About the research

Lumino is based on an online YouGov survey of 19,101 adults in England and Wales in October and November 2018. A boost sample of churchgoing Christians was included to ensure adequate response numbers for this group. The data were subsequently weighted back to national proportions to ensure results are representative of the overall population. The data were analysed using Multilevel Regression and Post-stratification, a technique that uses survey data to produce estimates for small geographies.

1.Quantitative survey

YouGov conducted an online survey of adults in England and Wales. 19,101 people were surveyed in total, including a boost sample of churchgoing Christians to ensure adequate response numbers for this group. The data was subsequently weighted by age, gender, social grade, ethnicity and region to ensure results were representative of the overall population. Additionally, weighting was applied by ‘churchgoing Christian’ status to correct for the proportional oversampling of this subgroup.

The survey covered attitudes to religion, Christianity and the Bible; behaviour in relation to the Bible and churchgoing; engagement; motivations and barriers to engaging with the Bible, among other areas.

Fieldwork was conducted between 11 October and 13 November 2018.


Following the survey, segmentation (cluster analysis) was conducted. This process brings together clusters of respondents who express similar attitudes or behaviours, based on responses to key questions/metrics. The key questions which define the segments are:

  • General satisfaction with life
  • Religion/faith status
  • Level of belief/non-belief in God/gods
  • Frequency of church attendance
  • Frequency of reading/hearing Bible
  • Level of Bible confidence (e.g. in describing the overall story, talking with others who may hold different views)
  • Likelihood to read the Bible in different situations in the future
  • Level of interest in finding out more about the Bible in the future
  • Attitude to Christianity in society
  • Attitudes to the Bible generally (e.g. relevance, veracity, value)
  • Attitudes to the Bible in church and in Christian life

Through this process we identified eight segments in the overall population.

3.Online focus groups

Online focus groups were conducted with all the segments (two groups per segment) to explore their attitudes, perceptions and behaviours in more detail and create personas. Online focus groups are conducted using an anonymous, chat-based platform.

4.Multilevel Regression and Post-stratification (MRP) 

Following the segmentation analysis, Multilevel Regression and Post-stratification (MRP) was used to produce estimates of survey findings for small geographies (parliamentary constituencies).  Even with the very large sample size included in the survey, on average each of the c570 constituencies in England and Wales would have had too few respondents for robust analysis, and MRP is therefore used to model results at constituency level.

MRP is based on the premise that similar people hold similar views and behaviours, irrespective of exactly where they live. We assess everything we know about the demographic make-up of a particular place and match individuals from our wider pool of respondents with the types of people that live there. For example, if a certain percentage of people in a particular constituency are female graduates in their thirties at a certain income level, we can look at the attitudes and behaviours of all such people in our wider sample and apply them to that constituency.

MRP was used to model four variables: respondents’ faith status and denomination, their church attendance behaviour, their level of interest in finding out more about the Bible, and the segmentation.