A bit of a rant. But I like talking about methodology.

I'm a little disappointed with the Questions feature of the website. A lot of the questions I've been answering are very loaded! Look at this one (drawn from memory):

Is it ever right to strike a child so hard it leaves a red mark for several hours?
Yes
No

My problem with the question is with the subordinate clause "it leaves a red mark for several hours?". If the purpose of the question is to determine whether someone approves or disapproves of spanking as a punishment there will be a significant bias in the result. Instead of allowing for the opinion "There are occasions where spanking may be appropriate", a yes answer would logically indicate that any degree of strike from a pat on the behind to a full-forced slap would be acceptable. The result (I hope?) ? Everyone clicks no.

This "no" is equally valid to every other question on the subject of child raising. But its a false negative. Every time it gets asked it reduces the accuracy of the database of answers people have made to that subject. This means the system lables you a little less in favor of spanking even if you were more or less neutral on the subject.

I'll stop here for a second. It's unfortunate that the offending question happened to be about a politically touchy subject. It is not my intention to suggest that anyone would want to appear more in favour of spanking. What I'm talking about here is the disruption of data on a site that makes comparisons based on that data.

I don't like bias. I guess I just think things should be accurate. It's always seemed to me that heavily loaded sentences like that are often the product of laziness and ignorance. One of the reasons I love linguistics is the focus on methodology and theoretical frameworks involved in creating "information" and these kinds of loaded questions and statements occur way too often in our media.

I remember reading an article about "the linguistics of blame" where the researcher looked at bias against battered women in newspapers in England. They looked at newspaper headlines like "Sex-starved hubby strangles wife and "Blonde attacked in park".

Anyway, the point of the article was that by providing an excuse in sentences like the first one they were in a sense creating a "well she had it coming" feel to the statement. In the second sentence they call her "blonde" instead of "woman" or even not including her at all and using a headline like "attack in park". Since everyone *knows* blondes are more fun, there is the idea that the woman was too much temptation, or in some way sexually available to the attacker. Notice the passive construction? They don't even say anything about the attacker!

This isn't about violence against women (or anyone else, for that matter). It's about recognizing the importance of the words used when you say something. I hope anyone who reads this is a little more careful with their words in the future.