LOUIS J. CECI, J.
This is a review of an unpublished per curiam decision of the court of appeals dated April 25, 1985, affirming a judgment for conviction and an order of the circuit court for Grant county, John R. Wagner, circuit judge, denying defendant postconviction relief.
There are two issues in this case. First, did the trial court err in admitting evidence of Paul J. Fishnick's (Fish-nick) alleged prior act of enticement involving a thirteen-year-old girl in the present action in which Fishnick is [250]*250charged with one count of first-degree sexual assault involving a three-year-old girl? Second, was the error of admitting into evidence Fishnick's alleged prior act of exposure involving a twelve-year-old girl harmless to his defense in the present action? We affirm the decision of the court of appeals that the alleged prior act of enticement was properly admitted into evidence pursuant to sec. 904.04(2), Stats.,1 and that the alleged prior act of exposure was erroneously admitted but harmless to Fishnick.
On June 18, 1983, at 11:30 a.m., three-year-old C.S. left her family's trailer home for about fifteen minutes. Upon her return, her parents noticed that she had a piece of red cinnamon candy. C.S.'s father testified that the candy was the same as the kind kept by Fishnick in his trailer home. Fishnick and C.S.'s family lived in the same trailer court.
C.S.'s father twice asked C.S. where she had gotten the candy, but C.S. did not respond. C.S.'s mother then asked her if she had received it from "Paul"; C.S. answered, "Yes." Her parents did not immediately pursue the matter. About one minute after this exchange, C.S. volunteered to her father that "that man licked my boo-boops." "Booboops" is a family term for the vaginal area. At noon that day, C.S. repeated the statement to a Lancaster police officer, and held herself in the crotch area. Her mother again asked, "Was it Paul?"; C.S. answered, "Yes," but would not state his name. Her father again asked her where she had gotten the candy. She responded by pointing in the direction of Fishnick's trailer and stating, "At that man's trailer." At no time did C.S. state the [251]*251name Paul Fishnick as a means of identifying the defendant.
On June 20, 1983, a criminal complaint was filed, charging Fishnick with first-degree sexual assault, pursuant to sec. 940.225(1)(d), Stats.2 An information filed on June 24,1983, accused Fishnick of "feloniously intentionally hav[ing] sexual contact with a person twelve years of age or younger, to-wit: the defendant intentionally touched the vagina of [C.S.] . . . with intent to become sexually aroused or gratified."
At trial, Fishnick denied giving C.S. candy, asking her to enter his trailer, or touching her vagina, although he did admit to having seen and briefly conversed with C.S. at approximately 11:20 a.m. on the day of the incident.
On cross-examination, Fishnick denied having enticed thirteen-year-old D.F. to show him her vaginal area or having exposed himself to twelve-year-old E.F., D.F.'s sister. During the state's rebuttal, D.F. testified that on June 11,1983, she was hanging the wash on a clothesline located near Fishnick's trailer. She stated that she heard Fishnick talking to her from his trailer and that he offered her $20 to go into his trailer and show him her vaginal area. She stated that Fishnick said, "I can't see too clearly, but it looks pretty good from here." She further [252]*252testified that she then returned to her family's trailer and informed her mother of the incident.
Also on rebuttal, E.F. testified that around 10:00 p.m. on June 10,1983, she glanced into the window of Fishnick's trailer and saw him standing nude in front of his window such that she could see his penis.
The trial judge, citing Whitty v. State, 34 Wis. 2d 278, 149 N.W.2d 557 (1967), and State v. Tarrell, 74 Wis. 2d 647, 247 N.W.2d 696 (1976), admitted the "other acts" testimony of both girls as probative of motive, intent, preparation, or plan. The court gave a limiting instruction to the jury immediately after the other-acts evidence: "[Such evidence] regarding other incidents involving the defendant for which he is not on trial . . . was admitted solely on the issues of motive, intent, preparation, or plan." Further, the jury was not allowed to consider such evidence to conclude that the defendant has a certain character trait or that he acted in conformity with that trait with respect to the offense charged in this case. The court essentially repeated this instruction in its general instructions to the jury after the closing arguments of counsel. Neither the state nor Fishnick objected to the court's jury instructions.
On January 11, 1984, a jury found Fishnick guilty as charged. He was sentenced to a prison term not to exceed five years. The trial court denied a postconviction motion regarding the admission of the other-acts testimony of D.F. and E.F.
On review, the court of appeals determined that D.F.'s testimony was relevant to show the identity of C.S.'s assaulter, but not to show motive, intent, preparation, or plan. It also held that E.F.'s testimony regarding the exposure incident was irrelevant and, therefore, was erroneously admitted by the trial court. It determined, however, that the error was harmless to Fishnick. The court of appeals thus affirmed the trial court's judgment convicting Fishnick of first-degree sexual assault.
[253]*253In our order granting Fishnick's petition for review, we limited the issues on review to the validity of the admission of other-acts evidence and to the application of the harmless error analysis.
OTHER-ACTS EVIDENCE
Section 904.04(2), Stats., states:
"Evidence of other crimes, wrongs, or acts is not admissible to prove the character of a person in order to show that he acted in conformity therewith. This subsection does not exclude the evidence when offered for other purposes, such as proof of motive, opportunity, intent, preparation, plan, knowledge, identity, or absence of mistake or accident.”
The general rule is to exclude evidence of other bad acts to prove a person's character in order to show that the person acted according to his character in committing the present act. Whitty, 34 Wis. 2d at 291-92. The purpose of the other-acts rule "is to exclude evidence which is relevant only for showing a disposition to commit a crime." State v. Spraggin, 77 Wis. 2d 89, 100, 252 N.W.2d 94 (1977). The rationale for the general rule was set forth in Whitty v. State:
"The character rule excluding prior-crimes evidence as it relates to the guilt issue rests on four bases: (1) The overstrong tendency to believe the defendant guilty of the charge merely because he is a person likely to do such acts; (2) the tendency to condemn not because he is believed guilty of the present charge but because he has escaped punishment from other offenses; (3) the injustice of attacking one who is not prepared to demonstrate the attacking evidence is fabricated; and (4) the confusion of issues which might result from bringing in evidence of other crimes." Whitty, 34 Wis. 2d at 292.
[254]*254Section 904.04(2), Stats., allows for uses of other-acts evidence when it is not used to show character conformity. Such evidence may be used to demonstrate a person's motive, opportunity, intent, preparation, plan, knowledge, identity, or absence of mistake or accident in committing the act with which he or she is charged. Thus, while we do not permit a jury to infer that because a defendant committed prior act X, he is therefore of such a character and disposition to commit present act Y, the law does allow, under the proper circumstances, a jury to infer from act X a defendant's motive, intent, etc. with respect to act Y.
Trial courts must apply a two-prong test in determining whether other-acts evidence is admissible. State v. Alsteen, 108 Wis. 2d 723, 729, 324 N.W.2d 426 (1982). First, the trial court must find whether the evidence of other acts fits within one of the sec. 904.04(2) exceptions. It must then determine whether any prejudice resulting from the admission of such evidence substantially outweighs its probative value. Sec. 904.03, Stats.3 Of course, a threshold question implicit within the two-prong analysis is whether the other-acts evidence is relevant to the case. Alsteen, 108 Wis. 2d at 729.
An underlying issue to the determination whether the trial court erred in admitting the D.F. and E.F. other-acts evidence is the significance of a certain passage in Tarrell, 74 Wis. 2d 647.
In Tarrell, the defendant was charged with sexually assaulting an eleven-year-old girl on August 25, 1974.
[255]*255Over defense counsel's objections, the trial court allowed the prosecution to introduce evidence that the defendant,
"[h]ad in 1969 been convicted of enticing a child for immoral purposes; had in February, 1973, unzipped his pants and masturbated in the presence of two young female hitchhikers. . and had in the summer of 1974 made an inappropriate or obscene remark to a girl." Id. at 657.
After considering the danger of admitting other-acts evidence, we upheld the trial court's admission of the evidence. We stated that,
"... the separate acts, ranging from inappropriate comments to a girl, to enticing a minor for immoral purposes, demonstrated [defendant's]propensity to act out his sexual desires with young girls and had a logical connection with the charged offense. They tended to show a general scheme or motive. While the admission of this evidence was prejudicial, it was extremely relevant and was appropriately admitted. The probative value exceeded the prejudicial effect and as such was admissible." Id. at 658. (Emphasis added.)
Fishnick and the state agree that the above-emphasized language should be withdrawn. Members of this court have expressed reservations about the "propensity" language. See, e.g., Tarrell, 74 Wis. 2d at 663-64 (Abrahamson, J., dissenting). The primary concern is that the emphasized language wrongly suggests that a trial court may admit other-acts evidence to show a defendant's propensity or disposition to commit the act for which the person is charged. Such an interpretation would be tantamount to admitting evidence to show a person's character to prove that the person acted in conformity with his character in committing the present offense. Again, other-acts testimony is inadmissible for this purpose. Section 904.04(2), Stats.
[256]*256We agree that the propensity language wrongly implies an excessively liberal interpretation in allowing the admission of other-acts evidence pursuant to sec. 904.04(2). Other-acts evidence is inadmissible to show a person's propensity or disposition to commit another act. The language of Tarrell which suggests otherwise is withdrawn.
The defendant next asks this court to address the admission of other-acts evidence in Tarrell for the purposes of showing a general scheme or motive of the Tarrell defendant. The basis of this request is to test the continuing validity of the " greater latitude of proof" standard in fitting other-acts evidence into one of the exceptions stated in sec. 904.04(2), Stats.
We have stated,
"A 'greater latitude of proof as to other like occurrences' is clearly evident in Wisconsin cases dealing with sex crimes, particularly those involving incest and indecent liberties with a minor child. This is not so much a matter of relaxing the general rule that it is not competent in a prosecution for one crime to introduce evidence of other offenses as it is a matter of placing testimony concerning other acts or incidents within one of the well established exceptions to such rule . . . ." Hendrickson v. State, 61 Wis. 2d 275, 279, 212 N.W.2d 481 (1973). (Footnote omitted.)
In Tarrell, this court affirmed the trial court's admission of other-acts evidence which, Fishnick argues, would not have fitted into the motive or scheme exceptions to sec. 904.04(2) but for the "greater latitude" standard.
We do not choose to reconsider the merits of the Tar-rell case with regard to the "greater latitude" standard. The state does not rely on the "greater latitude" rule in its argument that the D.F. enticement evidence is admissible. It merely argues that such evidence is admissible [257]*257under sec. 904.04(2), Stats., because it shows Fishnick's motive and plan in the act he allegedly committed. Because the state does not rely on a greater latitude standard in arguing for the admissibility of the enticement evidence, a discussion of the standard is unnecessary to reach a decision in this case. We reaffirm, however, our commitment to the principle that a greater latitude of proof is to be allowed in the admission of other-acts evidence in sex crimes cases, particularly in those involving incest and indecent liberties with a minor child.4 Certainly, however, the other-acts evidence must still be placed in one of the sec. 904.04(2) exceptions.
We now turn to the merits of the instant case. In reviewing evidentiary issues, the trial court's determination will be upheld where the court exercised its discretion according to accepted legal standards and in accordance with the facts of record. Alsteen, 108 Wis. 2d at 727. We will not find an abuse of discretion where a reasonable basis exists for the trial court's determination. Boodry v. Byrne, 22 Wis. 2d 585, 589, 126 N.W.2d 503 (1964). But the record must reflect that discretion was exercised, including evidence that the trial judge undertook a reasonable inquiry and examination of the facts as the basis for his decision. Alsteen, 108 Wis. 2d at 727-28. The inquiry on review is not whether this court, "ruling initially on the admissibility of the evidence, would have permitted it to come in. . . ." State v. Wollman, 86 Wis. 2d 459, 464, 273 N.W.2d 225 (1979).
[258]*258The state concedes that the E.F. exposure testimony is inadmissible because it is irrelevant to the issue of Fish-nick's guilt or innocence on the assault charge. We therefore concern ourselves at this point only with the D.F. enticement testimony.
Fishnick first argues that the trial court improperly admitted the D.F. enticement evidence for the purposes of motive or plan. Fishnick said his defense to the action was that the crime did not occur. He maintains that motive and plan are not in issue when a defendant contends a crime did not occur. To support his position, defendant cites State v. Sonnenberg, 117 Wis. 2d 159, 172, 344 N.W.2d 95 (1984).
As in the instant action, Sonnenberg dealt with a defendant charged with sexual assault on a minor. In Son-nenberg, the state on rebuttal sought to contradict the defendant's direct-testimony statement that he never had need to seek sexual satisfaction from any woman other than his wife. We concluded that it was "error [for the state] to introduce extrinsic rebuttal evidence on a collateral matter in an effort to impeach the testimony of a witness." Id. at 161. We stated that intent, motive, and plan were not in issue in that case because the defendant asserted that the alleged offense did not happen. Id. at 172. Fishnick here asserts that the offense did not occur; he concludes, therefore, that motive and plan — two of the four exceptions on which the trial court admitted the D.F. testimony — are not in issue.
The Sonnenberg decision may be distinguished from the present case on its facts. In Sonnenberg, the evidence which the state introduced was clearly intended merely to impeach the defendant's statement about never seeking extramarital sexual favors. The state called a twenty-five-year-old woman to the stand. She testified that the defendant had sought sexual favors from her. The alleged solicitation occurred after the assault on the minor. This collateral evidence was used only to impeach the defend[259]*259ant, in violation of sec. 906.08(2), Stats.,5 and in violation of the general rule against impeachment on collateral matters by extrinsic evidence. Sonnenberg, 117 Wis. 2d at 163. McClelland v. State, 84 Wis. 2d 145, 267 N.W.2d 843 (1978). The use of such collateral evidence in Sonnenberg was wholly irrelevant to show defendant's intent, motive, or plan regarding the sexual assault on the minor.
The other-acts testimony in Sonnenberg was inadmissible not because the defense posture that the incident did not happen necessarily rendered motive, plan, or intent not in issue, but simply because the other-acts evidence was irrelevant with respect to the sexual assault on the minor. In this case, however, D.F.'s other-acts testimony is arguably relevant to the matters of motive, plan, and intent, if the evidence fits into any of these exceptions.
No matter what line of defense the defendant intended to establish in Sonnenberg, that case has no bearing on whether motive or plan is in issue in this case. Son-nenberg, therefore, does not support Fishnick's argument that a defendant's denial of committing an offense necessarily makes the defendant's motive, intent, or plan not in issue. For Fishnick to prevail on review that the other-acts evidence was inadmissible, we must find that the trial court abused its discretion in applying either part of the two-prong test for other-acts evidence.
[260]*260The first prong — whether the evidence fits into one of the exceptions stated in sec. 904.04(2), Stats. — was analyzed by the trial court and the court of appeals. The trial court determined that the D.F. testimony fit into the classifications of motive, intent, preparation, or plan. The court of appeals found the evidence admissible because it was relevant to the issue of the identity of C.S.'s assailant. We find that the trial court did not err in admitting the D.F. testimony for the purpose of motive. It was also relevant to show identity. However, it was inadmissible to show plan, preparation, or intent to do the act.
Motive has been defined as the reason which leads the mind to desire the result of an act. Baker v. State, 120 Wis. 135, 145-46, 97 N.W. 566 (1903). In other words, a defendant's motive may show the reason why a defendant desired the result of the crime charged. Tarrell, 74 Wis. 2d at 662 (Abrahamson, J., dissenting). Here, the motivation for Fishnick's interaction with C.S. was sexual gratification. His attempted interaction with D.F. was also, arguably, sexual gratification. Whether the defendant acted with the purpose of becoming sexually aroused or gratified by the sexual contact is an element of the crime with which Fishnick was charged.6 Sec. 940.225(1)(d) and (5)(a), Stats. Other-acts evidence is admissible when probative of the elements of a crime, subject to the general rule excluding character evidence. Hough v. State, 70 Wis. 2d 807, 813, 235 N.W.2d 534 (1975). Because the purpose of the sexual contact is an element of the crime, and because the defendant's motive impacts upon his purpose [261]*261for committing the crime with which he is charged, other-acts evidence which tends to show Fishnick's motive is properly admissible. D.F.'s other-acts testimony, if believed by the jury, would tend to show Fishnick's purpose (i.e., motive) for the alleged sexual contact with C.S. The jury was properly instructed that the evidence was introduced on the issue of motive and not to show that the defendant has a certain character or that the defendant acted in conformity with that character with respect to the C.S. assault.
We must also review whether the trial court exercised its discretion in determining whether any prejudice resulting from the D.F. testimony outweighs its probative value. Sec. 904.03, Stats.
The probative value of other-acts evidence depends partially upon its nearness in time, place, and circumstance to the alleged crime or element sought to be proved. Whitty, 34 Wis. 2d at 294.
The incident between Fishnick and D.F. was very similar to the incident between Fishnick and C.S. The D.F. enticement incident occurred seven days before the C.S. assault. Both incidents occurred in the same trailer home park, either in or around Fishnick's trailer. Both C.S. and D.F. were extended enticements for the sexual activity: C.S. was given candy; D.F. was offered $20. The victims were both young girls and residents of the same trailer court. The object of the defendant's attention was the same: the girl's vaginal area. Because of the similarity of circumstances between the two acts, we find that the D.F. testimony has probative value with regard to Fish-nick's motive of sexual gratification in the C.S. assault.
Certainly the D.F. testimony hurt Fishnick's defense that he did not commit the C.S. assault. But the legal prejudice of which we speak here is the potential harm in a jury's concluding that because an actor committed one bad act, he necessarily committed the crime with which [262]*262he is now charged. Tarrell, 74 Wis. 2d at 657. A trial court must determine whether a jury will prejudge a defendant's guilt or innocence in an action because of his prior bad act.
As noted above, the trial court instructed the jury to consider the D.F. evidence only with regard to motive, intent, preparation, or plan:
"Such [other-acts] evidence was admitted solely on the issues of motive, intent, preparation or plan. You may not consider such evidence and conclude that the defendant has a certain character or a certain character trait and to further conclude that he acted in conformity with that trait or character with respect to the offense charged in this case. If you find that the defendant did such act, you must not conclude that it is proof that he committed the offense with which he is now charged.. . .You may not consider this evidence to determine whether the defendant is probably guilty of this offense because of his prior conduct."
This instruction went far to cure any adverse effect attendant with the admission of the D.F. evidence. The trial court did not abuse its discretion in concluding that the probative value of the evidence was not outweighed by any prejudicial effect.
Although the trial court did not expressly rule that the probative value was not substantially outweighed by its prejudicial effect, such a ruling was implicit in the trial court's admission of the evidence. See, State v. Pharr, 115 Wis. 2d 334, 347, 340 N.W.2d 498 (1983). The transcript reflects that the trial court took care to consider nearness of time, place, and circumstance with regard to the two incidents.
Because the D.F. testimony was properly admitted for motive purposes and because its probative value was not outweighed by prejudicial effect, it is not necessary to our holding in this case to determine whether the D.F. testimony was also admissible for identity purposes. Nev[263]*263ertheless, we conclude that the evidence was also relevant to show the identity of C.S.'s assaulter. Fishnick's counsel pointed out on cross-examination of one of the investigating police officers that C.S. never said "Paul Fishnick" to describe her assaulter and that although C.S. pointed in the direction of Fishnick's trailer to describe where the assault occurred, there were other trailers in that direction besides Fishnick's. Thus, the identity of C.S.'s assaulter was in issue. The court of appeals correctly ruled that the D.F. testimony was admissible for that purpose.
Because the trial court did not enter into an analysis to determine that the D.F. testimony was admissible for purposes of identity, we apply the second part of the two-prong test to determine whether any prejudice of the evidence outweighs its probative value. Where other-acts evidence is used for identity purposes, similarities must exist between the "other act" and the offense for which the defendant is being tried. Sanford v. State, 76 Wis. 2d 72, 80, 250 N.W.2d 348 (1977); Wis. J. L — Criminal 275. Similarities which tend to identify the defendant as the proponent of an act also tend to ensure a high level of pro-bativeness in the other-acts evidence. These similarities may be established, for example, where there is a discernible method of operation from one act to the next, State v. Rutchik, 116 Wis. 2d 61, 68, 341 N.W.2d 639 (1984), or where the other act and the crime charged and their surrounding circumstances are so similar that the incidents and circumstances bear the imprint of the defendant. Id. at 88 (Abrahamson, J., dissenting). In order for other-acts evidence to be admitted for purposes of identity, there should be such a concurrence of common features and so many points of similarity between the other acts and the crime charged that it can reasonably be said that the [264]*264other acts and the present act constitute the imprint of the defendant.7
For reasons discussed above under the motive analysis, we find the D.F. incident and the C.S. incident exhibit a concurrence of common features that establish the defendant's imprint; therefore, the D.F. incident is probative of the defendant's identity. Again, the acts were temporally and spatially similar. The aggressor's method was similar in each of the acts. For example, each act included an enticement of either money or candy. The victims were both young girls. The aggressor's object was the same: touching or viewing a girl's vaginal area.
The defendant argues that the jury was not instructed to consider the D.F. evidence for purposes of the assaulter's identity and that the evidence, therefore, was admitted erroneously. But this court will not reverse a trial court's ruling if the ruling is correct and the record reveals a factual underpinning that would support the proper findings. Mueller v. Mizia, 33 Wis. 2d 311, 318, 147 N.W.2d 269 (1967). The record reflects a factual basis for the admission of the D.F. testimony for identity purposes, as evidenced by the similarity analysis above.
We find no harm to Fishnick because of our holding that the D.F. testimony was admissible for identity purposes. The evidence, in fact, was not admitted for such purposes; therefore, there was no need for the trial court [265]*265to instruct the jury to consider the evidence in such a light. Because the jury did not consider the evidence for identity purposes, Fishnick was not harmed by the lack of an instruction discussing identity.8 The D.F. testimony was independently admissible for motive purposes, for which the trial court properly gave a limiting instruction.
ERROR IN ADMITTING THE E.F. TESTIMONY
We next address the admission of the E.F. evidence. The state concedes in its brief that the trial court erroneously admitted E.F.'s testimony that she saw Fishnick standing naked in the window of his trailer home. This court must consider whether the admission of this evidence was harmless to Fishnick.
The test for nonconstitutional harmless error is whether there is a reasonable possibility that the error contributed to the conviction. State v. Dyess, 124 Wis. 2d 525, 543, 370 N.W.2d 222 (1985). Where there is no reasonable possibility that the error contributed to the conviction, the error is harmless. The focus of the "reasonable possibility" analysis is whether the error " ’undermine[s] confidence in the outcome'" of the case. Id. at 545 (quoting from Strickland v. Washington, 466 U.S. 668, —, 104 S.Ct. 2052, 2068 (1984)).
Certainly the E.F. evidence did not satisfy either part of the two-prong other-acts analysis. It did not show motive in the assault on C.S., nor did the evidence fit into any of the other exceptions for which it was admitted. As the court of appeals reasoned, the E.F. testimony was simply irrelevant to the inquiry of the C.S. assault: "One can [266]*266be nude in one's home for innocent as well as deviant reasons. That [Fishnick] happened to be observed by [E.F.] does not make it more or less probable that he sexually assaulted a three-year-old a week later . . .
We are unconvinced, however, that the error of admitting the E.F. testimony was prejudicial to Fishnick. We are certain of the integrity of the outcome of the proceedings; our confidence in the outcome is not undermined by the admission of the evidence.
First, a review of the E.F. testimony suggests that the evidence was, at most, ineffectual in serving any relevant, probative purpose. For example, E.F. was uncertain of the distance between herself and Fishnick's window on the evening that she viewed him. She was unsure whether there was a light on in Fishnick's trailer at that time. She admitted that any outside light shining onto the Fishnick trailer might have been partially obscured by trees and other foliage. Although the E.F. testimony was ineffective for the state, we do not deem its admission prejudicial to Fishnick. The evidence most likely had no effect on the outcome of the trial.
Second, there is sufficient evidence untainted by the error in admitting the E.F. testimony, see, id. at 545, 546, which supports our confidence in the conviction. The evidence that C.S. was able to identify her assaulter is untainted by the error. D.F.'s testimony that Fishnick attempted to entice her is likewise untainted. These two items of evidence, together with the evidence that C.S. returned to her trailer after the assault with the type of wrapped, red cinnamon candy usually kept by Fishnick, are highly probative and would have resulted in the jury's returning a verdict of guilty even without the E.F. testimony. There is no reasonable possibility that" 'the result of the proceeding would have been different'" absent the error. Id. at 544 (quoting from Strickland, 466 U.S. at —, 104 S.Ct. at 2068).
[267]*267The strength of the untainted evidence confirms that there is no reasonable possibility that the erroneously admitted evidence contributed in any significant way to Fishnick's cohviction. The outcome would have been the same, absent the E.F. testimony. Our confidence in the integrity of the result of the proceedings is not eroded. The admission of the E.F. testimony, although erroneous, was harmless to Fishnick in the defense of his case.
By the Court. — The decision of the court of appeals is affirmed.