Young People in Scotland Survey 2021: attitudes to violence against women and girls

Methodology and analysis

Methodology

The YPIS is an omnibus survey carried out annually by Ipsos MORI in 50 local authority secondary state schools across Scotland. It contains questions on a broad range of topics. The module of questions on attitudes towards VAWG (see Annex D) was included in the self-completion section of the survey, which was completed by 2,285 pupils in 2014^[6] and 1,386 pupils in 2021.^[7]

Schools were selected to achieve a representative sample of 11 to 18-year-old pupils. Two of the questions (scenarios 8 and 9), which relate to commercial sexual exploitation, were only presented to older pupils aged 16-18 (993 pupils in 2014 and 686 pupils in 2021). The questions on VAWG are a sub-set of the questions asked of adults in the Scottish Social Attitudes Survey.

Pupils were presented with nine VAWG scenarios in which they were asked to imagine a specific behaviour and rate it on a 7-point Likert scale, where 1 was 'not wrong at all' and 7 was 'very seriously wrong'. Four of the scenarios were followed up with a question on how much harm the behaviour causes the victim. Perceived harm was rated on a 5-point scale between 'a great deal' to 'none at all'.

There were notable changes in the YPIS methodology between 2014 and 2021 that could have impacted the results of the survey. Due to the COVID-19 pandemic, the survey was administered online in 2021 compared to on paper in 2014. Moreover, most pupils (77%) completed the survey at home due to remote learning in 2021, whereas the completion had taken place entirely in schools in 2014. This change in completion location could have had an impact on the results as it represents a major methodological change.

Some key shifts to the societal context have also occurred since 2014, including the public discussion of violence against women and girls. These include prominent movements such as #MeToo, campaigns led by public and third sector organisations in Scotland, including publicising new offences around the posting of intimate images without consent, and the changing nature of social media. These wider contextual factors may have impacted pupils' attitudes towards the scenarios presented in the YPIS, but it is not possible to attribute these attitudinal changes to the shifting societal context or the methodological changes outlined.

Furthermore, one of the questions on commercial sexual exploitation changed in content from 2014 to 2021. Pupils in 2014 responded to the following question: "How wrong do you personally think it is for an adult (18 or over) to read magazines that feature topless women, or is it not wrong at all?" whereas in 2021 pupils responded to "How wrong do you personally think it is for an adult (18 or over) to watch pornography at home, or is it not wrong at all?". Due to the rephrasing of the question, it is not possible to produce robust and reliable comparison across time, therefore this question was excluded from the 2014 vs. 2021 comparative analysis presented in this report.

Finally, the survey from 2021 collected equality data on ethnicity, religion, health status, urban/rural and area deprivation (see Annex D for questions used). These variables were not collected in 2014, therefore comparison across years within these characteristics is not available.

Analysis

Confidence intervals were calculated at the 95% level to estimate the precision of results obtained. It is usual practice to create confidence intervals at the 95% level, which means that 95% of the time our confidence intervals should contain the true value found in the population.^[8] Statistical significance can also be established by using confidence intervals – where confidence intervals do not overlap, we may infer the two estimates to be significantly different. The term 'significant' is used to refer to statistical significance and is not intended to imply substantive importance. Only significant differences between groups or across time have been discussed in this report but full Data Tables can be found in the Supporting Documents.

Care must be applied when interpreting data points based on small sample sizes, as these produce wider and less reliable ranges. Only data points with sufficient sample sizes (N > 100) have been discussed in this report.

A cumulative response category was created to represent the proportion of pupils who rated a given behaviour as 'wrong' or 'harmful'. 'Wrong' refers to response options 5-7 on a 7-point Likert scale where 1 was 'not wrong at all' and 7 was 'very seriously wrong'. 'Harmful' refers to response options "a great deal", "quite a lot" and "some" on a 5-point scale between 'a great deal' to 'none at all'.

This allows for more robust interpretation of the data due to the pooled samples. This analytical approach is in line with that of the Scottish Social Attitudes Survey 2019: attitudes to violence against women.

Where data is presented in figures, data labels are presented for data points of 5% and above.