BOGGS, Circuit Judge.

Defendant Modesto Diaz appeals his conviction for possession, with intent to distribute, of one hundred pounds of marijuana, in violation of 21 U.S.C. § 841(a)(1), arguing that the district court erred in denying his motion to suppress evidence. He contends that an indication of the presence of marijuana by an allegedly improperly trained police dog did not constitute probable cause to search his vehicle. For the reasons discussed herein, we hold that the police properly may rely on a trained and certified dog as probable cause for a search. We therefore affirm the judgment of conviction.

I

Based on information obtained from a suspected drug courier at the Detroit airport, drug agents located Diaz's car at the Colonial Motel in Taylor, Michigan. A drug detection dog "alerted" on the car. Believing, he says, that he had no choice, Diaz later consented to a search of the car, and one hundred pounds of marijuana were found in the trunk. After an evidentiary hearing at which Diaz, Diaz's expert on drug detection dogs, the drug detection dog's trainer-handler, and one of the drug agents testified, the district court denied Diaz's motion to suppress the marijuana. Diaz entered a conditional plea of guilty and was sentenced to 24 months' imprisonment and three years of supervised release. Diaz contends that the government failed to establish the dog's training and reliability, and thus the agents lacked probable cause to search the car. Diaz further argues that the agents' entry into the motel's parking lot was unlawful.

II A

In this case, the dog sniff is determinative of the issue of probable cause to search Diaz's car; before the dog alerted on the car, probable cause for a search did not exist. A positive indication by a properly-trained dog is sufficient to establish probable cause for the presence of a controlled substance. United States v. Knox, 839 F.2d 285, 294 n. 4 (6th Cir. 1988), cert. denied, 490 U.S. 1019, 109 S.Ct. 1742, 104 L.Ed.2d 179 (1989). Diaz does not question the proposition that a dog alert can establish probable cause, but challenges the training and reliability of the drug detection dog, Dingo. For a positive dog reaction to support a determination of probable cause, the training and reliability of the dog must be established. See United States v. $67,220, 957 F.2d 280, 285 (6th Cir. 1992).

Courts have not definitively addressed the issue of the quality or quantity of evidence necessary to establish a drug detection dog's training and reliability. We look to analogous principles of evidence law for guidance on this issue. As with evidence generally, trial judges have broad discretion in determining the admissibility of expert evidence. See 3 Jack B. Weinstein Margaret A. Berger, Weinstein's Evidence ¶ 702[02], at 702-22 (1993); United States v. L.E. Cooke Co., 991 F.2d 336, 341 (6th Cir. 1993). Similarly, an expert's qualification is a question that lies within the trial judge's discretion. 3 Weinstein Berger, ¶ 702[04], at 702-45; Mannino v. International Mfg. Co., 650 F.2d 846, 849 (6th Cir. 1981). The court considers the proffered expert's education and experience in determining if he is qualified. Graham C. Lilly, An Introduction to the Law of Evidence § 12.2, at 485 (2d ed. 1987). Formal education is not always necessary to qualify an expert; practical skill and experience may suffice. Ibid. See also Mannino, 650 F.2d at 851 ("[T]he only thing a court should be concerned with in determining the qualifications of an expert is whether the expert's knowledge of the subject matter is such that his opinion will likely assist the trier of fact in arriving at the truth. The weight of the expert's testimony must be for the trier of fact."). When an expert has been qualified, other evidence, including the testimony of other experts, that contradicts or undermines the testimony of the expert affects that expert's credibility, not his qualifications to testify. See Davis v. Combustion Eng'g, Inc., 742 F.2d 916, 919 (6th Cir. 1984).

We find these principles to be useful guides in evaluating the training and reliability of a drug detection dog for the purpose of determining if probable cause exists based on the results of the dog's sniff. When the evidence presented, whether testimony from the dog's trainer or records of the dog's training, establishes that the dog is generally certified as a drug detection dog, any other evidence, including the testimony of other experts, that may detract from the reliability of the dog's performance properly goes to the "credibility" of the dog. Lack of additional evidence, such as documentation of the exact course of training, similarly would affect the dog's reliability. As with the admissibility of evidence generally, the admissibility of evidence regarding a dog's training and reliability is committed to the trial court's sound discretion.

We review the district court's decision on Diaz's motion to suppress evidence under "two complementary standards. First, the district court's findings of fact are upheld unless clearly erroneous. Second, the court's legal conclusion as to the existence of probable cause is reviewed de novo." United States v. Leake, 998 F.2d 1359, 1362 (6th Cir. 1993) (citations omitted). Because an alert by a properly-trained and reliable dog establishes probable cause, we will uphold the district court's decision here unless its findings as to Dingo's training and reliability are clearly erroneous.

At the evidentiary hearing on Diaz's motion to suppress, Wayne County Deputy Sheriff Kris Dennard, Dingo's trainer and handler, testified that she and Dingo successfully attended an eight-week training school in which both learned techniques for the detection of controlled substances, including marijuana, cocaine, and heroin; that as part of the training, Dingo was subjected to "live" search tests (in which drugs were present) and "dead" search tests (in which drugs were not present, but plastic bags and live animals sometimes were); that to gain certification, Dingo was required to successfully "indicate" narcotics on fourteen "live" targets; that Dingo would "indicate" by barking, biting, and scratching, but occasionally would "alert" by coming to a standstill in order to scent more intently; that Dingo was certified; that she and Dingo have passed recertification tests every year since their original training in 1989; that she and Dingo have had occasion to search for the presence of drug odors on approximately 1500 occasions; that on at least one occasion, Dingo indicated the presence of illegal substances but none was found, although there was evidence that drugs had been present among the items to which Dingo had positively responded; that she ran Dingo around a test car before scenting Diaz's car to avoid unduly suggesting to the dog a specific place to indicate; and that Dingo indicated on Diaz's car but not on the test car. The district court found Dennard's testimony to be credible.

Diaz presented testimony from Paul Piergiovanni, a former police officer who trains drug detection dogs and their handlers and who has testified a number of times regarding the reliability of dog sniffs. Piergiovanni relied on the transcripts of Dennard's cross, redirect, and recross examinations, but not on the transcript of her direct examination, which was unavailable (he relied instead on defense counsel's description of that testimony). He never visited Dingo's school, never spoke with Dingo's trainer or with Dennard, and had never seen Dingo in action. The district court thus noted that "the limited information upon which his opinion is based most assuredly detracts from his testimony."

Piergiovanni testified that barking and biting may indicate a dog's frustration over not detecting any drug odors; that Dingo's reliability was compromised by (according to Piergiovanni's understanding of Dingo's training) failing to be trained on "dead" targets; and that, because Dennard knew which car was suspected, she may have unconsciously cued Dingo, and thus Dingo's indication might have been tainted.

Diaz contends that the district court's finding that Dingo was a reliable drug detection dog was clearly erroneous because of Piergiovanni's superior qualifications (relative to Dennard's) and because the district court concluded, despite Dennard's apparent testimony to the contrary, that Dingo had never indicated a false positive. Diaz argues that the government could not establish Dingo's reliability because Dennard failed to bring the dog's training and performance records to court and so was unable to answer precisely how many searches Dingo had done and how many times drugs were or were not discovered when Dingo indicated. Diaz suggests that failure to require production of such records "would create a license for perjury on this crucial issue." Diaz further contends that the search was unreliable because of the possibility (indeed, he says, inevitability) of unconscious cuing by Dennard. Finally, Diaz argues that Dennard and Dingo were improperly trained, as evidenced by Dennard's taking Dingo's barking and biting as a positive indication when Piergiovanni says it may mean frustration at the failure to detect odors and by Dennard's failure to define "alert" and "dead target" in the same manner as Piergiovanni.

Dennard testified as to her and Dingo's training, certification, and experience. The district judge heard the testimony and made a credibility determination: Dennard was believable. Her testimony supports a finding that Dingo was trained and reliable. After reviewing the testimony, we are not left with a definite and firm conviction that a mistake has been made.

Furthermore, Piergiovanni's objections simply appear unpersuasive. The fact that Piergiovanni is a former police officer and now a drug detection dog trainer does not detract from Dennard's qualifications. As to the issue of false positive indications, Dennard admitted that there had been times when Dingo had alerted yet no drugs were found. She then described one such instance at an airport search. She explained that although there were no drugs present, the owner of the suitcase on which Dingo had alerted admitted that she had been smoking "weed" all weekend and that the scent could have remained in her clothing found in the suitcase. Diaz infers from this that there had been false positive indications and that therefore Dingo was not reliable. The district court was not clearly erroneous in concluding, rather, that based on Dennard's testimony, Dingo was reliable and the single incident described did not detract from this reliability. In any event, a very low percentage of false positives is not necessarily fatal to a finding that a drug detection dog is properly trained and certified. See United States v. Alvarado, No. 90-6058, 1991 WL 119265, at [**]2 (6th Cir. July 1, 1991) (unpublished) (95% accuracy in detecting narcotics); United States v. Spetz, 721 F.2d 1457, 1464 (9th Cir. 1983) (accuracy rates of 56 out of 61 and of 2 out of 6).

Regarding the failure to prove Dingo's training and reliability with training and performance records, this court has indicated that testimony is sufficient to establish a dog's reliability in order to support a valid sniff. See $67,220, 957 F.2d at 285. While training and performance documentation would be useful in evaluating a dog's reliability, here the testimony of Dennard, Dingo's handler, sufficiently established the dog's reliability.

United States v. Trayer, 898 F.2d 805, 809 (D.C. Cir.), cert. denied, 498 U.S. 839, 111 S.Ct. 113, 112 L.Ed.2d 83 (1990), noted that "less than scrupulously neutral procedures, which create at least the possibility of unconscious `cuing,' may well jeopardize the reliability of dog sniffs," but upheld the district court's determination that cuing had in fact not occurred. Here, similarly, although there was a possibility of unconscious cuing because Dennard knew which was the suspected car, the district court found that Dennard had not done anything differently between the sniffs of the test car and the suspected car and thus the chance of cuing was reduced. This finding was also not clearly erroneous.

Finally, Dennard did define "alert" and testified that Dingo was in fact trained with "dead targets," although she did not use the term "dead target." She also explained that Dingo was trained to bark, bite, and scratch when he indicated. In any event, she testified that, on this search, Dingo indicated by scratching alone. This supports the district judge's finding that Dingo was not frustrated as Diaz contends, but, instead, reliable.

In sum, the district judge's findings that Dingo was trained and reliable were not clearly erroneous. Therefore, there was probable cause to search Diaz's car, because a valid dog sniff indicated the presence of a controlled substance.

B

Diaz argues that even if the dog sniff adequately established probable cause, the sniff and search were fruits of an unlawful entry into the parking lot, because his car was in that part of the lot reserved for motel guests and thus he had a reasonable expectation of privacy. Diaz, relying on United States v. Carriger, 541 F.2d 545 (6th Cir. 1976), likens the lot to a common area of an apartment building; demands the amount of privacy accorded an overnight guest at a hotel or friend's house, see Minnesota v. Olson, 495 U.S. 91, 110 S.Ct. 1684, 109 L.Ed.2d 85 (1990); and concludes that only motel guests, and not the general public, were invited to enter that part of the lot in which his car was parked.

Although Diaz may have had an expectation of privacy in his motel room and (albeit less so) in his car, that expectation did not extend to the lot, which the officers could lawfully enter. See United States v. Brandon, 599 F.2d 112, 113 (6th Cir.) (government agents who entered used car lot and looked in windows of cars to determine odometer readings did not violate defendants' Fourth Amendment rights), cert. denied, 444 U.S. 837, 100 S.Ct. 72, 62 L.Ed.2d 47 (1979). This case is similar to United States v. Boden, 854 F.2d 983, 990 (7th Cir. 1988), in which the court held that tenants of a commercial self-storage facility had no reasonable expectation of privacy in the facility's common area, to which the tenants and their invitees had access. Here, the motel guests had no reasonable expectation of privacy in the motel's parking lot. See also United States v. Cook, No. 89-5947, 1990 WL 70703, at [**]2 (6th Cir. May 29, 1990) (unpublished) (expectation of privacy does not extend to the areaway outside a rented storage locker in a public facility). Thus, the drug agents could use the trained dog in the parking lot to sniff out drugs without implicating the Fourth Amendment. See United States v. Place, 462 U.S. 696, 707, 103 S.Ct. 2637, 2644-45, 77 L.Ed.2d 110 (1983).

Diaz's reliance on Carriger is misplaced, for the simple reason that the officers here lawfully could be in the parking lot. In Carriger, the government agent's entry into the apartment building was an illegal entry proscribed by the Fourth Amendment because the building was not open to the general public except on permission of the tenants and the agent had neither permission nor a warrant. 541 F.2d at 550-51. There was no unlawful entry here.

III

For the foregoing reasons, the district court's denial of Diaz's motion to suppress is AFFIRMED.