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
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
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.
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
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-
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 5 and Table 6
Table 7 and Table 8