Footnotes
1.
Isaiah
Berlin; “The Sense of Reality”; Farrar, Straus and Giroux; New York, N.Y.;
1996; p.p. 44-45.
2.
John
Kay and Mervyn King; “Radical Uncertainty”; W.W. Norton & Company; New
York, N.Y.; 2020; p. 320.
3.
Ibid.;
p.147.
4.
Ibid.;
p. 218.
5.
Ibid.;
p.p. 210-212.
6.
Ibid.;
p. 310.
7.
Ibid.;
p. 433.
8.
Study;
p.p. 22-23.
9.
Ibid.;
p. 37.
10.
On 12/12/24 we asked BARD the following question;
and received a sophisticated statistical answer:
Us: In an OLS (ordinary least squared)
business study, the authors found a very significant independent variable, but
they then defaulted the whole study to correlation rather than cause because
the equation R-squared was rather low. Why did they do this?
Bard: (It was a valid action.) Here’s
why:
·
Omitted
Variable Bias: Important variables that influence the dependent variable might
be missing from the model.
·
Heteroscedasticity:
(our observation of price behavior) Unequal variances in the residuals can
impact (general) model reliability. (But) a low R-squared doesn’t necessarily
invalidate the findings. It’s crucial to carefully consider the context, the
nature of the data, and the research question to interpret the results
appropriately.
In
essence, the authors exercised caution by recognizing the limitation of their
model and avoiding overstating the causal relationship based solely on the
statistical significance of one variable.
Therefore,
they also considered a qualitative instance of a significant specific person.