A colleague sent me a story by Cecilia Kang in the Washington Post: Survey finds gap in Internet access between rich, poor students. With my interest in programs to get connected computers into low income households, my friend knew I would be interested in the article which talks about a survey released Thursday by the Pew Research Center.

Indeed, I would commend the Washington Post article and the survey itself to you for reading.

I want to highlight the problem representing the survey results in the Washington Post. The fifth paragraph says:

Half of all students in higher income families have access to the Internet at home through a computer or mobile device. The figure drops to 20 percent for middle income children and just 3 percent of students from poor homes, according to the survey of 2,462 teachers by the Pew Internet & American Life Project in cooperation with the College Board and National Writing Project.

Something seemed off with those figures.

After all, I recalled that two weeks ago, I wrote about digital literacy programs trying to deal with the one-third of American households that aren’t on-line. How could it be that half of wealthy households with kids were without internet, if the national figures show two-thirds of households have internet access.

Something wasn’t right.

In going to the actual Pew report, I found the likely source of the Washington Post numbers. But Pew didn’t actually report on the availability of home internet by income. It was a different question.

The survey reported “% of teachers who say ALL or ALMOST ALL of their students have sufficient access to the digital tools they need [at home / at school] to effectively complete school assignments, by student socioeconomic status”.

This question may point to whether teachers have to adapt homework assignments; can the teachers assume that digital tools will be available?

The Washington Post appears to have treated these numbers as though the question ask “percentage of students who have home internet, by income.”

The Washington Post question is important to understand and address, but it was not addressed in the survey. And as a result, the numbers were just plain wrong.