Sie sind auf Seite 1von 19

Plain and Simple: On a Novel Feature of Amish

Michael Weight and Henry Harpending
Department of Anthropology
University of Utah
Salt Lake City, UT 84112, USA
October 10, 2014
Positive assortative mating for a heritable trait or collection of
traits, selectively neutral with respect to the population, can be equiv-
alent to directional selection from the perspective of subgroups. This
can rapidly lead to genetic dierences between subpopulations.
We examine consequences of this mechanism for North American
Amish, an Anabaptist population with little or no inward gene ow
but with substantial outward gene ow that has been declining in the
twentieth century. A standard personality assessment, the PF16, ad-
ministered to an Indiana population in 1970 shows nearly complete
divergence between Amish male adolescents and a sample of neigh-
boring non-Amish Indiana males. Without advocating or claiming a
genetic mechanism we show that a partially genetic model accounts
closely for the observed group dierences and provides a baseline for
comparison with other transmission models.
AmishMS/HCH/MW DRAFT Oct 2014
Assortative mating can be equivalent to directional selection. Assortment
for a trait generates either groups or a continuum within a population, those
with more of the trait at one extreme and those with less of the trait at the
other. To illustrate this consider a simple thought experiment of assortative
mating about stature in a population average with average stature of 69
inches for males and 64 inches for females, each with a standard deviation
of 3 inches. These gures are are close to observed values in North America.
Suddenly impose a mating rule such that those individuals taller than the
population mean mate only with others above the mean while individuals
shorter than the mean mate only with others shorter than the mean. The
mean of a half-normal distribution is .8 standard deviations, so after the
imposition of the new rule male and female stature in the tall group will be
71.4 inches and 66.4 inches, respectively, and in the short group 67.6 inches
and 61.6 inches.
The new subpopulations now mate at random within within groups. The
two sets of parents dier by 5 inches in average stature, 69 (the average of
the male mean of 71.4 inches and the female mean of 66.6 inches) in the
tall subpopulation vs. 64 (the average of the male mean of 66.4 inches and
the female mean of 61.6) in the short subpopulation. After reproduction,
assuming an additive heritability of .80 for stature, the group dierence will
be .85 = 4 inches. If the mating rule persists, the tall and short populations
will continue to diverge, but at a decreasing rate. It is important to notice
that after the initial generation there is no further regression to the mean
save for that in ospring of those who change groups. In other words, after
the rst generation of reproduction the genotypic group means dier by 4
inches from then on in the absence of further movement between groups.
Those who change groups prior to mating, for example tall ospring of the
short groups and short ospring of the tall group, will experience regression
to the mean (of their natal group) each generation.
Before the new mating system the standard deviation of stature was 3
inches within each sex so the variance was the square of that, or 9 squared
inches. After a single generation of the new mating system the between-
subpopulation variance is (3 0.8 0.8)
or 3.69 inches squared. (The de-
viation of each new subpopulation from the original mean, using males as
an example, is the old standard deviation within the population, 3 inches
multiplied by 0.8, to yield the change of the mean in the new subpopulation
AmishMS/HCH/MW DRAFT Oct 2014
before reproduction, +0.8 3 in the new tall subpopulation and 0.8 3
in the new short subpopulation.) After reproduction regression to the old
mean occurs in each subpopulation so the deviation from the old mean be-
comes 0.8 3 0.8 or 1.92 inches. The new squared deviation of the mean
from the grand population mean is 1.92
or 3.69 squared inches, the new
between-subpopulation variance. The new within-subpopulation variance is
9 3.69 = 5.31 inches squared. After one generation the fraction of variance
that is between groups is 3.69/9, or approximately 41%.
Richard Lewontin (1972) famously estimated the fraction of genic vari-
ance among continental human groups to be about 15% for a collection of
presumably neutral genetic markers. That gure has remained about the
same as the number of such loci available to has has grown from dozens to
hundreds of thousands. After a single generation of our thought experiment
the stature variance between groups at genetic loci inuencing stature is be-
tween two and three times as great as overall neutral genic variance among
major human populations. Even though there is no selection or dierential
reproduction in this experiment the consequence within groups is equivalent
to truncation selection.
Ethnographic Analogues: The Old-Order Amish
Are there concrete ethnographic examples of assortative mating about a trait
analogous to height with the above kind of strong eect? There are sev-
eral candidates. Cochran, Hardy, and Harpending (2006) proposed a similar
mechanism in the evolution of Ashkenazi Jews of Europe during the Middle
Ages: the Ashkenazim were a population closed to immigration, but with the
possibility of members leaving and they experienced strong selection for eco-
nomic and managerial success. The Parsi of India, a closed class of mangers,
scholars, musicians, and other economically and culturally successful individ-
uals (Nelson, 2012), are another distinct population in which assortment may
have been important in their evolution. Recently Charles Murray in Coming
Apart 2013 proposes that assortative mating by education, socioeconomic
status, and IQ is splitting America into distinct and dierent classesthe
coming apart of the title.
The Old-Order Amish in North America also follow the pattern of negli-
gible inward, but signicant outward, gene ow, making them an attractive
candidate for selection by assortment. We describe a model in which there
is a underlying trait that we call AQ or Amish Quotient by analogy with
AmishMS/HCH/MW DRAFT Oct 2014
IQ as derived from cognitive tests. We postulate in this model that AQ is
predictive of the probability of an individual remaining in the community
when the choice is made at adolescence to remain and be baptized or to
decline and emigrate. Specically our measure of AQ, a collection of self
reported personality traits, is associated with the plain and simple lifestyle
of the Amish, postulated to indicate an underlying anity for, or attractive-
ness of, membership in the community to an individual. Under the proposed
mechanism of selection the AQ would increase each generation as those with
lower AQs are truncated each generation through boilinig-o or defection.
We therefore predict declining boiling-o rates over time are a consequence
of this selection mechanism increasing overall population Amishness (AQ)
each generation.
The model performs well relative to alternative models of Amish defection
and retention (see Ericksen et al. (1980); Meyers (1991, 1994); Greksa (2002)
for review). While Old-Order endogamy is well described and understood,
the trade-os and constraints of defection are unclear. Reductions in defec-
tion rates over the last century (Greksa, 2002; Meyers, 1991, 1994; Kraybill,
2001; Hostetler, 1993) need explanation, so any comprehensive model should
also account for this trend. Previous models of Amish defection emphasize
rapid population growth in regions with nite resources, particularly expen-
sive farm land, and the resultant trade-o between culturally sanctioned high
fertility and the ability of families to support their own farms (Ericksen et al.
, 1980, 1979; Markle & Pasco, 1977; Wasao & Donnermeyer, 1996). Other
models focus on eects of occupational shifts consequent to economic con-
straints on farming. These models predict increasing exposure to non-Amish
lifestyles may tempt youth into defection or interfere with parents ability
to transmit Amish identity (Meyers, 1991, 1994; Greksa, 2002). Unfortu-
nately these more familiar approaches to the issue are dicult to quantify
and corroborate, and therefore wont be discussed further.
Old-Order Amish Endogamy
As the most traditional and plain Anabaptist group, the Old-Order Amish
are pacists who reserve baptism for believing adults and have the strictest
shunning practices of deviant church members (Hostetler, 1977, 1993; Roth,
2002). Such non-hostile ethnocentrism and social solidarity introduce a
near absolute boundary that culturally uncouples the Amish from the En-
glish(their word for non-Amish) (Huntington, 1957; Savells, 1988). Familiar
AmishMS/HCH/MW DRAFT Oct 2014
Amish norms like carrying intergroup discussions in Pennsylvania Dutch,
favoring horse-and-buggy over automobiles, traditional humble clothing, ab-
solute pacism, and forsaking appliances like telephones, televisions, refrig-
erators, and indoor toilets are means to signal this un-coupling from their
broader society (Kraybill, 2001; Fuchs et al. , 1990). Geneological and genetic
analysis reveals the Amish have been signicantly isolated since Old-Order
founding populations immigrated in the 18th and 19th Centuries (Lee et al. ,
2010; Pollin et al. , 2007; Hostetler, 1993; Kraybill, 2001; Hurst & McConnell,
2010). Arguably the Old-Order Amish are the most culturally isolated An-
abaptist group in the United States (Hostetler, 1993; Kraybill, 2001; Nolt,
Following this isolation the Old-Order substitute communal aide for sec-
ular public support. Initiatives like welfare, Social Security, life insurance,
barn insurance (Hurd, personal communication), government agricultural
subsidies, or public schools (if Amish schools are available) are therefore pro-
hibited or strongly discouraged (Hostetler, 1969, 1993; Kraybill, 2001; Hurst
& McConnell, 2010; Fuchs et al. , 1990) in the interests of communal self-
reliance. Amish substitutes for institutional support are primarily mediated
through kin networks so extended families bear the emotional and nancial
burden of the welfare of their members (Hostetler, 1993; Hurst & McConnell,
2010; Hurd, 1983). Children, especially males, rely on kin support to start
businesses and farms when they begin their own families (Ericksen et al.
, 1980; Wasao & Donnermeyer, 1996; Ericksen et al. , 1979; Hurd, 1981).
Highly fertile Amish mothers draw on relatives to help with their large fam-
ilies (Hewner, 1998; Hurst & McConnell, 2010; Kraybill, 2001). Costs of
disabled, ill, and senior members are mitigated by family and community
support (Hewner, 1998). The Amish are a healthy population with low inci-
dence of infectious disease, low overall infant and adult mortality rates, high
fertility, and high standards of living (Cross, 1969; Mckusick et al. , 1964;
Fuchs et al. , 1990; Hewner, 1998; Greksa, 2002; Cross & McKusick, 1970;
McKusick et al. , 1964).
In 1968 Paul Wittmer submitted a doctoral dissertation (Wittmer, 1968) at
Indiana State University in which he aministered two psychological question-
naires to 25 18-20 year old Amish young men and 25 young men of the same
age who were not Amish, all from Daviess County, Indiana. One question-
AmishMS/HCH/MW DRAFT Oct 2014
naire had to do with relations of these men to their parents, but it apparently
not popularly used and so we have no comparative data about it. The other
was a widely used personality assessment called 16PF (for 16 personality
factors) that was developed in the 1940s by Raymond Cattell and others at
the University of Illinois (Cattell and Eber, 1972) and went through several
revisions over the years. There are 180 questions on the form, from which are
extracted 16 numbers supposed to represent the personality of the subject
on 16 dimensions of personality. These are listed in Table 1 along with the
adjectives in common use that supposedly describe each measured trait.
We make no serious use of the details of the traits in this paper: we
simply regard the 16 numbers for each person in the study as self-reports of
interesting characteristics. The test and the scoring system were designed to
maximize the independence of the factors so the scores are not redundant
indicators of the same thing. In some of the literature factor B is not reported
nor used since it is regarded as an assessment of intelligence rather than of a
personality trait. We include B with the other 15 since we treat these data
in the same way that we would treat genetic markers such as SNPS, i.e. we
are agnostic about the interpretation and meaning of the factors.
Scores on personality assessments are substantially heritable (Turkheimer
et al. , 2014; Cattell et al. , 1985). This means that if there is selection
that aects any of them then they should change over time. Our textbooks
ordinarily dene additive heritability as the fraction of variance of a trait that
is caused by additive gene eects. Plant and animal breeders, on the other
hand, assess heritability by the response to selection. If the ideal quantitative
genetic model is a satisfactory description of the trait and the organism, these
denitions of heritability are the same.
Heritability in the human sciences has been curiously unapparent since
the middle of the twentieth century. With some exceptions it has been reso-
lutely ignored and occasionally viewed with revulsion, as if it would go away
if never mentioned. This was a consequence of the twentieth century strong
social science (Boasian) fad maintaining that human biology and human
behavior must be uncoupled. There were routine denunciations of heritability
studies (e.g. Kamin & Goldberger, 2002) picking at various possible weak-
nesses. For example much data on identical and fraternal twins, separated at
birth or not, was collected. These showed essentially that the environment
in which a child was raised had little eect on the development of the child
while shared genes had a large eect. A valid criticism was that twins had
shared the same uterine environment for the 9 months of mothers pregnancy
AmishMS/HCH/MW DRAFT Oct 2014
Trait Label High Low
A Outgoing Reserved
Warmhearted Detached
B Abstract thinking Concrete thinking
Fast learner Slow learner
C Unemotional Emotional
Calm Changeable
E Assertive Humble
Dominant Cooperative
F Cheerful Sober
Lively Taciturn
G Conscientious Expedient
Persistent Undisciplined
H Venturesome Shy
Socially bold Retiring
I Tough-minded Tender-minded
Self-reliant Sensitive
L Suspicious Trusting
Sceptical Accepting
M Imaginative Practical
Bohemian Conventional
N Shrewd Forthright
Discreet Straightforward
O Guilt-prone Resilient
Worrying Self-assured
Q1 Radical Conservative
Experimental Traditional
Q2 Self-sucient Group-dependent
Resourceful Aliative
Q3 Controlled Undisciplined
Compulsive Lax
Q4 Tense Relaxed
Driven Tranquil
Table 1: The 16 traits derived from the 16PF assessment with the adjectives that
conventionally describe each. We use these simply as anonymous markers of responses
to diverse questions by the subjects of the study.
AmishMS/HCH/MW DRAFT Oct 2014
and that this amounted to an important shared environment.
Within the last decade these criticisms of twin studies have essentially
been rendered moot by new data related to the human genome project. In
the past identical twins were of interest because they shared the same (more
or less) genome, but suspicious of shared environment persisted. Today the
state of technology is to type a million or so SNPs in a large sample of subjects
and to measure relationship between individuals by the number of SNPs they
share. Subjects, on the order of tens of thousands, are unknown to each other
and the level of kinship assessed is on the order or fourth cousins and lower,
more or less. This technology has instantly rendered nearly obsolete studies
of real know family relationship, even while validating the conclusions that
had been so widely denounced in the twentieth century.
Much of this innovation in quantitative genetics has been due to the
Australian quantitative geneticist Peter Visscher and his colleagues (Davies
et al. , 2011; Yang et al. , 2010; Visscher et al. , 2008). Eric Turkheimer, a
prominent sceptic of heritability of intelligence and personality in humans,
recently (Turkheimer, 2011) threw in the towel:
Thanks to the Visscher program of research, it should now be
impossible to argue that the whole body of quantitative genetic
research showing the universal importance of genes for human
development was somehow based on a sanguine view of the equal
environments assumption in twin studies, putting an end to an
entire misguided school of thought among traditional opponents
of classical quantitative (and by association behavioral) genetics
Wittmer provides 16PF scores for 25 Amish and 25 non-Amish young men
from the same rural Indiana county. There are 16 scores for each subject.
To make sense out of the array of scores we use a conventional for of data
reduction called principal components analysis (PCA). If we have data from
three subjects on three traits we could compute a distance among subjects
and plot those distances on a piece of paper. Distance between two things
can be plotted on a line, among 3 things in two dimensions, among four
things in three dimensions, and so on. We can compute a distance between
all pairs of the 50 subjects but such a picture would sit in 49 dimensions and
be of little use.
In parallel we could as well compute distances (dissimilarities) among
all pairs of the 16 traits, computed over all subjects, rather than distances
AmishMS/HCH/MW DRAFT Oct 2014
among subjects computed over all the traits. The distances would sit in a 15
dimensional space since there are only 16. Because there are only 16 traits,
the distances among the 50 subjects also sits in a 15 dimensional space, rather
than 49, and these spaces have essentially the same dimensions. They are
said to be dual to each other.
Since we cant visualize the data in all their 15 dimensional splendor we
lose some information by looking only at a two-dimensional slice through the
swarm. This slice is computed so that the disperal of subjects or traits along
its rst dimension is maximized and along its second dimension is maximized
after the rst dimension is removed. Of course we could build a model or
even a picture of the best 3 dimensions rather than just two but thse models
hardly even provide much insight.
This procedure is called PCA (Principal Components Analysis), but no-
tice that we want to look at two spaces rather than just one. The subjects
space plots principal components and the trait space plots principal coor-
dinates: we look at them simultaneously to visualize patterns in the data.
(Look at the two panels of gure 1 or gure 2, which is the easiest way to
make sense of the last sentence.) Further, by analogy with cognitive tests,
we will refer to the rst (largest) dimension as the AQ or Amish Quotient.
This is, for each subject, in our model, an estimate of how Amish that person
is, meaning what that persons score on our Amish quotient.
The process we are using is analogous to that of estimating IQ. Although
there is some disagreement, an important nding from cognitive testing is
that cognitive ability is essentially one-dimensional. People with high math-
ematical ability also have large vocabularies, and so on. One empirical result
of this is that almost anything we construct functions as an IQ test or a proxy
for a proper IQ test. SATs, GREs, MCATs, and AFQTs, are nearly inter-
changeable with IQ tests. The United States General Social Survey (GSS)
surveys a large sample of Americans every few years. The results are con-
veniently available on the web (General Social Survey, 2012). Among many
questions on the GSS, each round presents the respondent with ten words
and ask for their denitions. The tabulated variable, called WORDSUM,
varies from 0 to 10. This is a useful proxy for IQ in the sense that dierences
among groups in WORDSUM replicate closely reported dierences among
the same groups from more elaborate tests.
AmishMS/HCH/MW DRAFT Oct 2014
Results from 16PF
Scattergrams of groups and traits, that is principal components and principal
coordinates, from 16PF are shown in gures 1 and 2. Figure 1 shows patterns
among groups means, the Amish and non-Amish from Wittmers study along
with a collection of other populations culled from the literature. Details are
given in the caption of gure 1. The scatter of traits, in the left panel, is not
very informative. There is no clear pattern, reecting presumably the design
goal of independence among the 16 factors. The scatter among groups, in
the right panel, is more interesting. There are three groups that are on the
edge of the scatter: Amish, Chinese nurses, and candidates for high level
management positions in the UK. The outstanding (AN) and ordinary (EN)
Chinese nurses are nearly identical to each other on these rst two dimensions
(Zhang et al. , 2013). Generic English (UK) are world-average and are more
similar to Indiana farm boys than they are to high level managers in the UK
(Bartram, 1992).
AmishMS/HCH/MW DRAFT Oct 2014
Figure 1: PCA of 16PF score dierences among groups. The left panel portrays sim-
ilarities among the 16 traits (see Table 1). The right panel portrays dierences among
samples. AM are young Amish males from Indiana, IN are your non-Amish males from
Indiana, AN and EN are outstanding and ordinary Chinese nurses respectively, MK are
candidates for high level manager jobs in the UK, and UK are UK norms. Notice the
lack of pattern in the left panel, indicating the relative independence of the 16 traits.
AmishMS/HCH/MW DRAFT Oct 2014
The purpose of gure 1 is to convey a picture of global group dierences
but our real interest is in gure 2 showing in detail individual dierences
among young rural Indiana men. The scatter shows a nearly complete dis-
junction between the Amish and non-Amish along the horizonta axis, that is,
the rst principal component. For comparison the UK mean is shown in the
right panel as the large letter U in green. Remarkably the overall UK mean
is in the center of the non-Amish scatter implying that they are not very
dierent. The Amish, in the scatter, are strongly dierent. In our model the
dierences are the accumulated eects of Amish voluntary endogamy along
with boiling o of young Amish each generation who decide to leave the
Amish community.
The left panel of gure 1, the scatter of traits, helps identify those traits
that distinguish Amish and non-Amish. On the extreme left of the hori-
zontal axis, in the space of those with highest AQ, are traits G, Q3, and
I. The scoring of the 16PF questionnaire shows that Amish describe them-
selves, albeit indirectly, as conscientious and persistent (G), tough-minded
and self-reliant (I), and controlled and even compulsive (Q3). Correspond-
ingly they describe themselves as low in Q1 and E, that is as conservative
and traditional (low Q1) and humble and cooperative (low E). These im-
puted self-characterizations are in remarkable agreement with stereotypes of
the Amish that they hold themselves according to ethnographies. We our-
selves are agnostic about the meaning and uses of tests such as these except
as indirect self-descriptions by subjects.
A Model of (Self-)Selection
Under our model the horizontal axis in gure 2 is a scale, called AQ, esti-
mating Amishness meaning where on a numeric scale an individual is in
a suite of traits that dierentiate young Amish men from their non-Amish
neighbors. Since movement out of the Amish community is open to anyone,
especially around the age of adolescence, what would be the consequences
for the Amish community if individuals on the low end of AQ were especially
prone to leave the population?
Without proposing that Amish boiling o does reect gene dierences, it
is informative to bring a quantitative genetics model to the phenomenon as
a well understood baseline for evaluating competing models. We assume in
this spirit that AQ is a partially heritable trait. We have enough data on
behavioral or psychological trait heritabilities to use an estimate of additive
AmishMS/HCH/MW DRAFT Oct 2014
Figure 2: PCA of individual 16 PF scores. The left panel portrays similarities among
traits, the right panel among individuals along with the overall UK population. Amish
young me are shown as red As, non-Amish as blue Is, and the large green U shows
the centroid of the UK standard for comparison. The two panels are dual, so proximity
between a trait in the left panel and a subject in the right indicates a high value of that
trait in that subject. For example the Amish subjects have high values of 16PF trait
G, which is called conscientiousness and persistence in the literature. Non-Amish
young Indiana males have relatively higher scores on trait Q1, called Radical and
AmishMS/HCH/MW DRAFT Oct 2014
heritability of AQ, conventionally written h
of 0.5. Boiling o rates vary
among communities: a reasonable contemporary value is 0.10 meaning that
ten percent of youth chose to leave the community. From the ethnographies
it does not appear that this is youthful rebellion but instead a thought-out
change. Many join neighboring Mennonite groups for example.
Following standard practice in quantitative genetics we now assume that
AQ has an underlying normal or Gaussian distribution in the Amish pop-
ulation and that the ten percent boiling o rate is equivalent to losing the
bottom ten percent of the distribution each generation. We are assuming that
the boiling o mechanism is equivalent to truncation selection, but other se-
lection schemes will have an some equivalent truncation point that can be
With ten percent with the lowest AQ leaving each generation, the average
AQ of those who remain is 0.20 standard deviations if the the population
mean before emigratioin was 0. In other words this amount of self-selective
emigration leaves behind a population whose average AQ is 0.2 standard
deviations greater than it was before the emigration. We now assume that
mating occurs exclusively within the Amish. We assume that the heritability
of AQ is 0.5 meaning that the ospring will regress halfway back to the
parental mean.
The next generation of children will have an average AQ of 0.10 standard
deviations greater than their parents did before emigration. The process of
selective emigration repeats so that the mean Amish AQ increases by one
tenth of a standard deviation per generation. With 25 years per generation,
Amishness will increase by a full standard deviation in 10 generations
or 250 years. This is substantial social evolution on a time scale of a few
Our AQ of an individual is simply the value of that individual along the
x-axis of the right panel of gure 2. The computed standard deviations of AQ
for the Amish and non-Amish subjects are 1.33 and 1.31 respectively, so we
can take the within group standard deviation to be 1.32 units. The means
of the two groups are 1.4 and 1.4 units respectively and the dierence,
the distance between the two groups, is 2.8 units, slightly more than two
standard deviations. This may suggest that the heritability of AQ is higher
than our assumed 0.5, that the rate of boiling o was higher in the past,
for which there is support, or that the process has been going on longer
than ten generations. This last possibility is unlikely. Currently no Old-
Order exist in Europe. Persecution and political sanctions in the 17th and
AmishMS/HCH/MW DRAFT Oct 2014
18th Centuries had spread European Amish families too far apart to form
cohesive communities, forcing early European Amish to eventually assimilate
with non-Amish (Roth, 2002; Hostetler, 1977; Crowley, 1978; Nolt, 1992;
Gascho, 1937). Thus the magnitude of Amish endogamy was likely negligible
in Europe compared to Amish endogomous mating in the United States, so
the Old-Order tenure in Europe likely hasnt contributed to the dierences
reported here.
Quantitative genetic theory is well established in agriculture and in areas of
biology like biometrics and conservation biology. It deserves more attention
and use in the rest of evolutionary biology. We have shown that a simple
quantitative genetic model of a trait that is only partially heritable provides
a clean t to a small set of personality test data from over four decades ago.
Our t does not conrm the quantitative genetic model. Some other
mechanism may t the data as well or better. In particular purely cultural
transmission might generate this pattern, with no genetic consequence at
all. At this point the value of our approach becomes apparent since there
is no such cultural transmission model, only prose about plausibilities and
possibilities. In other words our model claries the mechanisms and magni-
tudes required of a purely cultural model. The dierence between the two
groups of subjects corresponds to the consequences of several centuries of
mild selection by choice of membership in an otherwise closed group.
AmishMS/HCH/MW DRAFT Oct 2014
Bartram, Dave. 1992. The personality of UK managers: 16PF norms for
short-listed applicants. Journal of Occupational and Organizational Psy-
chology, 65(2), 159172.
Cattell, Raymond B, Rao, DC, et al. . 1985. Heritability in the personality
control system: Ego strength (C), super ego strength (G) and the self
sentiment (Q3); by the MAVA Model, Q-Data, and maximum likelihood
analyses. Social Behavior and Personality: an international journal, 13(1),
Cochran, G., Hardy, J., & Harpending, H. 2006. Natural history of Ashkenazi
intelligence. Journal of Biosocial Science, 38, 659693.
Cross, Harold. 1969. Some demographic characteristics of Amish society.
Pages 126 of: Hostetler, John (ed), Conference on Child Socialization.
Cross, Harold E, & McKusick, Victor A. 1970. Amish demography. Biode-
mography and Social Biology, 17(2), 83101.
Crowley, William K. 1978. Old Order Amish settlement: diusion and
growth. Annals of the Association of American Geographers, 68(2), 249
Davies, G., Tenesa, A., Payton, A., Yang, J., Harris, S. E., Liewald, D., Ke,
X., Le Hellard, S., Christoforou, A., Luciano, M., et al. . 2011. Genome-
wide association studies establish that human intelligence is highly herita-
ble and polygenic. Molecular psychiatry, 16(10), 9961005.
Ericksen, Eugene P, Ericksen, Julia A, & Hostetler, John Andrew. 1980. The
cultivation of the soil as a moral directive: population growht, family ties,
and the maintenance of community among the Old Order Amish. Rural
Ericksen, Julia A, Ericksen, Eugene P, Hostetler, John A, & Huntington,
Gertrude E. 1979. Fertility patterns and trends among the Old Order
Amish. Population Studies, 33(2), 255276.
Fuchs, Janet A, Levinson, Richard M, Stoddard, Ronald R, Mullet, Mau-
rice E, & Jones, Diane H. 1990. Health risk factors among the Amish:
results of a survey. Health Education & Behavior, 17(2), 197211.
AmishMS/HCH/MW DRAFT Oct 2014
Gascho, Milton. 1937. The Amish Division of 1693-1697 in Switzerland and
Alsace. Mennonite Historical Society.
General Social Survey. 2012. ac-
cessed 05/08/14.
Greksa, Lawrence P. 2002. Population growth and fertility patterns in an
Old Order Amish settlement. Annals of Human Biology, 29(2), 192201.
Hewner, Sharon J. 1998. Fertility, migration, and mortality in an Old Order
Amish community. American Journal of Human Biology, 10(5), 619628.
Hostetler, John A. 1977. Old Order Amish Survival. Mennonite Quarterly
Hostetler, John A. 1993. Amish society. JHU Press.
Hostetler, John Andrew. 1969. Educational Achievement and Life Styles in
Traditional Society: The Old Order Amish. US Department of Health,
Education, and Welfare, Oce of Education, Bureau of Research.
Huntington, Gertrude Enders. 1957. Dove at the window: a study of an Old
Order Amish community in Ohio. Ph.D. thesis, Yale University.
Hurd, James. 1981. Mate choice among the Nebraska Amish of Central Penn-
sylvania. Ph.D. thesis, Pennsylvania State University.
Hurd, James P. 1983. Kin relatedness and church ssioning among the Ne-
braska Amish of Pennsylvania. Biodemography and Social Biology, 30(1),
Hurst, Charles E, & McConnell, David L. 2010. An Amish paradox: Diversity
and change in the worlds largest Amish community. JHU Press.
Kamin, Leon J, & Goldberger, Arthur S. 2002. Twin studies in behavioral
research: a skeptical view. Theoretical Population Biology, 61(1), 8395.
Kraybill, Donald B. 2001. The riddle of Amish culture. JHU Press.
Lee, Woei-Jyh, Pollin, Toni I, OConnell, Jerey R, Agarwala, Richa, &
Schaer, Alejandro A. 2010. PedHunter 2.0 and its usage to characterize
the founder structure of the Old Order Amish of Lancaster County. BMC
medical genetics, 11(1), 68.
AmishMS/HCH/MW DRAFT Oct 2014
Lewontin, R. C. 1972. The apportionment of human diversity. Evolutionary
biology, 6(38), 1398.
Markle, Gerald E., & Pasco, Sharon. 1977. Family limitation among the Old
Order Amish. Population Studies, 31(2), 267280.
McKusick, Victor A, Hostetler, John A, Egeland, Janice A, & Eldridge,
Roswell. 1964. The distribution of certain genes in the Old Order Amish.
Pages 99114 of: Cold Spring Harbor symposia on quantitative biology,
vol. 29. Cold Spring Harbor Laboratory Press.
Mckusick, Victor A, Hostetler, John A, & Egeland, Janice A. 1964. Genetic
studies of the Amish. Bulletin of the Johns Hopkins Hospital, 115, 203
Meyers, Thomas J. 1991. Population growth and its consequences in the
Elkhart-LaGrange Old Order Amish settlement. Mennonite Quarterly Re-
Meyers, Thomas J. 1994. The Old Order Amish: To remain in the faith or
to leave. Mennonite Quarterly Review, 68(3), 378395.
Murray, Charles. 2013. Coming Apart: The State of White America, 1960-
2010. Random House LLC.
Nelson, Dean. 2012 (October). Indias dwindling Parsi
population to be boosted with fertility clinics. www.
Nolt, Steven M. 1992. A History of the Amish. Good Books Intercourse, PA.
Pollin, Toni I, McBride, Daniel J, Agarwala, Richa, Schaer, Alejandro A,
Shuldiner, Alan R, Mitchell, Braxton D, & OConnell, Jerey R. 2007.
Investigations of the Y chromosome, male founder structure and YSTR
mutation rates in the Old Order Amish. Human heredity, 65(2), 91104.
Roth, John D. 2002. Letters of the Amish Division: A Sourcebook.
AmishMS/HCH/MW DRAFT Oct 2014
Savells, Jerry. 1988. Economic and social acculturation among the Old Order
Amish in select communities: Surviving in a high-tech society. Journal of
Comparative Family Studies, 123135.
Turkheimer, Eric. 2011. Still missing. Research in Human Development,
8(3-4), 227241.
Turkheimer, Eric, Pettersson, Erik, & Horn, Erin E. 2014. A Phenotypic Null
Hypothesis for the Genetics of Personality. Annual Review of Psychology,
65(1), 515540.
Visscher, Peter M, Hill, William G, & Wray, Naomi R. 2008. Heritability in
the genomics eraconcepts and misconceptions. Nature Reviews Genetics,
9(4), 255266.
Wasao, Samson W, & Donnermeyer, Joseph F. 1996. An analysis of factors
related to parity among the Amish in northeast Ohio. Population Studies,
50(2), 235246.
Wittmer, Paul Joseph. 1968. A comparison of the variability of perceived
parental behavior characteristics and personality traits of twenty-ve non-
Amish and twenty-ve Amish male youth, between the ages of eighteen and
twenty, from the same community. Ph.D. thesis, Indiana State University.
Yang, Jian, Benyamin, Beben, McEvoy, Brian P, Gordon, Scott, Henders,
Anjali K, Nyholt, Dale R, Madden, Pamela A, Heath, Andrew C, Martin,
Nicholas G, Montgomery, Grant W, et al. . 2010. Common SNPs explain
a large proportion of the heritability for human height. Nature genetics,
42(7), 565569.
Zhang, Li, Liu, Bo, Ren, Hui, Liu, Yu-Fu, & Zhang, Yan. 2013. The person-
ality prole of excellent nurses in China: The 16PF. Contemporary Nurse:
A Journal for the Australian Nursing Profession, 43(2), 219.