Thread by @DrJessTaylor

We have made some little videos explaining the key issues in the study. We also presented yesterday to 2000 people and explained the following to everyone:

1. We do not seek to generalise or extrapolate to the UK population from our data

2. This is the report of 22,419 UK women and their own experiences and that should be significant enough. We are not looking to make wider generalisations so we do not need to weight data or rake it.

3. Lots of people misunderstanding what a ‘self selecting sample’ is and how common that is in these kinds of studies. The UN 97% stat from last month was based on 1000 18-25yr self selecting sample too, but no uproar. Amazing.

4. We clearly posted to say that the study was about violent experiences of women since birth because it is vital that women are not deceived before they take part, in case it harms them. The only criteria was being a woman over 18 in UK. No other criteria was set out.

5. Recruiting participants using social media and referral techniques is very common, used in huge studies that people already cite in this field with no criticism and as long as we are clear and transparent about that, it’s okay. All studies have limitations. Including ours.

6. Every single study in which a participant can choose to take part, read full info beforehand and then consent or ignore has an element of self selection bias and that includes all surveys, questionnaires and interview studies. This is rarely raised until 22k women report.

7. Our sample is interesting in that it is huge, more diverse than the UK population and seems to be skewed towards women with more education than average. This is important when reflecting on findings. Women over 60 were not well represented at all in our sample, regrettably

8. We have much to consider and analyse over the coming months but it’s been shocking to see people attempt to discredit a study based on widely used methods - suddenly those common methods are not acceptable - maybe that’s because they are women talking about abuse and violence?

9. From a critical perspective, we have ended up completely distracted from the fact that for the first time ever, 22k women were able to clearly report their experiences anonymously - and instead have attacked contested methodology and even spread misinformation about sample

10. I hope that clears some things up. We haven’t and won’t extrapolate because we don’t want or need to. Recruitment method was common, not perfect, not RCT, but appropriate. Criteria to take part did not include experience of violence. Study was shared across many populations