Earlism #50 Do the Math People … Why I Hate Cable News, Polls, & Surveys
“Correlation & Causation are essential …
… if you plan to draw a realistic Conclusion …”
I am going to include a Long Form Post as an explanation, with this Earlism.
Now, I know that most people are “math averse.” More simply put, I know most people hate math!
But stick with me …
I am actually not math averse. I am pretty comfortable with math and statistics. Although I hide this fact from most of my general conversations, because I know I am among the few, and not the many.
Due to my level of comfort with math, many times when I watch Cable News, any of the channels, I get so OFFENDED when they discuss and cite polls & surveys as definitive facts. And try to use them as proof. I get offended, because I actually understand the math and statistics related to what the Talking Heads on Cable News are saying.
I know why mathematicians always say that:
“there are lies, (darned) lies, & statistics.”
They should have also thrown in “Polls & Surveys” at the end of that list of lies. Alas, they would never do that because most math averse Talking Heads love citing (stupid) surveys!
Anyway, the Cable News Talking Heads, and others who watch and unknowingly repeat them, throw specious statistics at you based on survey data. Then they improperly apply the survey results. And use the results to draw a false conclusion or a “False Positive” (Type I Error).
See Earlism #51: Apples, Oranges, & False Positives, which I will post tomorrow …
They will say; since all, or a majority of “all people” stated on the survey that they like or dislike someone or something …
… that Cause …
… means that this person or thing is …
… in Effect …
… liked or disliked by most or “all people.”
Causal, or Cause & Effect reasoning.
What they will do is take some public affairs issue or idea that they espouse.
Then, based on a survey that they have taken, without regard to appropriate random sampling technique or respondent size.
They say that “the survey shows that the public likes or dislikes, this person, policy, or public official.”
However, they cannot definitively say that the “public” likes or dislikes something or someone, when they do not know how many people, randomly chosen, from all walks of life, actually shared their opinion on the survey question.
No human being, surveyor, or Cable News Talking Head, can definitely relate two things without proof.
Nor can they “Co-Relate” (Correlate) the majority public opinion, which is generally based on a “feeling” anyway.
A “feeling” which will likely change within the next fifteen minutes.
As opposed to the survey respondent basing their stated response to the survey question on a critically analyzed and well reasoned “thought.”
No one who understands Surveys, Statistics, & Math can Correlate two separate entities, without showing how they overwhelming relate to each other.
And without showing how they collected a representative sample of the opinions of most or all people.
Then the Cable News anchor, or the people who conduct and report their survey. Or a member of the viewing audience who repeats this nonsense, will cite that survey and its poorly derived conclusion, to “support” their point of view.
You cannot draw a proper Conclusion about “what ALL or most people think” …
… if you have received poorly collected and analyzed survey data.
AND you cannot definitively draw a Conclusion about what “most or all people think” …
… unless you ask “all of the people” …
You have to ask them what THEY ALL believe. Or at least, get an actual, randomly selected sample of responses from “all types of people.”
Because, you did not ask “ALL PEOPLE.”
Nor did you ask a “random sample” of “ALL PEOPLE.”
You only asked:
all of the people who were home that day, or those who watch your Cable News channel, or who receive your emails, or those who decided that they were having a BORING enough day that they could waste five minutes and respond to your survey. Or those people who just saw the survey online, on YOUR website and clicked it. Or those people who signed up with your survey group to take YOUR surveys!
That is not “All People” … People!
That is a survey of “all of your people.”
Thus, you cannot draw a definitive conclusion, that (Co) Relates to what “all people” think …
… unless you ask “all people.”