I wonder if there is a correlation between twitter use and the length of time owning a smart-phone. Most people that I know don't tweet from their work or personal computers, but use their smart phone.

I'd hazard that some of the cause for the two uneven distributions arise from this trend.

Although why it may be associated with race, I haven't a clue.

"...but Whites (who get less education than Blacks and Hispanics)". Really? 

So people with more education are more likely to use Twitter,
but Whites (who get less education than Blacks and Hispanics) are less
likely.  There’s something really interesting going on here.  Any
idea what?

It is methodologically unsound to make the association as you did here. Without cross-tabulation, it's not clear whether there is such an association between college education and race to the extent that you are drawing it. I'm not saying that it's not there: just that you can't make the conclusion without supporting evidence.

Maybe this is wrong of me, but I always assumed that white people got more education than Blacks and Hispanics.

Theory: Twitter use goes up with education (as well as general "plasticity" to new tech).

This means that a lot of white demographics are either too young (teenagers) or too old (seniors) to have the college education needed to flock to twitter effectively, whereas Blacks and Hispanics skew younger (which means more of them are less likely to have the aging-out problem).

I'm not very satisfied by the income-table. Groups income ist not equally distributed and there is no comparision with the actual number of ppl. in this group. Next ist the fact, that the whole study is nonsense for non-US as the distinction between black (P.O.C.), white, hispanic (P.O.C.) an other groups is not that easy to make. Where is misc? Is see a lot of unexplainable numbers that mean nothing to me.

My suspicion would be that Twitter use among Blacks and Hispanics was relatively flat with educational achievement, but strongly skewed to the college educated among Whites, leading to these trends.  But without a breakdown by both race and education it would be hard to tell.

This post is...peculiarly written. I went to look at the Pew study, and one thing that's not clear from the tables included in this article is that the % is not "% of all twitter users", it's "% of people in category [X] who report using twitter".  So 25% of black, non-Hispanic American internet users surveyed use twitter, while only 9% of white, non-Hispanic American internet users use twitter.  The tables as excerpted seem to imply "25% of twitter users are black, non-Hispanic Americans" which would be a truly bizarre finding (given how widely twitter is used outside the US).

It does seem clear from the original report that, for some reason or reasons, non-white Americans are adopting twitter at higher rates than white Americans. I don't see any way to get from the data provided to a definite conclusion about any demographic group's rate of education, except that the Methodology section does indicate that the selection group was weighted on a number of parameters, including education, to conform to the US general population - and that non-Hispanic whites were weighted separately (it's not clear to me from the wording if that was done by correcting collected data, or at the selection phase).  Based on that info, I would assume that white, non-Hispanic Americans surveyed tended to have higher levels of education than the other groups, because that's what the general population looks like.

These tables are incredibly confusing and not well explained at all.  The first table is % of total internet users who use twitter AND fall into those income brackets?  Maybe?  it doesn't add up to 100% . . .

A more interesting statistic would be % of TWITTER users and their income bracket.  Although maybe the distribution would be the same?  Ditto for the rest of the charts.

I'd prefer not to see a post about this than see one that is this unclear.

1) An age crosstab is really needed.  The median age of African-Americans and Hispanics is significantly younger than the median age for White Americans. Twitter skews young.

2) I assume this doesn't explain the whole difference. My guess is that Twitter is more popular in cities than in rural areas, because it is complimentary with real life interaction (as opposed to Facebook which is in many ways a substitute).  My image of the average Twitter user is an urban post-college white person or an urban teenager/young person of any race or creed.

So, Twitter has a 54% market penetration?!  I sorta find that hard to believe... Was the survey conducted on Twitter? :-)

I think this needs to be compared to other social networking sites, as well as general internet use, to be terribly sound.  Would like to see the comparison if it exists, though.

I don't know about the quality of the posts that Lisa Wade contributes to this website, the rage comic post, the 9/11 post and now this incoherent chart with poorly worded so-called "analysis"?

"There's something interesting going on here. Any idea what?"

No, because the charts don't make sense. Also, what do YOU think? Because that's half the reason I'm on this site, you know, for actual, well thought out analysis.

Polls are useless.  

For the landline sample, interviewers asked to speak with the youngest adult male or female currently at home based on a random rotation. If no male/female was available, interviewers asked to speak with the youngest adult of the other gender. gender. For the cellular sample, interviews were conducted with the person who answered the phone.

Here's one bias: The sample did not attempt to select from 'internet users', it selected from 'young adult internet users' and 'adult internet and cellphone users'. That fact was misrepresented in the summary.

Polls are always bias. They never give the correct information.