Cookies on the Gambling Commission website

The Gambling Commission website uses cookies to make the site work better for you. Some of these cookies are essential to how the site functions and others are optional. Optional cookies help us remember your settings, measure your use of the site and personalise how we communicate with you. Any data collected is anonymised and we do not set optional cookies unless you consent.

Set cookie preferences

You've accepted all cookies. You can change your cookie settings at any time.

Skip to main content

Report

Gambling Survey for Great Britain - Annual report (2023): Official statistics

Gambling Survey for Great Britain - annual report (2023): Official statistics

  1. Contents
  2. Response to the survey

Response to the survey

Address-level response rates

Table A.1: Address-level response rates

Table A.1: Address-level response rates
Response of issued addresses Total addresses (number) Total addresses (percentage) Total eligible (percentage)
Issued addresses 37,554 100% Not applicable
Assumed ineligible 3,380 9% Not applicable
Assumed eligible (residential addresses) 34,174 91% 100%
Refusal or unable to take part 47 0% 0%
No response 27,491 73% 80%
Productive addresses 6,636 18% 19%

In total, 37,554 addresses were issued. In remote surveys (that is where participants complete the survey independently without any involvement from an interviewer), no information is known about the reason for non-response in individual addresses. However, it was assumed that around 9 percent of addresses in the sample (3,380) were not residential and were therefore ineligible to complete the survey1.

In total, 6,636 addresses were productive. The adjusted address-level response rate, that is the proportion of assumed eligible addresses where a questionnaire was completed by at least one adult, was 19 percent (slightly lower than the target of 22 percent). There was no response from 27,491 addresses and in a further 47 addresses, an individual contacted the office to say they did not wish or were unable to take part.

Table A.2 Address-level response, by country

Table A.2 Address-level response, by country
Country Total issued addresses (number) Total issued addresses (percentage) Total assumed eligible addresses (number) Total assumed eligible addresses (percentage) Total productive addresses (number) Household response rate (percentage)
England 32,219 86% 29,319 86% 5,714 19%
Scotland 3,468 9% 3,156 9% 599 19%
Wales 1,867 5% 1,699 5% 323 19%
All addresses 37,554 100% 34,174 100% 6,636 19%

Table A.2 Address-level response, by country shows the breakdown of the issued sample in England, Scotland and Wales and the household response rate in each country. In total, 86 percent of the issued addresses were in England, 9 percent in Scotland and 5 percent in Wales. The adjusted household response rate was 19 percent in all 3 countries.

Individual-level response rates

Following the process of removing duplicate responses2, cases deemed to have completed the online questionnaire too quickly to have properly engaged with the questions and cases above the maximum 2 completions per household3, it was assumed that all responses in the dataset were from 9,742 unique individuals who had completed the questionnaire.

In total, 44 percent of the responding unweighted sample were male, and 56 percent were female, as shown in Table A.3: (Individual-level response, by age and sex). This under-representation of males is similar to that seen in the latest published results for other surveys with the same completion modes, for example, the British Social Attitudes (BSA) 2022 survey - BSA 40 Technical details (natcen.ac.uk) (opens in new tab) (PDF). In total, 44 percent of the BSA 2022 unweighted main sample were male and 56 percent female.

Those in the youngest age group were the least likely to take part: 5 percent of responding adults were aged 18 to 24, whereas this age group makes up 10 percent of the adult population of Great Britain. Conversely, 18 percent of responding adults were aged 64 to 74, whereas this group makes up 13 percent of the population as shown in Table A.3: Individual-level response, by age and sex.

Table A.3: Individual-level response, by age and sex

Table A.4: Individual-level response, by age and sex
Age group (years) Total participants male (percentage) Total participants female (percentage) Total participants all (percentage) Total Great Britain population adult males (percentage) Total Great Britain population adult females (percentage) Total Great Britain population adult all (percentage)
18 to 24 4% 7% 5% 11% 10% 10%
25 to 34 13% 17% 15% 17% 17% 17%
35 to 44 15% 17% 16% 16% 16% 16%
45 to 54 14% 15% 15% 17% 16% 17%
55 to 64 18% 18% 18% 16% 16% 16%
65 to 74 20% 16% 18% 13% 13% 13%
75 and over 16% 10% 13% 10% 12% 11%
All participants 44% 56% 100% 48% 52% 100%
Unweighted bases (number) 4,262 5,460 9,722 25,005,805 26,712,827 51,718,632

This age variation was particularly pronounced for males: 4 percent of responding males were aged 18 to 24, compared with 11 percent of the total male adult population of Great Britain who are in this age group. The equivalent proportions for females aged 18 to 24 were 7 percent and 10 percent. The national percentages are based on the 2022 mid-year population estimates for Great Britain: Estimates of the population for the UK, England, Wales, Scotland and Northern Ireland by the Office for National Statistics (opens in new tab).

Table A.4: Individual-level response, by mode of completion and sex

Table A.4 Individual-level response, by mode of completion and sex shows response rates by mode of completion, overall and for males and females separately. Overall, 65 percent (6,296) completed the survey online and 35 percent (3,429) completed a postal questionnaire. These results are broadly in line with expectations, based on the development stages of the GSGB.

Table A.4: Individual-level response, by mode of completion and sex
Sex Individual-level response: Online completions (percentage) Individual-level response: Postal completions (percentage) Individual-level response: Total completions (percentage) Individual-level response: Total completions (number)
Male 64% 36% 44% 4,265
Female 65% 35% 56% 5,460
All participants 65% 35% 100% 9,725
Unweighted bases (number) 6,296 3,429 9,725 9,725

Table A.5: Individual-level response, by mode of completion and age

There was little difference between males and females in the mode of completion. In all age groups, except the oldest 2 (aged 65 to 74 and aged 75 and over), a higher proportion of participants completed the survey online than filled in the postal questionnaire. In total, 46 percent of those aged 65 to 74 and 33 percent of those aged 75 and over completed the survey online compared with between 58 percent and 88 percent of those in the younger age groups as shown in Table A.5: Individual-level response, by mode of completion and age.

Table A.5: Individual response, by mode of completion and age
Age group (years) Individual-level response: Online completions (percentage) Individual-level response: Postal completions (percentage) Individual-level response: Total completions (number)
18 to 24 85% 15% 531
25 to 34 88% 12% 1,482
35 to 44 82% 18% 1,589
45 to 54 72% 28% 1,440
55 to 64 58% 42% 1,723
65 to 74 46% 54% 1,734
75 plus 33% 67% 1,240
All participants 65% 35% 9,742
Unweighted bases (number) 6,306 3,436 9,742

Weighting strategy

The data was weighted to take account of non-response, bias, and improve representativeness. This was done separately for each of the 2 waves and then combined into a single weight for this report. As there was no disproportionate sampling, selection weights were not required. The weighting method consisted of 2 stages4, repeated for Wave 1 and Wave 2:

  1. A logistic regression model for number of responses within a household (run for households with more than one eligible adult).

  2. A calibration to population estimates.

For the first stage, forward and backward stepwise logistic regression models were used to test which variables were predictive of the number of responses within a household. These models were run only for households with more than one eligible adult. Area-level variables (from the 2021 census for England and Wales and the 2011 census for Scotland) and household-level variables were tested. Where both a household level and area level version of a given characteristic was available, the household-level version was used. For example, if household income and Index of Multiple Deprivation (IMD) were both predictive of number of responses, only household income was used. The final regression models included all variables that were significant in the stepwise regressions for that given wave (see the Gambling Survey for Great Britain - Year 1 (2023), wave 1 report and the Gambling Survey for Great Britain - Year 1 (2023), wave 2 report for detail of the variables included). Region of residence was also included in each model to control for any regional differences in survey response.

The predicted probabilities from these models were used to create response weights for households with more than one eligible adult. Weights were checked for outliers and trimmed if necessary. Weights for responding households with only one eligible adult were set to one.

The response weights were then calibrated to estimates of the eligible population, that is, residents of Great Britain aged 18 years and over. Calibration weighting adjusts the weights so that characteristics of the weighted achieved sample match population estimates, reducing bias. The following variables were included in the calibration for each wave: age categories by sex, region, IMD percentiles (quintiles for England and bitiles for Wales and Scotland), tenure, and ethnicity.

Estimates of the Great Britain population by age, sex, and region of residence were taken from Office for National Statistics (ONS) mid-year population estimates, which were for 2022 in England and Wales Population estimates - Office for National Statistics (opens in new tab) and for 2021 in Scotland Population estimates for the UK, England, Wales, Scotland and Northern Ireland - Office for National Statistics (opens in new tab). Population estimates for IMD percentiles within each country were taken from ONS England and Wales (opens in new tab) and National Records of Scotland (opens in new tab). Population estimates for tenure and ethnicity were taken from the most recent Labour Force Survey data available, which was gathered between April and June 2023 Labour Force Survey performance and quality monitoring report: April to June 2023 (opens in new tab).

After calibration, the weights were checked for outliers and retrimmed if necessary. To create the annual weight, each wave’s weights were rescaled to sum to the same value, to ensure that each wave equally contributed to analysis. The final weight for the 9,742 productive individuals has a design effect of 1.25, an effective sample size of 7,820, and efficiency of 80 percent.

References

1When estimating the proportion of ineligible addresses in an online survey, it is best practice to assume the same ineligibility rate as a recent face-to-face survey which uses the same sample frame and sampling approach and for which detailed outcomes are known for the entire issued sample. Ineligibility rates in Postcode Address File face-to-face surveys tend to fall between 8 percent and 10 percent. 9 percent is the rate recorded in the most recent face-to-face British Social Attitudes Survey (2019) and so has been used as an appropriate default for this survey.

2Where a single participant completed the survey a second time.

3For example, two completing the survey online and two different people completing postal questionnaires.

4This same method was also used to weight experimental statistics phase data, with the notable difference that highest level of education has not been included in the calibration variables for official statistics data collection. This is because the qualification questions in the GSGB are too different to those included in the Labour Force Survey (LFS) to be confident that they are measuring the same thing. Both the experimental statistics phase and GSGB Year 1 response datasets show significant divergence in education profiles compared to LFS estimates. Therefore, calibration to LFS estimates of education would not be reliable and has the potential to increase bias rather than reduce it. Alternative high-quality estimates of education levels are not available

Previous section
GSGB Annual report - Introduction
Next section
GSGB Annual report - Gambling participation
Is this page useful?
Back to top