Defendant was found guilty of kidnapping under 18 U.S.C. § 1201. The victim of the alleged kidnapping testified that on November 16, 1972, while walking near her school dormitory in Kansas City, Missouri, at approximately 9:30 p. m., the defendant forced her into a car and drove to a secluded place where he raped her. The FBI confirmed that fibers similar to those within the carpet in the defendant’s car were found on the girl’s underclothing.
The sufficiency of the evidence to sustain the conviction is not challenged on appeal. The sole error alleged is the exclusion by the district court of testimony of three witnesses offered by the defendant Oliver to show the prosecuting witness’ poor reputation for truth and veracity.1
The defense rested its entire case on the credibility of the prosecutrix. She testified that on the evening in question, she left the dormitory, after school curfew hours, to go to a nearby shopping area. She related that the defendant stopped his car at a lighted intersection in a residential area and asked her for directions to a restaurant. While she was pointing toward Main Street, she said the defendant got out of the car and forced her into the car on the driver’s side. On cross-examination she denied that she had gotten into the car voluntarily. She originally said that the time interval which elapsed from when he first asked her directions to when he forced her into the car was three minutes. On redirect examination she changed this to six seconds. She admitted visiting with the defendant in his car about her family, the fact that she had lived with her father, an Air Force *945officer, overseas, and that she was thinking of getting a job at the telephone company. She said she did this to calm the defendant. Defendant’s counsel asked her about other details of conversation relating to her going back to Nebraska to get her car, quitting school, moving out of the dormitory, etc., but she denied she had told him those things. She described how she was abducted when the defendant forcibly grabbed her with one hand behind the neck and pushed her into the car. She admitted she had originally told the FBI that he had grabbed her by both arms. She testified that when the defendant grabbed her she did not say anything to him and was too afraid to scream. She admitted she had told the FBI that she had screamed at the defendant and asked to be let go. She said the defendant at all times had held her firmly around the neck and drove her a long distance to the secluded area where she was raped. The path the car traveled was through busy intersections and streets with stop lights, but she said the defendant drove recklessly and never stopped his car at any time. After the car stopped she said defendant continued to hold her forcibly around the neck with one hand, pulled her blue jeans and pants just below her knees and raped her. She said she struggled throughout. After the rape, she testified that on the way back she got out of the car when defendant stopped at a stop sign and she went to a private home. A Mr. Hartley came to the door. She first called her boy friend and then Mr. Hartley called the police. Hartley did not testify, but his wife, who was upstairs when the girl came to the door, testified. She said she heard the girl crying, saying she had been raped. She said the girl was composed and not hysterical. The police came and took her to the University of Kansas Medical Center. Dr. Brandwine, a resident in obstetrics and gynecology, examined her. He found two small red rings on both arms, but no evidence of any bruises on any part of her body. There was no trace of blood nor any evidence of trauma on her lower pelvic area. There was no evidence of marks or trauma about the neck area. She had no torn clothing.
The only witnesses offered by the defense were the complaining witness’ former roommates. Two of the three girls still lived with her on November 16, 1972, the date of the alleged rape. None of the witnesses had known the complaining witness prior to the start of school in September, 1972. Defense counsel, out of the presence of the jury, conducted a voir dire examination of the three witnesses. After the offer of proof, the trial court ruled that none of the witnesses was qualified to testify as to the prosecutrix’ reputation for truth and veracity in the community for the following reasons:
(1) The reputation about which they sought to testify was not in the community in which she lived, but rather was her reputation in the dormitory.
(2) The witnesses had only known her for approximately seven weeks.
(3) The testimony regarding community reputation consisted in large part of the personal opinions of the witnesses.
The trial judge allowed them to testify only as to the prosecutrix’ tendency to bruise easily. They each testified that the prosecutrix had previously told them that she bruised easily. She had earlier denied this. The girls had previously seen black and blue marks on her when she had had minor bumps. She had told them she took vitamins plus iron because of her condition.
We find the court erred in excluding their testimony as to the prosecutrix’ reputation for truth and veracity and require a new trial.
The rationale for limiting reputation testimony to the “community in which he lives” is to restrict evidence of repute to reputation among the people who know the person best. McCormick, Evidence § 44 at 92 (2d ed. 1972).
*946However, this rule must be applied in a realistic and practical manner. Professor McCormick observes:
But as an exclusive limitation it would not be appropriate in this country today, where a person may be little known in the suburb or city neighborhood where he lives, but well known in another locality where he spends his working days or in several localities where he does business from time to time. Thus, today it is generally agreed that proof may be made not only of the reputation of the witness where he lives, but also of his repute, as long as it is “general” and established, in any substantial community of people among whom he is well known, such as the group with whom he works, does business or goes to school.
McCormick, Evidence § 44 at 92-93 (2d ed. 1972) (emphasis added).
Professor Wigmore echoes this understanding :
But in the conditions of life today, especially in large cities, a man may have one reputation in the suburb of his residence and another in the office or the factory at his place of work; or he may have one reputation in his place of technical domicile in New York and another in the region of the mines of Michigan or the steel-mills of Ohio where his investments call him for supervision for portions of time. There may be distinct circles of persons, each circle having no relation to the other, and yet each having a reputation based on constant and intimate personal observation of the man.
There is every reason why the law should recognize this. Time has produced new conditions for reputations. The traditional requirement about “neighborhood” reputation was appropriate to the conditions of the time; but it should not be taken as imposing arbitrary limitations not appropriate in other times. “Alia témpora, alii mores.” What the law, then as now, desired was a trustworthy reputation; if that is to be found among a circle of persons other than the circle of dwellers about a sleeping-place, it should be received.
5 Wigmore, Evidence § 1616 at 488 (3d ed. 1940) (emphasis added).
In accordance with this reasoning, courts have readily extended the concept of community to include the community in which one works, as well as where one lives. See, e. g., United States v. Parker, 447 F.2d 826 (7th Cir. 1971); State v. Jackson, 373 S.W.2d 4 (Mo. 1963); People v. Cobb, 45 Cal.2d 158, 287 P.2d 752 (1955) (en banc); People v. Kronk, 326 Mich. 744, 40 N.W.2d 788 (1950). Cf. United States v. Salazar, 425 F.2d 1284 (9th Cir. 1970).
In People v. Colantone, 243 N.Y. 134, 152 N.E. 700 (1926), the New York Court of Appeals held that an instructor and other students at defendant's school should have been allowed to testify as to his reputation. The court noted:
The only place where he can make a reputation, good or bad, is among his associates in his particular activities or in the personal contacts of his life where he actually lives it. * * *
We think that the Appellate Division was right in ruling that to confine the evidence of the defendant’s reputation in this case merely to the place where he resided was too limited an application of the rule of good character evidence, and that the testimony offered should have been received.
Id. at 702.
We conclude that those persons who have had daily contact with the complaining witness, even though it was for a short period of two months, are competent to testify as to her general reputation at the time.2 The fact that *947the witnesses had known her for a short period of time does not by itself render their testimony inadmissible. This factor goes to the weight of the evidence and is for the jury to evaluate. Cases finding character witnesses incompetent to testify relate to lack of foundation for reasons other than short periods of acquaintance.3
The third reason the district court gave for exclusion of the testimony of the three character witnesses related to its concern that the three girls were giving their personal opinions as to the truth and veracity of the complainant rather than testimony reflecting her community reputation.
We agree that much of the voir dire examination, conducted outside the presence of the jury, reflected improper interrogation and answer. Defense counsel, undoubtedly straining to lay foundation, went beyond his intended proof. Specific incidents may only be inquired into on cross-examination of a witness testifying as to character for truthfulness. See Proposed Rules of Evidence, Rule 608(b), 51 F.R.D. 315, 389 (1971). Nevertheless, we think it manifest that the defendant made clear the purpose of calling the witnesses and that a sufficient foundation was laid for them to testify as to the complainant’s reputation for truth and veracity in the community in which she lived. Once foundation was demonstrated, the court should have made clear to both counsel and the witnesses the limited extent to which the questions and answers would be restricted on direct examination.
We recognize the wide latitude generally accorded the trial judge in determining the admissibility of evidence. However, as we have set forth, these witnesses were qualified to testify as to the prosecutrix’ reputation for truth and veracity. This was not a question of relevancy resting in the trial court’s discretion. The evidence was rejected here because the trial court erred in its holding as to what constituted a proper foundation. The totality of the defense offered here rested on challenging the credibility of the complaining witness. The jury deliberated over ten hours before rendering the guilty verdict. Questions of credibility are peculiarly within the province of the jury. This being so, it is critical that all material evidence relating to truth and veracity be admitted for the jury’s consideration.
The issue before us is whether the exclusion of such testimony amounted to prejudicial error. See Michelson v. United States, 335 U.S. 469, 480, 69 S.Ct. 213, 93 L.Ed. 168 (1948); Panger v. Duluth, W. & P. Ry., 490 F.2d 1112 (8th Cir. 1974). We find it did. On the record before us, it is impossible to say that there is no “reasonable possibility” that the exclusion of the evidence might not have contributed to the conviction. See Kotteakos v. United States, 328 U.S. 750, 765, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946); Fahy v. Connecticut, 375 U.S. 85, 86-87, 84 S.Ct. 229, 11 L. Ed.2d 171 (1963). On the record, the evidence of guilt as to the kidnapping was not so “overwhelming” as to render harmless the exclusion of the testimony. See Harrington v. California, 395 U.S. 250, 254, 89 S.Ct. 1726, 23 L.Ed.2d 284 (1969).
As noted in Kotteakos v. United States, 328 U.S. 750, 66 S.Ct. 1239 (1946), the fact that this is a criminal case is significant:
[T]he appellate court [cannot] escape altogether taking account of the outcome. ... In criminal causes that outcome is conviction. This is *948.different, or may be, from guilt in fact. It is guilt in law, established by the judgment of laymen. And the question is, not were they right in their judgment, regardless of the error or its effect upon the verdict. It is rather what effect the error had or reasonably may be taken to have had upon the jury’s decision.
Kotteakos, supra at 764, 66 S.Ct. at 1247.
Recently this court affirmed its adherence to the Kotteakos standard of determining prejudicial error:
“If, when all is said and done, the conviction is sure that the error did not influence the jury, or had but very slight effect, the verdict and the judgment should stand * * *. But if one cannot say, with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed by the error, it is impossible to conclude that substantial rights were not affected. The inquiry cannot be merely whether there was enough to support the result, apart from the phase affected by the error. It is rather, even so, whether the error itself had substantial influence. If so, or if one is left in grave doubt, the conviction cannot stand.”
United States v. Stabler, 490 F.2d 345 (8th Cir. 1974), quoting Kotteakos v. United States, 328 U.S. 750, 765, 66 S.Ct. 1239, 1248 (1946).
Applying this standard to the facts and circumstances of the present case, we cannot say that the exclusion of the testimony of the complaining witness’ three roommates as to her reputation for truth and veracity was harmless. We reverse and remand for a new trial.
. Defendant alleges additional error in the district court’s refusal to allow him to submit voluntarily to a polygraph examination at his own expense. Defendant’s counsel urges that he was denied the opportunity to lay the predicate for an attempt to convince the district court that the status of the art and science of polygraph examinations had progressed to the point where results of such examinations should be admitted into evidence. In view of our reversal on other grounds it is unnecessary to pass on the question presented. Presumably counsel can renew his motion on remand before the second trial. We simply note that no prejudice to the government can flow from allowing the defendant the opportunity to at least attempt to lay a foundation for the admissibility of a polygraph test at trial. Although this court has in the past held such tests inadmissible, we note recent court decisions have found under certain circumstances a polygraph examination may be admissible. See, e. g., United States v. Ridling, 350 F.Supp. 90 (E.D.Mich.1972); United States v. Zeiger, 350 F.Supp. 685 (D.D.C.1972); cf. United States v. Wainwright, 413 F.2d 796 (10th Cir. 1969), cert. denied, 396 U.S. 1009, 90 S.Ct. 566, 24 L.Ed.2d 501 (1970).
. Uniform Rules of Evidence, Rule 63(28) refers to “the community in which he then resided or in a group with which he then habitually associated.” Proposed Federal *947Rules of Evidence, Rule 803(2.1), 51 F.R.D. 315, 422 (1971), admits evidence of reputation “among his associates or in the community.”
. For example, in United States v. Trollinger, 415 F.2d 527 (5th Cir. 1969), the charaeter witness had known the defendant one month, but had not lived in either of the cities in which the defendant had recently resided and to which inquiry was directed. See also United States v. Straughan, 453 F.2d 422 (8th Cir. 1972); United States v. Salazar, 425 F.2d 1284 (9th Cir. 1970).