The 1960 census was the first census mailed to households. The 1950 census had tested self-enumeration on a limited scale in Michigan and Ohio, but 1960 was the debut for this technique as a primary method for the collection of population and residential data. The postal service delivered questionnaires to every occupied housing unit. Householders were asked to complete the questionnaire and hold it for an enumerator to pick up.
Enumeration efforts were divided into two stages:
- Stage 1 concentrated on the quick collection of a few data items for every person and dwelling unit- information available from responses to questions on the questionnaire that had been delivered to every household.
- Stage 2 focused on the collection of more detailed economic and social information from a sample of households and dwelling units. Second stage questionnaires were hand-delivered by enumerators when they came to pick up the first form. Households receiving the second questionnaire were asked to complete the form and mail it to their local census office in postage-paid envelopes provided by enumerators. In areas of low population density, the two-staged enumeration was combined to allow enumerators to collect and record sample data at the same time they came to fill out the general questionnaire.
In the years just before 1960, the Census Bureau gave considerable thought and effort to making census questions and instructions easier to understand, providing better training to enumerators, and obtaining more accurate results from the population. Test censuses (“pretests”) were conducted in numerous locations. The test schedules were not retained, but “lessons learned” from those activities were kept in the record series, “Eighteenth Decennial Census (1960) Background Materials” (National Archives Identifier 5634068), which NARA is currently digitizing.
The census depends on people as all human activities do. The census schedules from 1790 to the present day contain plenty of erroneous information, both by mistake and on purpose (such as that discussed in "Census Fun Fact #2 - Fictional Names: Just Call Me Another Time.") We can learn much from “observer reports” of practice censuses ("pretests") about how unintended error creeps into census results.
In 1957, a special census of Yonkers, New York, provided a good opportunity for a test census. In Harlem, New York, on May 19, 1957, Frank S. Kristof observed an unnamed enumerator working a 15-unit tenement on 144th Street west of Lenox Avenue in Harlem. The enumerator was a middle-aged woman who had previously been a New York City Welfare Department investigator. Although the people were wary of opening their doors to strangers, she had no trouble dispelling suspicion. However, it appeared likely that the number of nonfamily residents (lodgers) in each unit was underreported (by people who didn’t want the landlord or anyone else to know there were extra people living there). The person answering questions for the household was often unsure how old their fellow household members were.
The citizenship question “was variously worded by the enumerator and some rather dubious responses obtained,” said Mr. Daugherty. “Many people do not understand the meaning of “naturalized.” This indicates a need for some guidance on the question.”
The woman was a careful enumerator but usually failed to mark the space for “dilapidated” or “not dilapidated” housing. Mr. Daugherty concluded the word “dilapidated” didn’t mean very much to her. In addition, based on his observation of the apartments, he thought that inaccurate answers about the number of rooms in the unit was given about half the time. (Frank S. Kristof, Assistant Chief, Housing Division, to Wayne F. Daugherty, Chief, Housing Division, March 26, 1957, frames 19-20, Roll 17, “Eighteenth Decennial Census (1960) Background Materials, images 21-22 at 22 at https://catalog.archives.gov/id/311023476?objectPage=21).
Howard G. Brunsman accompanied three unnamed enumerators somewhere in Yonkers on March 20, 1957. He was impressed by their enthusiasm but they, too, had problems with the citizenship question, getting correct ages, the concept of dilapidation, and other issues. (Howard G. Brunsman to Conrad Taeuber, March 22, 1957, frames 21-23, Roll 17, “Eighteenth Decennial Census (1960) Background Materials, images 23-25 at https://catalog.archives.gov/id/311023476?objectPage=23).
Ivan G. Munro observed training for timekeepers, enumerators, and crew leaders in Yonkers on March 18-20, 1957. Enumerator training was inadequate. “The period of approximately four hours left much unlearned (untaught), particularly in the areas of housing,” he said, and “With reference to dilapidation, the enumerator seemed to define the term in accordance with the degree of housekeeping inside.” (Wow!) 
The enumerator failed to ask Irish-born respondents whether they had come from Northern Ireland or the Irish Free State. That requirement had been taught, but there was no reminder of that requirement on the census schedule itself.
Crew leader training (75 minutes) was also inadequate. Munro stated, “We again concentrated too much material in too little time, that we overestimated the rate of absorption of a group of lay recruits whose knowledge yesterday of the Bureau of the Census in most cases approached zero.” He noted that common census terms like enumerator, enumeration district (ED), ED boundaries, and so forth, were “like another language to non-Census people” and “even maps may be puzzling.” (Ivan G. Munro to Robert B. Voight, March 27, 1957, frames 24-28, Roll 17, “Eighteenth Decennial Census (1960) Background Materials, images 26-30 at https://catalog.archives.gov/id/311023476?objectPage=26).
Wayne F. Daugherty observed enumerators Mrs. Ellen Brown in Harlem and Peter Fetzko in Yonkers on March 22, 1957. Two people were not enumerated because they told Mrs. Brown that “he or she did not live in the unit but was only visiting … [although] It was apparent in both cases that these persons either had no other place to live or did not want to be recorded as living in the particular unit.”
Both Mrs. Brown and Mr. Fetzko assumed foreign-born residents were naturalized without asking them! This was a mistake since a foreign-born person could live in the United States their entire life without becoming a naturalized U.S. citizen. (Wayne F. Daugherty, Chief, Housing Division, to Conrad Taeuber, Assistant Director, March 28, 1957, frames 29-30, Roll 17, “Eighteenth Decennial Census (1960) Background Materials, images 31-32 at https://catalog.archives.gov/id/311023476?objectPage=31).
These examples are just a few of the observations made by Census Bureau personnel in 1957 during preparation and planning for taking the 1960 census. Hopefully, the forms, instructions, and training used for the actual 1960 census were improved!
However, the problems noted in 1957 were not unique to the not unique to the 1957 Yonkers census pretest. Every enumerator since 1790 faced difficulties in asking questions and getting and recording accurate answers. Being an enumerator was a demanding short-term job.
Image: Example of "dilapidated" from Binder 29-B, Enumerator's Workbook (Short Form) (October 1949), Seventeenth Decennial Census (1950) Reference Materials. National Archives Identifier 205683228).