CNN is reporting that a chart used in a report on the 2016 election has been found to be too close in its histogram, making it less representative of the entire data set.
The data was gathered from a series of state polls conducted by Edison Research and the Public Religion Research Institute, which were combined to create a graph for the CNN article.
The report also notes that a large percentage of the data is not reported in the chart’s source code, leaving many users to infer that the actual percentages are much higher than they actually are.
The report notes that the data comes from a survey of 1,876 registered voters conducted by the Public Interest Research Group, which was conducted between November 15 and November 22, 2016.
It notes that while the poll was conducted in real time, there is some lag in reporting the results of the poll.
For example, the data shows that while Clinton has a 3.1 point lead in the poll, the actual lead is only 0.1 percentage point.
The exact margin of error for the poll is 0.3 percentage points.CNN is reporting the data on its website.
In a statement, Edison Research spokesperson John Boudreaux said, “The data we used in our analysis is not representative of a representative sample of registered voters in the U.S. The Edison poll sample has a margin of sampling error of plus or minus 4.7 percentage points.”
The data analysis is a clear violation of the guidelines for how data from surveys should be reported in publications.
According to guidelines from the National Institute on Money in State Politics, published by the Pew Charitable Trusts, reporting on data from such surveys should include “a breakdown of the sample, including characteristics, including race and ethnicity, religion, gender, age, education, and household income, among other variables.”