a similar view is held by the public at large. Plaintiff did not rest upon such an inference, but offered empirical evidence in the form of a survey conducted by the Becker Research Corporation of Boston, Massachusetts, concluding that New England residents associate the name “Colby College” with plaintiff. Becker interviewed by telephone 1500 New England residents, 500 from Maine, 500 from New Hampshire, and 500 from metropolitan Boston, chosen as a representative cross-section of the population in these areas. Seventy-nine per cent of the Maine respondents, 40% of the Boston respondents, and 37% of the New Hampshire respondents identified “Colby College” as located in Maine, where plaintiff is located, while only 2%, 7%, and 18% respectively placed “Colby College” in New Hampshire, defendant’s location. Moreover, 72%, 57%, and 45% identified “Colby College” as a four-year institution like plaintiff, while only 4%, 9%, and 15% identified “Colby College” as a two-year institution like defendant. And 81%, 53%, and 48% characterized “Colby College” as coeducational, as is plaintiff, while only 5%, 13%, and 19% believed that “Colby College” primarily educates students of only one sex, as does defendant. In sum, an average of well over 50% of all respondents identified “Colby College” as a four-year, coeducational institution, located in Maine, thus distinguishing plaintiff from defendant according to three major characteristics by which the two institutions may be distinguished.
The importance of qualified survey evidence in establishing secondary meaning is well recognized. See, e. g., Holiday Inns, Inc. v. Holiday Out in America, 5 Cir., 1973, 481 F.2d 445, 447; Standard Oil Co. v. Standard Oil Co., 10 Cir., 1958, 252 F.2d 65, 75; Zippo Mfg. Co. v. Rogers Imports, Inc., S.D.N.Y., 1963, 216 F.Supp. 670, 682-686. Nor can it be doubted that the percentage of respondents associating plaintiff with its name is sufficiently “appreciable,” Food Fair, ante, 177 F.2d at 185, or “significant,” Carling Brewing Co. v. Philip Morris, Inc., N.D.Ga., 1967, 277 F.Supp. 326, 332, to indicate secondary meaning. See Zippo, ante, 216 F.Supp. at 687 (42.6%).
Becker was an experienced surveyor. The court, however, rejected his findings, not because defendant offered contrary evidence, but largely because of criticisms of the survey offered by defendant’s expert, a business school professor who had done work in market research.' The witness voiced two complaints, one of which was that he found one of the posed questions ambiguous.5 The second was that the witness, who taught at Dartmouth, found it incredible that among the 500 respondents in the Boston area there were about as many who knew of Bowdoin and Colby Colleges as knew about Dartmouth (although they did not give them as high a reputation). Finally, the district court faulted the survey for failing to indicate “the confidence level of [its] validity .,” in light of its attempted extrapolation from 1500 respondents to at least two million New England residents.6 (374 F.Supp. 1145). This, in effect, questioned the surveyor’s basic expertise. While, of course, expert testimony need not be accepted, we find little reason for rejection.7 In any event, the court’s doubts do not remedy the
5
Accepting this criticism, the court felt that the survey’s respondents might have been confused as to the true character of defendant’s institution, and thus might have intended to identify defendant as “Colby College.” We have difficulty in following the reasoning.
7
Defendant’s counsel seeks to come to the court’s aid by saying that the Maine, Boston and New Hampshire target regions do not comprise a large enough area; that the survey excludes those “who have the good sense not to use the telephone,” and that the interviewers were inexperienced; criticisms not voiced even by defendant’s own witness.