Survey of Public Participation in the Arts (2022) User Guide

The Survey of Public Participation in the Arts (SPPA or PPA) is conducted by the U.S. Census as a supplement to the Current Population Survey (CPS). It is sponsored by the National Endowment for the Arts (NEA) and has previously been conducted in 2002, 2008, 2012, and 2017.

For detailed information about the Current Population Survey and the Public Participation in the Arts supplement, please see the technical documentation available from the Census Bureau and in the study-level documentation on the SPPA home page. You may also be interested in the NEA Research Brief Arts Participation Patterns in 2022: Highlights from the Survey of Public Participation in the Arts and the National Arts Statistics and Evidence-based Reporting Center’s Measuring the Arts series B.1: Who Attends Arts Events in Person and B.3: Who is personally Creating or Performing Art? reports based on these data.

Methodology

The Current Population Survey (CPS) is conducted monthly, in a face-to-face or telephone interview format, using computer-assisted interviewing. The CPS sample includes approximately 60,000 eligible housing units (containing at least one individual in the civilian noninstitutionalized population) in the US each month. This survey is the source of information about the country’s labor-force participation and economic activity, along with other demographic characteristics, and is meant to be representative of the US, individual states, and other specified areas. The 2022 SPPA was conducted as a supplement to the July CPS and respondents were interviewed between July 16-25, 2022. Respondents were randomly selected individuals, aged 18 or older and not in the armed services, from approximately half of the selected CPS households.

There were approximately 127,000 respondents to the July 2022 CPS, of which just under 13,000 completed the SPPA supplement.

Content

The content of the SPPA reflects respondents’ participation in various types of arts activities during the period July 2021-2022. The questions are grouped, based on topic, into two cores and four modules. Each respondent was randomly assigned to one core and two modules (A or B and C or D). Therefore, questions from each core and module were asked of approximately half (~6,500) of the SPPA respondents. The full set of questions can be found in a questionnaire, available on the SPPA homepage.

Core 1: Arts Attendance and Literary Reading contained questions about in-person arts attendance (e.g., music performances, dance performances, museum or gallery visits, art or craft fair attendance) and reading behaviors.

Core 2: Arts Attendance and Venue Types questions focused on attendance at specific types of events (e.g., music, theater, or dance performances; book reading, poetry, or storytelling events; art exhibits), where they were held, and how respondents learned of them. Information about movie attendance, including documentaries and films, and visits to historical or architectural attractions is also captured.

Module A: Consuming Art via Electronic or Digital Media contained questions about accessing streamed, recorded, TV or radio broadcasts of music, dance performances, or theater productions as well as any arts-related podcasts, programs, documentaries, or visual art through technology.

Module B: Performing Art, Creating Visual Art, and Writing included questions about performing (singing, playing an instrument, or dancing) or creating art oneself during the previous 12 months (including writing; photographs; films/videos; visual art; and games, software, or tools for computers, mobile devices, or other platforms).

Module C: Other Artistic, Cultural, and Civic Activities respondents answered questions about attendance and participation related to cultural activities including sports, hunting/fishing and other outdoor activities, volunteering, playing video games, cooking, or visiting a library.

Module D: Arts Education contains questions about whether individuals ever took lessons or classes for, or learned in some other way, skills such as photography, acting, music, and art (including digital).

Note that the dataset also contains basic CPS labor force and demographic items.

Weights

For any analyses of the 2022 SPPA to be generalizable to the US population, the data must be weighted. The weights account for oversampling of subgroups in the CPS as well as for non-response to the CPS and/or supplement. The supplement weight, PWSUPWGT, is included in the data and must be multiplied by a factor of 2, 4, or 8, depending on whether the analysis includes variables from just the core or a single module (multiply PWSUPWGT by 2), a core and a module or two modules (multiply by 4), or a core and two modules (multiply by 8). For examples, see pages 3-2 – 3-3 in the technical guide linked above.

Historically, weights in the SPPA (and CPS more generally) included implied decimals. The weight variables in the 2022 survey and replicate weight files have been formatted so that there are no implied decimals.

Using the 2022 SPPA

A few tips are helpful to keep in mind when using the 2022 SPPA.

Using the SPPA in ICPSR’s online analysis tool: The Survey Documentation and Analysis (SDA) tool is available on the SPPA, but it should be used with care. Specifically, using SDA to explore the data prior to download can be extremely helpful. When evaluating data to determine whether they will help answer a particular research question, it is important to know what variables are included and have an idea of the distribution of responses on your key concepts and demographic measures. If you wish to look at sub-groups (such as age groups, people with a given level of education, or those who identify with specific races/ethnicities), it is best to look at unweighted (actual) counts. If your focal group is made up of just a few individuals, while they may be nationally representative when weighted, comparisons based on small numbers can be much more tenuous, produce larger standard errors, and require more caution in interpretation than if the groups are closer in size. (A single individual’s responses will have a much greater influence on the overall results if there are a small number of individuals to begin with, which is problematic if that individual’s answers are unusual in any way.)
Conducting analyses using SDA and the 2022 SPPA is complicated by the way weights are handled for the supplement. If you want to generalize your results to anyone beyond those in the sample, you must use the PWSUPWGT variable, but it must be multiplied by a factor that depends which part(s) of the supplement contains your variables. SDA cannot handle a customized weight variable such as this. So basic descriptions of the people who participated in the supplement are fine, trying to describe any kind of national patterns is not.
Given the way weights are to be used, it is imperative that you keep track of whether your variables originated in a core or a module and, for modules, whether one or more. Variable names will help in this regard. The supplement variable names begin with “PE,” followed by either a “C” and a “1” or “2” (core 1 or core 2) or an “M” and an “A,” “B”, “C,” or “D” (module A, module B…), followed by the question number within that core or module. For example, PEC2Q1B is from core 2, question 1b, “With the exception of elementary, middle, or high school performances, did you go to a live book reading or a poetry or storytelling event?” Variables whose names do not follow this format are from the base CPS (full questionnaires here).
A replicate weights file is also available. The purpose of replicate weights is to “mimic” multiple samples from the same sample to allow analysts to compute more precise estimates of the standard error (a measure of variability) of a given statistic. These standard errors can then be used to generate more precise confidence intervals around the statistic and more nuanced significance tests. This is especially important when working with smaller samples. A good source for background information about using replicate weights conceptually for another Census survey (the American Community Survey) can be found at https://usa.ipums.org/usa/repwt.shtml. For more information about replicate weights in the CPS/SPPA, please see the SPPA Technical Documentation and the guides to using person-level and household-level replicate weights referenced therein.
For more detailed information about the conduct of the Current Population Survey, see the Design and Methodology technical working paper.