All Public Release data files have been processed and edited, and should meet the research needs of all users.
Over the past several years the PSID staff, using Computer Assisted Telephone Interview (CATI) technology and companion processing software, have significantly improved the quality and reliability of the timely release of data files. We now refer to the files posted for each new wave as Public Release Data. Note that:
1. Longitudinal data are subject to revision based on the most recent information received from individuals and families. New information that we find during family composition and economic editing in one wave may require revisions to previous waves. As additional data are collected through time on our two year collection cycle, prior files may be edited in light of the new information. Both the values of the variables themselves and the relationships of individuals to the families to which they are connected may be edited. Normally such changes are made only for a small number of cases.
2. An extensive set of computed or generated variables are included in the Public Release Data. As time and resources allow we occasionally add selected new generated variables for later release.
Since the PSID data files, as with the data files from any complex longitudinal study, are subject to minor changes and subsequent updated releases, due primarily to economic and family composition editing activities, it is therefore highly recommended that users retain and save all data files that are downloaded from this site and upon which individual research analysis is dependent. Only the most current data files are retained by PSID staff for distribution.
The term "Public Release I" used to refer to files released for general public use after they have been reviewed for data quality checks and consistency in both the reported family listing and the relationships among family members (this review process is called "family composition editing").|
The term "Public Release II" was previously used to refer to files which had undergone additional data checks to correct a very small number of cases and had been formatted in a more convenient form.
Because of successive improvements in our Computer Assisted Telephone Interviewing (CATI) software that PSID began using in 1993, the quality of the Public Release I files improved in recent waves, allowing the use of these data with confidence. There is now no longer a necessity to release two versions of the Public Release files.
A reinterview family is a family unit that was interviewed in the prior wave. |
A main family is one that is the source of a splitoff family (a new study family formed by a sample member who moves out and forms his or her own family unit). In some divorce or separation situations, both resulting families will contain sample members, so both will be interviewed. We interview the first spouse we are able to contact as the main family, while the other spouse will be in the splitoff family. In the case of children leaving home, the main family is almost always the parental family.
A split-off family consists of a person or group of people (at least one of whom is a "follow" person of any age) who moved out from a main family since the prior wave's interview to form a new, economically independent family unit living in a separate housing unit. Several criteria must be met for a split-off to occur. In addition to having moved out since the prior wave, and to being 'followable', the person or group of people in general may not have moved to an institution such as college or prison or to another family unit within the panel study. Moreover, the person or group of people who have moved out and formed their own family unit must be economically independent from the family unit from which they split off. These are general rules, however, and sometimes unique situations arise that determine whether a person or group of persons becomes a split-off. For example, while moving to an institution such as college does not generally meet the criteria for becoming a split-off, if the person is working, paying their own living expenses, and paying their own educational expenses in addition to attending school, then this person could be interviewed as a split-off. The living situation and interview data for each and every possible split-off case are first reviewed before split-off status is granted. Note that a splitoff family is only designated as a splitoff in the wave in which the family is newly formed and interviewed for the first time. In subsequent waves, they are considered a reinterview family.
In the PSID study, we are attempting to learn about our sample members, and the families in which they live. Each of these families is called a family unit (FU). The FU is defined as a group of people living together as a family. They are almost always related by blood, marriage, or adoption. And they must all be living in the same HU (see below).
Occasionally, unrelated persons can be part of an FU. They need to be permanently living with the family and share both income and expenses.
Any person in a study family is a family unit member. The term "other family unit member" (OFUM) is used of members who are not the Head or Wife/"Wife".
The household unit (HU) is the physical dwelling where the members of the FU reside. It can be a house, townhouse, apartment, a room in a rooming house, even a tent or a car.
Not everyone living in an HU is automatically part of the FU. There may be other people living in the HU temporarily who do not meet the criteria of relatedness and economic integration. The PSID data is about FU Members only.
Sample Members are individuals who were living in the original FU at the time of the very first interview their lineal descendants born after 1968. (For subsequent samples, such as the immigrants, the year of the first interview serves as the base for determining who is an original sample member, and all individuals present in the family at that time qualify.)
Follow status indicates whether we are interested in continuing to interview an individual. In general, sample members are always considered Followable. Non-Sample Members can be Followable too, if they represent a population of current interest. For example, we have in the past, followed such people as Non-Sample parents of sample children who were aged 25 or younger.
You can tell who is a sample member by looking at the individual's Person Number and Follow Status. Original Sample Members who were living in the original study FU in the first year of interviewing were given Person Numbers in the range of 001-019. Any Head's Spouse in the original interviewing year who was living in an institution was given a Person Number of 020. In addition, children of the Head (and Wife if present) who were under age 25 and in an institution the first year were considered Original Sample members and given Person Numbers in the range 0021-029. All of these people are followable.
Individuals who were born into a sample family after the first interviewing year and have a sample parent are considered "born-in Sample Members" and receive Person Numbers in the range of 030-169. All born in sample members are followable.
Some individuals who qualify as sample members (because they have a sample parent) are not born into a study family, but move in later. These "Moved in Sample Members" have Person Numbers of 170 or greater and are Followable.
All other people who have ever lived in a PSID family are not sample individuals. They also receive Person Numbers of 170 or greater, but are not Followable.
Response family unit members are those residing in an interviewed family at the time of interview. Nonresponse family unit members are those not residing in an interviewed family at the time of interview; they may have attrited, not yet appeared in the study, or not yet been born by a particular wave.
The phrase "main family nonresponse" means that both the individual and his or her family have at that time become lost to our study, although either or both may reappear in the study in subsequent waves. In the wave just prior to becoming nonresponse, the individual was connected with a family interviewed by our study; thus, both family and individual data are available for that prior year, and the individual's Sequence Number at that time was 01-59. However, data were collected for neither the individual nor his or her family in the nonresponse wave. The data for the wave in which nonresponse occurs (and all subsequent waves if and until the individual reappears as a member of a responding family unit, including a recontact family) are zeroes excepting the variables for type of individual record and reason for nonresponse, and if an individual was selected for recontact, follow status and reason for following the individual.
In contrast, mover-out nonresponse individuals have left a family that was still in the study. Since such individuals were usually present in that family for at least part of the calendar year preceding nonresponse, they have some additional nonzero data for the wave in which they became nonresponse, such as part-year income information. In later waves, mover-out nonresponse individuals are treated in two ways, depending on why they left the family. Those who moved out to institutions have several variables (Sequence Number, age, sex, Relationship to Head, type of individual and reason for nonresponse) with nonzero values, although income, housework, and other individual-level variables are filled with zeroes. Eventually, such an individual may (a) become response by moving into a family or by becoming a splitoff, (b) move from the institution and remain mover-out nonresponse (shown when Sequence Number=71-89), or (c) become main family nonresponse because the family itself became nonresponse. (See the preceding paragraph for an explanation of main family nonresponse data records.) The other type of mover-out nonresponse individual has either moved out, but not to an institution, or died. Later waves of data contain zeroes, as described above for main family nonresponse, unless they subsequently rejoined a responding family or were selected for recontact.
The data are released as one file, which includes not only those individuals with nonzero data records in the current data collection year (i.e., current response plus mover-out nonresponse), but also all other individuals-those who have zero data records for the current year (i.e., current year main family nonresponse and all nonresponse of either kind from earlier waves.
|Within each wave of data, each FU (family unit) has one and only one current Head. Originally, if the family contained a husband-wife pair, the husband was arbitrarily designated the Head to conform with Census Bureau definitions in effect at the time the study began. The person designated as Head may change over time as a result of other changes affecting the family. When a new Head must be chosen (see conditions for selecting a new Head below), the following rules apply:
The Head of the FU must be at least 16 years old and the person with the most financial responsibility for the FU. If this person is female and she has a husband in the FU, then he is designated as Head. If she has a boyfriend with whom she has been living for at least one year, then he is Head. However, if the husband or boyfriend is incapacitated and unable to fulfill the functions of Head, then the FU will have a female Head.
Husbands of Heads are extremely rare in the study. Early on, an FU might have a Female Head and a Husband of Head (instead of Head and Wife) if the Head was incapacitated in some way. (He may be still in the FU or he may have moved to an institution.) There are also a few cases where the female half of a married couple insists on being the Head, or where the male half of a married couple is adamant about not wanting to have his information included in the study. A Husband of Head has the Relationship to Head code 9 or 90.
In the PSID, an opposite sex romantic partner who has moved into an FU less than 1 year prior to the interview is labeled a boyfriend or girlfriend (code 88) in that first wave that he or she appears in the study. If the cohabitor has moved in at least one year before the interview, the couple will be coded as Head and "Wife" (code 22 from 1983 on). In the next wave, if the boyfriend or girlfriend is still living in the FU and the couple is still unmarried, they are recoded as Head and "Wife" (That is, a male head will remain head but his girlfriend will be labeled "Wife" or a Female Head will become "Wife" while her boyfriend will become Head).
Boyfriends and girlfriends are treated like family members who are not Heads or Wives/"Wives" — considerably less information is obtained about them. In waves since the late 1970s, information typically gathered for Wives has been gathered as well about "Wives".
Starting in 1983, the Relationship to Head (RTH) code allowed for differentiation between legal Wives and long-term female cohabitors. However, first year cohabitors can be detected prior to 1983 with a little bit of work. For example, their RTH would be 8 (nonrelative), their gender would be opposite that of Head's, and in subsequent years they may become Wives or Heads, while the Head would stay as Head or become a Wife. Anyone fitting this pattern can be decisively identified as a cohabitor. PSID has not distinctively labeled same sex cohabitors.