I heard a new report being discussed this morning and I was sort of listening with one ear until I suddenly heard mention of the sample size. Then I started giving it my full attention. The report, which made claims for the amount of violence perpetrated against women was pretty shocking, but as remarkable was that 42, 000 women had been interviewed for the survey. That is a huge sample. I often see published reports where 20 individuals have been interviewed and that isn't unremarkable.
I've suggested to students in the past that sample size is a rough and ready measure of how robust a report's findings might be. Clearly, there are lots of ways that research can be not so good, but when I see exaggerated claims being made for very small scale samples my bad research detector starts going off.
The current research (which looks extremely robust) published by the European Union Agency for Fundamental Rights (FRA) reports that roughly half of women in the UK have been victims of physical or sexual abuse. A more detailed breakdown of what the report offers can be found here.
For further details, have a look how the Independent newspaper reported the research from here.
Sorry decline of English social housing
3 days ago