Lies, damn lies and statistics (And NHS studies)

I like studies, some are good, based on solid evidence, yet some are really bad, manipulating the data to fit a certain agenda.

But the important thing when looking at a study is to appreciate that you're very often not looking at the data itself, rather, you're looking at someone else's interpretation of that data. So lets analyse how someone interprets data from what essentially isn't a very good study (not asking good enough questions)

Twitter user Greta McLachlan, who is apparently a General surgical trainee is a prime example in reading the top sheet only. I always find it best that if you're going to quote something, at least have a look underneath what you're actually quoting. But it seems to me that Greta is only seeking to qualify her own preconceived ideas.

Tweet begins

You cant deny the experience of 1400 people

89% of women and 81%of men witnessed sexual misconduct

63% of women and 23% experience sexual harassment

29% of women and 7% of men are sexually assaulted.

Tweet ends

So lets look at the actual data from the study itself

This is specifically about the UK surgical workforce (just for reference)

There are 1703 participants (seems a bit on the low side but still an OK sample)

I love these types of studies! Not only have they presented the date in a clear and concise manner they have also laid out the actual questions that were asked because some of the first questions I wanted to ask were

1. What constitutes sexual misconduct?

2. What constitutes sexual harassment?

3. What constitutes sexual assault?

But I have the answers.

Sexual harassment (non-exhaustive)

1. Jokes about sex

2. Uninvited comments on body

3. Asked for a date after previous refusal

Sexual assault

1. Touching (not genitalia or breasts)

So, if someone tells a dirty joke, this is sexual harassment, if someone puts a comforting arm around you, this is not only sexual harassment, its also a sexual assault.

So lets put it more bluntly, if someone touches you, that would be a sexual assault. I'm struggling to take this study seriously now. There are some really good questions in here, but its all undone by the presence of these ludicrous ones.

Traditionally, its men who would make the first move when (they hopefully) start a relationship, so its little wonder that in this study (with the questions being asked) that its men who are more likely to be the perpetrators of sexual harassment. Woe betide any man who deigns to make a physical pass at any woman at the office Christmas party. But this is human nature. 

Supposing a man asks a woman on a date and is refused because she is seeing someone, but then he asks again because he has now heard she is single? Well according to the confines of this study that would be sexual harassment.

But here's where the manipulation gets really ugly. 

Rape is a particularly heinous crime, even typing the word makes my stomach turn. But it would be very easy to file rape under sexual assault, because after all, rape is a sexual assault, but according to this study, so is touching somebody on the arm.

I cant help thinking the people who compiled this study are of a very progressive, left leaning mindset. So I looked them up (by the power of the internet) It turns out, they are ALL of a very left leaning, progressive mindset. Not that there's anything inherently wrong with someone's political views (as long as they're not an extremist) 

A more balanced member of the team may have encouraged an expansion of the study to better incorporate important substance and nuance. Or at the very least rooted out the asinine questions.

So to sum up.

I feel a decent study on workplace harassment would have tremendous benefits. 

But this isn't it

One more thing. If an article doesn't cite its source, then you're probably looking at propaganda.

Source