College and Research Libraries JOHN WILKINSON, KENNETH PLATE, and ROBERT LEE A Matrix Approach to Position ·Classification This article presents the results of a . study prepared in connection with the development of a classification system for librarians in the University of Western Ontario. The study emp.Zoys the formulation and application of four "general" and five "ranking" factors in posi- tion classification; the development of a matrix using the above fac- tors on a vertical axis and the desired number of positions on a hori- zontal axis; and the analysis of all existing professional positions in terms df the matrix. Administrative implementation of the findings is also described, including the appeals procedures established. EARLY IN 1971 Dr. Robert Lee, univer- sity chief librarian of the University of Western Ontario, Canada ( UWO), in- vited Dr. John Wilkinson and Dr. Ken- neth Plate of the University of Toron- to Faculty of Library Science to consult with his administrative group and with representatives of the Librarians' Asso- ciation of the UWO Library System in order to develop a classification program for librarians in that system. The fol- lowing paper is a result of the ensu- ing cooperation between the two consul- tants, Dr. Lee, and many members of his professional staff. THE PuRPOSE OF REVIEWING A CLASSIFICATION PROGRAM FOR PRoFEssioNAL PosiTIONS This paper presents in brief a some- what different approach to position clas- sification for librarians. This approach _ -while developed specifically for the ] ohn Wilkinson and Kenneth Plate are members of the Faculty of Library Science, University of Toronto, and Robert Lee is university chief librarian, University of Western Ontario, London, Ontario. UWO system, which is considered to be relatively advanced in its management by objectives concepts and in the strength of its management at all levels -can be applied to all types of medi- um-sized and large library organizations. Indeed, the approach involved the for- mulation and application of a schemata of "general" and "ranking" position characteristics which may have consid- erable theoretical as well as practical in- terest for librarians. Insofar as the anal- ysis is of interest, it will also be contro- versial; but controversy is the crucible of philosophy, and librarianship, we may at least agree, is long overdue for a philosophy of its own. Nothing is more central to the health of a library organization than the mo- tivation of its staff. Nothing is more detrimental to that motivation than an irrational position classification. Given that relatively large and complex organi- zations apparently require at least a modicum of hierarchical stability, it does not follow that such stability can be achieved through administrative fiat alone. Indeed, in a society more and more influenced by concepts of egalitar- / 351 352 I College & Research Libraries • September 1975 ianism and .specialization, the legitimacy of externally imposed power is being in- creasingly challenged. No longer is the dictate that a superior position exists by virtue of its superiority-that a depart- ment head is a department head is a de- partment head-wholly acceptable to a subordinate possessing specialized skills .and the knowledge of having made an important contribution to an organiza- tion. Nor should administrative au- thority as such be acceptable in a healthy organization-particularly if that organization includes a strong "pro- fessional" element. Some other ration- ale for position classification and hier- archical development is clearly needed. We will not here become involved in the increasingly tiresome debate as to the nature of professionalism, except to note that, however we define "pro- fessionals," the concept of a specialized body of knowledge and consequently of "expertise" seems inevitably to de- velop. It is fmportant at this point to note also that nothing in the definition of "professional" nor in the applica- tion of the general factors proposed in this paper should be construed as being derogatory to support positions. The constants used are merely delineators. A support position may clearly be more valuable to an organization than a pro- fessional position in a given context. A defensible position classification must attempt to optimize the use of hu- man resources. A defensible profession- al position classification must rational- ize, first, the optimum use of profes- sional personnel as distinct from sup- port staff and, second, the ranking of professional positions on an ordina·l scale in terms of key factors of profes- sional service. First, unless the classification program clearly distinguishes between profession- als and support staff, it cannot be · justi- fied as a professional position classifi- cation; and its application may indeed destroy the professional nature of the organization. Second, unless the pro- gram leads to a rational, viable, and easily justifiable ordinal ranking of pro- fessional positions, it can have little val- ue as a classification tool. The following approach meets the first criterion in that it identifies those characteristics which should be common to all professional positions and which, in total, delineate such positions from support positions on the one hand and from administrative positions on the other. These factors are here termed "general factors." The approach meets the second cri- terion in that it identifies those factors, distinct from the general ones, which should show greater development as the levels of positions in the classification schedule rise, and which may, therefore, be used to rank positions in · a broad classification scheme. These factors are here termed "ranking factors," and they are significantly characterized by their quality of continuum growth which en- ables them to alter in degree but not in kind as the continuum nears the upper levels of the schedule. Ranking factors, therefore, may be viewed as develop- mental stages in a given continuum. There are a number of studies in the literature of librarianship useful in constructing a position classification pro- gram. Two studies, the first by Rothen- berg et al. and the other by Ricking and Booth, provide definite statements use- ful in distinguishing between profes- sional and nonprofessional tasks. 1• 2 These ground-breaking studies do not, however, attempt to enunciate distinc- tions between the levels of professional positions. In an article published in 1965, Downs and Delzell also emphasize the differences between professional and nonprofessional tasks. 3 In addition, they describe what librarians do at vari- ous levels, using examples from the uni- versities of California, Michigan, and Illinois and the U.S. Civil Service Com- mission. The elements of position classi- fication · for librarians as presented in the Downs and Delzell article were use- ful in our preliminary planning at the University · of Western Ontario. How- ever, in terms of the present study, two major drawbacks were identified: ( 1) Because · of the modified management by objectives philosophy of library man- agement at the University of Western Ontario, such elements as "supervisory responsibility~' had necessarily to be de- emphasized. · Furthermore, provision had to be made for a team approach to problem solving, with emphasis on plan- ning and development and with distinc- tions made between projects and pro- grams. ( 2) The elements or factors used in the examples presented by Downs and Delzell were not always present in differing amounts in the position levels described in the present study. The best report of a library position classification program to date is the ar- ticle by Tompkins describing the meth- od by which positions were classified at the University of Michigan. 4 Tompkins and her committee opted for the- point rating system, a ·quantitative system that yields a numerical score which can be translated into a classification level. The principal advantage of such a quantita- tive approach is that it apparently de- velops a "precision of applicability,~' i.e., that it can be applied in a scientific manner. 5 Appearances may however, of course, be deceptive and ''research re- sults . . . show that less complex systems yield results almost identical with com- plicated systems .... "6 Bearing in mind that the principal disadvantage of a quantitative system such as that used at Michigan is the amount of time re- quired for its conception and imple- mentation, one must decide whether the extra time and expense in terms of the relevant precision achieved can be justi- fied. Tompkins notes that an average of thirty interviewer's hours was necessary to describe each position using this sys- tem for 105 positions! She notes further Matrix Approach I 353 that ~'twenty-one months after its incep- tion, the committee held its forty-eighth and final .meeting."7 It was soon decided at UWO that, based on Torripkin's ex- perience, we could not justify such an extended and costly program without more conclusive evidence that superior results would ensue. The factors used atWestem do, however, bear some simi- larity to those selected for use at Michi- gan with ,the important difference that the management by objectives approach at UWO precluded using some of the traditional yardsticks of position classi- fication. For example, Michigan's "inde- pendence of performance~' factor is de- veloped through the five grades in terms of freedom from supervision, whereas the "independence and freedom of ac- tion" factor in the present matrix is de- veloped in terms of systems constraints (projects, programs, and objectives) as these relate to individual positions. An- other example involves problem solving and decision making. Here Michigan's classification assumes a larger number of. problems which become more com- plex at the higher end of the scale. While this is undoubtedly true of most organizations, the introduction of man- agement by objectives diffuses the locus of complex problems through many lev- els. What determines the level of a po- sition in the present study in this regard is whether general participation in prob- lem solving is required or whether a po- sition carries specific responsibility for problem solving and planning and whether for tasks or projects, programs or policy, and ultimate approval or veto. IDENTIFYING AND DEFINING THE F AcroRs IN PosiTION CLASSIFICATION General Factors The general attributes which distin- guish all professional positions are four in number. Basic academic qualifica- tions form one such factor. The basic library science degree represents basic 354 I College & Research Libraries • September 1975 professional expertise and potential for promotability (in the case of a position requiring a professional specialist who is not a librarian, the appropriate de- gree would be the equivalent of the B.L.S. or M.L.S.). Arguments concern- ing the quality of relevant professional education do not apply in this instance, because no other generally accepted baseline for professional expertise has thus far been developed. However, as with all the general factors in this ap- proach, the requirement of basic aca- demic qualifications cannot alone be considered as justification for ranking a position as professional. Judgment is a second general factor. Judgment, in the professional context, involves the application of expertise to the reasoned and defensible analysis of alternatives in order to determine a preferable course of action. A position must require frequent application of such expertise in order to be regarded as professional. It is recognized, of course; that some degree of judgment is concomitant with survival, on the job or off it; but, whereas the clerical's exer- cise of judgment is normally circum- scribed and assisted by established rou- tines, the judgment of the professional should be less dependent upon prece- dent than upon comparative knowledge and expertise. Unless decision making involving the adjudged choice between alternatives as described above is present as a normal element of a position, the need to exercise the basic expertise al- ready specified as the rationale for in- cluding basic academic qualifications as a general factor is absent. Judgment, therefore, is that factor which exercises the latent force of relevant substantive education. Client relationships, the third general factor, are a hallmark of all established professions. Without such relationships, which carry with them a mentor conno- tation, professional expertise is not uti- lized on a one-to-one service basis. The client relationship, embodying as it does interpersonal change and enrichment, has indeed traditionally been one of the qualities distinguishing professional ex- pertise from creative skills. Once again it is important to realize that the client relationship is a general factor in part because it implements other general fac- tors. The requirement of client relation- ships in a position does not alone rank that position as professional. Voluntary involvement in profession- al activities is a general factor common to all professional positions. It would be unusual to find such involvement as a formal requirement of a position; but, leaving aside the question of the adequacy of professional associations, it is highly unlikely that professional development in a position can occur without the type of formal external contacts represented by professional as- sociations and other forms of continu- ing self-education. A position that does not involve from time to time external professional activities as an enriching and necessary element is unlikely to be a viable professional position in terms of the other general factors discussed above. All the preceding factors-basic aca- demic qualifications, judgment, client relationships, and involvement in pro- fessional activities-are general to all professional positions. Their joint pres- ence distinguishes such positions from support positions; but alone they are not viable instruments for ranking posi- tions because they are not themselves capable of being conceived of on a con- sistent growth continuum. Thus, for ex- ample, while additional academic and professional degrees represent evidence of additional expertise and may indi- cate suitability for promotion, the mini- mum number of degrees required by a position-stated as such and without reference to the degree of expertise re- quired by a position-is not a ranking factor because it is a given constant. (By analogy, we may determine that a lamp must give light to be a lamp; but we cannot usefully extrapolate from this general factor to rank all lamps by the degree of light they give. An object that does not give light, we may agree, is not by definition a lamp, but for cer- tain purposes a dim light may be much "better" than a bright light. ) Ranking factors are context-oriented; whereas general factors are arbitrary. Thus, scope for judgment may increase as the scope of responsibility and ac- countability increases; but the quality of judgment is an unrankable constant. The same is true of client relationships, although the level of external contacts can, as we shall see, be used as a criteri- on for ranking. Ranking Factors The development of ranking factors is based on four assumptions. One we have already noted-that medium and large-sized organizations do need the sta- bility provided by hierarchical ranking. The second assumption is that ranking must be based upon demonstrably rele- vant factors which can be severally ex- tended on a logical continuum. The third assumption is that arbitrary ad- ministrative designations are not, as such, demonstrably relevant to the ob- jectives of modem library organiza- tions, and that such designations are not viable over an extended continuum. The fourth assumption is that ranking factors for professional positions must relate to those general factors which identify the positions as professional. The degree of expertise required in a position to exercise assigned responsi- bility and accountability is a ranking factor. At the base level this factor equates to the basic academic qualifica- tion character upon which it rests. Be- fore the expertise factor can be de- veloped, therefore, it must include the variables of relevant in-service training, experience, professional activities, and Matrix Approach I 355 further formal education. The addition of these variables, capable of quantita- tive and qualitative analysis, provides the means of raising this ranking factor through the continuum. The greater the quantity and quality of relevant ex- pertise required by a position, the great- er the value of that position to the or- ganization and the higher the rank given to the position. In short, with re- spect to the quality of expertise, a basic principle of the approach developed in this paper is that the highest classifica- tion shall be reserved for those positions requiring the expertise necessary to ex- ercise the broadest scope of responsibil- ity and accountability. This principle in no way conflicts with current management trends toward more participative "team approaches" in decision making and the implementa- tion of library objectives. As responsi- bility and accountability are diffused throughout a system, expertise is at the same time diffused and increased and more positions will receive a higher ranking in this respect. Moreover, it should be stressed that management ex- pertise is itself a legitimate form of professional expertise (albeit perhaps less recognized in libraries in terms of formal education than in many other types of institutions). Thus a position requiring advanced administrative ex- pertise would rank high in terms of ex- pertise, as would a position requiring advanced bibliographical and subject expertise. The independence and freedom of action inherent in a position is a second ranking factor. The de facto "span of discretion," which is a quantitative tem- poral measure of the degree of inde- pendence and freedom which the or- ganization can accord to a position, is a viable indicator of the level at which the position should be classified on the continuum. The level of formal external contact which a given position requires to be ef- 356 I College & Research Libraries • September 1975 fective is a ranking factor in classifying the position. The importance of a sub- unit to the system of which it forms a part is largely determined by the degree of effective formal communication be- tween the unit and the system. This is true whether the unit is a department of a library in communication with the library, or a library in communication with the parent body or system of which it is a part. Clearly, however, the importance of a communication is di- rectly related to the expertise, the sig- nificance of error, and the span of dis- cretion involved on the part of both the sender and of the receiver. Thus a necessary formal communication be- tween a university president and mem- ber of a library staff has broader or- ganizational implications (and hence more "importance") than has a similar communication between two subordi- nate positions. The responsibility for communication with the president in- herent in a position places that position in a higher rank than that of a position charged with lower level communica- tions. Inherent in the level of formal ex- ternal contact is the level of responsi- bility and accountability for any posi- tion, which can be determined qualita- tively largely by the scope and depth of the responsibility involved and by the significance to the organization of an error made · .at the level of involvement in question. The greater the significance of error the higher the level of respon- sibility. In one sense, it is true, the sig- nificance of error reflects the positional impact upon the organization; but the stress is upon the degree of expertise re- quired to avoid the error, as well as upon the hierarchical freedom from constraint. De facto rather than de fure responsibility and accountability assist in measuring the level of external con- tact. Formal external contact is the ranking counterpart of the general factor of client relationships, the latter being, however, nonnally informal, self-ini..: tiated, and characteristic of all levels of professional performance. It is impor- tant to note, moreover, that formal ex- ternal contacts must be· necessary to the achievement of positional objectives in order to be considered a ranking factor, and that such contacts must demonstra- bly draw upon the appropriate level of expertise. Planning and development form a ranking factor. The degree to which a position necessarily carries with it the need to plan and to develop programs is an indicator both of required exper- tise and of span of discretion. All pro- fessional positions should, as we have noted, require the use of judgment; but greater judgment in the use of compara- tive data is required in the development of programs than of tasks. Indeed, in terms of planning and development, stress should be laid as much upon the importance to the organization of the projects planned and upon their com- plexity as upon the planning and de- velopment activity itself. In assessing this factor, therefore, the distinction between tasks (which are normally sin- gle operations involving either repeti- tion or a short span of discretion) and programs (which consist of the relating of a number of tasks) is important. The latter category presumably involves greater expertise, span of discretion, judgment, and significance of error. Supervision is a ranking factor, though not if interpreted merely in terms of the number of personnel su- pervised. The number of individuals di- rectly supervised by a position (span of control) is not an indicator or relative rank since the span of control normally shortens as the responsibilities of those supervised increases. The number of in~ dividuals indirectly supervised may, however, be of some value in assessing the rank of a position; but this criteri- on, in fact, relates more to the arbi- trary administrative scheduling of a hierarchical position than to its inherent value to the organization. Nevertheless, because this factor has quantitative val- ue in a hierarchical system; because it will undoubtedly continue for some time to retain a traditional aura; and because it normally reinforces other ranking factors such as independence (span of discretion) and level of for- mal external contact (significance of er- ror), supervision has some . value in ranking a position . . THE: MATRIX The above four general and five rank- ing factors may be used to form the vertical axis of a. matrix (Figure 1 ) . The general factors do not appear ·as such on the vertical axis . of the rna trix because, although implicit in many of the ranking factors and though neces- sary to the ''professional" nature of a position, they have little or no ranking capability. The five ranking factors do appear on the vertical axis, however, be- cause they can be used to rank positions, although ''supervision" may at times be regarded as ~'not applicable." The horizontal axis of the matrix I Matrix Approach I 357 specifies the positions in the classifica- tion-Professional I, Professional II, Professional III, etc. The development of a ranking factor must be capable of being plotted along this horizontal axis (a necessary characteristic of any rank- ing factor since those factors operate on a continuum and can be divided horizontally at any number of con- venient, and clearly defined, points ) . Thus, to take expertise as an example, it remains discretionary within the or- ganization to determine 'now much" ex- pertise is required within one class of position and at what point on the con- tinuum the expertise involved tran- scends a given class and raises the posi- tion in question to a higher class. Con- ceivably an organization could choose to incorporate only two classes of posi- tions (or possibly only one), in which case the span of expertise demanded within each position class would be very wide. The only caveat to the above is, of course, that the lowest category of professional positions cannot include criteria ''below" those specified in the description of the general factors which establish the professional nature of all positions. I Positions in the Classification ( P n) I General Factors I "Growth" continuum Degree of I Expertise I Independence ' "Growth" continuum and Freedom I of Action I -• Level of Formal ; "Growth" continuum External Contact I Planning and I "Growth" continuum I Development I Supervision I . "Growth" continuum ( ? ) I of Personnel I Fig. 1 The Position Classification Matrix • • • • • 358 I College & Research Libraries • September 1975 APPLICATION OF THE MATRIX TO THE UNIVERSITY OF WESTERN ONTARIO LIBRARY SYSTEM The factorial principles underlying the matrix, when applied to the Uni- versity of Western Ontario Library Sys- tem, led to the development of the five ranking factors on a six-level contin- uum (Table 1). As has been indicated earlier, the four general factors in the schemata were used to determine the suitability of a post for inclusion as a professional position. The first task of the consultants was, therefore, to obtain job descriptions for all positions classi- fied as "professional" by the UWO Li- brary System, the questions pertaining to the descriptions being prepared and pretested by the consultants and the answers being supplied by the incum- bent in conjunction with the relevant assistant director or department head. Those posts which could not, by applica- tion of the four general factors, be considered ''professional" were then "declassified"; but it was made clear to the incumbents that the position and not the person was at issue. No individ- ual was "demoted" because the position held was declassified, and professional librarians found to be performing non- professional duties were promised real- location to professional posts as soon as possible. Next, the five ranking factors were applied to the remaining job de- scriptions to determine the level at which the position should be placed in the overall six-level classification. Only the ranking factors were used in this de- termination, and the level of any given position in the existing hierarchy was not per se a consideration. For the University of Western On- ~ario .Library System, the following cod- Ification of the matrix continuum was established. Professionall This level predicates, as do all other levels, the four general factors-aca- dem~c qualifications, judgments, client relationships, and voluntary involve- ment in professional activities-re- quired for inclusion of a position in the schemata. However, grade 1 posi- tions carry minimal responsibility and accountability, are subject to frequent supervision so that there is a minimal span of discretion, do not normally in- volve levels of formal external contact do not involve primary responsibili~ for planning and development, and do not involve the supervision of profes- sional personnel or independent proj- ects. Professional 1 positions are profes- sional positions and should not routine- ly include duties normally assigned to support staff; but this level is regarded a~ a stepping -stone or internship for higher levels and is not to be considered as a career grade. Professional 2 This level, the first professional ca- reer grade, is characterized by a distinct though limited span of discretion and a demonstrable expertise. Positions in this grade carry with them the responsi- bility and authority for the implemen- tation of a specific task or tasks, though the scope of the task( s) is limited. The planning and development of the inter- nal development of each assigned task is part of Professional 2 responsibility, but the scope of the task is determined externally. Levels of formal external contact are normally limited to those members of the library, university, or community whose cooperation is neces- sary to the implementation of the as- signed task. Supervision of limited proj- ects is implicit at the Professional 2 lev- el; but supervision of personnel is not characteristic of this level and, if pres- ent, is normally restricted to support staff and Professionalllibrarians. Professional3 This level is characterized by the strength of the superv1s10n factor, ei- ther of projects or personnel. With re- spect to personnel, Professional 3 posi- tions carry the responsibility for the di- rect supervision of support staff and of Professional 1 and 2 librarians who are normally the largest proportion of a li- brary's personnel complement. Account- ability is to a sharply decreased sector of the hierarchy, and span of discretion is correspondingly increased. Planning and development become relatively im- portant at this level because of the ne- cessity for integrating projects and per- sonnel. Professional 3 positions are nor- mally primarily concerned with internal administration; and formal external contacts, when present, are with the level of the library, university, and com- munity which has decision-making func- tions. The degree of expertise required to exercise assigned responsibility and accountability at the Professional 3 lev- el must be such that personnel or proj- ect units of complexity and consider- able magnitude can be handled inde- pendently and successfully. Professional4 This level is characterized by the strength of the external contact factor. Professional 4 librarians are responsible for effectively presenting the library to policy advisory levels of the university and of the community and for assimi- lating external considerations into the library system. The degree of expertise necessary to meet this responsibility may be found either in management skills, highly developed professional compe- tence, or specialized subject knowledge. Major project management not involv- ing subordinate personnel, as well as the management of major personnel units, meets grade 4 requirements. The span of discretion for this level may be lim- ited only by annual reports to Profes- sional 5 or 6 personnel, and Profession- al 4 responsibility for planning _ and de- Matrix Approach I 359 velopment may include the initiating of conceptual schemata. Professional5 This level is characterized by a re- sponsibility for the broadest areas of homogenous activity and by direct ac- countability to sixth level positions. All ranking factors reach their culmination at Professional 5; and the general fac- tors distinguishing professional posi- tions should be present in an unusually high degree to develop fully the ele- ment of expertise. Supervision at this level is to be interpreted as coordina- tion. Professional 6 This level is characterized as the ini- tial repository of the authority and re- sponsibility delegated by the university to the library. Accountability at this level is to individuals or bodies external to the library. Professional 6 span of dis- cretion is subject only to the regulations of the parent body, and the application of expertise rather than the expertise itself distinguishes this grade from the Professional 5 level. ADMINISTERING THE INSTRUMENTS The basic instruments in the applica- tion of the matrix discussed above to the professional staff of the UWO Li- brary System were the codified matrix and the job description questionnaire. The codification was discussed fully with the UWO Library System admin- istrative group in order to ensure that the application would realistically re- flect any constraints of the UWO sys- tem. The administering of the question- naire was equally carefully prepared for, both by the administrative group and by the consultants. As has already been indicated, the questionnaire was pretested and modified where necessary. The modifications were then discussed with the administrative group. An orien- tation session was held with the profes- TABLE 1 CLASSIFICATION FOR PROFESSIONAL POSITIONS Pl P2 P3 P4 P5 P6 f$ 0 B.L.S. or basic Normally requires Normally requires Requires equivalent Requires varied and Requires the equiv- --M.L.S. or its degree two years profes- the equivalent of of five years at the synthesized experi- alent of the grade CJ equivalent. sional experience five years at the grade 3 level and ence leading to a 5 level including c ~ No experience and ability to func- grade 2 level and the advanced degree comparative knowl- . senior administrative ~ (J'q required. tion independently specialization in in order to carry edge base. experience in a ~ and effectively in management skills, conviction in ef- Requires the ad- comparable aca- c:- Degree client relationships, subject knowledge, fective external con- vanced degree. demic library which ~ of to supervise small or a very high level tacts and highly demonstrates su- ~ Expertise projects and/ or of professional com- developed profes- perior skills in c., ~ small service units. petence. Advanced sional competence planning, organizing, ~ .., degree not required necessary to give controlling, com- ~ ~ but often important effective advice. municating, and in ~ to carry conviction human and public & with clients, col- relations. ~ leagues, and external .., ~. contacts. ~ c., Time span of discre- Time span of discre- Time span of discre- Independence and Independence of Independence and tion is minimal. tion is great with tion is great with freedom of action action is limited freedom of action (./) Since emphasis is on respect to client re- respect to client is limited only_ to only by the stated is limited only by ~ ~ professional growth lationships within relationships but the degree that the objectives of the the regulations of ~ ~ and learning, fre- policy limits but with respect to unit, program, or library system. the parent body. ~ quent review of with respect to projects and project project concerned ~ ~ Independence work with clients projects, is limited supervision is limited involves other .., and and/ or projects is to completion of the to duration of pro- jurisdictions. '-' <:o Freedom required. project. Judgments grams. Judgments """l of Independent projects required involve required involve Cit Action with scope defined some intangibles but intangibles present externally may be are usually made on at the grade 2 level assigned. the basis of knowl- but in addition the edge and experi- integration and co- ence. ordination of pro- grams arid job elements. Normally does not Normally limited to Normally limited to Responsible for Responsible for Represents .the uni- involve formal levels those members of those members of effectively presenting initiating contacts versity library system of external contact. the library, univer- the library, univer- the library to policy and for coordinat- to other chief li- sity, or community sity, or community advising levels of ing interlibrary co- brarians. Formal re- Level of Formal whose cooperation whose cooperation the university and operation, as well porting relationship External Contact is necessary to the is necessary to the of the community, as representing the is to the appropriat~ ;::.... . ~ ;::.... • ~ cL Q) ~ ... o.-o-£-as ... ~'"0 .... 8 ;§ ~~co;S.s ;::.... .._.as ... "' ~ ~bO§~ ~ s .:a .8 8 ~ .~ 2 ~ s g. e ~~ O,.$'Q3,.o ::l tl) u ~ :>.:.:::: Cii I ~ ~ 55 "' ... ..... Q) ~~ ·~ .8 ·E~ Q) ..... • ~~~ tn~;n~ Matrix Approach I 361 sional supervisors involved to ensure that the purpose of the questionnaire and the questions themselves were un- derstood. Members of the UWO Li- brary System professional staff were in- vited to meet individually with one of the consultants to discuss the filling out of the questionnaire, and a number of staff did so. In addition to a summary job descrip- tion, including percentages of time spent on principal duties, the following are selected examples from the twenty- three item questionnaire: Indicate the kind and level of edu- cational requirements and , professional experience necessary to perform ef- fectively in your position. Why? To what extent does your position require regular contacts with library personnel outside of your own depart- ment or unit? With university person- nel outside of the university library? Illustrate the types of problems which you are expected to deal with on your · own and those which you would refer to a higher authority for solution. Give an illustration of the type of long-range policy planning normally associated with your position. Are the duties performed by the people you supervise primarily of a routine or of a discretionary nature? Give examples. After all of the questionnaires had been filled out and approved by the su- pervisors, each consultant took a group of responses and analyzed them in terms of a grid profile sheet, composed of the four "general" and five ~~ranking" fac- tors as shown in Table 1. The consul- tants then exchanged groups and con- ducted independent reevaluations. For most ranking factors there were at least three questions which dealt directly with each; but one factor, ~'supervision of personnel," was assessed in only two questions, and one factor, ~~degree of expertise,,, had nine questions specifical- ly related to it. 362 I College & Research Libraries • September 1975 The use of the responses to several questions as <