Letter to the editor: Bias in numbers

Regarding Lori Falce’s column “Truth, math and politics” (April 30, TribLIVE): Sorry, numbers do lie, numbers are biased, numbers are political and numbers do have agendas. You forget people with biases and agendas construct, collect and report the numbers.

America’s one-third of all worldwide cases despite having 4% of the world’s population is scary on the surface. But it is an “apples to oranges” comparison if America is classifying and counting deaths and cases in a different manner than the rest of the world.

From Google’s official covid-19 statistics page, May 4, Russia reported 1,356 deaths and 145,000 cases. That is a death rate one-sixth of the United States. Are Russia’s numbers to be taken as fact or propaganda?

Polling numbers are easily manipulated. Newspapers do not routinely state the composition of polling samples, just “a sample of likely voters.” By skewing the sample (maybe 75% Democrats, 25% Republicans, or vice versa), it is easy to get whatever polling results desired. Just changing the wording in surveys can introduce result bias.

Even basic math should be questioned. I recently found a math error Google cited as an example of calculating accuracy, of all things. The website corrected its error the next day.

Numbers have an undeserved aura of fact. We are best off questioning every single number we are presented with. It is not enough to look at numbers without bias. We must first look for the bias in the numbers.

Reed Pederson

Hempfield