Richard D. Dodd, convicted in Iowa state court of first degree kidnapping and sentenced to life imprisonment, filed a petition for a writ of habeas corpus pursuant to 28 U.S.C. Sec. 2254. The magistrate judge1 recommended that the petition be denied. Dodd appeals from the district court's2 order adopting the magistrate judge's recommendation and denying the petition. We affirm.

The kidnapping victim in this case was a ten-year-old girl. On July 9, 1984, the victim was spending the night at a friend's house in Des Moines, Iowa. Sometime after 2:00 a.m., she was abducted by a man who had been hiding in a bedroom closet. The abductor took the girl from the home and drove her to several locations, where he twice compelled her to perform oral sex upon him, following the latter of which such acts he engaged in an act of sexual penetration. Upon being released sometime around 4:00 a.m., the victim ran to a nearby home for help.

After the police arrived, the victim gave a full description of her abductor and his vehicle and was then taken to the hospital. Shortly thereafter, the victim went to the Des Moines police station where she looked at several photographic notebooks. She was not able to identify her abductor from any of the photographs. Over the course of the next three weeks, the victim was shown at least three photographic arrays containing five to seven pictures per array, but again she was unable to identify her abductor.

On July 31, 1984, a police detective went to the victim's home. After being shown a photograph of Dodd, she identified him as the man who had abducted her. The parties dispute whether the victim was shown a single photograph of Dodd (her testimony) or whether she was presented with an array containing seven photographs (the detective's testimony).

The victim's out-of-court identification was the state's primary evidence against Dodd at trial. During the trial, Dodd attempted to account for his whereabouts at the time of the abduction. He presented evidence that he had left his mother's house at approximately 2:30 a.m. and that he had picked up his girlfriend, Linda Petersen (whom he subsequently married), from work at approximately 3:30 a.m.

Following his conviction, Dodd filed an application for post-conviction relief, challenging, among other things, the out-of-court identification. Following an evidentiary hearing, the application was denied. In a consolidated appeal, the Iowa Court of Appeals affirmed the conviction and the denial of post-conviction relief. State v. Dodd, 414 N.W.2d 677 (Iowa Ct.App.1987). Dodd then filed the petition giving rise to this appeal.

Dodd first argues that he was denied effective assistance of counsel at trial. To prevail on his ineffective assistance of counsel claims, Dodd must establish that counsel's performance was deficient and that he was prejudiced by the deficient performance. Strickland v. Washington, 466 U.S. 668, 687, 104 S. Ct. 2052, 2064, 80 L. Ed. 2d 674 (1984). Counsel's performance was deficient if it fell "outside the wide range of professionally competent assistance." Id. at 690, 104 S.Ct. at 2066. Counsel's performance was prejudicial if "there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different." Id. at 694, 104 S.Ct. at 2068.

An ineffective assistance of counsel claim presents a mixed question of law and fact; we review the district court's factual findings for clear error and its legal conclusions de novo. Wilkins v. Iowa, 957 F.2d 537, 540 (8th Cir.1992).

Dodd argues that counsel was ineffective for failing to move to suppress evidence of the victim's out-of-court identification because it resulted from an allegedly unnecessarily suggestive showup involving a single photograph of Dodd. Alternatively, he contends that even assuming the victim was shown a photographic array, the identification procedure was nonetheless impermissibly suggestive because 1) the police failed to conduct a lineup; 2) the photographs shown to the victim were presented in a stack rather than in an array; and 3) Dodd's photograph was suggestive.

We will set aside a state court conviction based on an out-of-court identification "only when pre-trial identification procedures were so impermissibly suggestive that they give rise to a very substantial likelihood of irreparable harm." Trevino v. Dahm, 2 F.3d 829, 833 (8th Cir.1993).3 The critical inquiry is whether, under the totality of the circumstances, the out-of-court identification was reliable despite any suggestive or inappropriate pre-trial identification methods. Id. Factors bearing on the reliability of the identification include:

the opportunity of the witness to view the criminal at the time of the crime, the witness' degree of attention, the accuracy of the witness' prior description of the criminal, the level of certainty demonstrated by the witness at the confrontation, and the length of time between the crime and the confrontation.

Neil v. Biggers, 409 U.S. 188, 199-200, 93 S. Ct. 375, 382, 34 L. Ed. 2d 401 (1972).

Applying the Biggers factors to the present case, we conclude that the out-of-court identification did not give rise to a very substantial likelihood of misidentification. The victim had an ample opportunity to view her assailant. She spent a considerable amount of time with him, up to two hours. She testified that she was able to see his face on three separate occasions during the abduction--in the house, in the car, and once outside of the car. "She was no casual observer, but rather the victim of one of the most personally humiliating of all crimes." Id. at 200, 93 S.Ct. at 382-83. The degree of her attention was high. In addition, the victim testified that, after identifying Dodd's photograph, she responded affirmatively when asked by the police detective if she was sure that he was the man who abducted her. Finally, the length of time between the kidnapping and the confrontation (three weeks) was not unreasonably long.

Dodd argues that the identification was unreliable because of inconsistencies in the victim's description of him and his car. The victim described her assailant as a white male, thirty to forty years old, tall and thin, with short and straight black hair, a mustache, and a pushed-in chin. She further described his car as a blue, full-sized, four-door sedan that was rusty and dirty on the outside, with a white interior. Dodd argues that he has a very pronounced chin, that he had a goatee beard and tattoos on his arms and hands which the victim failed to include in her description, and that his car--later identified by the victim as her assailant's vehicle--was a brown, two-door sedan, with a tan interior.

It is true that there were discrepancies between the victim's description of the assailant and Dodd's appearance, and between her description of the assailant's vehicle and Dodd's car. Although these discrepancies negatively impact the out-of-court identification, we do not agree that they render the identification evidence unreliable. These discrepancies must be considered in the light of all the circumstances. Other than the victim's description of Dodd's chin (including the lack of any goatee) and her failure to describe his tattoos, her description of him was largely accurate. Cf. United States ex rel. Bonner v. DeRobertis, 798 F.2d 1062, 1066-67 (7th Cir.1987) (identification reliable notwithstanding victim's failure to notice moustache after having described only a goatee). The description "might not have satisfied Proust but was more than ordinarily thorough." Biggers, 409 U.S. at 200, 93 S.Ct. at 383. Moreover, the victim made no previous identification prior to the time she identified Dodd as the man who abducted her. "Her record for reliability was thus a good one, as she had previously resisted whatever suggestiveness" that surrounded the out-of-court identification. Id. at 201, 93 S.Ct. at 383. Any remaining concerns about the suggestiveness of the identification procedure or the reliability of the out-of-court identification were for the jury to resolve. Graham v. Solem, 728 F.2d 1533, 1548 (8th Cir.), cert. denied, 469 U.S. 842, 105 S. Ct. 148, 83 L. Ed. 2d 86 (1984). Accordingly, we hold that there was no substantial likelihood of misidentification and that, even assuming counsel's failure to move to suppress the out-of-court identification constituted deficient performance, there is no reasonable probability that had counsel objected to the identification, the outcome of Dodd's trial would have been different.

Dodd next argues that counsel was ineffective for failing to call Mike Keeran as a witness at trial. Keeran was a cook at the restaurant where Linda Petersen was working on the morning of the abduction. At trial, Linda testified that Dodd had picked her up from work at 3:30 a.m. Stephanie Schroder, also a co-worker, testified, however, that Dodd did not arrive at the restaurant until 5:30 a.m. and that Linda had complained about his late arrival.

Dodd contends that Keeran could have supported Linda's testimony and contradicted Schroder's testimony. However, at the post-conviction hearing, Keeran could not remember when Linda left the restaurant. He could only state that she left "quite a while" before Schroder and that Schroder "should have" left around 6:00 a.m. We agree with the Iowa Court of Appeals that there is no evidence that Keeran's testimony would have been helpful. Accordingly, we cannot say that counsel was ineffective for failing to call Keeran as a witness.

Dodd also argues that counsel was ineffective for failing to call Linda Webb as a defense witness. He contends that Webb could have testified to the fact that, a few days after the kidnapping, another man fitting the description of the assailant broke into the same house from which the victim was abducted. At the post-conviction hearing, counsel testified that he had interviewed Webb, a fifteen-year old girl, and decided against calling her as a witness because he was fearful of her disclosing that Dodd had allegedly made sexual advances towards her on a prior occasion. The Iowa Court of Appeals found that counsel's decision not to call Webb as a witness was a matter of trial strategy. Absent contrary evidence, we accord this finding of fact a presumption of correctness under 28 U.S.C. Sec. 2254(d), and we decline to second-guess counsel's strategic decision on collateral review. See Wilkins, 957 F.2d at 541.

Dodd argues that he was denied due process because the prosecutor manipulated and intimidated Webb and deterred her from testifying as a defense witness. Dodd further argues that the prosecutor attempted to deter defense counsel from calling Webb as a witness. After defense counsel had interviewed Webb, the prosecutor met with Webb in Webb's home without obtaining the permission of Webb's parents. The prosecutor and a police detective later interviewed Webb a second time. The prosecutor decided against calling Webb as a witness because she did not find Webb to be credible and advised defense counsel of that decision.

"We will reverse for prosecutorial misconduct only if the conduct, even if improper, so prejudiced [Dodd] that he was unable to obtain a fair trial." Carlson v. State, 945 F.2d 1026, 1029 (8th Cir.1991). Although we agree with Dodd that Webb's testimony was potentially relevant, there is no evidence in this record that the prosecutor manipulated or intimated Webb, or otherwise prevented her from testifying. As we have already noted, counsel chose not to call Webb as a witness because of his own interview with her rather than because of any statements that the prosecutor allegedly made. Accordingly, we conclude that the prosecutor's actions, even if improper, did not constitute a denial of due process. See Harrington v. Nix, 983 F.2d 872, 875 (8th Cir.1993) (per curiam) (no denial of due process where prosecution did not prevent witness from testifying).

Dodd next argues that the prosecutor engaged in misconduct by giving the victim presents (a hat and some leg warmers) prior to the post-conviction hearing. Although the prosecutor's action may have been ill-advised (she stated that she and the victim had become friends through the course of the case and that the items of clothing were in the nature of a Christmas present), we cannot say that the conduct so prejudiced Dodd as to deny him a fair hearing.

Finally, Dodd argues that the evidence at trial was insufficient to support his conviction. In support of this claim, he again points to discrepancies in the victim's out-of-court identification.

Our review of this claim is limited to determining "whether, after viewing the evidence in the light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt." Jackson v. Virginia, 443 U.S. 307, 319, 99 S. Ct. 2781, 2789, 61 L. Ed. 2d 560 (1979). After careful review of the record, we conclude that the evidence, although not overwhelming, was sufficient to permit a rational jury to convict Dodd of the crime with which he was charged. Although there was conflicting evidence presented to the jury concerning the victim's description of her abductor, we must presume that the jury resolved any conflicting inferences in favor of the prosecution. Jackson, 443 U.S. at 326, 99 S.Ct. at 2793.

The order denying the petition for a writ of habeas corpus is affirmed.