GSS 2023 Data Collection

Introduction The General Social Survey (GSS) provides information on the wellbeing of people in Aotearoa New Zealand. It covers a wide range of life domains, including health, material living standards, paid work, social connectedness, housing, safety and security, civic and human rights, trust, culture and identity, and subjective wellbeing. The target population is the usually resident New Zealand population aged 15 years and over, in private dwellings in the North and South Islands and Waiheke Island. The objectives of the GSS are to: provide a picture of the changes in wellbeing for New Zealanders understand and monitor the distribution of wellbeing across population groups of interest contribute to understanding the interrelationships between different aspects of wellbeing. We release information from the GSS on Tatauranga Aotearoa – Stats NZ’s website, under ‘Wellbeing Statistics’. Survey content We use computer-assisted personal interviews (CAPI), which last around 50 minutes. Computer-assisted video interviews (CAVI) were also introduced as an option during the 2023 GSS collection. The survey consists of three questionnaires: the household questionnaire (HQ), the personal questionnaire (PQ) and the demographic questionnaire (DQ). One individual in the household completes the HQ, which collects information about the household and all the usually resident people in it, such as family relationships and household income. We then randomly select one eligible individual in the household aged 15 years or over to answer the PQ and DQ. An eligible individual is someone aged 15 or over who usually lives at the address of that household. The PQ collects information on a range of wellbeing related topics, such as overall life satisfaction, self-rated health, social contact, and loneliness. The DQ is the last questionnaire, and collects demographic information, such as ethnicity, qualifications, sex at birth, and gender. It is usually completed by the respondent themselves (self-complete mode) but can also be assisted by the interviewer depending on which format is preferred by the respondent. Primary content Topics in the primary content of the 2023 GSS include: subjective wellbeing labour market satisfaction with free time health disability (Washington Group Short Set, WGSS, and Washington Group Short Set - Enhanced, WGSS-E) te reo Māori culture and identity trust material standard of living satisfaction with the natural environment housing emergency preparedness safety and security acceptance of diversity discrimination family wellbeing and functioning social connectedness social identity demographics Survey supplements The General Social Survey has rotating supplementary content. The supplement topics for 2023 were: digital inclusion and voluntary work. The digital inclusion supplement includes questions about people's satisfaction and use of the Internet, feelings of safety when using the Internet, and barriers to accessing the Internet. Voluntary work was repeated from the 2021 set of supplements. Changes to the survey content: new additions. Some questions and modules have been changed or expanded in this GSS. These changes include new questions in te reo Māori module, which ask those who speak at least some te reo Māori how they learned it, and how much te reo Māori they speak to adults and children in their home. A new question about ‘receiving help in the form of food, clothes, or money from a welfare / community organisation such as a church or foodbank’ was added to the material standard of living module. The sense of belonging to Aotearoa New Zealand question was included in the core content as part of the social identity module. Previously, this question was part of the ‘Social Identity’ supplement collected in 2016 and 2021. Finally, the demographics module includes some updates, such as sexual Identity question which was asked to all respondents aged 15 years and over, unlike 2021 GSS where it was asked to those aged 18 years and over. Also, the income bands in the questionnaire were updated to match the categories used in the 2023 Census. Data collection The data collection for 2023 GSS started on 15 May 2023 and finished on 21 April 2024. The survey interviews were carried out by Reach Aotearoa on behalf of Tatauranga Aotearoa – Stats NZ. Wellbeing statistics are usually collected biennially in the New Zealand General Social Survey (GSS). These collections have been: GSS 2008 GSS 2010 GSS 2012 GSS 2014 GSS 2016 GSS 2018 GSS 2021 (delayed due to the COVID-19 pandemic) GSS 2023 The GSS has typically relied on face-to-face interviews in people’s homes. In the 2023 GSS, some respondents were offered an opportunity to complete the PQ and DQ via a computer-assisted video interviewing (CAVI) platform. This opportunity was made available during the collection. Sample design information The GSS uses a three-stage sample selection method, similar to our other household surveys. For the first stage, we select a total of 1,590 primary sampling units (PSUs). The PSUs selected for GSS were a subsample of PSUs used for Household Labour Force Survey (HLFS). These PSUs are available from the household survey frame (HSF), which is the standard sampling frame we use to select samples and to manage overlap control for all our household surveys. The HSF lists PSUs with attributes determined by data from the census. We then assign PSUs to standard strata based on these attributes. The second stage of sample selection consists of selecting eligible dwellings within the selected PSUs. In the third stage, we select one eligible individual within each selected household. The eligible individual is chosen at random from all eligible individuals in the household. The GSS is designed to provide estimates at a national level. Sample population The target population is the entire group you would ideally like to get information from. The target population for the GSS is the usually resident population aged 15 years and over in private dwellings in the North Island, South Island, or Waiheke Island of New Zealand. The target population includes: New Zealand usual residents temporarily overseas New Zealand usual residents temporarily staying elsewhere in New Zealand (including other permanent and temporary private dwellings, institutions, and non-private dwellings; and people who have no fixed abode, but were found in private dwellings on the household enumeration date) people in the New Zealand armed forces if they reside in a private dwelling young adults at boarding schools (young adults who fall into this category are not surveyed in the personal questionnaire but are included as members of the household in the household questionnaire). The target population excludes: overseas visitors and international students who expect to be resident in New Zealand for less than 12 months people living in non-private dwellings such as hotels, motels, boarding houses, hostels, and homes for the elderly patients in hospitals, or residents of psychiatric and penal institutions people living on offshore islands (excluding Waiheke Island) members of the non-New Zealand armed forces and their dependants non-New Zealand diplomats or diplomatic staff members and their dependants The survey population consists of the group members (from the target population) who have a chance of being selected as part of the sample (they can be identified through the sampling frame). The survey population is the target population with the following exclusions: New Zealand usual residents temporarily overseas who don't return within the survey period New Zealand usual residents temporarily staying elsewhere in New Zealand (including other permanent and temporary private dwellings, institutions, and non-private dwellings; and people who have no fixed abode, but stay at private dwellings) who don't return within the survey period New Zealand usual residents who live in remote areas that are costly or difficult to access. Response rate and sample population The achieved sample rate (ASR) was 65.6 percent (unweighted), and the survey response rate was 74.4 percent (weighted). The ASR measures how many eligible households responded to the survey, as a proportion of all dwellings sampled. The survey response rate is the weighted percentage of eligible individuals who responded to the survey, divided by the estimated number of eligible individuals. The achieved sample rate differs from the response rate because it includes the ineligible dwellings in the denominator. This difference means that the response rate is particularly sensitive to the classification of household eligibility. As a result, the achieved sample rate is more stable over time than the response rate. The 2023 GSS achieved sample size was just over 7,800. This sample was representative of the total population in geographic and key demographic terms. The 2023 data can be compared to previous iterations of the GSS, although extra caution should be taken when comparing to the 2021 GSS due to the smaller sample size due to the shortened collection window as an impact of COVID-19 pandemic. Data processing Non-response and imputation Despite our best efforts, there will always be a proportion of households that cannot be contacted or do not respond to the survey. There are two types of non-response: total or unit non-response (when no information is collected on a sampled unit or where the amount or quality of the information supplied is insufficient to be a response) partial or item non-response (where the absence of information is limited only to some variables) – with respect to unit non-response in general, the effects of this are minimised by applying a non-response adjustment within our weighting procedure. We also use a non-response indicator to identify and remove ‘low-quality’ responses. To be counted as a ‘response’, a person must have given a valid answer for at least 4 of 8 key variables: overall life satisfaction, life worthwhile, self-rated health status, generalised trust, experience of discrimination, crimes committed against you, income adequacy, and housing condition. Respondents who do not satisfy this criterion are not assigned a final weight in the estimation process and are not included in further analysis. With respect to item non-response, we impute for certain key variables in the GSS that are critical for either the weighting procedure (in which a complete record is required among the respondents) or outputs. The process of imputation replaces missing or residual values (such as ‘don’t know’) with values present within the data of similar respondents. We imputed for the following variables in this GSS: age, sex at birth, gender, ethnicity, and income. Weights Weighting is used in the GSS to ensure that estimates account for the sample design, adjust for non-response, and align with expected population totals or benchmarks. The benchmarks used in the GSS are estimates of the total population by age, sex, region, Māori by broad age groups, and household composition (number of two-adult households and number of other households). Benchmarking ensures the estimates reflect the total population distribution for key groups. This process adjusts for any groups with less coverage in the sample selected due to non-response or sample composition. The survey has two sets of weights attached, one for the person and one for the household. We use the person weights to describe the attributes of a person; for example, how many people are ‘very satisfied’ with their life overall. Household weights are used to produce estimates of households (the number or proportion of households) and can also be used to produce some estimates about all people in the household. Output New additions The 2023 output tables include new breakdowns of wellbeing statistics by cisgender and transgender status. These columns are included alongside the sexual identity and LGBT+ columns. In 2021, the number of people of trans gender status was too small to produce reliable estimates. Reliability of survey estimates Two types of errors are possible in estimates based on a sample survey: sampling error and non-sampling error. Sampling error quantifies the variability that occurs by chance because a sample rather than an entire population is surveyed. Sampling error can be measured. Non-sampling errors are all errors that are not sampling errors. These errors are difficult to quantify, and include unintentional mistakes made by respondents or interviewers, variation in the respondent’s and interviewer’s interpretation of the questions asked, and potential errors in programming, recording and coding data. We aim to minimise the impact of these errors by applying best survey practices and monitoring known indicators (such as non-response). We also complete a quality assurance process to investigate and mitigate any known concerns. We estimate sampling errors using a jack-knife method, which is based on the variation between estimates and on taking 100 mutually exclusive subsamples from the whole sample. Sampling errors are quoted at the 95 percent confidence level. For example, if the estimated total proportion of people is 27.1 percent, and the estimate is subject to a sampling error of plus or minus 1.4 percent (measured at the 95 percent confidence level), that shows there is a 95 percent chance that the true proportion lies between 25.7 and 28.5 percent. The information about these errors are available in published tables, and are labeled as absolute sampling errors (ASE). The ASE is displayed in the same format as the estimate and can be directly applied to the estimate (+/-) to give the confidence interval. Relative sampling errors (RSEs) are displayed as a percentage and show the size of the error relative to the estimate. As the size of the sampled group decreases, the RSE will generally increase. Larger relative sampling errors mean estimates are less sensitive to change, and differences between groups are less likely to be statistically significant. In particular, estimates for smaller population groups may have particularly large relative sampling errors associated with them. For example, estimates for Pacific peoples would have a larger RSE than those for Europeans. Changes that are not statistically significant (even if large) may be due to differences in the samples selected, rather than real-world change. In the published tables, estimates with high RSEs are flagged as follows: Between 30 and 49.9 precent are flagged with a single asterisk *. Between 50 and 99.9 percent are flagged with double asterisks **. 100 percent or higher are flagged with triple asterisks ***. We advise caution in using estimates with large RSEs, particularly those over 100 percent, as they are considered unreliable for most uses. Suppressed estimates Data in the Wellbeing Statistics 2023 information release are suppressed if the weighted count is less than 1,000 or if the unweighted cell count is five or fewer people or households. This ensures the privacy of the respondents is maintained. Customised data The tables in the wellbeing statistics information release do not contain all possible analyses of GSS data. Data requests can be customised to customers’ specifications. Contact [email protected] for more information, or see the Information centre website. en-NZ