On November 3, 1977, defendant Harold Perryman was jury convicted of breaking and entering an unoccupied dwelling with intent *519to commit larceny, contrary to MCL 750.110; MSA 28.305.
The evidence adduced at trial revealed that on the evening of May 4, 1977, complainant Lucile Fairchild returned to her home to find broken glass and a previously locked door ajar, whereupon she ran to a neighbor’s house in order to call the police. She returned in the company of two men, one of whom was armed with a pistol. Subsequently, gunfire was exchanged between one of the complainant’s escorts, Irving Jackson, and those in the house. Complainant’s other companion, Dock Daley, testified in court that he observed Jackson shoot the defendant twice. Daley then saw defendant limp to the rear of the house and climb out through a back window. Jackson himself, however, could not testify that his shots struck anyone.
Approximately 45 minutes after discovery of the crime, Officer Phineas Williams arrived at the scene with Schultz, a police tracking dog. Upon being told that two people had jumped a backyard fence, Officer Williams placed Schultz on a scent at the fence where he had found evidence of blood and a footprint, a distance of 75 feet from the house. Schultz was able to locate defendant lying in a park a few blocks away, with two gunshots in his leg.
After the jury’s verdict, defendant was sentenced to a term of 7 to 15 years imprisonment. He now appeals as of right pursuant to GCR 1963, 806.1, and raises three issues.
Defendant first assigns as error the lower court’s decision to preclude defense counsel, during cross-examination at trial, from questioning a testifying police officer as to the substance of an exculpatory statement made by defendant to the officer, after defendant was arrested but before arraignment. *520Following a bench conference, the court ruled that, in accordance with MCL 764.26; MSA 28.885,1 it would suppress any declarations made prior to defendant’s presentment before a magistrate. In this the court erred, as decisions interpreting that provision are premised upon a concern for the voluntariness or coercive quality of statements made by a defendant in police custody prior to arraignment. See People v Antonio Johnson, 85 Mich App 247; 271 NW2d 177 (1978), and cases cited therein. Here, defendant himself sought to introduce the statement into evidence, thus rendering irrelevant any inquiry into the statement’s compulsive nature.
Nevertheless, we do not reverse, inasmuch as the statement offered (that defendant took no part in the breaking and entering, and was innocently shot by a third person) was excludable as hearsay not qualifying under any exception. Our disposition is not affected by the fact that the trial court assigned the wrong reason to the correct result. Durbin v K-K-M Corp, 54 Mich App 38, 46; 220 NW2d 110 (1974), lv den 394 Mich 789 (1975), Scott v Saupe, 32 Mich App 503, 508; 189 NW2d 159 (1971).
Defendant further alleges that the trial court erred in admitting the tracking dog evidence, maintaining that: (1) such evidence is scientific in nature, and lacks the requisite unimpeachable validity necessary to justify its admission, and (2) the introduction of tracking dog evidence deprived defendant of his constitutional right to confront *521witnesses.2 Neither of these objections were presented to the court below,3 therefore, we decline to review them. The court did, however, rule the evidence admissible as it satisfied the foundational guidelines enunciated in People v Norwood, 70 Mich App 53, 55; 245 NW2d 170 (1976), lv den 397 Mich 884 (1976), which followed People v Harper, 43 Mich App 500, 508; 204 NW2d 263 (1972), lv den 389 Mich 759 (1973), in holding tracking dog evidence admissible in Michigan,4 but only if the following four safeguards are established:
"(1) the handler was qualified to use the dog; (2) the dog was trained and accurate in tracking humans; (3) the dog was placed on the trail where circumstances indicate the alleged guilty party to have been; and (4) the trail had not become so stale or contaminated as to be beyond the dog’s competency to follow it.”
Defendant concedes on appeal that a proper showing was made that the trail Schultz pursued was fresh. Our own review of the testimony discloses adequate compliance with the remaining criteria sufficient to justify its evidentiary admission.
Officer Williams testified that he had trained extensively with Schultz for 13 weeks (which training included urban as well as rural tracking) and described in detail the dog’s manner of investigation, and how he had learned to "read” Schultz, *522 i.e., to interpret the canine’s actions. He stated that he had worked with one other dog before Schultz, and further that he had been associated with the K-9 Unit of the Detroit Police Department for five years.
Williams asserted that upon graduation with Schultz from training school they had together successfully tracked humans in at least 11 of 36 attempts, and explained the 25 nonsuccesses as involving those situations where a person had entered a motor vehicle, or where the tracking was attempted in a building that disclosed no occupants, etc. On cross-examination, he testified that, to his knowledge, Schultz had never made a mistake.
This testimony was sufficient to establish that (1) Officer Williams was qualified to handle Schultz, and (2) that Schultz had some record of accuracy in tracking humans. It was then up to the jury, in light of his record, to weigh the credibility of the dog’s tracking.
As to the third requirement, that the dog was placed on a trail where the circumstances indicate the suspect had been, Williams stated that upon arriving at the scene he was informed that two individuals (one of whom had been shot) had escaped over a rear fence. Upon determining that no officers or civilian witnesses had been in the area, he found a footprint near a bloody portion of the fence and had Schultz begin tracking from that point.
The officer’s account of what he found at the scene corresponds with testimony from other witnesses as to what did occur. His action in placing the dog on the trail from the footprint, rather than from the house, where several nonsuspects had been, was reasonably calculated to find a suspect.
*523In viewing this evidence as a whole, we are convinced that the prerequisites of People v Norwood, supra, and People v Harper, supra, were met in the present case, and therefore, are unable to conclude that the lower court’s evidentiary ruling was erroneous.
Defendant also contends that it was error for the court not to give a cautionary instruction on the unreliability of tracking dog evidence. However, defendant made no request at trial for the instruction, and counsel declared itself satisfied with the court’s intended jury charge, which included no such reference. As we are not persuaded as to the existence of manifest injustice, the failure to object to the jury instruction bars reversal. People v Dixon, 84 Mich App 675, 685; 270 NW2d 488 (1978).
Nevertheless, we are of the opinion that this type of evidence, though properly admissible, is to be treated with caution. Recognition of its fallibility can be found even in the opinions of those jurisdictions which permit its introduction at trial. For this reason, most hold, as does Michigan, People v McPherson, 85 Mich App 341, 344-347; 271 NW2d 228 (1978), that tracking dog evidence, standing alone, is insufficient to support a conviction.
We also perceive the danger that a jury may place undue importance on evidence of this type:
As McPherson noted:
" 'The principal objections to the use of bloodhound evidence appear to be: (1) That the many variable factors involved, such as the training and experience of the dogs, the manner in which they are handled, the circumstances surrounding the trailing, etc., as well as the inability of the court to check up on the dogs and tell why they acted as they did, make such evidence too *524uncertain and unreliable to warrant its admission in the trial of a criminal case; (2) that the jury, due to the superstitious awe with which people regard the actions of animals, are likely to give more weight to such evidence than it is entitled to, particularly where it is introduced in connection with some heinous crime about which there had been much public indignation and excitement.’ ” 85 Mich App at 345, quoting 18 ALR3d 1221, p 1223.
We view these hazards as genuine ones, and therefore hold, prospectively, that a court has a duty, even absent a request by counsel, to inform the jury that tracking dog evidence: must be considered with caution; is of slight probative value; and if found reliable, cannot support a conviction in the absence of other direct evidence of guilt. See People v Centolella, 61 Misc 2d 723; 305 NYS2d 279 (1969).
Finally, we find no merit in defendant’s assertion that his sentence was illegally imposed because the trial judge considered defendant’s maintenance of innocence and lack of remorse in determining punishment. In People v Yennior, 399 Mich 892 (1977), the Michigan Supreme Court, reversing this Court5 in lieu of granting leave, stated that "A court cannot base its sentence even in part on defendant’s refusal to admit guilt”. Here, while the court mentioned these factors, we note it did so not only after sentence had been imposed, but also in direct response to defense counsel’s invitation to reconsider the issue of defendant’s minimum sentence and allegations of innocence. Cf., People v McDaniels, 38 Mich App 174, 177; 196 NW2d 25 (1972), lv den 387 Mich 787 *525(1972). On these facts, we do not find reversible error.
Affirmed.
D. E. Holbrook, Jr., J., concurs in result only.