College and Research Libraries MARJORIE JOHNSON Performance Appraisal of Librarians- A Survey pERFORMANCE APPRAISAL TOUCHES ON one of the most emotionally charged ac~ tivities in business life-the assessment of a person's contribution and ability. This is true whether the business is that of running a university library or oper- ating a commercial organization. In the spring of 1971, a study was un- dertaken by the author to compile in- formation on performance appraisals of libr.arians in college and university libraries. The objectives of the study were to : ( 1) determine approaches used in ap- praising librarians, together with the apparent success, or lack of success; of these approaches; ( 2) compare the results with per- formance appraisal concepts ex- pressed in recent literature; ( 3) draw conclusions :which could be helpful to those responsible for appraisal of professional person- nel in libraries. All university libraries in the United States and Canada having more than fif-' teen librarians on the staff were · sur- veyed. Out of this total of 185 libraries, responses were received from 138. The majority, almost 95 percent, indicated that some form of appraisal was used. Ttte methods ranged from .a casual ob- servation of staff members by the direc- tor with no written record made, to lengthy interviews with staff members Ms. Johnson is cataloger at the Memphis Public Library and Information Center, Memphis, Tennessee. discussing their performance based on the . results of a form. The frequency of appraisal varied from several times during the first year, to annually until tenure was achieved. The type of form used showed the most divergence: of the seventy-four sample forms received, only one fqrm was duplicated. The forms ranged from mediocre to excel- lent in each of the three categories irito which they were separated-rating-sheet types, forms identical to those used to evaluate teaching faculty performance, and forms designed to evaluate the spe- cial competencies of librarians. There were samples of thirty-five rating-sheet types, seventeen of the kind used to evaluate teaching faculty performance, and twenty-two: forms especially de- signed to evaluate librarians. Not all li- braries stating that they conducted ap- praisals and used .a form included one with their questionnaire, and some li- braries used two or three forms of vari- ous types. A consensus of ·experts in the field suggests that no . organization . has a choice of whether it should appraise its personnel and their performance.· Every time a promotion is made or a salary in- crease is given, an appraisal of some kind takes place. The question is not then whether there · should be an ap- praisal, but rather it is· a question . of method. The . fact that performance appraisal is increasingly used is borne) out in the literature by many writers. Thompson and Dalton .point out that the signals a person receives about the assessment his I 359 360 I College & Research Libraries • September 1972 supervisor is making about his contribu- tion and ability has a strong impact on his self-esteem and on his subsequent performance. 1 Sloan and Johnson stress that the scope of performance appraisal is growing. 2 Its traditional focus has been enlarged to include not only the individual's on-the-job behavior but also his functioning as an integral part of the organizational system. Applying the systems approach to personnel appraisal, the individual stands as an integral part of a unit or department and his per- formance should also be evaluated by the degree to which he accomplishes spe- cific results that contribute to depart- mental and organizational achievement. 3 Performance is evaluated by the degree to which a person achieves explicit ob- jectives in terms of measurable per- formance or results. Kellogg points out that the appraisal system has also become .a basis for coun- seling and coaching subordinates, if used properly. 4 Wallace lists the three functions of performance appraisal as an individual growth tool, to distribute rewards among a group, and to £le as a personnel reference history. 5 He con- cludes that these are distinct, different needs, and no one appraisal system de- signed for one need should be used for another. Performance appraisal in libraries has, in the past, been closely tied to pe- riodic library planning activities; it has been used primarily as a written justifi- cation for salary action. The personnel director of a large university library in Pennsylvania stated that, in the past, li- brarians were immune from appraisals of any kind; however, with severe budg- et cutbacks, she felt that some kind of appraisal system would have to be initi- ated. Others who were using rating scale appraisals, indicated that they were try- ing a new technique. The new . method was often a form similar to that used to evaluate teaching faculty, particular- ly if the librarians had just recently achieved faculty status. Other library directors, possibly more comfortable about the security of their librarians' status, were designing forms which spe- cifically measured the competencies of librarians, or were experimenting with management-by-objectives, or mutual goal-setting techniques. Of the respon- dents indicating that they used apprais- als, 43 stated that they were considering changing their present technique; an ad- ditional 22 respondents indicated that their appraisal method was either in its first or second year of use. FREQUENCY OF APPRAISAL The majority of those libraries sur- veyed which conduct appraisals per- · formed these evaluations periodically, 80 percent on an annual basis. Continu- ous evaluation of performance is an es- sential part of supervision states Harold Mayfield, and many others advocate a relationship between supervisor and sub- ordinate where performance is discussed as a part of day-to-day operations. 6 Black suggests that the appraisal inter- view is primarily a teaching device. 7 The objective is to help the employee help himself by persuading him to recognize and co~rect his deficiencies. The follow- up of the appraisal interview is the training program. Talking to a subordi- nate about how he does his job is a vital part of his training, and successful managers recognized for their talent to develop people use every opportunity to give their employees individual coaching. This includes regular critiques of their performance, not just a once- a-year, get-it-over-with-for-another-year approach. Daily, on-the-job contact with subordinates is a natural time to point out the specifics of job performance. Praise or criticism in such circumstances does not take on exaggerated impor- tance. The employee more clearly un- derstands the deficiency when it is point- ed out to him immediately than if it is discussed six months later in a formal Performance Appraisal of Librarians I 361 interview. If he has done something well and receives instant recognition, he is far more pleased than if his merits are recited at the .annual review. Practitioners as well as theorists seem to agree that the greatest disadvantage of the single periodic appraisal is in ex- pecting it to accomplish too many ob- jectives.8 The salary action issue so dom- inates the .annual comprehensive per- formance appraisal interview that nei- ther party is in the right frame of mind to discuss plans for improved perform- ance. Some supervisors indicate that set, periodic appraisals tend to cause them to save up material in order to have enough to discuss. Although the majority of libraries surveyed conducted appraisals on an an- nual basis, several specified modifica- tions, such as more frequent appraisals where past deficiences were recorded or elimination of appraisals after tenure was .achieved. The inference under lying this response seemed to be that some di- rectors felt that appraisals were to be avoided. The confident reply of one di- rector fairly well sums up this attitude: "It has been our experience that, by ju- dicious selection of applicants and by careful screening during the probation- ary period, we have a staff of such qual- ity that appraisals are not needed for those who attain tenure." It is easy to find administrators who feel that they know their people so well that apprais- als are unnecessary; however, it is much more difficult to find staff members who would not welcome a chance for a bet- ter understanding of the over-all goals of the library and a reassessment of the contributions they are making. The at- titude of some administrators seems to be that appraisals are intended only to point out shortcomings and that when you have a tenured staff, appraisals are no longer necessary. This interpretation completely eliminates the functions of counseling, praise for jobs well done, a measure of how to distribute rewards, an individual growth tool, or .an oppor- tunity to talk with individual staff mem- bers to compare their self-goals with the over-all objectives of the library. For the majority of libraries conduct- ing appraisals annually, fall, winter, and spring were specified with equal frequency. Significantly, the .appraisal time was almost always selected to tie in with the annual salary review, budget preparation, or determination of reten- tion, tenure, or rank for the coming year. Thompson and Dalton, as noted earlier, strongly advocate resisting the temptation to devise one grand per- formance appraisal system to serve all management needs. Culbreth points out that · the appropriate time for evaluating an employee-in terms of his individual development-rarely, if ever, coincides with a rigid timetable; he recommends holding an appraisal whenever one is needed to maintain a good relationship between the supervisor and subordinate or to advance employee development.9 UsE OF FoRMS FOR APPRAISALS Of those university libraries which do have appraisals, 78 out of 130 used some kind of form or set of guidelines to perform the appraisal. Those not using forms stated that they held in- formal discussions, or prepared written summary reviews (with no apparent cri- teria followed) . The use of a well-designed appraisal form is no guara~tee of an effective ap- praisal method. An administrator who possesses the proper skills could conduct satisfactory appraisals with or without a form. The information gathered sug- gests, however, that the use of a written guide assured a greater degree of suc- cess, even in the hands of unskilled evaluators, than would an evaluation system with no guidelines. PRos AND CoNs oF V Aruous APPRAISAL TECHNIQUES Cangemi suggests that when using a form based on .a rating scale, the evalu- 362 I C allege & Research Libraries • September 1972 ator should be careful to avoid four common errors: personal bias, central tendency, halo effect, and logical error. Personal bias errors result when an eval- uator rates all individuals consistently too high or too low. The error of cen- tral tendency signifies that the evaluator seldom ever gives ratings at the extremes of the scale. The halo effect is usually found in those operating under the presence of strong personal biases. In this situation, the rater judges a person to be the same or nearly the same in all characteristics. Logical error results from .a misunderstanding of the charac- teristic to be 'measured, and happens more often when no definitions of the characteristics are given. 10 Another reason for dissatisfaction with forms is the difficulty of fitting an individual's performance into pet phrases such as "Knowledge of Biblio- graphic Resources-Excellent, Good, Fair, Poor (check one)." A measure such as this may be irrelevant to the work performed, yet an evaluation must be given; therefore, a decision is made on probable performance. 11 Once the evaluation of that characteristic has been transformed into a rating such as "Good," it is considered equal to other ratings with more reliable bases of mea- surement. Even the phrase "Quantity of Work" can be difficult to evaluate fair- ly. Books cataloged can be counted, but there is extreme variation in effort re- quired. Certain kinds of reference work defy quantification, yet many reference librarians are regularly evaluated on the quantity of work produced. A frequently stated advantage of the checklist is that it promotes consistency between managers; but different inter- pretations of the levels of rating, and different opinions of what proportions of the employees should be given above- average ratings often negates this appar- ent advantage. In some of the traditional methods of appraisal, the subordinate sometimes reacts adversely to criticism. To accept criticism-even if it is deserved-from a comparative stranger is a bitter pill. One c.an imagine the rated person think- ing "Let he who is without fault cast the first criticism." In order to avoid the role of "judge," other more objective techniques have been designed. One is the management-by-objectives approach popularized by Peter Drucker .12 This method centers on the assessment of performance by contrasting it with goals set mutually by the supervisor and the subordinate. The individual partici- pates in the goal setting, makes a com- mitment, and then evaluates his own achievements. McGregor suggests that this type of management encourages the professional to bring his talent, train- ing, and creativity to his job. If a man- ager uses this approach, he listens, ad- vises, guides, and encourages the subor- dinate to develop his own potential. 13 One important advantage of the self- evaluation or participatory approach is that it can often provide the supervisor with useful insights to the person being appraised. Macoy advocates appraisal in- terviews as an opportunity to take sig- nificant steps forward in understanding subordinates, and also as an opportunity for meaningful self-examination by the employee. 14 Employee development is, in the last analysis, self-development. With rating scales, there is .an implied requirement that all qualities must be evaluated. In many cases, this forces judgments on characteristics which have no relationship to job performance. The most frequently used qualities on rating scales are: initiative, leadership, quantity of work, dependability, atti- tude toward others, cooperativeness, ac- curacy, judgment, loyalty, organization of work, and quality of work. Several of these traits might be easy to measure and evaluate in a cataloger but impos- sible to consider fairly in .a reference librarian. The quality of leadership could be assessed in a person with super- Performance Appraisal of Librarians I 363 visory responsibilities but not in a per- son who had never had an opportunity to lead others. The various grades, or levels, of each item on the rating sheet present another opportunity for inconsistency. If five levels are allowed, such as 1, 2, 3, 4, and 5, with 5 being the top of the scale, how meaningful is it to give someone a 4 in accuracy? How many errors would one need to make in order to be rated at a 4 instead of a 5? Unless ridiculously specific guidelines are ~et up, the entire process becomes little more than an ex- . ercise. The raters themselves differ in the way they evaluate people. A very conservative person may rarely give a 5 to anyone, while another supervisor may rate everyone at the top. About the only real value of the rating scale kind of appraisal is that it is done periodically and provides an opportunity to think about an employee's progress and dis- cuss with him more than merely why he was given a 3 in loyalty this time. One strange requirement on many ap- praisal forms is that of the employee's signature. As Maier points out, this re- quirement is inconsistent with the goals of an appraisal except where a warning is given. 15 In this case, the purpose of the signature is to preclude the employ- ee saying at a l~ter time that he was nev- er told his work was unsatisfactory. It is somewhat incongruous to come to the clos~ of an otherwise pleasant evalua- tion interview and be reminded, "Now we must have your signature on the form." One other possible explanation for the signature requirement is that it proves to the administration that the su- pervisor actually held an evaluation in- terview with the employee. Culbreth states that there is no better way to ensure a department's maximum efficiency than to give its personnel prop- er, continuing evaluation.16 With this thought comes the question of person- nel v~. organization appraisal. It has been suggested that a person is respon- sible for and can control only certain aspects of his performance. 17 Perhaps we should be evaluating performance problems rather than people. Does a supervisor, as he appraises each employ- ee, ask himself the question, "Did I, as the supervisor, in any way contribute to a performance problem of the subordi- nate?" Haynes suggests that an employ- ee's effectiveness depends on four things: the employee, the job, the super- visor, and the work environment. 18 Therefore, a performance discussion aimed at increased effectiveness should not be limited to just the employee. LIBRARIAN APPRAISAL IN THE AcADEMic ENVIRONMENT "How many professionals in the li- brary have academic rank? Is your ap- praisal method for librarians the same as for teaching faculty?" An analysis of · the questionnaires showed that many librarians possess a status somewhat equivalent to faculty, whether or not they had academic rank. An analysis of those responses in which an appraisal form of some kind was used showed that 38 out of 78 used a form identical to that used for teaching faculty, or one similar which had been designed to evaluate the special compe- tencies of librarians. Performance rat- ing, or rating scale forms, were used in 35 libraries. A few libraries relied main- ly on summary-type or self-evaluations. Several used more than one technique. Eldred Smith has discussed the merits of, and prospects for, academic status for librarians. 19 Most librarians desire academic status but a conflict can occur when the librarian is rated on the regu- lar faculty appraisal form in which the criteria are related more to teaching or research faculty activities. A librarian who catalogs eight hours a day may suf- fer when measured against such criteria. The logical conclusion is that the ap- praisal must be for performance in a specific position, while still allowing 364 I College & Research Libraries • September 1972 credit for scholarly achievement beyond the call of duty. The differences and similarities of faculty and librarian re- sponsibilities must be considered specifi- cally for each person and position. WHO APPRAISES? In .almost all cases where appraisals were conducted, the person responsible for the appraisal gathered opinions from others or reviewed the results with others. If the immediate supervisor was not responsible for the appraisal he was usually consulted. The findings are sum- marized as follows: percent The director or administrator ei- ther prepares or participates in preparing the .appraisal 7 4 One or more committees, coun- cils, or peer groups are in- volved in the appraisal process 29 Personnel officer is involved in the process 6 Only the director is involved in the appraisal 4 Only the immediate supervisor is involved 6 Only a committee, council, or peer group is involved 3 Some libraries used more than one method, resulting in a total of more than 100. When appraisal is performed by someone other than the immediate su- pervisor, .and the results are not dis- closed to the employee, the main ad- vantages of the process are lost. It is es- sential for people to know if their work is satisfactory and if not, how and why it fails to meet the mark. Even an out- standing employee should be informed of specific instances both of superior performance and of areas where he might improve. Grandy cites .an example of an out- standing employee completing her first year in an organization. At her apprais- al interview her supervisor mentioned no particular areas of excellence but merely told her how pleased he was with her work. Following the evaluation, her work became quite erratic. Unless the supervisor is able to point out exactly what the employee is doing well, the employee will devise his own tests to dis- cover which behavior pattern will gain approval and which will not. 20 Although only a few university li- braries in this survey assigned primary responsibility for appraisal to a peer group or committee, many respondents indicated participation by such a group. The apparent objective in the use of a peer group is to attempt to increase the reliability of the appraisal by drawing upon several opinions. French states that when the ratings are made by a group or committee, .actual knowledge of the person's performance by each member is vital. Without objective data, a committee may be simply pooling their collective ignorance. 21 In a large library, all members of a committee usually do not have firsthand knowl- edge of everyone's work. French further indicates that a major problem in the use of peer, or subordinate, ratings is the potential danger that the ratings may be unintentionally made on criteria which are . useful to the rater but not necessarily to the enterprise. For exam- ple, a reference librarian participating in the peer group might give excessive weight to incidents involving reference room activities. French concludes, how- ever, that ratings by peers do show promise of being useful in promotional decisions. Hollander found peer ratings to be high in reliability in a study of officer candidates where leadership po- tential was an important factor. 22 TYPICAL RATINGS "On the basis of your experience, what percentage of your personnel get ratings that are in the following cate- gories-Above Average, Average, Below Average?" Performance Appraisal of Librarians I 365 The majority of respondents indicat- ed that they gave above-average ratings to 70 percent of their staff. Two gave this rating to 95 percent of their staff. Below-average ratings were given to only about 5 percent of the personnel. The report of the Conferenc~ on Per- formance Appraisal and Review held in 1957 at Michigan, sponsored by the Foundation for Research and Human Behavior, pointed out that appraisal rat- ings are usually bunched at the good end of the scale, and that rater differ- ences are sometimes more marked than differences between those rated. One rea- son is that the supervisor has to justify his evaluation to the subordinate, and few supervisors possess the desired in- sight and tact to tell a subordinate how to improve. The easiest way out is to limit the evaluation discussion to the employee's good points. Often the ad- ministrator rationalizes his all-above-av- erage ratings by stating that the library does not hire or retain average or below- average people. Kirchner and Reisberg conclude that better supervisors tend to show more spread and variation in their ratings, and also tend to give more regard for independent action and creative think- ing on the part of their subordinates. More effective supervisors also tend to check a greater number of least strengths, whereas the less effective su- pervisors tend to rate subordinates much more alike. 23 Thompson and Dalton found one manager who expressed the belief that low performers should be given extremely low scores-possibly even lower than they deserved-to en- courage them to leave. It was found however, that this approach was mor~ likely to influence them to stay to try to gain security. 24 The supervisor should not deliberately adjust his ratings to try to achieve a particular purpose but should be able to defend his evaluation with specific instances of good and poor performance. EFFECTIVENESS OF APPRAISALS- SURVEY FINDINGS The questionnaire asked several ques- tions about the effectiveness of the ap- praisal technique in the opinion of the director (or person responding to the survey questionnaire). The questions in- cluded: How effective do you think your appraisal method is in .. . pointing out employee shortcomings? giving employees encouragement to improve performance? helping make reliable judgments upon which to base salary increases? giving the person .an opportunity to discuss problems? helping eliminate bias and favoritism in ratings? Regardless of the appraisal system used, the replies indicated that 26 percent felt that their method was "very effective," 64 percent thought their technique of appraisal was "somewhat effective," 8 percent considered it "ineffective," and 2 percent rated theirs as "completely in- effective." Bias cannot be eliminated by the choice of a particular type of appraisal system according to a majority of writ- ers. Rather the rater must become aware of bias tendencies so that he can at- tempt to · avoid them in his ratings. There is more likely to be bias in an ap- praisal system where the supervisor does not have to confront the person ap- praised, since the discussion and ques- tioning by the person appraised tends to bring out any bias in the ratings. Failure of appraisal programs is due to inept techniques, ineffectual commu- nication between rater and ratee, and the role conflict experienced by the . rater·. The type of form used has much less influence on the final results. CoNCLUSION Whether formal or casual, appraisals are conducted in every organization. There is seldom any difficulty in identi- 366 I College & Research Libraries • September 1972 fying extreme patterns of behavior- the very good or the very poor-and not much time or money needs to be spent on the exercise. Appraisal systems are more essential for those who fall be- tween these extremes. Appraisal forms are useful at least for assuring a periodic review and as- sessment of the employee's contribu- tions. Rating scale forms have many shortcomings, and it is difficult to judge a librarian's work performance by ap- plying the same criteria used to evaluate teaching faculty. The best yardstick for measuring librarians is a form specifi- cally designed to consider the special competencies of librarians. Ideally, two forms should be used: one to be com- pleted by the individual, permitting him to list activity in research or publica- tion, professional or personal develop- ment, and participation in professional, .academic, or community affairs; the sec- ond form, to be completed by the li- brarian's supervisor, requiring essay-like summaries of the person's on-the-job performance, personal traits, and atti- tudes. The supervisor has to accept the re- sponsibility to judge the performance of other people. Often this responsibili- ty is hesitantly taken because he feels uncomfortable in his role as a judge. It is this psychological barrier which un- derlies the failure of most evaluation systems. When a specific system fails, a common solution is to adopt a different evaluation instrument, but the underly- ing fault is with the people making the evaluations. REFERENCES 1. Paul H. Thompson and Gene W. Dalton, "Performance Appraisal: Managers Be- ware," Harvard Business Review (Jan.- Feb. 1970), p. 150. 2. Stanley Sloan and Alton C. Johnson, "New Context of Personnel Appraisal," Harvard Business Review (Nov.-Dec. 1968), p. 14- 16. 3. Robert Gagne, ed., Psychological Principles in System Development (New York: Holt, Rinehart & Winston, 1962). 4. Marion Kellogg, What to Do About Per- formance Appraisal (New York: .American Management Association, 1965), Chapters 3, 4, and 5. 5. William H. Wallace, "Performance Ap- praisal by Non-Self Directed Personnel," Personnel Journal (July 1971). 6. Harold Mayfield, "In Defense of Perform- ance Appraisal," Harvard Business Review (March-April1960). 7. James M. Black, How to Get Results from Interviewing (New York: McGraw-Hill Book Co., 1969), Chapter 5. 8. Herbert M. Meyer, Emanuel Kay, and John R. P. French, "Split Roles in Performance Appraisal," Harvard Business Review (Jan.-Feb. 1965), p. 124; Norman R. F. Maier, "Three Types of Appraisal Inter- view,'' Personnel Journal (March-April 1958), p. 28-29; Rensis Likert, "Motiva- tional Approach to Management Develop- ment," Harvard Business Review (July- Aug. 19·59), p. 75. 9. George Culbreth, "Appraisals That Lead to Better Performance," Supervisory Man- agement (March 1971), p. 8. 10. Joseph P. Cangemi, "Personnel Evaluation Rating Scale," Personnel Journal (Aug. 1970), p. 665. 11. David Peele, "Performance Ratings and Li- brarians Rights," American Libraries (June 1970)' p. 595. 12. Peter F. Drucker, The Practice of Manage- ment (New York: Harper & Row, 1954). 13. Douglas McGregor, "An Uneasy Look at Performance Appraisal,'' Harvard Business Review (May-June 1957), p. 89. 14. Wells A. Macoy, "Conducting a Perform- ance Appraisal Interview," Supervisory Management (Feb. 1970), p. 13. 15. Norman R. F. Maier, "Three Types ... .'' 16 .. George Culbreth, "Appraisals .... " 17. Philip R. Kelly, "Reappraisal of Apprais- als," Harvard Business Review (May-June 1958), p. 63. Also, Harold Mayfield, "In Defense .... " 18. Marion E. Haynes, "Improving Perform- ance Through Employee Discussions," Per- sonnel Journal (Feb. 1970), p. 140. 19. Eldred Smith, "'Academic Status for Col- lege and University Librarians-Problems and Prospects," CRL (Jan. 1970), p. 7-13. 20. C. C. Grandy, "Do Your Performance Ap- Performance Appraisal of Librarians I 367 praisals Backfire?" Supervisory Manage- ment (Nov. 1971), p. 3. 21. Wendell French, The Personnel Manage- ment Process (Boston: Houghton Mifflin Company, 1970), p. 300 and 308. 22. E. P. Hollander, "The Reliability of Peer Nominations Under Various Conditions of Administration," Journal of Applied Psy- chology (April 1957). 23. Wayne K. Kirchner and Donald J. Reis- berg, "Differences Between Better and Less-Effective Supervisors in Appraisal of Subordinates," Personnel Psychology (Nov. 1962), p. 297. 24. ~aul H. Thompson .and G~?e W. Dalton, Performance Appraisal. ...