HENLEY, Senior Circuit Judge.

Kenneth Huie Davis was convicted of raping Mary Cammack Hanks in Hot Springs National Park. On direct appeal, his conviction was affirmed by this court in United States v. Davis, 785 F.2d 610 (8th Cir.1986). Davis has petitioned for habeas relief, 28 U.S.C. § 2255, claiming that his counsel was ineffective for not discovering that the report of the doctor who examined Hanks indicated a positive gonorrhea culture. He also claims that the magistrate and district court erred in relying on the testimony of a public health official, and that the district court erred in considering evidence of juror bias in the second mistrial. The district court 1 denied the petition for habeas corpus, and defendant appeals. We affirm.

The doctor’s report, which was entered into evidence at trial by stipulation, indi *166 cated that the victim did not behave like a typical rape victim, and that there was no evidence of physical trauma. The medical report also indicated a “positive Nissera culture.” “Nissera” is a misspelling of “neisseria.” The physician who wrote the report, Dr. Gibson Dan Kimberlin, testified at the evidentiary hearing on the habeas petition that the term was used to denote neisseria gonorrhoeae, the organism that causes gonorrhea.

On his direct appeal, Davis contended that the government should have informed the defense that the victim had gonorrhea, but this court held that the meaning of the word “neisseria” was easily discoverable, and that counsel’s failure to discover it was due to failure to prepare and investigate thoroughly. 785 F.2d at 618. Davis claims his counsel was ineffective in failing to discover that the medical report indicated that Hanks had gonorrhea.

In order to establish that he is entitled to relief on ineffective assistance grounds, Davis must establish, first, that counsel’s performance was deficient, and second, that he was prejudiced by counsel’s errors. Strickland v. Washington, 466 U.S. 668, 687-88, 104 S.Ct. 2052, 2064-65, 80 L.Ed.2d 674 (1984). In this case, we need not determine whether the first prong of Strickland is satisfied, as we find that Davis was not prejudiced by counsel’s asserted error. Id. at 697, 104 S.Ct. at 2069 (object of ineffective assistance claim is not to grade counsel’s performance; “If it is easier to dispose of an ineffective assistance claim on the ground of lack of sufficient prejudice ... that course should be followed”); United States v. Lee, 782 F.2d 133, 135 (8th Cir.1986).

In order for the appellant to prevail, he must affirmatively prove prejudice. Strickland, 466 U.S. at 693, 104 S.Ct. at 2067. “The defendant must show that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different; a reasonable probability is a probability sufficient to undermine confidence in the outcome.” Id. at 694, 104 S.Ct. at 2068. Thus, the relevant inquiry is whether a reasonable probability exists that the factfinder would have entertained a reasonable doubt as to the defendant’s guilt in the absence of counsel’s errors. Id. at 695, 104 S.Ct. at 2068. We are to consider the evidence in its totality, accepting the factual findings unaffected by error as a given and taking into account the effect of the errors on the remaining findings. Id. at 695-96, 104 S.Ct. at 2068-69. We must examine the error in light of the strength of the prosecution’s case; if the government’s case is strong and would not have been substantially weaker had the error not occurred, then prejudice did not result. Walker v. Lockhart, 807 F.2d 136, 140 (8th Cir.1986).

Here, the government’s case against Davis may fairly be characterized as strong. Davis claimed he did not rape Hanks and in fact had never seen her before the charge. However, the victim’s testimony was corroborated by several items of physical evidence. She testified that the appellant had picked her up while she was hitchhiking, and that he drove her into the park where he initially turned onto a dirt berm. At some point he backed up, throwing gravel. National Park Service Rangers later observed an area where gravel had been thrown from the berm. Hanks testified that Davis had grabbed a pocketknife from her hand and had thrown it into the woods; the pocketknife was later found at the scene. A tampon was also found, which was consistent with the victim’s testimony. Hanks was able to give the police the license number of Davis’s truck. A pubic hair was found in the cab of the truck, where the rape occurred; it was similar to the defendant’s. Semen was present in the vaginal washings taken from the victim at the hospital, and was present on the underwear taken from the defendant some few hours after the rape allegedly occurred. While the doctor’s report indicated that the victim seemed not to be suffering from emotional distress, and that her reactions were atypical of rape victims, he stated that he could not say she had not been raped. After Davis was arrested, Hanks picked his photograph from a dis *167 play of six photographs and identified him as her attacker.

With the foregoing facts in mind, we must attempt to evaluate the effect that the failure to introduce evidence that the victim tested positive for gonorrhea might have had on the jury. Conflicting evidence was introduced at the evidentiary habeas hearing with respect to the likelihood of transmission of the disease during intercourse. Dr. Kimberlin testified that the likelihood of transmission from a female to a male was fifty to seventy-five per cent. However, Steve Boone, a public health investigator for the Arkansas Department of Health, testified for the government that the female-to-male rate of transmission was approximately twenty per cent. Boone testified that this figure was based on his practical experience and was corroborated by literature on the subject. Thus, had Dr. Kimberlin testified at the criminal trial, the jury would have heard testimony to the effect that the possibility of transmission of gonorrhea from a female to a male was anywhere from twenty to seventy-five per cent. 2

We do not find that petitioner has met his affirmative burden of proving prejudice. There is no guarantee that the jury would have accepted Dr. Kimberlin’s figures, and it is quite possible that his testimony could have been successfully impeached. Although he admitted he had not made any systematic study, Boone’s testimony was based on personal observation of hundreds of cases and was substantiated by his reading of a recognized medical treatise. Although Dr. Kimberlin stated that there were articles contradicting the twenty per cent figure, he did not cite any. While such articles may exist, the absence of any citation to them in the record to some extent weakens petitioner’s proof. Moreover, a survey of the cases in which this issue has arisen tends to support a figure of less than fifty per cent. See Logan v. Marshall, 680 F.2d 1121, 1123 (6th Cir.1982) (defendant’s proffered scientific evidence showed twenty-five per cent probability that a male will contract gonorrhea from an infected female after one act of intercourse); State v. Carmichael, 240 Kan. 149, 152, 727 P.2d 918, 923 (1986) (study showed that less than twenty-five per cent of noninfected males would contract gonorrhea as a result of nonconsensual contact with infected female); Miller v. State, 67 Md.App. 666, 669, 509 A.2d 135, 136 (nurse with twelve years of experience in venereal disease clinic testified that statistics showed less than fifty per cent of males would contract disease during intercourse), cer t. denied, 307 Md. 260, 513 A.2d 314 (1986); State v. Ervin, 723 S.W.2d 412, 415 (Mo.App.1986) (less than one-third); Smith v. State, 737 S.W.2d 910, 915 (Tex.App.1987) (twenty per cent); see also R.C. Noble, Sexually Transmitted Diseases: Guide to Diagnosis and Therapy 24 (1985) (risk of contracting gonorrhea from single sexual exposure to infected person probably less than twenty-five per cent). It may be that Dr. Kimberlin’s higher figures were derived from studies involving repeated exposure. See H. Handsfield, Gonorrhea and Uncomplicated Gonococcal Infection, in Sexually Transmitted Diseases 205, 207 n. 11 (K. Holmes, P. Mardh, P. Sparling & P. Wiesner 1984) (risk increases from twenty per cent after single episode of intercourse to between sixty and eighty per cent after four exposures) (citing Hooper, et al.,) Cohort Study of Venereal Disease: The Risk of Gonorrhea Transmission from Infected Women to Men, 108 AmJ.Epidemiol. 136 (1978).

Moreover, evidence of the victim’s infection with gonorrhea has been excluded as either irrelevant, or as relevant but overly prejudicial or inflammatory when compared with its low probative value. E.g., Ervin, 723 S.W.2d at 415; Smith, 737 S.W.2d at 915 (evidence “immaterial”). Other cases *168 have excluded such evidence as irrelevant on the ground that, like the petitioner in this case, the defendant failed to offer proof that he did not contract the disease. State v. Terry, 582 S.W.2d 337, 341 (Mo.App.1979); Milenkovic v. State, 86 Wis.2d 272, 286, 272 N.W.2d 320, 327 (App.1979). We need not decide whether the same result would be proper under the Federal Rules of Evidence, as that issue is not squarely presented to us for review. We do, however, take the cited decisions as some indication of the weakness of the proffered evidence when contrasted with the physical evidence in this case, and there is no guarantee that it would have been admitted into evidence. Quite possibly it would not have been admitted. But if admitted, in the circumstances of this case the proffered evidence would have had little exculpatory value, if any.

Relatedly, Davis argues that Boone, the public health investigator, was not qualified to testify as to the probabilities of transmitting gonorrhea. He points out that Boone was a university graduate with a degree in journalism, had never been to medical school, and based his testimony largely on practical experience as opposed to documented surveys.

The Federal Rules of Evidence allow a witness to give expert testimony relating to “scientific, technical, or other specialized knowledge [if it] will assist the trier of fact to understand the evidence or to determine a fact in issue.” Fed.R.Evid. 702. The witness may qualify as an expert by reason of “knowledge, skill, experience, training, or education....” Id. We have no difficulty in finding that Boone’s qualifications meet these requirements. The decision to permit a witness to testify as an expert lies within the discretion of the trial court, and will not be disturbed on appeal in the absence of an abuse of discretion or clear error of law. Sweet v. United States, 687 F.2d 246, 249 (8th Cir.1982). Rule 702 does not state a preference for academic training over demonstrated practical experience. Circle J. Dairy, Inc. v. A.O. Smith Harvestore Products, 790 F.2d 694, 700 (8th Cir.1986). Here, the magistrate properly stated that she would admit the testimony and give it the weight to which it was entitled. See generally United States v. Gilliss, 645 F.2d 1269, 1278-79 (8th Cir.1981).

Boone testified that he had been investigating and managing cases involving sexually-transmitted diseases for over eight years. He received specialized training “through the Center for Disease Control as well as periodic updates on current information as it changes.” He further testified that he reads current information in his field as time permits. Prior to the evidentiary hearing, he had read a named recognized reference which he stated substantiated his practical experience. Rule 703 of the Federal Rules of Evidence permits an expert to testify to an opinion or inference based on “facts or data ... perceived by him or made known to the expert at or before the hearing. If of a type reasonably relied upon by experts in the particular field ... the facts or data need not be admissible in evidence.” Davis makes no argument that the literature which Boone used to substantiate his testimony did not meet this test. Again, Davis’s challenge in this regard goes to the weight of the evidence rather than to its admissibility. See Cashman v. Allied Products Corp., 761 F.2d 1250, 1255 (8th Cir.1985). Accordingly, we affirm the magistrate’s decision to admit Boone’s testimony.

Finally, Davis objects to the magistrate’s finding that the second mistrial led to a criminal contempt conviction of a juror based on his failure to reveal his bias in favor of Davis. Davis claims that this evidence was irrelevant. The government contends that the evidence is relevant inasmuch as petitioner had argued, apparently in order to demonstrate prejudice as part of his ineffective assistance claim, that he had been “successful” in two trials because the juries were unable to reach a verdict. The government contends it was entitled to rebut this argument, and we agree. Appellant’s only authority for his argument is Federal Rule of Evidence 606(b), which in general prohibits inquiry into the validity *169 of a verdict. However, the rule does permit inquiry into whether “outside influence was improperly brought to bear on any juror.” Apparently this issue was resolved against Davis when the juror was found guilty of contempt, and it is too late in the day to raise it now. Since Davis first broached the issue of the jury’s verdict, we see nothing improper in the government’s attempt to show that the verdict was the result of improper influence rather than a result of weakness in the government’s case.

In conclusion, given the overall strength of the government’s case and the low probative value of the proffered evidence that Hanks had gonorrhea, we find that Davis has not demonstrated the Strickland requirement of a reasonable probability sufficient to undermine confidence in the guilty verdict. Nor do we find error in the admis-. sion of the public health investigator’s testimony or in the introduction of the evidence of juror misconduct. 3 Accordingly, we affirm the denial of appellant’s petition for habeas corpus.

. The Honorable Oren Harris, United States Senior District Judge, Eastern and Western Districts of Arkansas.

. We do not completely agree with the magistrate’s characterization of Dr. Kimberlin’s testimony as stating that the probability is "more like SO [per cent]” for female-to-male transmission as opposed to seventy-five per cent generally. On cross-examination by the United States Attorney, Dr. Kimberlin denied that twenty per cent was an accurate figure, and in response to leading questions agreed that it was "more like fifty [per cent] ... and possibly seventy-five.”

. We further reject appellant’s contention that he was prejudiced by counsel’s failure to call Dr. Kimberlin to testify at trial as to the victim’s demeanor during her medical examination. This information was introduced into evidence by way of the medical report, and defense counsel relied on the report in her argument to the jury.