Thursday, March 29, 2018

Eroding credibility one error or misrepresentation at a time.

The continuing campaign to advocate for numeracy in journalism.

Ip is implying that there must be some political payback in project selection because 66% of projects are going to counties which voted for Trump.

If political payback is the causal mechanism for project selection, then 66% is too low. Trump won 2,654 of the 3,141 counties in the US. In other words, he won 84% of all counties. The numbers fail to support Ip's implied hypothesis that projects are selected as political payback.

Trump supporting counties are roughly 25% under-represented in terms of where projects are being funded.

Ip's implication seems unsupported. It appears that projects are being selected on factors other than how a county voted.

Whether Ip is deliberately omitting numeric context or is instead ignorant of the context cannot be known. What is known is that his communication is misleading by implying that project selection is solely determined by whether or not a county voted for the president.

No comments:

Post a Comment