{$1513. - I , AN EXPLORATORY ANALYSIS 6? THE PREFERENCES OF FIRST-YEAR THEOLOGICAL STUDENTS DESIGNATED AS LIBERAL AND CONSERVATIVE TOWARD DIRECTIVE AND NON-DIRECTIVE RESPONSES IN THE PASTOR— PARISHIONER COUNSELING RELATIONSHIP Thesis for the Degree «5 'Ph. D. MICHIGAN STATE UNIVERSITY Ralph LIeweIIyn Miller 1963 THESE; LIBRARY Michigan State University This is to certify that the thesis entitled AN EXPLORATORY ANALYSIS OF THE PREFERENCES 0F FIRST-YEAR THEOLOGICAL STUDENTS DESIGNATED AS LIBERAL AND CONSERVATIVE TWARD DIRECTIVE AND NON-DIRECTIVE RESPONSES IN THE PASTOR-PARISHIONER COUNSELING presented by RELATIONSHIP RALPH LLEWELLYN MILLER has been accepted towards fulfillment of the requirements for Ph. D, degree in Guidance and Counseling Date August 9, 1963 0-169 -,_#___———— ”##fi ._—— _.—~ AHS [ITAC'I’ A N [X PLO RA II ) R Y ANALYSIS 01“ THIS PRISFI‘JI‘ZI'INCICS (IF FIRST-YI'AR 'lIIlC(_)II.)(-§ICAL S’l'IIIHN'IS DESIGNATED AS LIBERAL AND CONSERVATIVE TDNARD DIRECTIVE AND NON-DIRECTIVE RESPONSES IN THE PAS'II)R—PARISIII()NF.R COINS [TLI NC RIiLATI ( )QJSIII P by Ralph Llewellyn Miller THE PROBLEM The objective of the study was to investigate the preferences of first-year seminary students designated as liberal and conservative for directive and non-directive pastoral responses in the pastor-parish- ioner counseling relationship. The investigation was designed to focalize the basic information upon the problem of theological implications in counseling method in order to assess any possible relationship. The sample was drawn from the population of first—year seminary students in ftmrteen representative theological seminaries in the United States. Ihe following null hypotheses were tested: 1. 'fhere \dill be IN) siggiificarn «lifftntance in the tnrnber'tif directive and non-directive pastoral responses preferred by liberal first—year seminary students. 2. 'There.t4ill be IN) signit icant «Jitfenwsnce> in the turnber «M7 directive and non-directive pastoral responses preferred by Conservative first-year seminary students. 3. There will be no significant difference between liberal and conservative first—year seminary students in their prefereHCes for directive and non-directive pastoral responses. Ralph Llewellyn Miller 4. There will be no significant difference between liberal and conservative first-year seminary students in their preferences ftn7txs rele\1nrt fer wuynbers ed? the Rtmunl Cattu)lic (fliurch. llunl the members of all groups were asked to respond anonymously to a self-explan— atory questionnaire which consisted of seventy—two Likert—type statements. -32- The statements were so worded that they could reply as members of the Roman Catholic Church, members of the psychology class or members of other groups. Confirmation was found for the hypothesis that subjects in the experimental group would use the Roman Catholic Church as their reference group and would therefore be more likely to respond to state— ments in a manner prescribed for Roman Catholics while subjects from the control groups would respond in terms of reference groups other than the Roman Catholic Church.(5:ulS—u20l Charters and Newcomb's findings are of importance to this study in that there is an indication that when subjects are placed in situa— tions that provoke awareness of a particular reference group affiliation, they are then likely to behave in a manner congruent with the values of that reference group. In Coleman‘s study, The Adolescent Society, an examination was made of the youth culture of ten schools in various types and sizes of communities. When a study was made of the direction of loyalty when students were caught in a cross—pressure of values between peer and parents in regard to joining a club, it was found that the elites of the schools gravitated more toward the values of their peers. lhus, Coleman concluded that those who set the standard are more oriented than their followers to the adolescent culture.(6) These findings support Newcomb's hypothesis that the greater the ego—involvement in a reference group, the stronger the influence it has on the shaping of an individual's beliefs and behavior. This being the case, it can also be assumed that the minister will have greater ego-involvement in the reference groups that shape his religious beliefs and behavior than does the layman. Thus, the increased ego—involvement can result in more rigidly snaped patterns of beliefs and behavior. -33- lott,(3:259) in the process of studying a number of urban fam- ilie", concluded that social class can better be understood as a reference group rather than as some objective entity such as a member- U) ship group. This approach to the understanding of social class seem to add interpretive depth to the findings of Centers in his studies on the psychology of social classes. In Center’s major work, The Psychology of Social Classes, he found that subjective class identification served as a global index of conservative-liberal political, social and ecoromic attitudes. In a sample of 1100 white males representing a cross-section of the population of the United States he found that the higher the class identification, the more conservative were the social, economic and political attitudes, while Conversely, the lower the class identi— fication, the more liberal were these attitudes. In addition, those who would be objectively rated as being in one class but who rated themselves as being in another class shared the attitudes of the sub— jectively rated class.(4) Interpreting these findings from the view- point of reference group theory, it would be stated that the subjective class rating represented one's reference group while the objective class rating was equivalent to one's membership group. On this basis, the added interpretation could be made that reference group affiliation serves as a better indication of social, political and economic at- titudes than does membership group affiliation. Hartley examined the relationship between perceived values and acceptance of a new reference group using a sample of 146 unselected male freshman students in a municipal tuition—free college in an urban setting. The college community was considered the new group; off— campus groups and associates were treated as the prior groups making up -34- the subjects' established hierarchies of reference groups. By means of a ranking technique, scores were obtained showing the relative congruity between subjects' personal values, the values they perceived as typical of the new group (the College), and those they perceived as characteristic of their established groups. It was then determined whether the subject considered the new group as a reference group. Support was found for the hypothesis that “the greater the compatibility between the articulated values of the individual and the perceived values of the new group, the more likely the individual is to accept the new group as a reference group.(ll:87—95) The significance of this study is the fact that the college can become a reference group when the values of the student and the perceived values of the college have a high degree of congruence. In like manner, a student's selection of a seminary could be based on the perceived values and beliefs of the institution. As will be pointed out in Chapter III, this is, in fact, the case for students with conservative beliefs,who with few exceptions, attended conservative seminaries while students with liberal beliefs attended liberal seminaries. This gives support for the belief that the seminary becomes a reference group for the student. In summary, it is felt that the five studies just reviewed repre- sent the direction of research findings in regard to reference group theory. It is apparent from the research that the reference group af- filiation rather than the membership group affiliation more adequately explains the source of acquisition of values, beliefs and patterns of behavior. In addition, it justifies the use of seminaries rather than denominational affiliation as the basis for sample selection in that the former represents the reference group affiliation while the latter represents the membership group affiliation. -35- Finally, it is felt that reference group theory points to the possibility of behavioral differences in regard to seminary students of divergent theological points of view. If this is the case, then there is the additional possibility that a particular behavioral difference will be observed in the counseling methods of seminary students holding liberal and conservative theological beliefs. Thus, it is felt that reference group theory and its accompanying research is particularly cogent to the study under consideration. Summary to the Review of the Literature: In this chapter a number of studies have been reviewed which were considered relevant to the objective of this thesis. First, the citations of leaders in the fields of pastoral counseling and educational guidance pointed to a growing coneensus in regard to the influence of religious beliefs on counseling techniques. Second, research studies were re- viewed that represented attempts to determine what variables are associated with different types of counseling techniques. Third, several studies were cited which point to differences in the motivationahd per— sonality of conservative and liberal seminary students. Fourth, the research of reference group theory indicates that it serves as a more adequate explanation of the acquisition of values, attitudes, beliefs and behaviors . With the exception of Mannoia's study which bears direct reference to this study, the remaining research is primarily of an inferential or theoretical value. It is within the context of this literature twat this study has been written. CHAPTER 111 METHODOLOGY AND PROCLDHRE The development of this study's design was based upon the research that has been reviewed in Chapter ll. Included in the research design are: a description of the sample, measuring instruments, and the analysis procedures. The Fourteen Theological Seminaries: The population from which the sample was dtawn is represented by fourteen traditionally Conservative and liberal theological seminaries in the United States. All fourteen seminaries are accredited by the American Association of Theological Schools. Six of the seminaries have affiliatitni with denominations who hold membership in the National Council of Churchs; five of the seminaries are affiliated with denomi- nations holding membership in the National Association of Evangelicals; one seminary is affiliated with a denomination that holds no membership in either group: and two seminaries are inter-denominational. The four— teen seminaries comprise a total of 1489 students working for Bachelor of Divinity degrees. Table 3:1 indicates the participating seminaries, the denominations they represent and the inter—denominational organiza- tions of which the denominations are members. The geographical represent- ations of the seminaries are indicated on Table 3:2. -36.. TABLE -37- 3:1 DENOMINATIUNAL AND lNllR—DhNUMINAlTONAL‘AYFILIAT[UNS (H7 PARllXZIPAIlhKJ SlMlNARilS Inter-denominational Seminary Affiliation Denominat iona l Affiliation Anderson School of Theology Andover—Newton Divinity School Andrews Theological Seminary Asbury Theological Seminary Bethel Theological Seminary Bexley Hall Divinity School Boston University School of Theology Calvin Theological Seminary Concordia Seminary Harvard Divinity School Lincoln Theological Seminary Moravian Theological Seminary Union Theological Seminary Western Theological Seminary NCC NCC NAB NAB NCC NCC NAB NAE Church of God United Church of Christ 7th Day Adventist Methodist, Wesleyan, Free Baptist General Conference Protestant Episcopal Church Methodist Christian Reformed Missouri Lutheran Synod Inter—denominational Christian Church Moraviarl Inter-denominational Reformed Church of America NCC: National Council of Churches TABlJi NAl-i: National Association of Evangelicals 3:2 GEOGRAPHICAL REPRESENTATIONS OF THE SEMINARIES State Numbe Semin r of aries Illinois 1 nd iana Kentucky Massachusetts Michigan Minnesota Nebraska New York Ohio Pennsylvania D—Jt—Jt—‘r—Jt—‘WUJh—‘r—Jo—I -38- The seminaries from which the sample was drawn were systematically selected from advertisements in the Christian Century. In the spring of each year the Christian Century publishes a special issue limited to topics on theological education. A portion of this issue is set aside for seminaries to advertise. A seminary known to be either traditionally conservative or liberal was selected from alternate pages in the adver— tising section. A letter was then sent to the Dean, Registrar or Presi— dent of each of the selected seminaries, this depending upon whose name was listed in the advertisement. The letter consisted of the following parts: a cover-letter of introduction, an abstract of the proposed study, a copy of the two instruments used as the basis for the study, a check—list response sheet, and a self-addressed return envelope (see Appendix). 0f the fifteen Seminaries that were contacted only one refused to cooperate, this being Southern Baptist Theological Semi— nary in Louisville, Kentucky. Thus, 93% or fourteen out of the fif— teen seminaries contacted, cooperated in the study. Sample Selection from the Fourteen Seminaries: The initial step in the procedure of sample selection was to limit the sample to first-year seminary students in each of the four- teen participating seminaries. Delimiting the sample to first-year students was justified as a control factor for the variable of cur- ricular differences. It was felt that if second or third—year seminar- ians were used as the sample for this study there would be a greater variance of curricula within the sample than if first—year students were used. This rationale is based upon the controlling effect that -39- the American Association of Theological Schools has as the accrediting agency of all participating seminaries. During the first year of semi— nary education, the curriculum is generally limited to foundation courses in Biblical languages, history and theology. Subsequent study during the remaining two years leading to the Bachelor of Divinity degree per- mits greater lattitude in the selection of electives such as pastOIal counseling courses. Thus, while some of the students within the sample had taken pastoral counseling courses, this numberczould have been even greater if second or third—year students had been the basis of the sample. lhe second step in the selection of the sample was to obtain an estimate from each of the participating seminaries as to the population of their first-year seminary classes. With this information in hand, the appropriate number of questionnainmswas sent to each seminary, accompanied by instructions as to their distribution and administration to the students. The latter was for the purpose of controlling for any error variance that might result from differences in presentation and administration of the two instruments. The final step in the selection of the sample was the distribution of the questionnaire instruments. "the qUestionnaires were distributed to every member of the first-year classes of the fourteen seminaries. Questionnaires were placed in the mailboxes of each student with a cover letter from the seminary’s administrative office requesting the coopera- tion of the students. The nature of the cover letter was specified in the instructions for administration of the instrument. Consequently, no indication was given to the students as to the purpose or objectives of the study. -40- Dimensions of the Sample: 0f the #89 questionnaires that were mailed to the seminaries, 391 were returned and all within a five—week period. Consequently,the sample obtained for this study represents 80% of H19 population of the first-year classes of the fourteen seminaries. Of the 391 returned questionnaires, four were considered incomplete thus reducing the sample to 388 subjects. The remaining 388 subjects represented thirty— three denominations and every state in the Union with the exception of Alaska. The number of subjects representing each denomination is in- dicated on Table 3.3. The level of education for the sample was held constant with all subjects having completed four years of college and approximately six months of Seminary. It was assumed that since none of the sample had met requirements for ordination that the amount of formal pastoral experience was negligible. lhus, amount of actual experience in the ministry was controlled for in the study and held constant for the sawnflx:. 'Ihe mean age (J tine sample was 25.6 and 9147?: fell in the age range of 22 to 32. Since the age range for 9/4”..- of the sample represents one generation of students, it was felt that the effect of age upon the findings was kept to a minimum. -41- TABLE 3:3 NlDBER 0F STUDENlS REPRESENTING EACH DKNOMINATION Denomination Number of Students American Baptist Convention 29 Assemblies of God 1 Christian Church 13 Christian and Missionary Alliance T Christian Reformed 21 Church of the Brethern 1 Church of Christ 2 Church of Christ, Scientist 2 Church of God 13 Church of God Reformed Movement 1 Congregational Christian 4 Conservative Baptist 1 Disciples of Christ 13 Evangelical Convenant 2 Evangelical United Brethern 2 Free Methodist 16 lndependent or no membership 7 Lutheran (American) 3 Methodist 65 Missouri Luthern Synod 44 Moravian 6 Presbyterian, U.S. (Southern) 2 Protestant Episcopal l7 Quaker k Reformed Church of America 28 Seventh Day Adventist 29 Unitarian 3 United Church of Canada 2 United Church of Christ 34 United Missionary 1 United Presbyterian, U,S,A, 2O Unive fsalist l Wesleyan Methodist -42- Liberal and Conservative Criterion Samples: It will be recalled that the theological Seminaries were systema- tically chosen for sample selection on the basis of their being tradi— tionally either conservative or liberal. Assuming that for the subjects the seminary serves as the reference group and denominational affilia- tion serves as the membership group, then reference group theory sup— ports the selection of the seminary rather than the denomination as the best approach to obtaining students of similar belkafs. To ob- tain a sample of conservative and liberal seminarians, reference groups that were known to be conservative and liberal were used as the basis for sample selection. The findings in Table 3:5 indicate the wisdom of this approach and support the assumption that the semi— nary rather than the denomination is the more accurate index to the beliefs of the sample. However, it is feasible that within the larger seminary reference group there are smaller reference groups which represent beliefs at variance with the larger reference group. Thus, two indices were in- corporated into the study to sharpen our estimate of who were the con- servative or liberal subjects. First, each student was asked on the Preliminary Data Sheet to make an estimate of the conservatism or liber— alism of his religious beliefs. Second, the Religious Belief Inventory was used to identify the mean-split and extreme conservatives and liber— als in the sample. A simple correlation was then made between subjects' self-rating and the mean-split and extreme groups of the sample as iden- tified by the Religious Belief Inventory which indicated a near perfect congruence between self—rated and measured beliefs. -43- As a result of these three approaches to identifying the liberal and conservative students, it was felt that a very precise dichotomizing of the sample was achieved. The Religious Belief Inventory: The Religious Belief Inventory was developed by Anderson and Toch during the past three years as an instrument to describe the content of religious belief. It is designed to differentiate four religious class— ifications within two major divisions — liberal and conservative. Liberalism is comprised of secular and liberal types while conservatism is comprised of fundamentalist and orthodox types. Since this study is only concerned with conservative and liberal types, the Inventory was scored in a manner which combined the two sub—scales for each major division, thereby yielding two scores for each student rather than four. The original Religious Belief Inventory was developed from state- ments of belief that had been compiled by the authors on the basis of the selective criterion of how conservative or liberal the beliefs were. These statements were formulated under the heading of God, Jesus Christ, the Bible, the Church, Epistemology, and Metaphysics. After an informal screening process, 146 items remained that were formally pre—tested for the final selection. The sample which was used for the purposes of the pre—test consisted of twenty—one ministers in Lansing and East Lansing, Michigan. Each minister was asked to “check every item to indicate whether you believe it tends to reflect more of a conservative belief or a liberal belief."(3l:9h) 0n thelnsis of the judgments of the seventeen ministers who completed the pre—test, the number of items was -44- reduced from 146 to 101, The remaining items represented unanimous agreement(included no more than two abstentions) among the seventeen ministers as to the conservatism or liberalism of the items, As a check on the reliability of the instrument, an independently obtained sample of ministers in Jackson, Michigan, yielded similar findings. A short form of the Inventory, consisting of 60 items, was constructed by the authors, These sixty items represented relatively pure expressions of the four dimensions. It was this instrument that was used in this study (see Appendix). In order to validate the instrument in regard to the purposes of this study correlation coefficients were obtained comparing the subjects' self—rating of conservatism/liberalism of their beliefs with the measured estimate of their beliefs from the Religious Belief Inventory, The correlation coefficients for the mean—split and extreme groups were indicated in Table 3:4, TABLE 3:4 COEFFICIENTS BETNEIIN Slilf—lbkl'lNC AND MEAN—SPLIT AND EXTRENE GROUPS Self-rating Mean-split Extreme Conservative .92 .96 and Liberal -45- In light of the correlation coefficients yielded from a com— parison of self—rated and measured conservatism or liberalism of be— liefs, it is felt that the Rtligious Relief Inventory is not only well suited for this study but also achieves a highly accurate measure of the I: dichotomized criterion. Religious Belief Sample Classification: Four incomplete returned questionnaires were eliminated from the total responses of 391. Ihe remaining 388 returned were analyzed to determine whether the subject was to be classified as liberal or con— servative according to the Religious Relief Inventory. The mean of the conservative responses was 12.227 and the stan— dard deviation was 6.122. The mean of the liberal responses was 4.871 and the standard deviation was 4.009. On the basis of means and stan— dard deviations, the sample was then dichotomized according to mean— split and extreme group classifications. Mean—split group: To determine who were liberal and conservative, tfliose who scored above the liberal mean and below the Conservative mean hmare classified as liberal, and those above the conservative mean and befilow the liberal mean were classified as conservative. Vsing the mean- 5P1.it division, there were 116 (29.89%) liberals and 179 (45.9?) con- seW‘vatives. Ninety—three (24,33) were classified as neither. Extreme group: For the purpose of intensifying the dichotomy 0f tweligious belief, the extremes of both groups were selected. This was aachieved by selecting those subjects who were above one standard deviation in one classification and below the mean in the other. Accord- i”81)/, there were 54 (13.9%) liberals and 89 (22.97) conservatives. -46- Preliminary Data Sheet: Attached to the two instruments was a short questionnaire giving information about age, denominational affiliation, amount of pastoral counseling experience and self—rating of one’s religious beliefs (see Appendix). With the exception of the self—rating estimate, a comparison of the data from the data sheet with the mean-split and extreme group classifications is summarized in Table 3:5. TABLE 3:5 SUMMARY OF AGE OF THE LIBERAL AND CONSERVATIVE SAMPLE FROM THE PRELIMINARY DATA SHEET SAMPLE Mean—Split Frequency Extreme Frequency Number of Number of Age Liberal Conservative Liberal Conservative 20-21 3 3 3 1 22—23 75 64 3O 30 24-25 16 38 ll 19 26-27 7 28 h 13 28-29 7 19 3 11 30-31 4 8 2 5 32-33 1 7 5 34-35 1 36-37 1 3 2 38-39 3 2 40-41 3 1 42—43 1 1 44— 2 TABLE SUMMARY OF DENOMINATION OF FROM THE PR -47- 3:6 THE LIBlZRAL AND C(D‘JSERYA'IIVII ELIMINARY DATA SHI‘IET SAMPLE Denomination SAMPLE Mean-Split Frequency threme Frequency Liberal Number of Conservative Liberal Number of Conservative American Baptist Convention Assemblies of Cod Christian Church Christian Christian Church of Church of Church of Church of God Church of God Reformed Movement Congregational Christian Conservative Baptist Disciples of Christ Evangelical Convenant Evangelical United Brethern Free Methodist Independent or no membership Lutheran (American) Methodist Missouri Lutheran Synod Moravian Presbyterian, U.S. Protestant Episcopal Quaker Reformed Church of America Seventh Day Adventist Unitarian United Church of Canada United Church of Christ United Missionary United Presbyterian, U.S.A. Universalist Wesleyan Methodist Reformed the Brethern Christ Christ, Scientist and Missionary Alliance 15 7 9 l 13 19 2 1 l l 5 l 3 1 2 1 8 3 l 1 12 l 2 l 1 29 23 19 M3 1 2 l 6 2 2 2 2 1 15 1 21 3 3 2 2 26 l 7 1 13 t—‘ t—l 12 34 SUMMARY OF SEMINARY OF THE LIBERAL AND F ROM '11 l E -48— TABLE 3:7 CONSERVATIVE SAMPLE PRELIMINARY DATA SHEET SA}PLE Mean—Split Frequency Extreme Frequency Number of Number of Seminary Liberal Conservative Liberal Conservative Anderson 1 S 1 Andover—Newton 37 2 13 Andrews 19 6 Asbury 47 21 Bethel 8 Bexley Hall 3 2 Boston University 27 18 Calvin 19 12 Concordia 43 34 Harvard 11 11 Lincoln 15 6 Moravian 1 2 Union 35 14 Western 1 15 8 lABLE 3:8 SUMMARY OF COUNSELING IRAII.‘£1N(J OF THE LIBERAL AND OINSERVATIVE SAMPLE 1’ ROM THE PRELIMINARY DATA SHEET Counseling Training SAMPLE Mean—Split Frequency Extreme Frequency Number of Liberal Conservative Number of Liberal Conservative 18 43 16 _qg_ The Interview Sets: The Interview Sets were developed by Mannoia as the basis of his doctoral thesis at Michigan State University.(21) It is designed to differentiate preferences for directive or non-directive counseling responses in four majOrcuun$£hfi‘ problem areas: emotional, spiritual, ethical, and marital. As a basis for the final instrument twenty inter— view sets were constructed which were believed to represent some of the counseling problems faced by today's ministers. Each interview set consisted of a client's statement of a problem accompanied by six to eight alternative pastoral responses. The twenty sets were then sub- mitted to eight expert judges in the field of counseling for the pur- pose of rating the items "directive,” "non—directive" or “does not apply.” The judges' ratings yielded sixteen validated sets, four in each problem area, each of which had one pair of directive and non- directive responses that had received agreed judgments from six or more of the eight judges. All of the directive foils for the final sixteen sets represented the unanimous agreement of the judges (i.e., none of these were rated as non—directive). Eight of the non—directive foils for the final six- teen sets represented the unanimous agreement of the judges. Five of the eight remaining non—directive responses were rated by all but one of the judges as being "non—directive.’ The last three non—directive respons*s were rated by all but two of the judges as being ”non-directive.“ In addition, the sixteen counselee statements were rated by the eight judges as to the types of problems that they represented. Eleven of -50- the sixteen counseling problems were categorized according to problem area with unanimous agreement. Of the remaining five, two counseling sets were categorized with all but one agreeing, two sets were categorized with all but two agreeing, and one was categorized with only three agreeing. Using this technique and on the basis of the resultant ratings, sixteen interview sets were selected for incorporation into the final form of the instrument. It is the instrument in this form that is used in the present study as the basis for determiring preferences for direc— tive or non—directive counseling responses (see Appendix). Analysis Procedures: The data from the Interview Sets was tabulated by handscoring to group the ministers into two classifications: directive or non-directive. These classifications were based upon the number of directive and non— directive responses that were selected. lhose subjects who scored ten or more of either type were regarded as “significant;’ those who scored seven, eight, or nine responses of either type were considered as being "inconsistentt" and those who had five or more omissions were regarded as "omissions." In regard to the scoring of the "omissions,” V two—thirds of either type of responses was considered a 'significanf' score and the return was included in the final sample. If either Part I or Part II was left completely unmarred, the return was considered “in— complete" and was not used in the final sample. -51- When all the data was scored for the Preliminary Data Sheet, the Religious Belief Inventory and the Interview Sets, it was then converted into numerals and recorded on master sheets. In addition, the counseling reSporses of the subjects in the conservative and liberal groups were then converted into percentages to later be reported as frequency distribution histograms. Having completed these procedures, the data yielded a sample that was dichotomized both directive or non-directive and Conservative or liberal. Since the data is dichotomized and the focus of the study was to determine the existence or non-existence of a relationship between the two groups of dichotomies, chi square is the appropriate statistic. A 2 X 2 contingency table was used in conjunction with the appropriate chi square formula. Ihree chi square formulas were used, all with one degree of freedom. The formulas and the reasons for their usage are as follows: 1) When all expected frequencies were over ten: x2: 02 — N E 2) When three frequencies were over ten and one frequency was less than ten but greater than five: N x2: (iAD - BC: — 3wL (A+B)(C+D)(A+C)xB+D) 3) When two frequencies were over ten and two were less than five: g2: (AD — BC)2 Nt TA+BT(C+D(A+C)(B+D) In addition, -52- two chi square values were converted to Phi coeffi— cients for the purpose of clarifying the strength of relationships. The formula for the Phi coefficient is: Q: \ X2. J N The chi square values were determined according to the following dichotomies. A. C] Mean—split group on the Religious Belief Inventory versus: Global score on counseling Intelview Sets U'I-F'LAJMv—a Score from Score from Score from Score from Emotional sub-area Spiritual sub—area Ethical sub—area Marital sub-area Extreme group split on the Religious Belief Inventory versus: Global score on counseling Interview Sets 0 outcome—a Score from Score from Score from Score from Self—rating of m-PwMH Emotional sub-area Spiritual sub—area Ethical sub-area Marital sub—area theological position versus: Global score on counseling Interview Sets Score from Score from Score from Score from Emotional sub—area Spiritual sub—area Ethical sub—area Marital sub-area Self—rating versus Mean—split group on the Religious Belief Inventory (Test of agreement, converted to Phi) Self-rating versus Extreme group on the Religious Belief Inventory (Test of agreement, converted to Phi) Self-rating versus Mean—split group on the Religious Belief Inventory (Test of accuracy of self—rating of theological beliefs) Self—rating versus Extreme group on the Religious Belief Inventory (Test of accuracy of self-rating of theological beliefs) -53.. H. Comparison of Conservative Mean-split and Extreme groups with global score on counseling Interview Sets I. Comparison of Liberal Mean-split and Extreme groups with global score on counseling Interview Sets The four null hypotheses can be summarized by one broad statis— tical null hypothesis to be tested: H0: The two groups of first—year seminary students, liberal and conservative, will not differ significantly in the proportion of each group with directive and non—direc— tive pastoral response preferences in various combin- ations. The .05 level of significance was selected as the appropriate level for rejecting or accepting the stated hypotheses. This level of significance was selected in that it represents the best compromise between making a Type I or a Type 11 error. Summary: The method and procedure of the study was delineated by describing the sample, the measuring instruments, and the analysis. The sample for the study was drawn from fourteen theological seminaries in the United States. Questionnaires were distributed to the population of each first- year seminary class and the cooperation of the students was requested. Of the 489 students receiving the questionnaire, 391 or 80% responded by completing and returning it. The questionnaire included the Prelimi- nary Data Sheet, the Religious Belief Inventory, and the Interview Sets, The justifications for the use of these instruments were discussed. The data yielded by these instruments was analyzed by the use of the chi square statistic. The analysis of the data is found in the following chapter. CHAPTER IV AN ANALYSIS OF DIFFERENCES IN Illli RELIGIOIYS BELIEFS AND COINSlleN-G RESPONSES OF FIRST—YFAR SICMINARY STUDENTS Chapter IV conta ins an analysis of the response performance of liberal and conservative ministers. The two groups are compared accord- ing to the indices of Mean—split and Extreme group classifications as discussed in Chapter III. In addition to these two indices, a third index, the subject's self—rating. is compared with response preferences, including the whole sample of 388 subjects. The analysis of the response preferences of the two religious belief groups falls into five different comparisons: total counseling reSponses, and the four problem areas: emotional,spiritual, ethical, and marital (see graphs 4:1—hz20). The analysis was directed toward accepting or rejecting the null hypotheses at the .05 level of significance. The several hypotheses are summarized into a general hypothesis stated in the previous chapter: H0: The two groups of first year-seminary students, liberal and conservative, will not differ significantly in the proportion of each group with directive and non—direc— tive pastoral response preferences in various combina— ‘ticnls. -54- -55_ Frequency of Liberal and Conservative Counseling Responses: Table 4:1 indicates the frequency of responses by liberal and conservative seminary students according to the mean—split and extreme group classifications. In addition, the responses were further analyzed on the basis of percentage in order to indicate proportionate differences. This was deemed advisable in that the totals for each group were dif— ferent. The conversion of frequencies into percentage distributions is illustrated in the graphs of this chapter. An overview of the descriptive data indicates several factors of pertinence. First, there is an overall trend for conservative seminary students to have a greater preference for directive responses than the liberal seminary students. In the mean—split group this is the case with but two exceptions, namely items 13 and 14. In the extreme group there is one exception, item 13 (see graphs h 1-4:u). Second, it will be noted that regardless of religious belief, interview sets numbers 1, 2, 3, 6, 8, 11, 13, and 15 Consistently have higher frequencies of non—directive responses and conversely, consis— tently lower frequencies of directive responses. Interview sets numbers 7 and 15 have consistently higher frequencies in directixe responses and conversely, consistently lower frequencies of non-directive res— ponses. Therefore, while the conservatives have a greater preference for directive responses than do the liberals, generally both groups have a greater preference for non-directive responses than directive responses (see graphs 4:1—4:4). —56— TABLE 4:1 FRliQ'TlZNCY 01’ RESPONSE. SELECTIONS 017 1.181lesz AND CONSERVATIVE FIRST— YWAR SEMINARY STUDENTS ACOWRDING 10 TWO COMPARISON SCHEDULES MllAN—SPLI T CW )1}? EX'I'RICMH ($120171) Interview Liberal Conservative Liberal Conservative Set llir. N-Dity 1)ir. N—JiJ‘ llir. N-Dir. llir. N—Dir. 1 28 88 49 129 12 42 .36 52 2 29 86 55 122 13 41 32 55 3 6 110 15 164 4 50 10 77 4 52 63 100 78 21 32 57 31 5 21 95 87 9O 9 45 50 38 6 49 67 82 96 23 31 41 48 7 74 42 116 62 36 18 6O 29 8 2O 96 43 134 13 4O 28 61 9 48 68 105 73 21 33 55 33 10 39 76 8O 99 22 32 50 39 11 16 99 32 146 ll 43 22 66 12 33 82 52 126 17 36 36 52 13 15 100 14 165 11 42 10 79 14 73 42 89 89 34 19 64 24 15 '16 97 36 141 6 47 23 64 16 6 108 44 35 7 46 3O 59 N=116 N=179 N=54 N289 (Note: When Directive and Non-Directive cells do not total N, an omission has occurred in that Interview Set.) -37- Comparison of Formal Counseling Training with Response Preferences: It was believed that formal counseling training may have had an influence on the type of counseling responses selected by the. sample. Since information concernmg amount of training was included on the Pre— liminary Data Sheet, 8 comparison of training versus counseling pre- ference could be conducted, 0n the basis of this comparison, it was found that of the forty—three respondents that were classified as Con- sistently "directive,' only six (14%) had formal training. On the other hand, of the 233 that were classified as Consistently "non directive,“ sevent y-five (303’) had training. The proportionate difference is significant at the .05 level (see Table #:2). Consequently, it can be concluded that formal training was apparently related to preference for non—directive responses. It might be added, however, that this finding may have been influenced by the greater preference of the total sample for non—directive responses. Comparison of Self—Rating with Religious Belief Inventory — Test of Agreement: As a means of determining the validity of the Religious Relief Inventory in regard to its categorization of subjects as conservative or liberal, a self—rating item of religious beliefs was included on the Preliminary Data Sheet. As was mentioned in Chapter 111, the Phi coefficients of .92 and .96 for the mean—split and extreme groups, respectively, indicate that, for the purpose of this study, the Religious Belief Inventory indicated a high degree of concurrent validity. The X2 values from which the Phi coefficients were computed are found in Table 4:2. —58- Comparison of Self—Rating with Religious Relief Inventory — Accuracy of Self-Rating: Since the data was available, it was thought that a comparison of the liberals and conservatives in the accuracy of their self—rating would be of value, even though it does not relate directly to the pur- poses of the study. Since the X2 values were not significant, it can be said that both the conservatives and liberals in the mean—split and extreme groups rated themselves with equal accuracy(see Table hz2). In summary, several ancillary comparisons have been made and dis- cussed which are considered to be of importance in the interpretation of the data in Chapter V. It is well to note that while these comparisons do not influence the accepting or rejecting of the null hypothesis, they do attempt to add interpretative depth to the conclusions of this study. TABLE 4:2 ANCILLARY mMPARISONS: lil'l‘l‘Cl 017 ’I'RA'INI NU ON RILSPtllNS}: PRllhljl-ZFI‘JCES , "ll 9" (ll: AGREEMENT Oir Sl‘LF—RATTNCI (3F BELIEFS Wllll INYIIN'IURIEI) BELIEFS, ANI) ACCURACY OF SELF—RATING WHIN GIMPARU) WITH INYI'N'HHII 1],) BILLIEFS Comparison X2 Value Mea£§SpiiEeGr0up Exiregzligoup Effect of Training 4.492 Test of Agreement 2Q8.722 273.948 Accuracy of Self Rating 3.77C 1.85C *significant at .05 level Ccorrected formula -59- Analysis of the Interview Sets In the analysis of the Interview Sets, counseling response pre- ferences will be compared with the three indices to sample grouping: Mean-split, Extreme, and Self-rating groups. The 2 X 2 contingency table with the computed chi square value will serve as the statistic for the analysis. The summary of the findings are contained in Tables 4:3 and 4:4. Total Responses: lbpothesisl: There will be no significant difference between liberal and conservative fifst-year seminary students in their preferences for directive and non-directive pastoral responses. On the basis of the comparisons indicated in rows 1 and 2 of Table 4:3, there was a significant difference, and the null hypothesis of no difference between conservative and liberal first-year seminary students in total responses was rejected. The conservative seminary students with but two exceptions, consistently preferred more directive responses than did the liberal seminary students. This was true in the case of comparing the raw totals of directive and non—directive responses of the three indices (mean-split, extreme, and self-rating groups), as well as when the comparison was made on the basis of those in each religious belief category who were consistently directive or non—dir— ective. -60- Problem Area Responses: Hypothesis II: There will be no significant difference between liberal and conservative first—year seminary students in their responses in the area of emotional, spiritual, ethical and marital problems. The null hypotheses regarding the spiritual andwmrital problem areas were rejected and the null hypotheses with respect to the emotional and ethical problem areas were accepted. These findings are again based on the three indices to sample grouping, mean-split, extreme and self— rating groups. When the null was rejected, the chi square values in these three groups were all significant. When the null was accepted, the chi square values, with one exception, were not significant. Re— garding the exception, in the emotional problem area, the extreme group chi square value did achieve statistical significance but since the other two group values were non—significant the null hypothesis was still accepted. TABLE 4:3 A SUMIVIARY OF THE RESITL'I’S OF CHI SQUARE ANALYSIS OF SlX DIFFERENT AREAS ACCORDING TO lHREE COMPARISON SCHEDULES Mean-Split group Extreme group Self—rating group Interview Sets (X2 value) (\2 value) (X2 value) Total of subjects with 6.091 7.331 4.310 consistent direction of pastoral responses Total Pastoral Respon- 30.174 41.763 10.371 ses (all four areas) Emotional Area 1.780 7.330 .038 Spiritual Area 38.699 27.666 33.538‘ Ethical Area .79 2.01 2.191 Marital Area 6.647’ 12.40 4.60h Significant at the .05 level -61- Comparison of the Counseling Preferences of Conservative Mean-split and Extreme Groups: Since the chi square of this comparison was significant, it is apparent that the conservative extreme group had a greater preference for directive responses than did the conservative mean—split group. Thus, it appears that the more conservative the students were, the greater they gravitated toward directive responses. Comparison of the Counseling Preferences of the Liberal Mean-split and Extreme Groups: The chi square for this comparison was not significant, thus there is an indication that being moderately or extremely liberal in religious beliefs is not related to preferences for directive counseling responses. TABLE 4:4 COMPARISON OF COI'NSLL INC PREFERENCES WITHIN EACH RELIGIOYS BELIEF GROUP Comparison Mean—split/Extreme vs. Conservative response preferences 6.203' Liberal response preferences 1.195 “Significant at the .05 level —62- Bar—Graph Presentation of Data: To illustrate the interpretation of the differences between the Mean~split and Extreme groups in their preferences for counseling responses, the data has been summarized in bar-graph form (see Graphs 4:1—20). -63.. wh.-‘ a--- .- .- ___——.——— 3H .. _- sawomriomztu No.75: mH NH _ i n _ w H m E . . w m _ _ w _ t _, _. _ i . . __ _ ”To Em; _ fl _ . w . 11 ’\ L..__ ....__. _ o. 3) g .. . mm: o m .m o. m d ., T...-i.l+.l:t _ . 7 . _. u ._ . i... .w _ _ ... ._u .* w. n: _ , . i i. m .1 .1. . u . . _i .... _ u .. .. W M. m _ m . t. _ .. . s... _ fl. . 7* - g . H” “ TWA . _ .mm . . m“ .0» _. . :. _ u m u “ _dd ,. m. L M . _ _ . _ _ m. M "--u .- 77 fi 0 --——_——.—...-_ _ - - on mm A9 _ CNS ...:,_x_x.,r x; V WHZmaéwm >2~.lemzam :p.HH<./::mz~b 272 ASEEHA m0 mumzcmmgz 4.249053: :0 ...._C<.Hzm_m:fia 1:2}; N n d :m<~._3 :HE U 9.3 a my: omzcu pom 39w>eoqu Hmsmnoo es ms as Ms NH-F Ha _ as a w A e m . s . m . N n H 1 A . at a 4.. d O m : W _ :- 2 n M . . _ m. m A” Mn” . ca _ E _ .7 u e m: 7 _. M. u __ r. . .“m E i w i .._... j i . . ._ s m. m _ . m .... .t 12 l _ i. i W . 9 . m w i ._ i i. y... .fm om . _ E i w _ i .. M L. em L em W . m w w m . . r as M .lqi, w M ... H.. .W .fl Om A [wfllv *4 M .1 . 4 mm am M i - m J w .. I co m Om . mm ... ”h H..— H m a .. _.. a . . I . w i _ u w 00 s M 1.0m _ ON NR # M _ we om . . i mm mm mm ow mm om rt loo . ~¢ II No a s? e p e _ _ _ . _ _ L _ _ _ _ 2: mm. 772.: E. m >zz-emeoe o>He<>eemzco :z< othe individual needs of the students so that their counseling behavior Can approach congruence with their personality structure. 2. A second implication relates to the incongruity between the Rogerian view of man and the view of man held by the conservative, It is difficult to understand how a conservative minister Could, on the one hand, believe in the total depravity of man and, on the other hand, accept for the counseling situation the Rogerian belief in the salient goodness of man. This study has pointed out that it is more difficult for the conservative seminary student to accept a type of behavior based upon the Rogerian philosophy than it is for H10 liberal. Is it possible that the seminary student, in his relative ignorance of the Rogerian theory of man's goodness, senses an inCongruency that some seminary educators themselves do not see? In any case, this study has indicated that non—directive counseling behavior goes far beyond the level of technique to the level of beliefs. Therefore, it is imperative that. educators espousing non—directive Counseling in conservative seminaries take a Closer look at theory, lest they see it only as a technique un— related to beliefs. 3. A third implication relates to the training of therapists. In light of the prior discussions it is feasible that the modus-operandi of every counselor—trainee is affected by his beliefs about the nature of man, whether these are secular or religious beliefs. If one's style of counseling is a correlate of a belief system, then should not the counselor—tra iner consider permitting the trainee to adapt. a theory of Counseling that is most congruent with his beliefs about man'.‘ Otherwise —85— the counselor-trainee may verbally behave, say, in a non—directive manner while non-verbally he may be communicating to his client a contrary attitude. If such were the case, movement in therapy would be greatly hindered for the client. In conclusion, belief and behavior can be seen as two sides of the same coin, namely the individual's needs. Attempting to impose upon the coin a face incongruent with its value is to commit a fradulent act. In a similar way, attempting to impose a type of behavior or a particular belief upon a person that is incongruent with his needs is to commit him a disfavor. ConseqUently, the cmnral implication of this study is the integral relationship of beliefs with behavior especially in the counseling relationship. Keeping this in mind, the training of seminary students in particular and therapists in general will be enhanced as they run the gamut of their training period. BIBLIOGRAPHY U) 10. ll. 12. BIBLIOGRAPHY Arbuckle, Douglas. "Fixe Philosophical Issues in Counseling,“ Journal of Counseling Psychology, 5:211—215, No. 3, 1958. Blizzard, Samuel. "The Role Perceptions of Protestant Ministers," unpublished speech, Princeton, 1958. Bott, Elizabeth. ”The Concept of Class as a Reference Group," Human Relations, 7:259—283, 1954. Centers, Richard. The Psychology of Social Classes, Princeton, Princeton University Press, c. 1949, Charters, w. W., and Theodore M. Newcomb. ”Some Attitudinal Effects of Experimentally Increased Salience of a Membership Group,” Readings in Social Psychology, Guy E. Swanson, ed., Henry Holt and (30., 1\’—ew York, c. 1952. Coleman, James S. The Adolescent Society, The Free Press, Clencoe, Illinois, c. 1961. Durnall, E. J., J. F. Moynihan, and C. C. Wrenn, ”Symposium: The Counselor and His Religion,” The Personnel and Guidance Journal, 362326-334, No.5, January 1958. Elder, James Lyn. "The Attitudes of the Pastoral Counselor," An Introduction to Pastoral Counseling, Wayne E. Oates, ed., Broadman Press, Nashville, c. 1959, Fielder, Fred E. "A Comparison of Therapeutic Relationships in Psycholanalytic, Non-Directive, and Adlerian Therapy,” Journal of Consulting Psychology, 14:436-445, 1950. Cump, Paul V. A Statistical Investigation of One Psychoanalytic Approach and 3 Comparison of It with Non—Directive Therapy,” Master’s thesis, Ohio State University, Columbus, Ohio, 1944. Hartley, Ruth E. "Norm Compatibility, Norm Preference, and the Accept- ance of New Reference Groups,” The Journal of Social Psychology, 52:87-95, 1960. Hiemstra, William L. "A History of Clinical Pastoral Training in the United States," Reformed Review, 17:30—46, 1963. 87— 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. -88— Hiltner, Seward. Pastoral Counseling, Abingdon Press, New York, c.1949. Hiltner, Seward. Preface to Pastoral Theology, Abingdon Press, New York, c. 1958. Hyman, H. H. "The Psychology of Status," Archives of Psychology, 19A2, no, 269. Johnson, Paul E. "Clinical Pastoral Training at the Crossroads," The Journal of Pastoral Care, N0. 2, 16:70-75, 1962. Johnson, Paul E. Psychology of Pastoral Care,Abingdon Press, New York, c. 1953. '— Kling, Frederick R., "The Motivations of Ministerial Candidates," Research Bulletin, Educational Testing Service, Princeton, 1959. Linn, L., and L. w. Schwartz. Psychiatry and Religious Experience, Random House, New York, c. 1958. Loetscher, Lefferts A. The Broadening Church, University of Pennsylvania Press, Philadelphia, c. 1954. Mannoia, V, James, ”A Pilot Analysis of the Preferences of Liberal and Conservative Ministers toward Directive and Non-Directive Res— ponses in the Pastor—Parishioner Counseling Relationship.’ un- published Ph.D. thesis, Michigan State University, E. Lansing, 1962. Merton, R. K. and A. S. Kitt, Studies in the Scope and Method of “The American Solider," Free Press, Glencoe, Illinois, c. 1950. Miller, Ralph L. "A Statistical Analysis of a Guidance Group Program," unpublished study, Princeton Theological Seminary, Princeton, 1959. Newcomb, Theodore, H. Social Psychology, The Dryden Press, New York, c. 1950. Ranck, James G. ”Religious Conservatism—Liberalism and Mental Health,” Pastoral Psychology, 112:34—40, vol. 12, 1961. Rosen, E. "Conf‘icting Group Membership: A Study of Parent—Peer Cross—Pressures,” American Sociological Review, 20:155-161, 1955, Saunders, David R. "Evidence Bearing on Use of the Myers~Briggs Type Indicator to Select Persons for Advanced Religious Training: A Preliminary Report," Research Bulletin, Educational Testing Service, Princeton, 1957. Sherif, Muzafer, An Outline of Social Psychology, Harper, New York, c. 1948. _— 29. 30. 31. 32. 33. 34. -89- Strupp, Hans, H. "Psychotherapeutic Technique, Professional Affilia— tion, and Experience Level," Journal of Consulting Psychology, 19:97-102, no. 2, 1958. Strupp, Hans, H. "An Objective CompariSon of Rogerian and Psychoanalytic Techniques," Journal of Consulting Psychology, 19:1-7, No. 1, 1955. Toch, Hans, "Religious Belief and Denominational Affiliation,” Religious Education, reprint from May-June, 1969 issue, pp. 193-200. Wise, Carroll A. Pastoral Counseling, Its Theory and Practice, Harper and Brothers, New York, c. 1951. Withrow, Quentin, ”A Study of the Possible Correlation Between Theolo— gical Orientations and Certain Variables of Personality,"unpublished Ph.D. thesis, University of Southern California, 1960. Wrenn, R. L. "Counselor Orientation: Theoretical or Situational," Journal of Counseling Psychology, 7:40-45, No. l, 1960. APPE N1) 1 X A -91- MICHIGAN STATE UNIVERSITY East Langigg College of Education I am writing you in regard to some current research that is being conducted here at Michigan State University in the area of pastoral counseling. The nature of the research is to egplore the relationship between religious belief and its effect upon the behavior of seminary students in the pastoral counseling situation. An initial study has been completed at Michigan State which indicates that religious belief has just about as much influence upon the counseling approach of ministers 'as does the degree and type of training he has had in counseling theory and practice.. We are now exploring the same relationship at the seminary level using Junio; clasg students from a number of leading representative seminaries across the country. Your seminary has been selected as one of the twelve whom we are request- ing to cooPerate in the study. Tb gather data for the latest aspect of the project it is necessary that we gain the assistance of your seminary in the distribution of a questionnaire to your Junior class students. Enclosed is a capy of the questionnaire and a more complete summary of the present piece of research. If, upon reading these materials, you extend your cooperating in the gathering of data from your Junior class, it will be most heartily appreciated. The inclusion of your students in the sample would contribute much to the study. And from this research we hope to make a small contribution to our knowledge concerning the training of ministers as pastoral counselors. One more word is needed. For convenience, a stamped return envelope and a brief form is enclosed by which you can indicate the response of your seminary. The questionnaire can be filled out by the student (in a half hour) and returned to the class. Thus the time factor for you involves simply distributing and gathering the questionnaire at the and of the class period. (Suggestions for administration of the questionnaire can be found in the enclosure.) Since the study deadline for gathering data has been set for May 20, we need to send out the questionnaires and have them returned as soon as possible. Consequently, an indication of your cooperation will result in our forwarding the questionnaires by return mail. If you have further questions, I would welcome your correspondence. I will be expectantly awaiting your response and decision. Sincerely yours, The Rev. Mr. Ralph L. Miller Instructor: Pastoral Counseling RLM:jic APPENDIX B -93- A.\' liXPLt’tRA'll’lRY ANALYSIS 01’ THE PRl-TFICRILNCFS ()1: FIRST-YEAR 'IHHHJKJICAL STUDENTS DYTQIGNATED AS LIBERAL AND (MNSI‘RVATI VP. THWARD 1)]: RHJ'I'IYE AND NON—DIRECTIVE RIJSPLA'SES 1N THE PASTOR- PARISHTONER CUFNSELTNG RELATIONSHIP 1. Purpose: The study is concerned with gathering primary source data from seminary juniors who are current students in twelve representative theological seminaries in the United States regarding their preferences of pastoral counseling responses. This investigation is designed to focalize the basic information upon the problem of theological impli- cations in counseling method in order to assess any possible relation- ship. The following hypotheses will be tested: 1. There will be a signifitmnrt difference in the number (fif