Year-round vacant units are those intended for occupancy at any time of the year. Seasonal vacant units are those intended for occupancy during only a season of the year.

Available vacant units are those which are on the market for year-round occupancy, are in either sound or deteriorating condition, and are offered for rent or for sale. The group "for sale only" is limited to available units for sale only and excludes units "for rent or sale." The group "for rent" consists of units offered "for rent" and those offered "for rent or sale." The 1960 category "available vacant" is comparable with the 1950 category "vacant nonseasonal not dilapidated, for rent or sale."

Other vacant units comprise the remaining vacant housing units. They include dilapidated units, seasonal units, units rented or sold and awaiting occupancy, units held for occasional use, and units held off the market for other reasons. This category is comparable with the 1950 category "other vacant and nonresident."

Tenure. A housing unit is "owner occupied" if the owner or co-owner lives in the unit, even if it is mortgaged or not fully paid for. All other occupied units are classified as "renter occupied," whether or not cash rent is paid. Examples of units for which no cash rent is paid include units occupied in exchange for services rendered, units owned by relatives and occupied without payment of rent, and units occupied by sharecroppers.

Color and ethnic group.-Occupied housing units are classified by the color or ethnic origin of the head of the household. (See sections above on "Race and color" and "Persons of Spanish surname and Puerto Ricans.")

Persons. All persons enumerated in the 1960 Census of Population as members of the household were counted in determining the number of persons who occupied the housing unit. These persons include any lodgers, foster children, wards, and resident employees who shared the living quarters of the household head.

In the computation of the median number of persons, a continuous distribution was assumed, with the whole number of persons as the midpoint of the class interval. For example, when the median was in the 3-person group, the lower and upper limits were assumed to be 2.5 and 3.5 persons, respectively. The median may be based on a sample or on the complete count of units (see table A). Persons per room.-The number of persons per room was computed for each occupied housing unit by dividing the number of persons by the number of rooms in the unit.

Year moved into unit.-Data on year moved into unit are based on the information reported for the head of the household. The question relates to the year of latest move.

Units in structure.-A structure is defined as a separate building that either has open space on all four sides, or is separated from other structures by dividing walls that extend from ground to roof.

Statistics are presented in terms of the number of housing units rather than the number of residential structures. However, the number of structures for the first two categories may be derived. For 1-unit structures (which include trailers), the number of housing units and the number of structures are the same. For 2unit structures, the number of housing units is twice the number of structures. For the remaining categories, the number of structures cannot be derived from the data as tabulated.

The categories for number of housing units in the structure in 1960 are not directly comparable with those in 1950, particularly for 1- and 2-unit structures. In the 1950 tract report, units in detached or attached structures werc shown separately but those in semidetached structures containing 1 or 2 units were combined into one category. Comparability between 1950 and 1960 data may also be affected by the change in concept from dwelling unit to housing unit.

Year structure built.—“Year built” refers to the date the original construction of the structure was completed, not to any later remodeling, addition, or conversion.

The figures on the number of units built during a given period

to the inventory during that period. The figures represent the number of units constructed during a given period plus the number created by conversions in structures originally built during that period, minus the number lost in structures built during the period. Losses occur through demolition, fire, flood, disaster, and conversion to nonresidential use or to fewer housing units.

Basement.-Statistics on basements are presented in terms of the number of housing units located in structures built with a basement, on a concrete slab, or in another way. "Basement" is defined as enclosed space accessible to the occupants beneath all or part of a structure and of sufficient depth so that an adult can walk upright. Units built on a "concrete slab" have no basement and no crawl space or air space below the first floor. The category "other" includes units in structures resting on a foundation of posts, those built directly on the ground, and those having only crawl space under the building to provide ventilation or accessibility for service or repairs.

Rooms. The number of rooms is the count of whole rooms used for living purposes, such as living rooms, dining rooms, bedrooms, kitchens, finished attic or basement rooms, recreation rooms, lodgers' rooms, and rooms used for offices by a person living in the unit. Not counted as rooms are bathrooms; halls, foyers, or vestibules; closets; alcoves; pantries; strip or pullman kitchens; laundry or furnace rooms; unfinished attics, basements, and other space used for storage.

In the computation of the median number of rooms, a continuous distribution was assumed, with the whole number of rooms as the midpoint of the class interval. For example, when the median was in the 3-room group, the lower and upper limits were assumed to be 2.5 and 3.5 rooms, respectively. The median was computed on the basis of the tabulation groups shown in the table. If the median falls in the category "8 rooms or more," it is shown in the table as "7.5+" rooms.

Condition and plumbing.-Data are presented on condition and plumbing facilities in combination. The categories represent various levels of housing quality.

The enumerator determined the condition of the housing unit by observation, on the basis of specified criteria. Nevertheless, the application of these criteria involved some judgment on the part of the individual enumerator. The training program for enumerators was designed to minimize differences in judgment.

Sound housing is defined as that which has no defects, or only slight defects which are normally corrected during the course of regular maintenance.

Deteriorating housing needs more repair than would be provided in the course of regular maintenance. It has one or more defects of an intermediate nature that must be corrected if the unit is to continue to provide safe and adequate shelter.

Dilapidated housing does not provide safe and adequate shelter. It has one or more critical defects, or has a combination of intermediate defects in sufficient number to require extensive repair or rebuilding, or is of inadequate original construction. Critical defects result from continued neglect or lack of repair or indicate serious damage to the structure.

In 1950, the enumerator classified each unit in one of two categories, not dilapidated or dilapidated, as compared with the three categories of sound, deteriorating, and dilapidated in 1960. Although the definition of "dilapidated" was the same in 1960 as in 1950, it is possible that the change in the categories introduced an element of difference between the 1960 and 1950 statistics. The category "with all plumbing facilities" consists of units which have hot and cold water inside the structure, and flush toilet and bathtub (or shower) inside the structure for the exclusive use of the occupants of the unit. Equipment is for exclusive use when it is used only by the persons in one housing unit, including any lodgers living in the unit.

Units "lacking only hot water" have all the facilities except

may not have hot water but lack one or more of the other specified facilities. Also included in this category are units whose occupants share toilet or bathing facilities with the occupants of another housing unit. The combination of "lacking only hot water" and "lacking other plumbing facilities" is presented as "lacking some or all facilities" in some census reports.

The categories of plumbing facilities presented in the 1960 report are not entirely comparable with those in the 1950 report. However, the 1950 category "no private bath or dilapidated" is equivalent to the following 1960 categories: "Dilapidated," "sound, lacking other plumbing facilities," and "deteriorating, lacking other plumbing facilities."

Bathroom. A housing unit is classified as having a bathroom if it has a flush toilet and bathtub (or shower) for the exclusive use of the occupants of the unit and also has hot water. The facilities must be located inside the structure but need not be in the same room. Units which have an additional toilet or bathtub (or shower) for exclusive use are classified as having "more than one bathroom."

Heating equipment.-The main type of heating equipment was to be reported even if it was temporarily out of order at the time of enumeration. For vacant units from which the heating equipment had been removed, the equipment used by the last occupants was to be reported.

"Steam or hot water" refers to a central heating system in which heat from steam or hot water is delivered through radiators or heating coils. "Warm air furnace" refers to a central system which provides warm air through ducts leading to the various rooms. "Built-in room units" are permanently installed heating units in floors, walls, or ceilings. They include floor, wall, or pipeless furnaces as well as built-in electrical units. Floor, wall, and pipeless furnaces deliver warm air to immediately adjacent rooms but do not have ducts leading to other rooms. "Other means-with flue" describes stoves, radiant gas heaters, fireplaces, and the like connected to a chimney or flue which carries off the smoke or fumes. "Other means-without flue" describes portable or plug-in devices not connected to a chimney or flue, such as electric heaters, electric steam radiators, kerosene heaters, and radiant gas heaters.

Automobiles available.-The count of automobiles available represents the number of passenger automobiles, including station wagons, owned or regularly used by the occupants of the housing unit. Passenger cars were to be counted if they were owned by a member of the household or if they were regularly used and ordinarily kept at home as are some company cars. Not to be counted were taxis, pickups or larger trucks, and dismantled or dilapidated cars in an early stage of being junked.

Value. Value is the respondent's estimate of how much the property would sell for on today's market (April 1960). Value data are restricted to owner-occupied units having only one housing unit in the property and no business. Units in multiunit structures and trailers were excluded from the tabulations. In rural territory, units on farms and all units on places of 10 acres or more also were excluded from the value tabulations.

For tracts where information on value was tabulated on a complete-count basis (usually in large cities and other urban places), the median was computed on the basis of the tabulation groups shown in the table and the terminal category was $25,000 or more. For these tracts, $25,000+ is the highest median which could be shown. For tracts where information was tabulated from a sample of units, the median was computed on the basis of more detailed categories than are shown in the table and the terminal category was $35,000 or more. Therefore, a median of $25,000+ is not to be interpreted as being necessarily lower than a specific median above $25,000; for example, a median of $25,000+ for a tract in the city may or may not be lower than $29,000 for a tract outside

In the 1950 tract report, value was not published separately for owner-occupied units (except for nonwhite owner-occupied units) but was shown in combination with vacant units available for sale. In rural territory, value data excluded values for farm units but may have included some units on places of 10 acres or

more.

Contract rent.-Contract rent is the rent agreed upon regardless of any furnishings, utilities, or services that may be included. Renter-occupied units for which no cash rent was paid were excluded from the computation of the median. In rural territory, rent data exclude rent for farm units.

In the 1950 tract report, contract rent was not published separately for renter-occupied units (except for nonwhite renteroccupied units) but was shown in combination with vacant units available for rent.

Gross rent. The computed rent termed "gross rent" is the contract rent plus the average monthly cost of utilities (water, electricity, gas) and fuels such as wood, coal, and oil if these items are paid for by the renter in addition to contract rent. Thus, gross rent eliminates rent differentials which result from varying practices with respect to the inclusion of heat and utilities as part of the rental payment. In rural territory, rent data exclude rent for farm units.

Renter-occupied units for which no cash rent was paid are shown separately in the tabulation but were excluded from the computation of the median. The median was computed on the basis of more detailed tabulation groups than are shown in the tables.

Farm residence. In rural territory, farm residence is determined on the basis of number of acres in the place and total sales of farm products in 1959. An occupied housing unit is classified as a farm housing unit if it was located on a place of 10 or more acres from which sales of farm products amounted to $50 or more in 1959, or on a place of less than 10 acres from which sales of farm products amounted to $250 or more in 1959. Occupied units for which cash rent was paid are classified as nonfarm housing if the rent did not include any land used for farming (or ranching).

In 1950, farm residence was determined by the respondent's answer to the question, "Is this house on a farm (or ranch)?" In addition, the instructions to the enumerators specified that a house was to be classified as nonfarm if the occupants paid cash rent for the house and yard only.

COLLECTION AND PROCESSING OF DATA Collection of data.-Several enumeration forms were used to collect the information for the 1960 Censuses of Population and Housing. A few days before the census date, the Post Office Department delivered an Advance Census Report (ACR) to households on postal delivery routes. This form contained questions which were to be answered for every person and every housing unit. Household members were requested to fill the ACR and have it ready for the enumerator. The census enumerator recorded this information on a form specially designed for electronic data processing by FOSDIC (Film Optical Sensing Device for Input to Computer). The information was either transcribed from the ACR to the complete-count FOSDIC schedule or entered on this schedule during direct interview.

In the densely populated areas, containing over nine-tenths of the census tracts in the Nation, the enumerator left a Household Questionnaire to be completed by each household (or person) in the sample and mailed to the local census office. The population and housing information was transcribed from the Household Questionnaire to a sample FOSDIC schedule. When the Household Questionnaire was not returned or was returned without having been completed, the enumerator collected the missing information by personal visit or by telephone and entered it

areas, when the enumerator picked up the ACR, he obtained all the information by direct interview and recorded it directly on the sample FOSDIC schedule. For vacant units, the enumerator collected the information by direct interview with the owner, landlord, or neighbor.

Soon after the enumerator started work, his schedules were examined in a formal field review. This operation was designed to assure at an early stage of the work that the enumerator was performing his duties properly and had corrected any errors he had made.

A more detailed description of the 1960 Census practices in the collection of data is given in a report entitled United States Censuses of Population and Housing, 1960: Principal Data Collection Forms and Procedures, Government Printing Office, Washington 25, D.C.

Electronic processing.-Several steps were required to process the data. First, the enumerator recorded the information by marking appropriate circles on the FOSDIC schedules. These schedules were later microfilmed and the information was read by FOSDIC, which converted the markings to signals on magnetic tape. The tape, in turn, was processed in an electronic computer, which was used extensively to edit and tabulate the data and to produce the publication tables.

Editing. In a mass statistical operation, such as a national census, human and mechanical errors occasionally arise in one form or another, such as failure to obtain or record the required information, recording information in the wrong place, misreading position markings, and skipping pages. These were kept to a tolerable level by means of operational control systems. Nonresponses and inconsistencies were eliminated by using the computer to assign entries and correct inconsistencies. In general, few assignments or corrections were required, although the amount varied by subject and by enumerator. Whenever information was missing for a housing item, an allocation procedure was used to assign an acceptable entry, thereby eliminating the need for a "not reported" category in the tabulations. An acceptable entry was assigned also when the reported information was inconsistent. A similar procedure was followed when information was missing or was unacceptable for a population item, although for several items a "not reported" category was retained.

The assignment of an acceptable entry was based on related information reported for the housing unit or person or on information reported for a similar unit or person in the immediate neighborhood. For example, if tenure for an occupied unit was omitted but a rental amount was reported, the computer automatically edited tenure to "rented." Another technique is illustrated by the procedure used in the assignment of age in the complete-count tabulations: the computer stored reported ages of persons by sex, color or race, household relationship, and marital status; each stored age was retained in the computer only until a succeeding person having the same characteristics and having age reported was processed through the computer; this stored age was assigned to the next person whose age was unknown and who otherwise had the same characteristics. This procedure ensured that the distribution of ages assigned by the computer for persons of a given set of characteristics would correspond closely to the reported age distribution of such persons as obtained in the current census.

The extent of the allocations for nonresponse or for inconsistency is shown for States, places of 10,000 inhabitants or more, and other areas in appendix tables in chapters B, C, and D of 1960 Census of Population, Volume I, Characteristics of the Population, and in 1960 Census of Housing, Volume I, States and Small Areas. Specific tolerances were established for the number of computer allocations acceptable for a given area. If the number was beyond tolerance, the data were rejected and the original schedules were re-examined to determine the source of the error. Correction

Accuracy of the data.--Tract statistics provide data for relatively small numbers of housing units and persons; hence, information for one tract almost always represents the work of only a few enumerators (sometimes one or two). Moreover, such items as the delineation of living quarters into housing units and the classification of the condition of a housing unit were always determined by the enumerator. Therefore, users of the data should bear in mind that misinterpretation of the instructions or variation in interpretation of responses may lead to a wider margin of relative error and response variability in data for census tracts than for larger areas. The systematic field review early in the enumeration corrected some of the errors arising from misunderstandings on the part of the enumerator.

To the extent that answers to the census questions were entered on the ACR and on the Household Questionnaire by household members, the responses were not affected by any misunderstanding on the part of the enumerator. The self-enumeration forms provided brief but uniform explanations and called attention to the response categories in a uniform manner.

Some innovations in the 1960 Censuses reduced errors in processing and others produced a more consistent quality of editing. The elimination of the card-punching operation removed one important source of error. The extensive use of electronic equipment ensured a more uniform and more flexible edit than could have been accomplished manually or by less intricate mechanical equipment. It is believed that the use of electronic equipment in the 1960 Censuses has improved the quality of the editing compared with that of earlier censuses but, at the same time, it has introduced an element of difference in the statistics.

SAMPLE DESIGN AND SAMPLING VARIABILITY Sample design. The unit of sampling was the housing unit and all its occupants; in group quarters, the sampling unit was the person. On the first visit to an address, the enumerator was instructed to assign a Sample Key letter (A, B, C, or D) to each housing unit sequentially in the order in which he first visited the unit, whether or not he completed the interview. Each enumerator was given a random key letter to start his assignment, and the order of canvassing was indicated in advance, although the instructions allowed some latitude in the order of visiting individual units at an address. Each housing unit which was assigned the key letter "A" was designated as a sample unit for housing data, and all persons enumerated in the unit were included in the sample for population data. In group quarters, the sample for population data consisted of every fourth person in the order listed, and no information was collected on housing. Information for the persons and housing units in the sample was recorded on a sample FOSDIC schedule. For population data, there was one form of the sample FOSDIC schedule. For housing data, every fifth sample FOSDIC schedule carried questions comprising the 5-percent sample items; the other four-fifths carried questions comprising the 20-percent sample items. Items which appeared on both types of housing schedules comprised the 25-percent sample items. Thus, the population sample consisted of approximately 25 percent of the population, while the housing sample was 5, 20, or 25 percent of the housing units. In some situations, the same item was tabulated from various samples, as indicated in table A.

The sample rate for "automobiles available" generally was 20 percent of the housing units in large urban places and 5 percent in all other places. For balance of county, places with fewer than 25,000 inhabitants, and individual tracts where the sample rate was 5 percent of the housing units, data on automobiles are not shown. Totals for the SMSA and component counties usually were based partly on a 20-percent sample and partly on a 5-percent sample.

For value of property in table H-2, the data generally were tabulated on a 100-percent basis for the individual tracts in large urban places; 2 for all other individual tracts, and for the totals of the SMSA, city, county, balance of county, and other areas, the sample rate was 25 percent of the housing units. The figures for the individual tracts in large urban places, therefore, do not necessarily add to the figures in each of the categories in the total columns. For table H-3, value data were tabulated from a 25percent sample for individual tracts as well as the totals for all places.

Data on automobiles and value of property were usually tabulated at different sample rates in the case of "split tracts," where the city boundary divides a tract. This situation occurs when part of a tract is located inside and part outside a large urban place according to the boundaries at the time of the census. Data on value of property are shown for split tracts, but data on automobiles are suppressed for the part of the tract outside the city as well as for the tract total.

Although the sampling procedure did not automatically insure an exact 25-percent sample of persons or 25-, 20-, or 5-percent sample of housing units in each tract, the sample design was unbiased if carried through according to instructions. Generally, for large areas, the deviation from the estimated sample size was found to be quite small. Biases may have arisen, however, when the enumerator failed to follow his listing and sampling instructions exactly. According to preliminary estimates, 25.07 percent of the total population in the United States as a whole and 24.95 percent of the total housing units were designated for the 25percent samples.

Ratio estimation. The statistics based on samples of persons and housing units are estimates that were developed through the use of a ratio estimation procedure. For population items, essentially this procedure was carried out for each of 44 groups of persons in each tract. The groups are as follows:

For each of the 44 groups, the ratio of the complete count to the sample count of the population in the group was determined. Each sample person in the group was assigned an integral weight so that the sum of the weights would equal the complete count for the group. For example, if the ratio for a group was 4.2, one-fifth of the persons (selected at random) within the group were assigned a weight of 5, and the remaining four-fifths, a weight of 4. The use of such a combination of integral weights rather than a single fractional weight was adopted to avoid the complications involved in rounding. In order to increase the reliability, where there were fewer than 50 persons in the complete count in a group, or where the resulting weight was over 16, groups were combined in a specific order to satisfy these two conditions.

For housing items, a similar ratio estimation procedure was carried out for each of 7 groups of housing units in the tract, separately for the 25-, 20-, and 5-percent samples. The 7 groups are as follows:

The ratio estimates achieve some of the gains of stratification which would have been obtained if the sample had been stratified by the groups for which separate ratio estimates were computed. The net effect is a reduction in the sampling error and in the bias of most statistics below that which would be obtained by weighting the results of the 25-percent sample by a uniform factor of four (the 20-percent sample by 5 or the 5-percent sample by 20). The reduction in sampling error is trivial for some items and substantial for others. Further, as a byproduct of this procedure, estimates from the sample are generally consistent with the total numbers of persons and housing units obtained from the complete count in each tract.

Sampling variability.-The figures from sample tabulations are subject to sampling variability, which can be estimated by using the factors shown in table D in conjunction with table B for absolute numbers and with table C for percentages. These tables do not reflect the effect of response variance, processing variance, or bias arising in the collection, processing, and estimation steps. Estimates of the magnitude of some of these factors in the total error are being evaluated and will be published at a later date. The chances are about two out of three that the difference due to sampling variability between an estimate based on a sample and the figure that would have been obtained from a complete count is less than the standard error. The chances are about 19 out of 20 that the difference is less than twice the standard error and about 99 out of 100 that it is less than 21⁄2 times the standard error. The amount by which the estimated standard error must be multiplied to obtain other odds deemed more appropriate can be found in most statistical text books.

For most population characteristics, the use of the household as a sampling unit increases slightly the standard error above what would be expected for a simple random sample of persons taken

Estimates of characteristics of the population from the sample for a given tract are produced using the formula:

44

Y

where r' is the estimate of the characteristic for the tract obtained through the use of the ratio estimation procedure,

I is the count of sample persons with the characteristic for the tract in one (1) of the 44 groups,

yis the count of sample persons for the tract in the same one of the 44

groups, and

Y is the count of persons in the complete count for the tract in the same one

Table D provides a factor by which the estimates proportionate to the standard errors in tables B and C should be multiplied to adjust for the combined effect of the sample design and the estimation procedure. Table B shows estimates proportionate to the standard errors for estimated numbers of persons or housing units. Table C shows estimates proportionate to the standard errors of estimated percentages of persons or housing units.

To estimate a standard error for a given characteristic, locate in table D the factor applying to the item; multiply this factor by the estimate proportionate to the standard error given for the number shown in table B. The product of this multiplication is the approximate standard error. Similarly, to obtain an estimate of the standard error of a percentage, multiply the figure as shown in table C by the factor from table D. For most estimates, linear interpolation in tables B and C will provide reasonably accurate results.

The sampling variability of the medians, presented in some of the tables, depends on the size of the base and on the distribution on which the median is based.

The standard errors estimated from tables B and C (using the factors given in table D) are not directly applicable to differences between two sample estimates. These estimates are to be applied differently in the following three situations:

1. For a difference between the sample figure and one based on a complete count (e.g., a difference arising from comparisons between value of property from the 100-percent tabulation for one tract and value based on a 25-percent sample in another tract), the standard error of the difference is identical with the standard error of the estimate which is based on the sample.

2. For a difference between two sample figures (e.g., one from 1960 and the other from 1950, or both from the same census year), the standard error is approximately the square root of the sum of the squares of the standard error of each estimate considered separately. This formula will represent the standard error quite accurately for the difference between estimates of the same characteristic in two different tracts, or for the difference between separate and uncorrelated characteristics in the same tract. If, however, there is a high positive correlation between the two characteristics, the formula will overestimate the true standard

error.

3. For a difference between two sample estimates, one of which represents a subclass of the other, the difference should be considered as the sample estimate and an estimate of the sampling error of the difference obtained directly.

Illustration: Let us assume that, for a tract, table P-1 shows that there are an estimated 800 persons 25 years old and over who had completed 4 years of high school. Table D shows that for "years of school completed" the appropriate number in table B should be multiplied by a factor of 1.0. Table B shows that the estimate proportionate to the standard error for an estimate of 800 is about 46. The factor of 1.0 times 46, or 46, means that the chances are approximately two out of three that the results of a complete count would not differ by more than 46 from the estimated 800. It also follows that there is only about 1 chance in 100 that the results of a complete count would differ by as much as 115, that is, by about 21⁄2 times the standard error. Assume also that table H-1 shows an estimated 50 units with more than one bathroom. The factor of 1.2 from table D multiplied by 15 from table B gives an estimated standard error of 18.

Statistics in the tract report may differ from those in other reports from the 1960 Censuses of Population and Housing. In some reports an item may be tabulated on a 100-percent basis, whereas in other reports it was tabulated from a sample and the figures will differ because of sampling variability. Differences arise also through errors of processing and enumeration, some of which are discovered in early reports and are corrected in subsequent reports.