission, and this court will not substitute its judgment for that of the Commission even if this court would have made a different initial conclusion.” Ransburg v. Great Plains Drilling, 22 S.W.3d 726, 732 (Mo.App.2000) (emphasis omitted).
In Ransburg, the appellant appealed the Labor and Industrial Relations Commission’s affirmance of the Administrative Law Judge’s award of fifty percent disability for a back injury. The court explained that its review depended upon classifying the issues appealed as those of law or fact. An appellate court may review questions of law and make holdings as if it were the court of origin. Ransburg, 22 S.W.3d at 730. However, when reviewing questions of fact the appropriate standard of review mandates that the appellate court consider the evidence supporting the Commission’s award only to determine whether or not there is sufficient and competent evidence to support it. Id. If sufficient and competent evidence is found, then the appellate court must examine all the evidence to decide whether the award is “plainly against the weight of that evidence.” Id. Finally, and most importantly, the Court in Ransburg found that the determination of a percentage of disability is a finding of fact within the province of the Commission. Id. at 732. Further, the Court said, that since the determination of a disability percentage and an award thereunder are findings of fact, an appellate court cannot substitute its judgment for that of the Commission even if its conclusion would have been different. Id. In summary, when a finding of fact is supported by sufficient and competent evidence and is not plainly against the weight of the evidence, an appellate court cannot change the finding of fact and must affirm the Commission’s decision. Id.
Thatcher thus claims that we are limited to a review of whether there is sufficient and competent evidence to support the Commission’s finding of a 34 1/3 percent disability. If we find that there is evidence to support the Commission’s award, and the award is not against the weight of the evidence, Thatcher argues we must uphold its decision. Thatcher also points out that “[t]he determination of the percentage of disability is within the special province of the commission. It is not bound by the percentage estimates of medical experts, and it may consider all of the evidence, including the testimony of the employee, and draw all reasonable inferences in arriving at the percentage of disability.” Doria v. Chemetron Corp., 784 S.W.2d 323, 325 (Mo.App.1990).
*542Thatcher additionally argues that there was widely varying evidence as to his hearing loss. Most of that variation occurred, however, because each expert also rated Thatcher’s condition under the American Medical Association’s (AMA) Guide to the Evaluation of Permanent Hearing Impairment, 4th Ed. (1993). Dr. Korunda found Thatcher’s binaural impairment to be 39.4 percent under the AMA Guide versus 23.08 percent under the statutory formula. Dr. Thedlinger found Thatcher’s impairment to be 47.2 percent under the AMA Guide and 21.83 percent under the statutory formula. Two principal differences between the two formulas are the AMA direction for testing at 3000Hz in addition to the lower frequencies specified by § 287.197. Another difference is the statute’s requirement for an age correction or adjustment that is not contained in the AMA guidelines.8
Neither party directly addresses the issue of whether the rule that determination of the percentage of disability is a fact question within the special province of the Commission applies to the same extent where the statute sets forth a precise mathematical formula for determining the percentage of disability. We find it unnecessary to resolve that issue because all parties, including the Commission, assumed that the statutory formula applied and rejected reliance upon the AMA guidelines as a basis for determining the percentage. More importantly, we find that the Commission’s re-calculation of the percentage of disability was based on an erroneous interpretation and application of the law. We review such error de novo. White, 857 S.W.2d at 280.
Should the Commission have used a reference level of 15 dB rather than 26 dB for calculation of the amount of hearing loss?
The Commission concluded that both doctors had erroneously relied upon 8 CSR 5-50.060(15) in calculating their measured hearing level based upon the amount by which it exceeded 26 dB. The Commission found that the regulation, to the extent it permitted the calculation to based upon the level of 26 dB, was in conflict with that portion of § 287.197.4 providing, in part, “[f]or every decibel of loss exceeding fifteen decibels an allowance of one and one-half percent shall be made up to the maximum of one hundred percent which is reached at eight-two decibels.”
Thatcher correctly claims that the language of the statute takes priority over the regulation. Johnson v. Labor & Indus. Relations Comm’n, 591 S.W.2d 241, 244 (Mo.App.1979). He contends that there is no statutory authority for the use of the 26 dB figure and that the regulation authorizing the use of 26 dB must be rejected and concomitantly the calculations of both doctors based on that figure. The correctness of this argument depends, however, upon several assumptions of the science of hearing testing, which, on the basis of the record here, this court is uncomfortable in endorsing. If the 1969 and 1989 ANSI reference levels (apparently 25 dB and 26 dB) are in some manner mathematically the equivalent of 15 dB under ASA-1951, then there may be no need to recalculate Thatcher’s hearing threshold levels (by using 15 dB rather than 25 or 26 dB) as done by the Commission.
Thatcher’s conclusion, although facially understandable, may be erroneous. *543Thatcher argues and the Commission found that the 8 CSR 50-5.060(11) was “obsolete” because it required use of au-diometric instruments with reference zero levels as set by standards for calibration established either by the American Standards Association (ASA-1951) or the International Standards Organization (ISO). The Commission correctly noted that the ASA-1951 standard is no longer used. Currently used standards are those set by the American National Standards Institute (ANSI 1969 or 1989 versions). These were the standards used by the two expert doctors.
We believe the Commission misunderstood the meaning of the reference level of 15 dB under the ASA-1951, the level of 26 dB ISO specified in 8 CSR 50-5.060, and the establishment of 15 dB as the threshold for hearing loss disability under § 287.197. We must first examine and determine the meaning and intent of the legislature in establishing a 15 dB hearing level as the threshold for calculation of occupational hearing loss. The primary rule of statutory construction is to determine and give effect to the intent of the legislature. Spradlin v. City of Fulton, 982 S.W.2d 255, 261 (Mo. banc 1998). Proper analysis also requires consideration of the context in which words are used and, very importantly, the problem the legislature sought to address with its enactment. Mabin Construction Co. v. Historic Constructors, Inc., 851 S.W.2d 98, 100 (Mo.App.1993). The rules of statutory construction require that a statute be given a reasonable interpretation, State ex rel. Rowland Group, Inc. v. Koehr, 831 S.W.2d 930, 931 (Mo.1992), and should avoid absurd results. Murray v. Missouri Highway and Transp. Comm’n, 37 S.W.3d 228, 233 (Mo. banc 2001).
What, therefore, did the legislature mean by the specification of 15 dB in § 287.197? As discussed earlier, Missouri first recognized the viability of a claim for occupational deafness by case law. Marie, 319 S.W.2d 871. Many questions remained unanswered by that decision. How much hearing loss would be required for compensation? How was that to be measured? What types of tests for hearing loss should be used? What percentage of disability should be allowed for each decibel of hearing loss beyond the non-compensatory level? What frequencies should be used? What adjustments, if any, should be made for age and to convert monaural loss to binaural loss? These and other questions were addressed by § 287.197. The legislature established 15 dB as the hearing level below which losses would not be compensated. In other words, if the tested hearing level was 15 dB or below, then there would be no compensation. If the tested hearing level was 25 dB, then the loss at that frequency would be 10 dB (the difference between 15 and 25). But how did the legislature arrive at the level of 15 dB and how was it to be determined (15 dB as compared to what)?
The answer is found in the legislature’s specification that tests be those approved by nationally recognized authorities. Beginning in the late 1930’s the United States Public Health Service began examining thousands of persons for hearing levels. This and others’ work eventually resulted in establishment of average thresholds of hearing. Wayne O. Olsen, Historical Vignette: Pure-Tone Hearing Loss and Hearing Loss for Speech, American Journal of Audiology, March 1998, at 14 (available at http://jour-nals.asha.org/1059-0089/v7nl/014.htm). These values were incorporated into the ASA-1951 standards. Id. The ASA standard was a specification for the calibration of an audiometer reference level. A level *544of 15 dB, therefore, represented the consensus at that time of the level of normal hearing. ASA, however, was not the only-standard setting organization at the time. In the 1950⅛, the ISO had established as a calibration standard for audiometers the level of 26 dB. For all essential purposes, 26 dB as measured on an ISO standard calibrated machine was the equivalent of 15 dB on an audiometer calibrated according to ASA-1951 standards.
Beginning in the 1950’s, an effort was begun to standardize reference levels worldwide. This work culminated with the ISO recommended “Standard Reference Zero for the Calibration of Pure Tone Audiometers.” In 1969 ANSI issued new standards. Thus, both standards organizations established 26 dB for normal average hearing. The difference between the ASA 1951 and ANSI-1969 standards averaged 11 dB across the 500, 1000 and 2000 Hz frequencies. Id.
We conclude that the legislature in enacting § 287.197 intended the specification of 15 dB to be determined as then established under the ASA-1951 standard. Reference to an ISO standard of 26 dB in 8 CSR 50-5.060 was not, therefore, in conflict with the statute but recognized that some audiometers might be calibrated to ISO specifications and that the ISO decibel level was essentially equivalent to the ASA decibel level.
The question remains as to how we cope with the replacement of the ASA 1951 standard and its accompanying reference level of 15 dB with the 1969 ANSI standard’s reference level of 26 dB. If testing is done by an audiometer calibrated to 1969 ANSI specifications but 15 dB is used as the reference level, the apparent immediate effect is to raise the calculated hearing loss by 11 dB. It would be an absurd result to conclude that the legislature intended for an increase in hearing loss calculation to occur simply because the calibration specifications of the testing equipment had changed. The effect of the difference in the earlier and later standards is recognized in chapter eight of the United States Naval Flight Surgeon’s Manual:
If both ASA and ANSI audiograms appear in the medical record of an individual whose hearing had not changed, it would seem that the hearing has gotten worse. This is due to the different au-diometric “zero” standards and not to any organic change in [hearing threshold levels].
Naval Aerospace Medical Institute, United States Naval Flight SurgeoN’s Manual (3d ed.1991) (available at http://www.vnh.org/FSManual/08/07Mea-surement.html).
Our research suggests that it is possible to convert ASA audiometric findings so that they may be directly compared to ANSI findings. Id. An example of such a comparison is shown in the chart attached as Appendix A.9 We lack the scientific confidence to conclude on the inadequate record here that it is as simple to convert tests using ANSI 1969 or 1989 calibrated audiometers to the testing levels that would equate to the 15 dB threshold specified by the legislature in the statute. The Commission is in far superior position to this court to consider and examine the need for and manner of such conversions. There is not adequate evidence in the record below to provide a sufficient basis for resolving this issue. We hold, therefore, *545that the Commission erred in recalculating the audiometric findings by both doctors to measure compensable hearing loss above 15 dB rather than 26 dB. We, therefore, reverse the award below and remand so that the Commission may consider further evidence upon the issue of whether (and how) audiogram measurements under the ANSI standards may be converted so as to be equivalent to measurements under the ASA-1951 standard. It may ultimately be determined by the Commission that the reference level of 26 dB under the ANSI standards is essentially equivalent to the 15 dB ASA-1951 reference level used at the time § 287.197 was enacted. If that is the case then recalculation was not necessary to comply with the statutory intent.
What is the meaning of the term “lowest average loss”?
Although we must remand this matter to the Commission for further proceedings, we must also deal with the claim that the Commission erred in making a second recalculation of the hearing tests. TWA next complains that the Commission erroneously applied the law when it recalculated Mr. Thatcher’s hearing threshold level by using the highest decibel measurement at each frequency rather than the lowest decibel rating used by both doctors (as set forth earlier in the opinion in the section entitled “Calculation of Average Hearing Loss.”) The highest decibel measurement reflects the worst hearing level. The lowest decibel rating indicates that Mr. Thatcher in that particular test heard the tone at a lower (softer) level than in the other tests where the tone was not heard until a higher (louder) decibel was used.
TWA argues that the Administrative Law Judge erred in finding that the “lowest measured loss” meant the worst hearing level. They contend that “lowest measured loss” should mean the smallest or least amount of loss, not the least amount of hearing ability. Section 287.197.4 mandates that “the lowest measured losses in each of the frequencies shall be added together and divided by three to determine the average decibel loss.” The provision in 8 CSR 50-5.060(13) does not use identical language but rather mandates that “[t]he lowest hearing level at each of the three (3) frequencies shall be added together and the sum divided by three to determine the average hearing level.” The term “hearing level” is defined by the regulation as “the reading on an audiometer in decibels corresponding to the threshold of hearing of the individual being tested.” 8 CSR 50-5.060(2)(F). This definition is consistent, if not identical, with definitions commonly used in the field of audiology. See Note 2, supra. “Threshold” refers to the “weakest sound that can be heard.” 8 CSR 50-5.060(2)(C).
The Commission’s authority for its interpretation was II Mo. Workers’ Compensation Law § 7.23 (Mo.Bar.2d ed.1997), and specifically the words “[t]he higher the decibels, the lower the hearing level.” Based on this statement and the author’s opinion, the Commission went on to find that “the lowest decibel figures represented Mr. Thatcher’s highest (best) hearing levels, not his lowest measured loss.” In other words, the Commission thought that the higher decibel figure measured the lowest hearing level and, therefore, the “lowest measured loss” was reflected by the higher (worst hearing) decibel findings. Although this interpretation may have some seeming facial appeal, it ignores both general and specific guides to statutory construction. Under Missouri law, the plain and ordinary meaning of words should be used whenever possible. State v. Bouser, 17 S.W.3d 130, 138 (Mo.App.1999). Neither the Commission nor the parties believed that the terms “lowest *546measured loss” in § 287.197.4 and “lowest hearing level” in 8 CSR 5-50.060 (15) were inconsistent and not equivalent. Where a technical term is used, we should construe that term in a similar sense unless directed otherwise. City of St. Louis v. Triangle Fuel Co., 193 S.W.2d 914, 915 (Mo.App.1946).
“Hearing level” is a technical term that refers to the point (or threshold) in decibels when a testing sound is first detected by the listener. The “lowest hearing level,” therefore, represents best hearing not worst hearing. The “lowest measured loss,” therefore, is reflected by the lowest decibel rating at which the listener heard the test tone. The smallest amount of loss is the patient’s best hearing. In fact the interpretation urged by TWA is one that the Commission has itself followed in the past. Bishop v. Stupp Bros. Bridge and Iron Co., 1998 WL 809876 (Mo. Lab. Int. Rel. Comm’n Sep. 23, 1998). The Commission did not explain its departure from its previous position. Although we are not bound by the Commission’s interpretation of law, our own research has indicated that other states have used similar terminology in their statutes and Administrative rules. See Me.Rev.Stat. Ann. Tit. 39-A, § 612.5 (2001) (“lowest measured losses”); Iowa Code § 85B.9 (1997) (“lowest measured levels”)10; Wis. Admin. Code § DWD 80.25(9)(c) (2001) (“multiply the lesser loss”).
Our independent research has discovered only one case interpreting these terms.' Illinois Rev. Stat.1985, ch. 48, par. 172.42(c) provides that “[in] measuring hearing impairment, the lowest measured losses in each of the three frequencies shall be added together and divided by three to determine the average decibel loss.” In a case of apparent first impression, the Illinois Court of Appeals held that this statutory language meant that the lowest decibel (best hearing) reading recorded should be utilized. Wagner Castings Co. v. Indus. Comm’n, 241 Ill.App.3d 584, 182 IlL.Dec. 94, 609 N.E.2d 397, 407-408 (1993). As the Wagner court observed, the Illinois statute (like Missouri’s) does not provide a time frame for resolving which of differing test results should be used. Moreover, the Missouri administrative regulation requiring three separate tests on separate days creates the potential for six or more test results in each frequency if each party uses a physician who complies with the regulation. Also potentially an issue is whether a rating physician should use the lowest decibel ratings from just his own tests or whether he should look at the results of some other physician’s audiometric study and use some (or all) of them if those other results are lower. The Wagner court did not have to more fully consider these questions because there were apparently only two au-diometric tests, one by the employer and another by claimant’s physician, to compare.11 Id. at 407-08. Nor do we need to consider any of these questions to resolve the issue in this case. The Commission rejected Dr. Thedlinger’s apparently single test result as “incomplete” and TWA raises no complaint with that finding. The only audiometric results in each frequency that were before the Commission were those from Dr. Korunda’s three separate tests.
We find that the Commission misconstrued and misapplied the law by recalculating the physicians’ disability calculations *547based on the highest, rather than the lowest, decibel levels recorded in each frequency.
Retroactivity of 1998 Amendments to § 287.190.7
In its last point, TWA appeals the Commission’s retroactive application of the amendment in 1998 to § 287.190.1(27). Section 287.190 establishes the schedule of losses, in weeks of compensation for permanent partial disability under the workers’ compensation laws. In 1998, § 287.190.1(27) was amended to increase the number of weeks of compensation for total deafness from 168 weeks to 180 weeks. TWA argues that the Commission’s application of the 180-week compensation period was a retroactive application of the amendment prohibited by the Missouri Constitution. The first difficulty with TWA’s argument is that it fails to allege or cite to any provision of the Missouri Constitution that was allegedly violated. The failure to do so is generally considered a waiver of the constitutional objection. Giddens v. Kansas City S. Ry. Co., 29 S.W.3d 813, 823 (Mo. banc 2000). Even more fundamentally TWA waived its constitutional objection by failing to raise it before the Industrial Commission. The Administrative Law Judge originally applied the 1998 amendment in making his award. TWA then filed an Application for Review. The application set forth in writing the grounds upon which TWA sought review. Although it complained of the retroactive application of the 1998 amendment, it did so on the basis that it violated “governing authority.” Although it has been said that an administrative agency has no authority to declare a statute unconstitutional, Duncan v. Mo. Bd. for Architects, 744 S.W.2d 524, 531 (Mo.App.1988), TWA does not contend the statutory amendment was invalid. Rather it solely complains that the method by which the Industrial Commission applied the statute violated some unspecified portion of the constitution. The purpose of a rule requiring a constitutional argument to be raised before an administrative agency is to give the agency a chance to apply the statute in a way that TWA believes is constitutional. We hold that TWA waived its constitutional argument by failing to raise it before the Industrial Commission. Lemay Bldg. Corp. v. Director of Revenue, 889 S.W.2d 835, 836-37 (Mo.1994).
The award is reversed for a recalculation of the percentage of binaural hearing loss consistent with this opinion. The Commission may need to hear expert testimony to determine the proper method for comparing the results of modern day testing methods to the statutory specification of 15 dB under ASA-1951 as the base line for hearing loss compensability. In all other respects the award is affirmed.
THOMAS H. NEWTON, Presiding Judge, and JOSEPH M. ELLIS, Judge, concur.*548Appendix A
[[Image here]]
American Academy of Opthalmology and Otolaryngology, Guide for Conservation of Hearing in Noise 39 (1973).
. Tinnitus is a ringing in the ears or sense of noise in the head. It is separately compensa-ble under the Workers’ Compensation Law. Poehlein v. Trans World Airlines, Inc., 891 S.W.2d 505, 506 (Mo.App.1994).
.These definitions are derived from publications of the National Institute of Occupational Safety and Health (NIOSH) (available at http://www.cdc.gov/niosh/hpterms.html) and Hunter College Center for Occupational and Environmental Health, (available at htttp://www.hunter.cuny.edu/health/coeh/pub-lications/P acket — HTML/packet—index 1 .htm). Similar definitions are found in 8 CSR 50-5.060.
. The threshold of discomfort is usually noted between 85 and 95 dB and the threshold of pain is between 120 and 140 dB.
. A history of the development of hearing loss as compensable under workers compensation laws can be found at 3 Arthur Larson, et al, Larson’s Workers Compensation Law § 52.05 (2001).
. Since the enactment of Missouri’s statute, other states and authorities have suggested measurements at cycles above 2000. The AMA Guide to Medical Impairment recommends testing at 500, 1000, 2000 and 3000 cycles per second. Wisconsin adds measurement at 3000 Hz. Wis. Admin. Code § DWD 80.25 (2001).
. The term "corrected” does not refer to hearing tested with assistive devices, such as hearing aids, but rather to a numerical adjustment required when the age of the person tested is over 40. See § 287.197.6 and 8 CSR 50-5.060(14).
. It should be noted that Dr. Thedlinger's chart did not reflect that he performed more than one audiogram. 8 CSR 50-5.060(13) requires that three separate audiograms be made each on different days. The Administrative Law Judge described these records as "incomplete” because they did not reflect multiple tests.
. Thatcher's tested hearing loss in the higher frequency ranges was much greater than in the 500 or 1000 Hz ranges. Some states have changed their statutory requirements to in-elude testing at the 3000 Hz frequency or even higher and/or to utilize or incorporate evaluation according to the AMA guidelines. Missouri has not. See Note 5, supra.
. It might be noted that some more recent statutes use 25 dB rather than 26 dB as the average hearing level at which compensable loss occurs. The AMA Guide also uses 25 dB. The difference between 25 and 26 dB standards may result from the fact that the true mathematical difference between ASA 1951 and ANSI appears to be 10.6 rather than 11.
. This section was substantially revised in 1999, to explicitly detail the testing methodology-
. Additionally there was some question about the accuracy of the test with the higher decibel readings.