VACATED FOLLOWING SETTLEMENT

 In the Matter of the Complaint Against

 UNITED COMMERCIAL INSURANCE SERVICES,
 a corporation d/b/a, CHECKWRITER INSURANCE,
 CHECKWRITER INSURANCE AGENCY, and
 CHECKWRITER SECURITY AGENCY,

 and

 BURTON RAYDEN and
 JEFFREY RAYDEN,
 15250 Ventura Boulevard, at
 Sherman Oaks, CA 91403-3201

 P.S. Docket No. 19/18

 04/25/85

 Grant, Quentin E.  Administrative Law Judge

 APPEARANCES FOR COMPLAINANT:
 Thomas A. Ziebarth, Esq. 
 James A. Harbin, Esq. 
 Ben A. Kilgrow, Esq. 
 Consumer Protection Division 
 Law Department U. S. Postal Service
 Washington, DC 20260-1112 

 APPEARANCES FOR RESPONDENT:
 John W. Vardaman, Jr., Esq. 
 F. Lane Heard, III, Esq. 
 Williams & Connolly 839-17th Street, N.W. 
 Washington, DC 20006-3942

In a Complaint filed on March 16, 1984, Complainant alleged that United Commercial Insurance Services, (UCIS) a California corporation, and Burton Rayden and Jeffrey Rayden, individually and as officers of UCIS, are engaged in a scheme to obtain money or property through the mail by means of false representations in violation of 39 U.S.C. 3005. It is alleged that UCIS, by means of direct mail solicitations (attached as exhibits to the complaint), seeks the remittance of money through the mail to the trade names and addresses appearing on such exhibits. The two Raydens are alleged to be corporate officers of UCIS responsible for formulating, directing, and controlling the acts and practices of UCIS. The complaint alleges further that Respondents' direct mail solicitations "are in the form of and reasonably could be construed as bills, invoices, or statements of account due," but did not contain the notice required by 39 U.S.C. 3001(d) or by the regulations promulgated thereunder. Paragraph 7 of the complaint alleges that these solicitations make the following materially false representations:

a. The amount set forth on the face of the solicitation is owing and due to Respondents;

b. The recipients of Respondents' solicitations previously purchased products or services from Respondents or have otherwise established a business relationship with Respondents; and

c. The recipients of Respondents' solicitations have agreed to purchase the product or services described therein prior to receipt of the solicitations.

The Complainant requests the issuance of the false representation orders and cease and desist order attached to the complaint.

Respondents' answer to the Complaint admits that Burton Rayden and Jeffrey Rayden are officers of UCIS responsible for formulating, directing, and controlling its acts and practices, that the solicitations attached to the complaint are copies of portions of solicitations employed by Respondents, and that such direct mail solicitations are used to solicit remittances through the mail to the tradenames and addresses specified in paragraph 1 of the Complaint. They deny that their solicitations are in the form of, and reasonably can be construed as, invoices or that they make the representations set forth in Paragraph 7 of the Complaint. They also deny that such representations are materially false. Lastly, an affirmative defense of estoppel is asserted.

A hearing was held in Washington, DC on December 12, 13 and 14, 1984. Dr. Paul Scipione, an expert in surveys and consumer psychology, testified for Complainant. Complainant's other witnesses were Postal Inspector Larry B. Johnson who conducted the investigation leading to the initiation of this proceeding, and the following lay witnesses: Mr. Dana Miller, Mr. Raymond Watson, Ms. Ruth Nassi, Ms. Cynthia Rogers, Ms. Barbara Johns, Mr. Peggy Shivers, Mr. Quinten Burt, Ms. Miriam Lalum, Ms. Katie Martz, and Ms. Phyllis Shannon.

Respondents called Dr. John Parkington, an expert in direct mail marketing and survey research, and Jeffrey Rayden.

The parties have filed proposed findings of fact, conclusions of law, and written argument all of which have been considered in reaching this decision. To the extent indicated they have been adopted. Otherwise they have been rejected as irrelevant, immaterial, or contrary to the evidence.

Preliminary Ruling on Admissibility of SGI Study

At the hearing, ruling was reserved on Complainant's offer of a study made by Response Analysis Corporation (RAC) of Princeton, N.J. through the testimony of Paul Scipione, Ph.D., a vice-president of the corporation and an expert in conducting surveys and in consumer psychology (Tr. 137-143; CX-8a, b). The study is entitled, "Solicitations in the Guise of Invoices: The Experience, Attitudes and Perceptions of U. S. Businesspersons" (CX-10, 11, and 12 for ident., hereinafter sometimes referred to as the SGI study or the study).

The study was offered to prove that Respondents' solicitations reasonably can be interpreted or construed as bills or invoices. The study was made by RAC at the request of USPS for design, conduct, and report on a generic business-to-business solicitation-in-the-guise-of-invoice survey (CX-10, p. 2).

The report of the study is based on hearsay. The parties agree, however, that its receipt as an exception to the hearsay rule should be controlled by the principles governing the admissibility of surveys set forth in The Manual for Complex Litigation (1 Pt. 2 Moore's Fed. Practice 2.71) which emphasizes that inquiry should center on circumstantial guarantees of trustworthiness and that where, as here, the proffered survey records subjective data (i.e., impressions of interviewees based on the appearance of solicitations and bills) the question of trustworthiness should be more closely scrutinized because it may well be less (idem. p. 137, 138).

Respondents object to the admission of the survey on the ground that it fails to comply in the following respects with the requirements set forth in the Manual:

a. The survey was not based on a representative sample.

b. It examined the wrong universe.

c. It is incomplete and, therefore, inaccurate.

Respondents raise the further objection that the survey is infected with bias in that it violated the requirement that both interviewers and participants be unaware of the purpose of the survey (Pittsburgh Press Club v. United States, 579 F.2d 751, 758 (3d Cir. 1978).

After considering the many arguments of the parties as to the admissibility of the survey, I conclude that it should not be received for the following reasons:

a. It was not intended to, and does not, provide a definitive survey measurement for use in this case (Scipione, Tr. 200). It was meant to be, and is, more

qualitative than quantitive in nature, having been designed to shed light on the solicitation-in-guise- of-invoice phenomenon generically (CX-10 id., p. 3);

b. The survey failed, in part, to examine the universe most relevant to this proceeding in that many of the 100 persons interviewed did not have authority in their business capacities to make the decision whether to respond to a solicitation or pay a bill. (Tr. 432, 433) Approximately 90% of the interviewees would not, in their employment, have acted solely on the basis of their own impressions of the items used in the survey (which included one of Respondents' solicitations no longer in use) (Tr. 580-582). This fact casts substantial doubt on the survey's compliance with that part of its design which called for interviewing the person in each sample firm most responsible for accounts payable (CX-10 id., p. 3).

c. Doubt is cast on the accuracy of the survey results by the fact that comments of several interviewees omitted from the report show that while their first impression of the solicitation was recorded as "definitely a bill" or "probably a bill" their considered perception of it was not as a bill in the sense of a statement of a preexisting obligation to pay (Tr. 515-525). Also there are instances of probable erroneous coding of individual perceptions (Tr. 526-529).

d. The survey was exposed to serious bias in that RAC's instructions to supervisors and interviewers told them that the survey was being conducted for USPS to examine issues concerning fraudulent mail pieces, specifically solicitations made to look like invoices, for use in connection with several upcoming mail fraud trials (RX-24).

e. There was bias in the choices given interviewees in that each choice used the word "bill" with no option to find an item to be an ad or a solicitation (Tr. 436).

f. The study was designed to measure attitudes, not behavior. It was not intended to, and did not, produce statistics as to how many recipients of solicitations would actually pay them in the belief that they were invoices (Tr. 579, 580). The reported results reflecting attitude have not been shown to predict behavior (actual payment) (Tr. 187-189).

g. One of the stated objectives of the study was that it be conducted in such a way that results could be