College and Research Libraries Characteristics of 11 Success 11 among Academic Librarians John N. Olsgaard This study examines selected characteristics of U.S. academic librarians listed in Who's Who in Library and Information Services. Through use of a systematic sample of entries, a com- posite profile of attributes of successful academic librarians was delineated. Among the results that were tabulated in this study were: men had a greater chance of being deemed successful in the profession; approximately a third of the academic entries had advanced degrees in addition to theM. L. S.; and successful academic librarians tended to work in ARL-member institutions. A discussion of the problem of detennining II success'' for academic librarians and recommen- dations for further research are also given. n his book about the Mercury space program, writer Tom Wolfe attempted to define the factors that led to the selection of the successful astronauts. Unable to ad- equately describe these factors, he de- cided that the successful astronauts had "the right stuff"; they had whatever it took to succeed. 1 In a similar way, one of the more difficult problems in the field of academic librarianship is the measure- ment of what constitutes ''success' I in the profession. What makes the successful li- brarian different from the unsuccessful li- brarian? What may initially seem to be merely a speculative question becomes · more important when one is serving on a tenure committee or completing a perfor- mance evaluation. It is at the time that one is evaluating the success of another indi- vidual that the question becomes espe- cially pragmatic. The purpose of this study is to conduct a preliminary investi- gation into the characteristics of success among academic librarians. Unlike the business community, where success can be measured in terms of wealth, the nonprofit sector, including li- brarianship, must develop other mea- sures of achievement. Logically there are two methods of determining whether ali- brarian has been successful. The first is to set up a theoretical model of what consti- tutes success, and then to compare indi- viduals against that ideal state. The sec- ond method of determining success is to gather together a group of those individ- uals thought . to be successful and deter- mine what attributes they have in com- mon. This study will share the characteristics of both methods. In 1982 the American Library Associa- tion published a biographical directory en- titled Who's Who in Library and Infonnation Seroices, which listed some twelve thou- sand of the more successful contributors to the field of library and information sci- ence in the United States and Canada. Success was judged by the editors of Who's Who on the basis of a comparison of individuals against a theoretical model composed of the following guidelines: 1. Evidence of active participation in professional, educational, or service orga- nizations. 2. A record of activity in support of li- braries and information services. 3. Substantial experience as a practi- tioner or educator. 4. Contributions to the professional lit- erature. 5. Receipt of awards or honors from John N. Olsgaard is a doctoral candidate at the Graduate School of Library and Information Science, University of Illinois at Urbana-Champaign. 5 6 College & Research Libraries professional or educational organiza- tions.2 Assuming that Who's Who does provide the best available estimate of what consti- tutes success in the profession, it can be further presumed that certain attributes of those included in the listing can be mea- sured and calculated. Who's Who was used in much this same manner to examine the characteristics of ARL directors in an arti- cle by Ronald Dale Karr. 3 This study will measure selected personal characteristics of U.S. academic librarians listed in Who's Who in all types of academic institutions and job positions. A product of this exami- nation of the ''successful'' is to explore the possible direction the profession is mov- ing in several educational and employ- ment areas. METHOD Data for this study were collected by us- ing a systematic sample of the entries in Who's Who (in this case k = 31). The sam- ple size needed to validate a bound on the error of .05 was delineated by use of a pro- portional allocation model where, through use of a pre-sample, the propor- tion of academic listings to nonacademic listings was estimated at .4 to .6, respec- tively (see appendix A). The intent of the above formulation was to draw a valid sample of a subset (i.e., academic librari- ans) from the population (i.e., all listings in Who's Who) within statistically con- trolled bounds. In this case the sample size needed to validate the survey was 373. The number of entries collected to- taled 382. The general composition of the sample is given in table 1. The following rules were implemented to standardize entries and to make the results compatible with previous research. 4 1. To determine the sex of an entry, an analysis of first names was undertaken in conjunction with the following rules: (a) first names that could be of either gender or that were not recognizable as being at- tributable to either gender were listed as indeterminable; and (b) first names of en- tries represented only by initials were listed as indeterminable. 2. Age was calculated using 1980 as a base year. January 1984 3. The data entry for a college or for a university was taken directly from the in- stitutional name. 4. In the context of this study, the job designation of administrator includes the job titles of director, dean, and associate or assistant dean of an academic library. Ad- ministrative titles of departmental li- braries were listed as departmental job po- sitions. Tr£e Academic Special Public School TABLE 1 COMPOSITION OF WHO'S WHO SAMPLE N= 148 87 49 24 Library school faculty 17 Retired 7 Non-U.S. 23 Other 27 Total 382 %= 39 23 13 6 4 2 6 7 100 5. Geographic location was entered by grouping states into the following re- gions: (a) Northeast-Connecticut, Dela- ware, District of Columbia, Maine, Mary- land, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont; (b) Southeast- Alabama, Florida, Georgia, Kentucky, North Carolina, South Carolina, Tennes- see, Virginia, West Virginia; (c) Midwest-Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Ne- braska, North Dakota, Ohio, South Da- kota, Wisconsin; (d) Southwest-Ari- zona, Arkansas, Louisiana, Mississippi, New Mexico, Oklahoma, Texas;- (e) West-Alaska, California, Colorado, Ha- waii, Idaho, Montana, Nevada, Oregon, Utah, Washington, Wyoming. 6. For purposes of this study, a publica- tion was defined as a titled print article or monograph. Editorships and nonprint projects were not included. The data collected are presented within four general headings: I. Demographic Characteristics; II. Educational Degree Obtainment Characteristics; III. Employ- ment Characteristics; and IV. Publication Characteristics. I l Characteristics of "Success" 7 TABLE 2 GENDER OF ACADEMIC LISTINGS BY REGION Women Men Region (Percent) (Percent) N= d =· N- d = Northeast 43 .2 56:8 40 3 37 Southeast 52.2 47.8 24 1 23 Midwest 66.7 33.3 30 0 30 Southwest 46.2 53.8 14 1 13 West 44.2 55.6 40 4 36 Com~osite average 50.4 49 .6 148 9 139 PoEu ation average 61.5 38.5t *dis the number of entries for which data could not be determined. +Represents the percentage of all U.S. academic librarians. Association of College and Research Libraries, Salary Structures in Higher Education for the Academic Year 1975-1976 (Chicago : American Library Assn., 1976), p.6-12. I. DEMOGRAPHIC CHARACTERISTICS OF ACADEMIC ENTRIES The analysis of demographic character- istics of the academic library listings from Who's Who followed a three-level design. The first level of consideration was to de- termine the proportion of academic en- tries that were female or male (see table 2). Although the sample national proportions were roughly of equivalent size, the per- centage of female listings was ten percent- age points below their representative na- tional proportion of the population. However, it should be noted that the dif- ference between the proportions given from the Who's Who listing~ and the pro- portions of the population was only mar- ginally statistically significant (chi-sq. = 2.86, 0.10 ?: 2.71, d.f. = 1). That is, one can be 90 percent sure that the difference between the sample proportion and the population proportion was not caused by a chance distribution. The second level of consideration was to determine the geographic location of aca- demic listings in Who's Who. The results of this examination are presented in table 3. Every region of the U.S. was slightly un- derrepresented in comparison to their proportion of the population, with the ex- ception of the West region. For example, while 27 percent of the sampled listings in Who's Who resided in the Northeast, 31.3 percent of all academic librarians in the U.S. are in the Northeast region. This small difference between the geographic location of the Who's Who listings and the geographic location of the population of TABLE 3 GEOGRAPHIC DISTRIBUTION OF ACADEMIC ENTRIES Population Who 's Who Regional Listings Average• Region N= (in Percent) (in Percent) Northeast 40 27.0 31.3 Southeast 24 16.2 17.3 Midwest 30 20.3 24.7 Southwest 14 9.5 11.1 West 40 27.0 15.6 Total 148 100.0 100.0 *Represents percentages of all U.S. academic librarians . Data derived from U.S . Department of Health, Education and Wel- fare . National Center for Education Statistics, Library Statistics of Cplleges and Universities , Fall 1975: Institutional Data (Washing- ton, D.C. : Govt . Print . Off., 1977), p .221-78. academic librarians was statistically insig- nificant (chi-sq. = 5.89, 0.05 ?: 9.49, d.f. = 4). For general purposes this would in- dicate that the Who's Who listings of academic librarians, and by implication the sample drawn from it, are representa- tive of the general population of academic librarians in the case of geographic loca- tion. The third level of analysis examined the age of academic listings in Who's Who. The average age of all academic listings was 43.9 (N = 131). However, the average age for men was 45.6 (N = 69) and for women was 42.0 (N = 62). A histogram of the class frequency distributions in age is given in figure 1. This analysis demon- strates that a high proportion of the women academic librarians were in their thirties (48 percent), whereas only 28 per- cent of the men were in the same class fre- quency. Although the above distribution 8 College & Research Libraries % 50 40 30 20 10 0 48 20s 30s January 1984 Women rno Men ~ 30 40s 50s 60s FIGURE 1 Frequency Histogram of Academic Entries by Age and Gender may be arithmetically important, it does not demonstrate strictly statistically sig- nificant results (chi-sq. = 7.29, 0.20 ~ 5.99, d.£. = 4). II. EDUCATIONAL DEGREE OBTAINMENT OF ACADEMIC LISTINGS Several sections of analysis were used to quantitatively describe the educational degree obtainment of the academic list- ings in Who's Who. All sample entries had at least a bachelor's degree. The data in ta- ble 4 reflect the percentage of the sampled academic librarians that have been granted various professional and gradu- ate degrees. At the bachelor's degree level of educa- tional obtainment, an analysis was con- ducted of the subject majors given for aca- demic listings (see table 5). The vast majority (nearly two-thirds) of all the aca- demic listings had subject majors in the humanities. The most popular undergrad- uate subjects were in the fields of English and history, composing 25 percent and 21 percent of all listings, respectively. The percentages between men and women ac- ademic librarians were nearly identical in most broad categories of majors with the exception of the sciences, where men held TABLE4 EDUCATIONAL DEGREE OBTAINMENT OF ACADEMIC ENTRIES DeS!:eeT~£e N = %= * Professional B.L.S. only 7 4.7 M.L.S. 136 91.9 Neither B.L.S . norM.L.S. 5 3.4 Advanced M.L.S. & 2d Master's 47 34.6 M.L.S. & Doctoratet 10 7.4 M.L.S . & Doctorate in Library & Info. Sci- ence 2 1.5 *As a percentage of all academic entries . +Includes Ph .D ., Ed. D., D.L.S., and D.A. l l Characteristics of ''Success'' 9 TABLE 5 BACHELOR'S DEGREE MAJOR SUBJECT AREA FOR ACADEMIC ENTRIES Women Men All Entries Cate~o:r N= %= N= %= N= %= Humanities* 47 68.1 42 62.7 89 65.4 Social Sciences+ 8 11 .6 8 11.9 16 11.8 Applied Sciences+ 13 18.8 11 16.4 24 17.6 Sciences§ 1 1.4 6 9.0 7 5.1 Total 69 99.9 67 100.0 136 99.9 The number entries for which data could not be determined = 12. *Humanities includes the subjects : history, English, languages, music, philosophy, art, religion . +Social sciences includes the subjects: sociology, economics, political science, government, anthropology, communications, psychol- ogy. ;tA.pplied sciences includes the subjects : education, mathematics, geography, journalism, business, geology, forestry, health sciences (e.g ., nursing) . §Sciences includes the subjects : biology, physics, zoology, chemistry. a 9 percent to 1.4 percent superiority (chi- sq. = 3.99, 0.05 ~ 7.82, d.f. = 3). The second educational degree obtain- ment level analysis was conducted for those academic librarians that had re- ceived a master's degree in library science. As no surprise, the M.L.S. is the profes- sional degree of preference among aca- demic librarians; 92 percent have been awarded an M.L.S. versus 5 percent that have a bachelor's degree in library sci- ence, or the 3 percent that have neither the M.L.S. nor the B.L.S. On first examina- tion, one could be led to infer that the level of M.L.S. obtainment among the entries in Who's Who (i.e., 92 percent) was actually less than the M.L.S. obtainment level for the general population reported by Mi- chael D. Cooper in 1976 (i.e., 95 percent). 5 The difference of 3 percent was probably due to sampling error, rather than actual differences between the two groups (chi- sq. = 0.71, 0.05 ~ 3.84, d.f. = 1). How- ever, it should also be noted that there is no evidence that the rate of M. L. S. obtain- ment among the academic entries in Who's Who is any higher than it is in the normal population. Those academic librarians sampled that do have an M.L.S. are alumni of forty-two different library schools. An interesting sidelight to this examination is a fre- quency distribution ranking of library schools that was constructed of these list- ings and compared to a perception rank- ing of library schools by library adminis- trators conducted by Herbert S. White in 1981 (see table 6). 6 The top fifteen ranked schools generated from the listings in this study represent the granting institutions of the M.L.S. to 64 percent of all the aca- demic librarians sampled from Who's Who. Although the rank order of the library schools between the White study and this study's results are quite different, there is still a 53 percent (nine of seventeen) agree- ment rate between the composition of the lists. The next educational degree obtainment level surveyed was for other graduate de- grees in addition to the M.L.S. Approximately one in three of the aca- demic librarians listed in Who's Who had a master's degree in addition to an M.L.S. (34.6 percent). Of the group that had a sec- ond master's degree, 61 percent were men and 39 percent were women; a difference that was statistically significant at the 0.10 level (chi-sq. = 3.73, 0.10 ~ 2.71, d.f. = 1). The subject majors of the non-M.L.S. master's degrees followed a pattern quite similar to the data collected for bachelor's degrees. As with bachelor's degree data, nearly two-thirds of the non-M.L.S. mas- ter's degrees were in the humanities (see table 7). Among those that had received a master's degree in addition to the M.L.S., women held a small proportional superi- ority in both the humanities and the social sciences, while men held the proportional advantage in the applied sciences and the sciences. These subject differences were not statistically significant (chi-sq. = 2.66, 0.05 ~ 7.82, d.f. = 3). Of considerable interest to many in the 10 College & Research Libraries January 1984 TABLE 6 FREQUENCY DISTRIBUTION OF LIBRARY SCHOOLS ATTENDED BY ACADEMIC ENTRIES Who's Who Entries 1. Columbia (13)* 2. Michigan (11)* 3. illinois (7) * 4. Indiana (6)* Washington (6)* 6. Catholic (5) Denver (5) Peabody (5) Simmons (5)* 10. California, Berkeley (4)* Florida State ( 4) North Carolina (4)* Oklahoma (4) Pittsburgh (4)* Syracuse ( 4) usc (4) 53 additional entries distributed among 26 other schools. *Matches between the two lists . White's Perception Study+ 1. illinois (44) 2. Michigan (41) 3. California, Berkeley (38) Chicago (38) 5. UCLA (37) 6. Columbia (32) 7. Indiana (31) 8. North Carolina (29) 9. Pittsburgh (24) 10. Rutgers (19) 11. Simmons (15) 12. Drexel (9) 13. Washington (8) Wis ., Madison (8) 15. British Col. (6) Case Western (6) Texas, Austin (6) 39 additional responses distributed among 19 other schools . +Source : Herbert S. White, "Perceptions by Educators and Administrators of the Ranking of Library School Programs," College & Research Libraries 42:198 (May 1981) . TABLE 7 NON-M.L.S. MASTER'S DEGREE MAJOR SUBJECT AREA FOR ACADEMIC ENTRIES Category Humanities Social Sciences Applied Sciences Sciences Total N= 12 2 3 0 17 Women %= 70 .6 11 .8 17.6 100 The number of entries for which data could not be determined = 3. profession is the extent to which the doc- torate is becoming a prevalent require- ment in job listings for administrative po- sitions in academic libraries. Olsgaard and Olsgaard have reported that nearly 40 per- cent of the job listings for directors of uni- versity libraries preferred a doctorate. 7 Of those academic librarians sampled in Who's Who that had an M.L.S., only 7.4 percent also b.ad earned a doctorate (i.e., Ph.D., Ed.D., D.L.S., or D.A.) in any field, and only 1.5 percent had doctorates in library and information science. On a more comparable level, 13.6 percent of the sampled administrators in Who's Who had a doctorate in some field. The difference between the obtainment level of the doc- Men All Entries N= %= N= %= 17 63.0 29 65 .9 1 3.7 3 6.8 7 25.9 10 22.7 2 7.4 2 4.5 27 100 44 99.9 torate among the sampled administrators and the level requested in job listings was statistically significant (chi-sq. = 6.41, 0.05 ~ 3.84, d.f. = 1). The last level of consideration dealt with the average age that degrees were ob- tained at various educational levels (see table 8). The average age of men and women was quite similar at both the bach- elor's degree level and at the M.L.S . de- gree level. However, an interesting phe- nomenon does occur for those listings that have both an M .L.S. and a second mas- ter's degree. While women academic li- brarians who had both degrees received a non-M.L.S. master's degree an average of approximately two years after receiving Characteristics of "Success" 11 TABLE 8 MEAN AVERAGE AGE OF EDUCATIONAL DEGREE OBTAINMENT FOR ACADEMIC ENTRIES De~ee ________________ W~om~e~n ____ ~M~e~n ____ ~AU~E~n~rr~ies~--~N~= ______ d~=-·----~N~-~d~=-- Bachelor's 23.6 23.3 23.4 148 17 131 M.L.S. 30.0 29.4 29.7 148 27 121 M.L.S. + 2d Master's 31 .8 28 .3 29.5 47 7 · 40 *dis the number of enrries for which data could not be determined . an M.L.S., men academic librarians who had both degrees received their non- M.L.S. master's degree an average of one year before theM. L. S. However, the statis- tical significance of this difference is ques- tionable (chi-sq. = 0.10, 0.05 2:: 5.99, d.f. = 2). III. EMPLOYMENT CHARACTERISTICS OF ACADEMIC ENTRIES The analysis of the employment charac- teristics of the academic librarians listed in Who's Who will follow a two-level design. The first level of consideration was the dis- tribution of academic librarians by the type of institution in which they were em- ployed (see table 9). The highest distribu- tion frequency was for academic librarians employed in ARL-member U.S. academic libraries. 8 Nearly 46 percent of all the list- ings for academic librarians were em- ployed in ARL institutions. The lowest class frequency was for academic librari- ans employed in two-year colleges with 7.4 percent. A frequency histogram of the distribution of entries by institution is given in figure 2. The second level of analysis of employ- ment characteristics examined the type of job held by the academic librarians listed in Who's Who. The results, as given in table 10, are that nearly a third (31.7 percent) of all the academic listings are in an adminis- trative position. The next most frequent position was for the categories of refer- ence librarians or departmental librarians, both having approximately 18 percent of all the academic listings. It is interesting to note that while 42 percent of the entries for men were in administrative positions, the highest frequency for any job category among women was for the position of ref- erence librarian (24.3 percent). The differ- ence between the number of men in ad- ministrative positions and the number of women in administrative positions is sta- tistically significant (chi-sq. = 5.90, 0.05 2:: 3.84, d.f. = 1). IV. PUBLICATION ACTIVITY OF ACADEMIC ENTRIES The last area examined was the publica- tion activity of the academic listings in Who's Who. The analysis of this section again followed a two-level design and is presented in table 11. The first level of analysis examines the publication activity of men and women academic librarians. Although the overall rate for all entries was an average of 1.5 publications per en- try, the rate for men academic librarians was nearly twice that of women academic librarians (2.0 to 1.1, respectively). The second level of publication analysis examined the rate of publication by the TABLE 9 EMPLOYING ACADEMIC INSTITUTION Women Men All Enrries Tn:e N= % = N= % = d =* N= %= ARL 30 45.5 36 54.5 2 68 45.9 Non-ARL University 21 55.3 17. 44.7 4 42 28.4 Four-Year College ·u 44.0 14 56.0 2 27 18.2 Two-Year College 8 80.0 2 20 .0 1 11 7.4 Total 70 69 9 148 99.9 *dis the number of enrries for which gender data could not be determined . ... , .'t 12 College & Research Libraries January 1984 i. 50 40 30 20 10 0 46 ARL 28 Non- ARL Univ. 18 Four Year College 7 Two Year College FIGURE2 Frequency Histogram by Institutional Type type of academic institution in which the entries were employed. It was found that those employed in ARL institutions had the highest publication rate among the various types of academic institutions. However, the size of the institution does not necessarily accurately predict the pub- lication activity of the entries; for example, those employed in non-ARL university li- braries had a slightly smaller publication rate than those employed in two-year col- leges. TABLE 10 JOB POSITIONS OF ACADEMIC ENTRIES Women Men All Entries Tr£e N= %= N= %= N= %= Administration 15 21.4 29 42.0 44 31.7 Reference 17 24.3 8 11.6 25 18.0 Acquisitions/ Collection 3 4.3 4 5.8 7 5.0 Hecilth Sciences 4 5.7 1 1.4 5 3.6 Cataloging 5 7.1 2 2.9 7 5.0 De~artmental 12 17.1 15 21.7 27 19.4 Bib iographer 2 2.9 2 2.9 4 2.9 Law 5 7.1 3 4.3 8 5.8 Other 7 10.0 5 7.2 12 8.6 Total 70 99.9 69 99.8 139 100 *The number of entries for which data could not be determined = 9. TABLE 11 MEAN AVERAGE PUBLICATION RATE FOR ACADEMIC ENTRIES Category Gender Women Men Institution ARL Non-ARL University Four-Year College Two-Year College All Entries N= 70 69 68 42 27 11 148 DISCUSSION AND RECOMMENDATIONS FOR FURTHER RESEARCH Mean Rate 1.1 2.0 1.7 1.2 1.6 1.3 1.5 Using the above analyses, and assum- ing that the entries portray a set of attri- butes accepted as describing the ''success- ful'' librarian, it is possible to draw a com- posite profile representing a majority of those academic -librarians listed in Who's Who. I. Demographic Characteristics: 1. Gender. Although there are approxi- . mately the same number of men and women academic librarians who are suc- cessful, there is a better chance to succeed if one is male. 2. Geographic location. A majority of successful librarians reside in the North- east or West regions. 3. Age. The successful academic librar- ian is in his/her forties (the men being slightly older than the women). · II. Educational Degree Obtainment Charac- teristics: 1. The successful academic librarian has a bachelor's degree, preferably in the hu- manities. 2. The successful academic librarian has an M.L.S., preferably from a highly ranked school. 3. The successful academic librarian probably does not have an advanced de- gree in addition to the M.L.S., but a large Characteristics of "Success" 13 number of his/her colleagues do have one in the humanities. III. Employment Characteristics: 1. It is highly probable that the success- ful academic librarian will work in a large academic library rather than in a small in- stitution. 2. The successful male academic librar- ian will probably be an administrator. The successful female academic librarian could be an administrator, but would more likely be a reference librarian. IV. Publication Characteristics: 1. The successful academic librarian publishes in the professional literature, but does not publish very much. The purpose of this study was to ob- serve certain quantifiable characteristics of successful academic librarians, at least as judged by one set of standards for success. That is not to assert that these standards for success encompass an exclusive list, or that the characteristics of those judged to be successful are necessarily desirable. For example, while it is not necessarily desir- able that most administrators are men, it is important to observe that most adminis- trators are men . It could also be claimed that most suc- cessful librarians demonstrate certain qualitative characteristics such as having a "pleasant personality." While such claims might be true, it would probably be impossible to describe those characteris- tics either in a study such as this, or when dealing with the legalities of a tenure hear- ing. What is needed is a set of measurable standards for success that would guide performance evaluation committees and library school educators. This future study would establish a theoretical model of suc- cess that would be acceptable to the major- ity of the profession. After all, the profes- sion can hardly berate administrators for not demanding excellence, or educators for not teaching the "right stuff," if we don't know what constitutes the "right stuff." REFERENCES 1. Tom Wolfe, The Right Stuff (New York: Farrar, 1979) . 14 College & Research Libraries January 1984 2. Joel Lee, ed., Who's Who in Library and Information Services (Chicago: American Library Assn., 1982), p.ix. 3. Ronald Dale Karr, "Becoming a Library Director," Library Journal108:343-46 (Feb. 15, 1983). 4. Martha C. Adamson and Gloria J. Zamora, "Publishing in Library Science Journals: A Test of the Olsgaard Profile," College & Research Libraries 42:235-41 (May 1981); Bob Carmack and John N. Olsgaard, "Population Characteristics of Academic Librarians," College & Research Libraries News 42:141-42 (May 1981). 5. Michael D. Cooper, "A Statistical Portrait of Librarians: What the Numbers Say," American Li- braries 7:327-30 (June 1976). 6. Herbert S. White, "Perceptions by Educators and Administrators of the Ranking of Library School Programs," College & Research Libraries 42:191-202 (May 1981). 7. John N . Olsgaard and Jane Kinch Olsgaard, "Post-MLS Educational Requirements for Academic Librarians," College & Research Libraries 42:224-28 (May 1981). 8. U.S. academic institutions that were members of ARL were taken from Gordon Fretwell, comp., ARLAnnual Salary Survey, 1979-1980 (Washington, D.C.: Association of Research Libraries, 1980). APPENDIX A: PROPORTIONAL ALLOCATION MODEL n = sample size N = population size Npq n= _____ _ (N- 1) D + pq where B2 D= 4 p = estimate of the proportion that are academic librarians ·q = 1- p B = bound on the error of estimation; in this case .05