That Pew Online Harassment Survey

Suggests that lots of people have been harassed online:

Forty percent of American adult Internet users say they’ve been harassed online, according to a new survey by the Pew Research Center, which found the harassment is often by strangers on social-networking sites.
More than half of those who experienced online harassment did not know the person involved in the most recent incident of harassment. Some 38% said a stranger was responsible while 26% said they didn’t know the real identity of the people involved. Two-thirds of those who experienced harassment said the incident occurred on a social networking site or app.

What does the survey mean by harassment?  It didn't define harassment but

... asked respondents if they had witnessed or been subject to six behaviors: offensive names, physical threats, harassment for a sustained period, stalking, purposeful embarrassment and sexual harassment.

Because studies of online harassment are still in baby diapers all the surveys I've seen have various types of problems.  For example, this survey might classify vastly different levels of nastiness into the same group.

Take calling someone names.  That could cover labels such as "stupid," "c**t, "n***er" and "fa**ot."  The anger reflected in those is of a different nature.  The last three are class slurs, the first one probably isn't.

And what does purposeful embarrassment mean?

All this matters when the next stage compares percentages of people experiencing online harassment.

Another aspect which matters is how to control for the types of sites people visit.  Some of the finding that young people are especially likely to be harassed may be because of the types of sites they visit.  And yes, I get that the sites we visit are partly our own choices and not extraneous determinants of the level of harassment.

All this means that I'd take the most recent Pew survey with a dose of salt.  Still, it's probably true that the group which faces the highest level of online harassment consists of young women.  From the survey summary:

Online harassment is especially pronounced at the intersection of gender and youth: women ages 18-24 are more likely than others to experience some of the more severe forms of harassment. They are particularly likely to report being stalking online (26% said so) and sexually harassed (25%). In addition, they are also the targets of other forms of severe harassment like physical threats (23%) and sustained harassment (18%) at rates similar to their male peers (26% of whom have been physically threatened and 16% of whom have been the victim of sustained harassment). In essence, young women are uniquely likely to experience stalking and sexual harassment, while also not escaping the high rates of other types of harassment common to young people in general.

Other differences the survey found were between black and Latino users on the one hand and white user on the other:  The former were more likely to report having experienced harassment.  This suggests that the intersection effect for young black women and Latinas might be quite high.

I would like to see an online harassment survey which more carefully separates harassment by acquaintances from harassment by strangers and which also looks at the reason the harassment happens in a different way.

To explain the latter by an example:  It matters quite a bit if I'm harassed by a stranger because I wrote something controversial or even nasty or just because the harasser decides that I am female (insert here any other group category of the same type)  and therefore obvious prey for harassment.