PDA

View Full Version : question about methodology



how yes no 2
01-11-10, 00:14
I did some math few weeks ago, in order to try to very roughly determine different waves of settlement of two related populations, and I wonder how wrong can be method that I used. I took data from
http://onlinelibrary.wiley.com/doi/1...21235/suppinfo

haplogrupa____Srbija____Crna Gora
I1_____________7.8%_____6.2%
I2b1___________1.67%___1.73%
I2a2 __________38.5% ___29.2%
E1b1b _________ 17.3%___27.0%
R1a____________14.5%____7.4%
R1b____________ 4.5%____9.4%
J1______________0.6% ___0.5%
J2a ____________3.3%____4.7%
J2b ____________1.7% ___4.5%
G2a____________2.2%____2.5%
N ______________3.3% ___1.5%
Q _____________1.7%____2.0%
H______________2.2% ___1.5%
L_______________0.6%____1.2%

than I decided to very roughly determine separate waves of settlement on following way
I assumed that previous inhabitants were more or less homogeneous, and that
waves of settlement are more or less homogeneous. Idea is that if a same wave enters larger area and smaller area, imprints of parts of the wave on those two populations will still keep same proportion...

so, I did divide percentages of haplogroups in Serbia and Montenegro
as a result I could notice several clusters


haplogrup____Serbia____Montenegro__________Serbia/Montenegro


N ______________3.3% ___1.5%____________2.2
R1a____________14.5%____7.4%___________1.96

I1_____________7.8%_____6.2%___________ 1.26
I2a2 __________38.5% ___29.2%___________1.32


E1b1b _________ 17.3%___27.0%___________0.64
J2a ____________3.3%____4.7%____________0.7


I2b1___________1.67%___1.73%___________ 0.97
G2a____________2.2%____2.5%____________0.88
Q _____________1.7%____2.0%____________0.85

L_______________0.6%____1.2%___________0.5
R1b____________ 4.5%____9.4%___________ 0.48
J2b ____________1.7% ___4.5%____________0.38


H______________2.2% ___1.5%____________1.47

J1______________0.6% ___0.5%____________1.2


H and J1 I have excluded because H are gypses, and J1 could be influence of Turks, Avars or whatever...it's too small to consider it anyway...

I reasoned that N could not have been common among previous inhabitants...
it seems that R1a and N are in same cluster, that is they have mostly arrived in same wave..according to ratio of R1a and N this wave would be some north Slavs... in addition, we can see that this wave had 2 times larger impact in Serbia, which is very logical since Montenegro is mountains area and it is logical that newcomers settle plains while previous inhabitant retreat to mountains......

related to that we can see a cluster with J2a and E1b1b that represent previous inhabitants since it is stronger in mountains area

than there is J2b for which we know that it was spreading along sea costs and was thus already in previous inhabitants more spread in Montenegro ...logically J2b is more present in Montenegro than it is the case for previous inhabitants that were homogeneously spread...thus number 0.38 is logical...
from this we conclude that similar story holds for R1b and L, that they must have been more spread in Montenegro already before other waves arrived...
.

I2b1, Q i G2a have ~ same percentages and makes another cluster..My conclusion was that this was a wave from north: Q could be Avars, I2b1 Gepids and G2a Sarmats Iazyges all together from panonian plain

I2a2 i I1 seems to be separate wave of settlement......
here I was puzzled... Serbs or Goths? or two waves that were spread in similar ways between Serbia and Montenegro?

Note that number for R1a is larger than for N indicating it also came in other waves in some percentages....Avar wave had generally small impact, so even if R1a was very dominant there it would be only small part of R1a

So, I am curious what you think about methodology, and results...

LeBrok
01-11-10, 09:51
Not enough data to get right conclusions. Wait for more research into ancient haplogroups before going there.

how yes no 2
17-11-10, 23:36
sorry, the link above was broken...

http://onlinelibrary.wiley.com/doi/10.1002/ajpa.21235/suppinfo