Americans and the Arts [1973 - 1992] (ICPSR 35575)

Time Period(s) View help for Time Period(s)

Date of Collection View help for Date of Collection

Data Collection Notes View help for Data Collection Notes

Study Design View help for Study Design

In 1973, care was taken to avoid any bias in the wording or emphasis of the questions so that respondents would not feel pressure to provide "right" answers. As an additional step in ensuring unbiased responses, the interviewers did not at any time reveal the name of the study's sponsor, nor did they mention the National Research Center of the Arts. They identified themselves as representatives of Louis Harris and Associates, NRCA's corporate parent. The interviewers used in the field work were residents of the areas where they conducted interviews, and wherever possible they were of the same ethnic group as the people they called upon. Additionally, the questionnaire instructs the interviewer: "How to determine whom you are to interview in this household: (1) If only one woman or man lives here, interview that person, provided your quota for that sex has not been completed. (2) If more than one woman or man lives here, list below all the women or all the men -- not both -- according to age, the oldest first. Then, starting at the bottom of the list, moving upward until you get to the first "x" next to which you have entered a name (or described by position). The person next to this "x" is the only person you can interview." A total of 3,005 personal interviews averaging one hour and thirty four minutes in length were conducted in January 1973. For the Americans and the Arts 1975, data were collected from 1,555 respondents in interviews conducted in June 1975. The Americans and the Arts 1980 collected data from 1,501 respondents between July 17 and July 28, 1980. The 1984 Americans and the Arts was administered to 1504 respondents from March 5 to March 25, 1984. In 1987, in order to determine which adult member of the household will be interviewed, the interviewer asked for one of the following household members, in order of priority: youngest adult male at home, next youngest adult male at home, youngest adult female at home, or next youngest adult female at home. This procedure results in a good distribution of respondents by sex and age. It also avoids automatically accepting as a respondent the adult in the household who routinely answer the phone. Non-answering telephones were dialed two or more times over a three-day period. Callbacks were made on different days and at different times of the day. One callback was made to try to convert any designated respondent who refused or terminated an interview. In the event of a busy signal, a follow-up call was made 15 minutes later, followed by two callbacks on different days and at different times of day. The 1987 study was administered to 1,501 respondents from March 13 to April 6, 1987. For the 1992 study, "Frankel-Goldstein grids" were attached to the backs of cards for interviewers to use to select an adult if there is more than one in the household. An M or F designation was put on the front of the cards at a 66/33 ratio to help the completed interviews fall out 48% male and 52 percent female. Interviewers used this designation to determine which respondent to select if there is an eligible respondent of each sex residing in the household. An initial attempt plus three attempts will be allowed at each open number, with the following exceptions: If there is a definite "call back" on the fourth attempt, one additional attempt will be allowed; the number will be dead regardless of the result on the fifth attempt. The refusal conversion procedure that will be employed for this study is as follows: After the first refusal, one contact (a completed interview or a second refusal) is required in order for that number to be considered dead, unless the first refusal comes on the special fifth attempt. After the first refusal, a total of four (NA's, BSY's, CB's, etc.) attempts are required in order for that number to be considered dead. Remember, the attempted conversion should be with the respondent in the household that originally refused. Updates of open numbers should not be considered as an attempt. The 1992 Americans and the Arts collected data from 1,500 adults from February 6 to February 25, 1992. Survey sampling differences preclude comparisons with measures of arts participation between the 1973 and 1975 Americans and the Arts studies which used face-to-face interviewing and the 1980, 1984, 1987, and 1992 studies which used computer-assisted telephone interviewing.

Sample View help for Sample

For Americans and the Arts 1973 and 1975, random multi-stage cluster design was used. Regarding sampling error for 1973, a sample of 3,000 is considered a statistically sound base for nationwide surveys and is projectable to the entire United States population at a 95 percent confidence level and a sampling error of plus or minus 2 percent. For 1980 and 1984, multistage cluster sampling was used to select states in the continental United States, then counties, then minor civil divisions with probability proportional to census estimates of their respective household populations. For each of these primary sampling units, telephone exchanges and the next two digits in the telephone numbers were chosen on a random basis. Interviewers then added two randomly generated digits to complete a seven-digit number. For 1987 a multistage unclustered sampling was used to select the primary sampling units. Within the primary sampling units, random-digit-dialing was employed to select households. For results based on the total sample, the estimated sampling error is plus or minus 3 percent. Survey results based on subgroups of small size can be subject to large sampling error. For 1992, a stratified unclustered systematic random selection, random-digit dialing was used.

Time Method View help for Time Method

Universe View help for Universe

For Americans and the Arts 1973 and 1975, United States residents age 16 and over. For Americans and the Arts 1980, 1984, 1987, and 1992, United States residents age 18 and over.

Unit(s) of Observation View help for Unit(s) of Observation

Data Type(s) View help for Data Type(s)

Mode of Data Collection View help for Mode of Data Collection

Description of Variables View help for Description of Variables

The 1973 and 1975 studies look at topics including childhood arts exposure, current arts participation, leisure activities, and attitudes regarding arts and arts funding. The 1980, 1984, 1987, and 1992 Americans and the Arts studies measured topics including attendance, participation, art, and education, funding, individual artists, TV and the arts, children and the arts, and support for the arts.

Response Rates View help for Response Rates

Response rates were not calculated for this collection. However, information was provided which will allow researchers to calculate a response rate for American and the Arts 1987: A total of 7519 telephone numbers were called in order to obtain 1501 completed interviews. Of the 7519 telephone numbers attempted, 5040 were in service and were not business phone numbers.

Original Release Date View help for Original Release Date

2015-03-19

Version History View help for Version History

Weight View help for Weight

For American and the Arts 1973 weighting was applied to correct any sample distortions. The archived data set for the 1980 Americans and the Arts study does not contain weight variables, although published reports based on these data indicate that weights were applied. Researchers intending to analyze trends across the six waves of the Americans and the Arts studies are advised to develop appropriate weights if they plan to include the 1980 data in their analyses. For American and the Arts 1984, the archived data set also does not contain weight variables. According to published reports, weighting was carried out in a two-stage process. First, weighting was carried out on the age, sex, and race figures in order to bring them into line with their actual proportions in the population. Then the sample was weighted on the basis of the latest data on United States household income distribution. The 1987 study was weighted to the Census Bureau's latest population parameters on sex, race, age, and income. This adjusts these key variables, when necessary, to their actual proportions in the population. For Americans and the Arts 1992 the completed interviews will be weighted to the following variables: age, race, and sex; education or income (population). The weighting table is created by Data Processing from the March 1991 Current Population Survey from the United States Census Bureau.