Page 214 - Coincidences in the Bible and in Biblical Hebrew

193
CHAPTER 13 METALS AND OTHER SUBSTANCES
CHAPTER 13 METALS AND OTHER SUBSTANCES 193

A further analysis, which excluded tin, is displayed in Figure 13.2. The regres-
sion correlation jumped from 0.97858 in the previous analysis to 0.99113, though

the significance of the results is somewhat weaker (for n = 4, the model F-ratio is
111.2, with p = 0.00875).

In both analyses, the likelihood of obtaining by chance F-ratio values these
high (or higher) is smaller than 1%.
13.4.2 Substances

This analysis comprised all observations detailed in section 13.3. A first regres-
sion run stored a surprise. The results of this analysis are displayed in Figure 13.3.
Apparently, there are two distinct groups here: one group that comprises lime,
15

quicklime, silica, and sodium (where the latter is represented by borit), and

another subset that includes all the other observations (with sodium representable
14
15
either by neter or borit —the former was included in the analysis).
We will shortly allude to a possible explanation for that bizarre partition of
observations, which still exhibits an interesting pattern.

Analyzing separately the first group, we obtain the results displayed in Figure

13.4. The regression correlation is 0.9998, and for n = 4, the model F-ratio is
4372, which is highly significant (p = 0.00023).

Analyzing the second group, we obtain the results displayed in Figure 13.5.

The regression correlation is -0.9915, and for n = 4, the model F-ratio is 115.6,
which is highly significant (p = 0.008542).

Why the partition of the two groups? One possible explanation is that the
two groups contain elements from different chemical groups, as these are defined

in the periodic table . Thus, for the first group, we have composites of calcium

(alkaline metal), silicon (semi-metal), and sodium (alkali metal). For the second

set of substances, we have sodium (alkali metal), sulfur (nonmetal) and lead (basic

metal).

A linguistic explanation may be provided based on the size of the MaNV val-
ues in the two groups. All substances in the second group have extremely high
values of MaNV, which have put these substances in a set of their own. The high
values of the MaNVs still needs elaboration.
Both explanations are not satisfactory, and some more in-depth exploration is
needed. The high statistical significance is undebatable.