The purpose of this appendix is to present the statistical details that
Support the claims made in Chapter 2. Several statistical tables are presented
at the back of this appendix. This discussion is intended as an explanation
of those tables.
Table 1 shows the type distribution in the study of the Boston Church of
Christ. Type tables are displayed with the introverts in the top two rows
and the extraverts in the bottom two rows. The eight sensing types are shown
in the two columns on the left with the eight intuitive types in the two
columns on the right. The two outer columns contain the eight thinking types
and the two inner columns contain the eight feeling types. The eight judging
types are displayed in the top and bottom rows while the eight perceiving
types are in the two middle rows. Results are shown separately for males
and females because of differences on the thinking-feeling scale. Approximately
60% of males prefer thinking judgment and only 40% prefer feeling judgment,
but 60% of females prefer feeling judgment and only 40% prefer thinking
judgment. The three rows in each cell represent outcomes on the three different
forms of the MBTI.
Consider the ISTJ cell in the upper left corner as an example. Here is what
the figures mean. When answering the questions on the MBTI the way they
think they would have answered them before their conversion (or five years
ago for the few who had been members that long), 16.49% of the males and
11.68% of the females came out ISTJ, thus indicating preferences for introversion,
sensing, thinking, and judging. However, when they answered the questions
indicating present preferences, only 8.46% of the males and 6.69% of the
females
p193
came out ISTJ. Furthermore, when they answered the questions on the MBTI
the way they think they will answer them after five more years of discipling,
even fewer came out ISTJ--only 1.32% of the males and 1.30% of the females.
If you examine all 16 cells in Table 1, you will find that 10 of the psychological
types show a steady decline from past to present to future outcomes. Three
of the types--ISFJs, INFJs, and male ENTJs--show the largest percentages
in the present outcome. These appear to be transitional types. The changes
people are making move them into these types on their way to becoming something
else. Three typesESTJ, ESFJ, and ENFJ--show a steady increase from past
to present to future outcomes. The most popular type is ESFJ with 54.23%
of the males and 53.48% of the females indicating that type preference when
answering the MBTI questions the way they think they will after five more
years of discipling. The next most popular type is ESTJ with 20.37% of the
males and 23.04% of the females indicating that as their future preference.
The only other popular type is ENFJ with 14.81% of the males and 12.17%
of the females indicating that future preference.
Table 2 shows the deviations from a base population in this study. The purpose
of this comparison was to see which of the three distributions came closest
to population norms. Since most of the members of the Boston Church of Christ
are college students or college graduates, they were compared with a sample
of college students and college graduates who have taken the MBTI. Each
of the percentages in Table 1 was compared with a corresponding percentage
in the base population. What is shown in Table 2 are the percentage point
differences in the two figures. The mean percentage point deviation for
the total sample was closest to population norms when members of the congregation
p194
answered the MBTI questions the way they think they would have answered
them before their conversion (or five years ago for the few who had been
members that long). The present outcome showed a greater mean deviation.
The future outcome showed a much greater deviation from population norms.
Table 3 is a selection ratio type table showing the ratio of the percent
of each type among church members to the percent of that type in the base
population. A ratio of 1.00 would indicate a perfect match with exactly
as many of that type in the sample as would be expected based on population
norms. A ratio of 2.00 would indicate that the sample had twice as many
of that type as would be expected on the basis of population norms. A ratio
of 0.50 would indicate that the sample had only half as many of that type
as would be expected on the basis of population norms. Many of the cells
have significant under-representations in the future outcomes. The cells
with the significant over-representations in the future outcomes are ESTJ
ESFJ, and ENFJ. There were more than eight times as many male ESFJs and
more than three times as many female ESFJs as would be expected based on
population norms.
The significance levels indicate how confident one can be that the observed
differences do not result from chance and would be observed again in repeated
samples. At the .05 level, there is only a 5% probability that the observed
pattern resulted simply from chance. At the .01 level, there is only a 1%
probability of such error and thus one can be more confident. At the .001
level, there is only one chance in 1,000 of such error and thus one can
be still more confident. For any readers who are not familiar with statistics,
significance levels in this kind of study are usually based on a statistic
known as Chi Square. When some of the cells are empty or have very small
numbers, it is necessary to use an alternative statistic known as Fisher's
Exact Probability.
p195
Table 4 is another selection ratio type table. This time, however, the comparison
is not with population norms. Since the past distribution in this study
came closest to population norms, that was taken as the best estimate of
true type in the congregation. In Table 4, the present and future distributions
are compared with the past distribution. What this table shows is that the
changes in psychological type observed in the study of the Boston Church
of Christ are statistically significant. The past-to-present changes are
significant, but the past-to-future changes are highly significant.
Table 5 summarizes the changes on the four MB scales. Notice how the percentages
change from past to present to future outcomes. Notice also how many of
the members of the Boston Church of Christ show a future preference for
extraversion, sensing, feeling, and judging.
Table 6 shows the past-to-future MBTI scale changes by type. The 16 types
are listed in the left column following the usual type table order. The
second column shows the number who indicated each type preference when they
answered the questions the way they would have before conversion. The next
columns show the percent and the actual number who had no past-future changes,
then those who had one, two, three, or four changes. The column on the right
shows the mean number of scale changes for each type. The figures across
the bottom show the percent and the actual number who had no changes, one,
two, three, or four changes, and the mean number of scale changes for the
entire sample. What this shows is that the average member of the Boston
Church of Christ changed on at least two of the MBTI scales. Only 6.83%
had no past-future changes; 19.64% had one; 34.97% had two; 26.35% had three;
and 12.22% had four and thus experienced a total reversal of type.
Table 7 shows the past-future scale changes by
p196
preference. The figures on the left show the percent and the actual number
who started with each preference. The figures in the middle show how many
of those remained unchanged. The figures on the right show how many changed.
What this shows is that those who started with preferences for extraversion,
sensing, feeling, and judging tended to remain unchanged, but those who
started with the opposite preferences tended to change.
Table 8 shows the past-future changes by preference. The mean number of
scale changes was less for those who started with preferences for extraversion,
sensing, feeling, and judging that it was for those who started with preferences
for introversion, intuition, thinking, and perceiving.
Table 9 shows the past-future changes by combinations of preferences. In
each of the sets of four, one combination includes two of the ESFJ preferences,
two combinations include one of the ESFJ preferences, and the other combination
does not include any of the ESFJ preferences. In each of the five sets,
the combination that includes two of the ESFJ preferences shows the least
change and the combination that does not include any of the ESFJ preferences
shows the greatest change.
Table 10 shows the past-future changes by type. On the left side of this
table, the 16 types are arranged in order from the type that showed the
least change (ESFJ) to the type that showed the greatest change (INTP).
The ranking at the right side of this table is based on differences from
ESFJ. ESFJs, of course, have zero difference points and INTPs have four.
There is a Spearman rho rank order correlation of .91 between these two
ranking and that correlation is significant at the .001 level.
Tables 5 through 10 all make the same basic point: the group dynamics in
the Boston Church of Christ operate to influence a movement away from introversion,
intui-
p197
tion, thinking, and perceiving with a strong movement toward extraversion,
sensing, feeling, and judging.
Keep in mind that these statistical tables do not prove that any individual
is going to experience the psychological problems associated with falsification
of psychological type. The focus of this research was not on any individual,
but rather on the overall pattern observed in the group. This pattern, however,
dearly indicates a potential danger for the individuals subjected to this
kind of influence. Those who are already ESFJs when they come to the Boston
Church of Christ are likely to fit in quite well and not feel much of the
pressure toward conformity that others feel. The greater the difference
between a person's true type and the ESFJ model, the more likely that person
is to feel the pressure toward conformity. Those who come to the Boston
church as INTPs are in the greatest danger.
Table 1
Table 2
Table 3
Table 4
Table 5 and Table 6
Table 7 and Table 8
Table 9
Table 10
End
