Critical requirements of the University librarians job : methodological considerations in collecting incidents and weighing requirements.

Statement of the Problem The object of this study is to examine the effect of methodological variations in the Critical Incident Technique as it applies to job analysis. Three methodological areas were investigated in this study: I The questions asked the observers. It has been common practice in past studies to ask the observers to describe incidents in which a person acted in such a manner as to be particularly effective or ineffective with regard to performing the activity. It is possible that by asking additional kinds of questions which establish different criteria for reporting or ignoring incidents different kinds of behaviors would be reported. II The observers' perceptions of the aims of the activity. Flanagan (1954) suggests that the observers, in order to judge whether a person's behavior is effective or ineffective with regard to a particular activity must be aware of the general aims of the activity as laid down by the administrators of the activity. Few past critical incident studies have checked the qualifications of the observers used with regard to their awareness of the aims of the activity. No research has been carried out to determine the effect of observers perceptions of aims, which are in disagreement with the accepted aims, on the incidents they report. III The relative importance of the critical requirements. It may be possible to determine the relative importance of each critical requirement. Frequency of behaviors reported has been used as an indicator of the relative importance of critical requirements. Ratings by supervisors of the importance of each requirement, have also been used in an effort to establish their relative importance. Neither of these methods was proven to be valid. To date no effective, valid method has been discovered for determining the relative importance of the critical requirements. This study investigated these three areas in an attempt to test the following propositions: 1. Incidents collected from a "behavior pattern" questions will contain different kinds of behavior from those contained in incidents elicited from "isolated behavior" questions. 2. Observers who are in disagreement with the accepted aims of the activity under study will report different kinds of behavior from those reported by observers who are in agreement with these aims. 3. The relative importance of the critical requirements can be determined by using the observers' ratings of the criticalness of the behaviors they report. The Method Individual interviews were held with 126 observers (52 library staff members, 44 students and 30 faculty members) at the University of British Columbia. To elicit critical incidents the observers were asked two types of questions. "Isolated behavior" questions, asked for descriptions of behavior that was particularly effective or ineffective librarian performance. "Behavior pattern" questions, asked for descriptions of behavior that, while not particularly effective or ineffective in a single instance, did become significantly critical when repeated over a time. Critical requirement categories were generated and the types of behaviors elicited by each type of question and their relative contribution to each critical requirement were investigated. Each observer was asked to state his perception of the general aims of the librarians' job. This statement was compared with an "official" statement of aims. The behaviors reported by observers who were in disagreement with the "official" general aims were compared with the behaviors reported by observers who stated aims that were compatable with the "official" statement. Each observer was asked to rate the effectiveness or ineffectiveness of the behaviors he reported on a four-point rating scale. The average rating of the behaviors in each critical requirement category was calculated. These average ratings were used to rank order the critical requirements in terms of relative importance. This rank ordering was compared with the rank orderings obtained by calculating the frequency of behaviors in each critical requirement category and by having ten senior librarians rate the importance of each critical requirement category. Conclusions Thirty-seven critical requirement categories were generated by behaviors collected by both types of questions. The addition of the "behavior pattern" question resulted in a greater number of negative behaviors being reported and contributed greatly to the generation of 4 critical requirement categories. These findings support the hypothesis that additional kinds of behaviors can be elicited by varying the questions asked thus resulting in a more complete list of critical requirements. No observers were found who disagreed with the "official" statement of aims so no conclusions could be reached on the effect of such disagreement on behaviors reported. Subtle differences in emphasis within the framework of the "official" statement of aims did not influence behaviors reported to a significant extent. A rank ordering of critical requirements by averaging the observers' ratings of the behaviors they reported was achieved. The results did not prove that this rank ordering was a valid indicator of the relative importance of the critical requirements.Arts, Faculty ofPsychology, Department ofGraduat