TORONTO—A detailed statistical review of the 13 statistical graphs contained in the Conservative changebook platform document finds that not one of them conforms to the normal requirements of academic or professional practice.
That’s the conclusion of a detailed review of the platform by Jim Stanford, economist and Research Associate with the Ontario office of the Canadian Centre for Policy Alternatives.
Stanford finds that at least three of the graphs (which illustrate various statistical arguments related to the Conservative platform) present data that is clearly false. All of the others contain major errors in the labeling of variables or axes; internally inconsistent or manipulative scaling of bars and data; and misleading or incomplete references to source data.
For example, in several cases, the axes of graphs have been manipulated, without labeling, to artificially increase the size of particular data points and mislead readers about their relative importance. In other cases, source notes cite government reports which did not actually print the data being portrayed. One graph presents statistical claims that are calculated from outdated forecasts (rather than actual data), without indicating this to readers; when actual (rather than forecast) data is used, the graph’s conclusion is reversed completely.
“The statistical graphs in the changebook have been presented in ways that are clearly unacceptable in normal academic or professional practice,” Stanford says. “They consistently mislead the reader about the relative proportions of the variables being discussed, and reflect a consistent willingness to bend the statistical truth.”
“Whatever one thinks of the changebook’s policy proposals themselves,” Stanford concludes, “the consistently misleading and inaccurate statistical representations contained in the changebook are lamentable.”
–30–
For media inquiries, contact: [email protected].
