stamat.p65 322 College & Research Libraries July 1998 Effects of Library Instruction on University Students’ Satisfaction with the Library: A Longitudinal Study Anthony Stamatoplos and Robert Mackoy Consideration of satisfaction should be an important part of the evalua­ tion of library services. Satisfaction depends, to some extent, on patron expectations of services. This study evaluated changes in student ex­ pectations following library instruction and how they were related to over­ all, long-term satisfaction with the library. Satisfaction appeared to be related to student perceptions of information accessibility, staff compe­ tence and helpfulness, computer usefulness and ease of use, and skill level for using libraries. The study suggests that libraries may be well served by measuring patron satisfaction and learning what variables drive satisfaction at particular libraries. ssessment and service quality are two of the dominant themes in library research today. Over the past three years, College & Research Libraries has published numer­ ous articles on each of these themes. Yet no research has appeared to date in which empirical evidence is used to address both themes simultaneously. The study re­ ported here represents a preliminary at­ tempt to determine the impact of a library instruction program on patrons’ overall, long-term satisfaction with a large, urban university library. The ultimate goal of this research effort is to define the “driv­ ers” of patron satisfaction at the library. As such, the goal of this study was not to measure objective changes in skill or knowledge level resulting from library in­ struction but, rather, to evaluate percep­ tual changes following instruction and then to track whether these changes are related to patron satisfaction levels with library service. Although several studies have mea­ sured relevant variables pre- and post- instruction, none has evaluated the effect of post-instruction perceptions on longer- term patron satisfaction. Ought the effect of library instruction on patron satisfac­ tion be judged immediately following in­ struction or several months later? Each is valuable, but for different reasons. Short- term evaluations provide useful feedback for the instructor, assessment of the teach­ ing methods, effectiveness of meeting spe­ cific library instruction objectives, and so on. Long-term evaluations can determine lasting effects of instruction, retention of information, and the effects of instruction on patron perception of, and satisfaction with, overall library service. Anthony Stamatoplos is an Assistant Librarian at Indiana University/Purdue University-Indianapolis; e-mail: astamato@iupui.edu. Robert Mackoy is an Assistant Professor of Business at Butler University; e- mail: mackoy@thomas.butler.edu. 322 mailto:mackoy@thomas.butler.edu mailto:astamato@iupui.edu Effects of Library Instruction 323 Why should library administrators care about patron satisfaction? Some think patron satisfaction is of secondary impor­ tance—that the purpose of a library is to provide information services and that evaluation of the library performance ought to be judged from that perspective alone. However, others recognize that fail­ ing to satisfy the patron is failing to serve the patron. This is not about just making patrons “feel good.” There is a basic rea­ son for focusing on patron satisfaction: Who is the best judge of whether a pa­ tron has been served well or poorly? Pro­ fessional librarians may know whether they have provided accurate, timely in­ formation in response to a request. Is that enough? Clearly, accurate and timely in­ formation is a minimum requirement, but unless the patron is satisfied, in general, the service could have been better. Evaluating Satisfaction Researchers in the fields of consumer be­ havior and marketing have focused on understanding satisfaction only since the late 1970s. Satisfaction is considered to be an important topic because it is thought to be related to organizational objectives of repeat patronage and positive word of mouth. In the 1980s, the U.S. government sought to encourage and reward “qual­ ity” in U.S. organizations, most visibly through creation of the Malcolm Baldrige Award for Quality. The creators of this (and other) quality awards recognized that satisfaction plays a major role in defin­ ing what quality is. For example, satisfac­ tion is the single most heavily weighted component on an organization’s final score in evaluating Baldrige Award ap­ plications. After the importance of satisfaction was recognized, researchers began trying to define it, measure it, and determine what causes it.1 Although numerous defi­ nitions of the term exist, most research­ ers would agree that satisfaction is a reac­ tion to a patron’s subjective assessment of the degree to which the organization’s performance met or exceeded some stan­ dard internalized by the patron. This defi­ nition provides hints about how satisfac­ tion should and should not be measured. First, the measure needs to be subjective: Only the patron can judge whether he or she is satisfied. Second, satisfaction should be measured on a continuum rather than on a dichotomous scale: Most people experience satisfaction in greater or lesser degrees, depending on numer­ ous factors. The question of what causes satisfac- Professional librarians may know whether they have provided accurate, timely information in response to a request. Is that enough? tion is still being investigated.2 One thing that is clear from the above definition is that satisfaction depends to a certain ex­ tent on the standard internalized by the patron and on the perceived level of per­ formance of the organization. The stan­ dard that has been most widely tested and used has been predictive expectations— that is, the level of performance the sub­ ject actually expects to encounter.3 Evalu­ ation of library services, including library instruction, has not conceptualized or measured satisfaction in this manner. Evaluating Library Instruction Various studies point to the need for li­ brary instruction for college students. A Johns Hopkins University study com­ pared baseline measures of freshman li­ brary skills to upper-class students’ skills, concluding that exposure to a library does not necessarily improve those skills nor do students learn good library skills on their own.4 This study also found stu­ dents’ self-assessment of library skills to be “quite accurate.” Most studies agreed that some method of formal instruction is important. Evaluation usually focused on the efficacy of instruction or the supe­ 324 College & Research Libraries July 1998 riority of one method over others. Evalu­ ation criteria have been relatively nar­ rowly focused; some were objective, some were subjective, and some were both. Al­ though some looked at student satisfac­ tion with the instruction sessions or pro­ grams, none focused on the relationship between instruction and the library ser­ vice as a whole. Evaluation Using Objective Measures Numerous studies have used objective measures of student knowledge and skill performance to evaluate library instruc­ tion. These studies focused on the degree to which students’ testable knowledge increased or performance improved fol­ lowing instruction. The usual objective of such studies was to assess whether the instruction subject matter was being learned by the students (i.e., whether the short-term goals of instruction were be­ ing met). Numerous variables and meth­ ods were used. One method was to ex­ amine the effects of instruction on patterns of library use. In their study of undergraduates at the University of Illi­ nois, David N. King and John C. Ory found that instructed students used a wider variety of sources, made greater use of catalogs, and showed more use of vari­ ous libraries and services.5 Evaluation of the library instruction program at the Ohio State University showed that in­ struction produced significant improve­ ments in students’ tested library knowl­ edge and use.6 In a common approach to evaluating library instruction, Patricia Daragan and Gwendolyn Stevens used pre- and posttest measures to assess the library knowledge and skills of cadets at the U.S. Coast Guard Academy, and con­ cluded that instruction increased stu­ dents’ levels of information literacy and narrowed the range of information lit­ eracy among incoming cadets.7 Some studies have compared different instructional methods and objectively evaluated their relative effectiveness. At Southeastern Oklahoma State University, four groups of students each received a different method of CD-ROM instruction, then used PsycLIT to complete test assign­ ments that measured searching skills.8 Patricia F. Vander Meer and Galen E. Rike studied the effectiveness of a new multi­ media self-instructional tutorial com­ pared with the traditional method (a workbook and librarian introduction) and found higher posttest scores for both groups, but no significant differences in effectiveness.9 Although some looked at student satisfaction with the instruction sessions or programs, none focused on the relationship between instruc­ tion and the library service as a whole. Some studies combined objective ap­ proaches. For example, Trudi E. Jacobson and Janice G. Newkirk used question­ naires and examined printouts of search strategies to measure the impact of librar­ ian instruction and assistance on students’ CD-ROM search proficiency. 10 They found some evidence that instruction or assistance had a positive effect on CD­ ROM searching skill level, and a major­ ity of students also reported that their search results were “valuable,” which the authors referred to as “high satisfaction level.” Evaluation Using Subjective Measures Effects of instruction on measurable skills or knowledge has not been the only fo­ cus of studies; many studies also have evaluated subjective factors such as stu­ dents’ comfort level in the library or con­ fidence in their library skills. In addition, several studies addressed the degree to which student attitudes toward, and per­ ceptions of, the library have changed fol­ lowing library instr uction and/or whether the students were satisfied with instruction. The usual objective of such studies was to determine patron short- term response specifically to the instruc­ http:proficiency.10 Effects of Library Instruction 325 tion. The focus on subjective evaluations of library instruction presumably is the result of a growing recognition among librarians that “evaluation of library ser­ vice quality is based on customer percep­ tion.”11 Jacobson and Newkirk acknowl­ edged that whether or not instruction produces increases in skills, “if students learn little but become comfortable with CD-ROM technology or with the library in general, our work is clearly purpose­ ful.”12 A study at Southern Illinois University evaluated student “appreciation” of (i.e., “satisfaction” with) a full-term, one-credit course of bibliographic instruction, sug­ gesting that the course produced a higher comfort level and confidence in library skills among upper-class students.13 Penelope Pearson and Virginia Tiefel used subjective questions to measure changes in attitudes toward and use of libraries among Ohio State University (OSU) stu­ dents, and an independent OSU poll con­ curred with the library’s conclusion that positive increases were related to the li­ brary instruction program.14 Subsequent evaluation confirmed that the OSU library instruction program improved student attitudes toward librarians and library services.15 Evaluation of a freshman li­ brary instruction program at North Park University suggested that students’ con­ fidence in their library skills and useful­ ness of the instruction increased with the program’s new focus on learning styles.16 Researchers at Illinois State University used student “user panels” to evaluate in­ tegrated library instruction and discov­ ered that the library instruction program may not be producing significant effects on students in the general education gate­ way course.17 Evaluation Combining Both Objective and Subjective Measures In addition to the Johns Hopkins study cited earlier,18 many studies have used a combination of objective and subjective measures to evaluate instruction. Al­ though some may see these two perspec­ tives and their associated methodologies as being at least somewhat at odds, it is important to recognize that each perspec­ tive can contribute to the evaluation and ultimate improvement of library instruc­ tion as well as other programs. King and Ory also looked at changes in student perceptions, which indicated that library instruction increased stu­ dents’ sense of confidence and compe­ tence in their use of the library.19 A study of library literacy of undergraduate stu­ dents at the University of Northern Colo­ rado (UNC) found that students’ confi­ dence levels increased with increased ex­ posure to the library and its services (i.e., from freshman to senior years), although tested overall proficiency did not increase significantly. However, only seven per­ cent of UNC students believed that bib­ liographic instruction was the primary influence in their development of library skills.20 At the University of Idaho, research­ ers found that library instruction can in­ crease student knowledge as well as con­ fidence level, regardless of teaching method or student personality type.21 Connie J. Ury and Terry L. King supple­ mented objective questionnaire data with data from focus groups that discovered significant agreement between student and faculty perceptions of library instruc­ tion activities.22 Vander Meer and Rike used attitudinal questions on the posttest portion of their questionnaires and found no significant differences between stu­ dent ratings of the traditional and new instruction methods.23 Service Quality versus Satisfaction in Library Evaluation Recent attempts to evaluate overall library performance from the patrons’ perspec­ tive have focused on measuring service quality rather than satisfaction. These re­ searchers used SERVQUAL, or a modi­ fied version of it, as an assessment instru­ ment.24 This approach has greatly http:methods.23 http:activities.22 http:skills.20 http:library.19 http:course.17 http:styles.16 http:services.15 http:program.14 http:students.13 326 College & Research Libraries July 1998 improved the quality of this type of re­ search by (1) basing assessment on a model that is associated with relevant theory and (2) using an instrument that has itself been assessed for validity and reliability. The authors believe that the use of SERVQUAL has contributed sig­ nificantly to the development of a theo­ retically and methodologically rigorous approach to library research. Despite these strengths associated with the use of SERVQUAL, numerous limita­ tions exist. First, SERVQUAL was devel­ oped within the context of five industries, which were neither libraries nor not-for­ profit organizations. It was developed from excellent research conducted in the following five service sectors: product repair and maintenance, retail banking, long-distance telephone service, securities brokering, and credit cards. Although SERVQUAL may be considered a good starting place for evaluating generic for- profit service organizations, without sig­ nificant additional research and revision it is unlikely that it would perform ex­ tremely well in evaluating the quality of library service. Second, SERVQUAL items adminis­ tered in actual service settings have con­ sistently been shown to yield a factor structure inconsistent with that on which the instrument is based. In other words, researchers have found that the five ser­ vice dimensions of SERVQUAL (reliabil­ ity, responsiveness, assurance, empathy, and tangibles) rarely are replicated in a particular service organization context, including library contexts. Third, the overall service quality rat­ ing derived from SERVQUAL appears to be not as highly correlated with general measures of service quality (from a user ’s perspective) as one should expect. In a recent review of literature reporting on SERVQUAL’s use in library evaluation, Danuta A. Nitecki found correlations be­ tween an overall quality measure and an overall SERVQUAL score in the range of .58 to .80.25 The average correlation was .71. In practical terms, this means that about 50 percent of the variance in users’ own perceptions of library service qual­ ity is accounted for by the SERVQUAL instrument. The remaining 50 percent of the variance is unexplained within the structure of SERVQUAL. If SERVQUAL really measured overall service quality from the users’ perspective, one would expect a much higher correlation. Methodology This study focuses on the relationship between a specific service, library instruc­ tion, and user satisfaction with the library. Specifically, recognizing the key role of expectations in satisfaction formation, the authors evaluated the change in student expectations following library instruction and then examined how changes in ex­ pectations were related to overall satis­ faction with the library. Students in several sections of an intro­ ductory English composition course at a large, urban university were used as sub­ jects for this study during the fall semes­ ter. The students were required to com­ plete a library research project by the end of the semester. Library instruction ses­ sions were held in each section of the course during the third week of class. The sessions used an approach that combined lecture and demonstration with a hands- on workshop and were geared to use ac­ tive learning to teach students basic skills and concepts of library research. Students were taught basic techniques of using in­ dexes and abstracts, searching CD-ROM databases, and using the library’s online catalog. They worked in teams to research assigned topics chosen by English compo­ sition faculty. Class discussion followed the hands-on portion of the sessions. The authors collected data at three times during the semester: during the week pre­ ceding the library instruction (T1), during the week immediately following the li­ brary instruction (T2), and during the week the research assignment was due, near the end of the semester (T3). Effects of Library Instruction 327 The questionnaires were developed and pretested extensively in the six months preceding fieldwork and subse­ quently revised. In addition to the formal pretesting process, the authors solicited feedback from other members of the li­ brary staff.26 The T1 questionnaire contained ques­ tions about the students’ expectations of the library collection, library staff, and computer- and print-based information; their confidence level regarding library use; and basic demographic information. The T2 questionnaire differed from the T1 questionnaire only in that demographic in­ formation was not solicited; that is, the T2 questionnaire focused on patron expecta­ tions and perceptions. The T3 question­ naire asked about actual experiences and perceptions of library performance; in ad­ dition, it asked students to indicate their overall level of satisfaction with the library. Completed questionnaires were collected and matched for seventy-six individuals; that is, seventy-six students completed each of the three questionnaires. Analysis focused on addressing the following three questions: 1. Which perceptions of performance are most/least associated with patron sat­ isfaction? 2. Which perceptions of performance are or are not congruent with T2 expecta­ tions? 3. Which T2 expectations appear to have been influenced by the library in­ struction? It is important to avoid falling into the trap of using only library instruction-re­ lated variables to assess user satisfaction. Studies that only consider program-spe­ cific variables as antecedents of satisfac­ tion nearly always will find relationships between the program and satisfaction be­ cause the analytical procedures only have the program-specific variables available to explain variance in satisfaction. By using program-specific variables and other rea­ sonable measures (as developed during pretesting), the impact of the program rela­ tive to other possible antecedents can be assessed. This information is more useful to administrators than is program-specific information presented in a vacuum. Findings Analysis of the data is based primarily on two established procedures: the paired t- test difference of means test and correla­ tion analysis. The former test is useful when measurements are taken of a par­ ticular group of individuals at differ­ ent points in time; it is appropriate for determining whether there is any dif­ ference (1) in the mean value of a single variable measured at two different times or (2) between two different vari­ a b l e s w h e n t h e v a r i a b l e s c a n b e matched to a specific individual. Cor­ relations simply reflect the strength of the linear relationship between two variables and range from -1.0 to 1.0. Positive corre­ lations indicate a positive relationship, negative correlations reflect an inverse re­ lationship, and a correlation of 0.0 indi­ cates the two variables are unrelated. Table 1 lists selected study variables in column 1. Column 2 contains the differ­ ence of means in expectations/assess­ ments before (T1) and after (T2) the library instruction. Positive numbers indicate the degree to which average values for that variable increased following instruction, and negative numbers indicate the degree to which average values decreased fol­ lowing instruction. In column 3, the dif­ ference between perceived performance (T3) and post-instruction expectations/ assessments (T2) are presented. The as­ terisks by a number in columns 2 and 3 indicate that the difference is statistically significant. Column 4 contains the simple correlations between performance and overall satisfaction with the library. The asterisks in this column indicate those correlations that are statistically signifi­ cant. Positive numbers indicate a positive relationship between performance and overall satisfaction, and negative num­ bers indicate a negative relationship. http:staff.26 328 College & Research Libraries July 1998 TABLE 1 Results (N = 76) Change in Difference between Pre-instruction to Performance and Correlation between Post-instruction Values Expectations Performance and (T2 - T1) (T3 - T2) Satisfaction General Library Percent of needed materials 2.014 -4.045 .53*** available at the library Rarely has the material I need 0.030 -0.075 -.66*** Difficult to find materials at -0.074 -0.269** -.41*** the library Library Specifics Average wait time for books 0.313** -1.183*** -.07 Maximum wait time for books 0.194* -1.759*** -.08 Average wait time for articles 0.388*** -0.906*** -.09 Maximum wait time for articles 0.134 -1.328*** -.24* Computers useful for research 0.000 0.090 .40*** Computers difficult to use -0.149 -0.209 -.49*** Printed references useful for -0.119 -0.090 .16 research Printed references difficult to use 0.045 0.060 -.19 Library Staff Library staff competent -0.224 0.045 .47*** Library staff helpful -0.030 -0.448*** .58*** Patron Self-Assessment Skillful using libraries in general 0.164* 0.149* .21* Skillful using the university library 0.433*** 0.239** .32*** Hours spent conducting 0.149 -0.462*** .05 library research Levels of significance*p < .10 **p < .05 ***p < .01 Difference between Pre­ and Post­ Instruction Values Looking at the results in column 2, sig­ nificant differences appear in only two general categories: 1. Wait time for materials: Patron esti­ mates of average wait times for books and articles and the expected maximum wait time for books increased significantly fol­ lowing library instruction. 2. Self-assessed skill level: Patrons as­ sessed their skill level as being higher fol­ lowing library instruction. Somewhat surprisingly, there were no significant differences in expectations re­ garding the usefulness or difficulty in using computers or print references, in the general availability of material at the library, or in the expected helpfulness/ competence of the library staff. Difference between Performance and Post­ instruction Expectations Looking at column 3, assessed perfor­ Effects of Library Instruction 329 mance differed from patron expectations in five areas: 1. Perceptions of the difficulty of finding material in the library: Patrons experienced less difficulty in finding material than ex­ pected. 2. Wait time for articles and books: Patrons received materials faster than expected; average and maximum wait times were shorter than expected for both books and articles. 3. Perceptions of staff helpfulness: Patrons perceived the library staff to be signifi­ cantly less helpful than expected. 4. Self-assessed skill level: Patrons as­ sessed their skill level as being signifi­ cantly higher at T3 than they did at T2. 5. Hours spent conducting library research: Patrons spent significantly less time con­ ducting library research than expected. Correlates of Satisfaction Finally, looking at column 4, performance in four general areas appears to be corre­ lated with satisfaction: 1. Perceptions of information accessibility: The proportion of needed materials avail­ able at the library, the difficulty of find­ ing the materials, and the maximum wait time for articles all are significantly related to overall satisfaction with the library. 2. Perceptions of staff competence and help­ fulness: Perceptions that library staff are competent and helpful are highly corre­ lated with overall satisfaction. 3. Perceptions of computers: There is a moderate positive relationship between the perceived usefulness of library com­ puters and overall satisfaction, and a moderate negative relationship between the perceived difficulty of using comput­ ers and overall satisfaction. 4. Self-assessment of skill for using the li­ brary: There is a moderate positive corre­ lation between a person’s self-assessed skill level for using libraries and overall satisfaction with the library. Surprisingly, performance in three other areas appears not to be related to satisfaction: 1. perceived wait time for books; 2. perceived usefulness or perceived difficulty of using print references; 3. time spent in the library conducting library research. Discussion The primary objectives of library instruc­ tion sessions focused on students learn­ ing basic skills needed to complete their assignments. Closely tied to these were the objectives that students would gain realistic expectations of library resources and learn that library staff are available and willing to help. Librarians might ex­ pect that after an introductory session covering various sources and search tech­ niques, most students would have differ­ ent expectations of the library and more favorable perceptions of library staff. Stu­ dent expectations about the amount of time it takes to obtain library materials increased following instruction and, it is hoped, moved closer to the level at which the library usually performs. Student per­ ceptions of their skill levels also increased, consistent with previous studies that sug­ gested library instruction can increase stu­ dent confidence levels. However, some results were not so straightforward. Instruction did not ap­ pear to affect student expectations of com­ puters or print sources. Students knew that a main purpose of the sessions was to learn practical computer-centered li­ brary skills, which would benefit them in the class. Many of the students had pre­ vious experience with computers and probably were predisposed to positive attitudes and expectations about them. It may be that these students already had positive perceptions of computers in gen­ eral, which they applied to library com­ puter systems as well. The library con­ text of the computers may not have af­ fected students’ already high views of computer usefulness. It also is surprising that expectations about difficulty using computers did not change significantly. It seems reasonable that hands-on expe­ 330 College & Research Libraries July 1998 rience would either raise or lower expec­ tations, and librarians assume or hope that comfort and ease of use come with practice. The English composition teach­ ers focused more on journal sources than books or other formats, and much of the librarians’ instruction involved searching of indexes and abstracts via computer. It is possible that print sources may have been taken for granted by, or considered irrelevant by, the students in this sample. Another implied objective of library in­ struction was to increase student expec­ tations concerning the availability of ap­ propriate materials to meet their research needs. It is unclear whether instruction was unsuccessful in this or whether ex­ pectations were already high and difficult to move up. It also might be expected that following library instruction, students would have greater expectations of help­ fulness and competence of library staff. However, this was not the case, and it is unclear why students’ experience with librarian instructors did not inspire more confidence in library staff. Does this re­ flect a student perception of less-than­ adequate performances on the part of li­ brarian instructors? Does student assess­ ment of library staff decrease as self-as­ sessment of skills increases? Are students, in effect, comparing library staff compe­ tence and helpfulness to their own per­ ceived abilities and seeing themselves as closer in knowledge and skills to library staff? These are questions that might be addressed in further research. In the third wave of the survey, stu­ dents assessed their own performance, as well as that of the library and its staff. Not only did students find it easier to use the library than they had expected, but it did not take as long to receive needed mate­ rials as they had believed it would. In the students’ experience, the library per­ formed better in these areas. Students found the library staff to be less helpful than expected. It is difficult to know what this means, particularly because the li­ brary experiences of students exposed them to a much wider range of person­ nel than the librarian instructor in the in­ struction session. Moreover, it is unclear what students judged to be helpful or not helpful behaviors or perceived attitudes of staff. If students really found themselves using the library faster and more easily than they expected, perhaps they per­ ceived less need for help from staff. Given student assumptions, library instruction librarians should not find it surprising that successful performance, along with increased exposure or experience, might Student expectations about the amount of time it takes to obtain library materials increased following instruction and, it is hoped, moved closer to the level at which the library usually performs. result in students’ higher self-assessment of library skills. One might ask whether this is more a function of increased skills and knowledge or one of building self- confidence. It would seem difficult to separate the two, especially because other studies have shown students to be rela­ tively good judges of their library skills. Most likely, the finding that students spent less time doing their research than they expected is an extension of the per­ formance-related findings stated above. These findings suggest that certain fac­ tors in students’ library experience, such as access to information and perceptions of library staff, are considerably more im­ portant to their overall satisfaction than are other factors. Accessibility is valued by stu­ dents, and the more accessible materials were, the more satisfied the users were. Similarly, students valued library staff, and when they perceived staff to be com­ petent and helpful, they were more satis­ fied with the library as a whole. It is interesting that the perceived use­ fulness and ease of use of computers as a reference source are associated with pa­ tron satisfaction, but perceived usefulness Effects of Library Instruction 331 and ease of using print reference sources are not. The university library at which this research was conducted is a leader in the use of computers in a university library setting. PC and Macintosh work­ stations with full Internet and multiple database access are widely available throughout the library. It is unclear from this research whether the importance of computer-based reference material was correlated with patron satisfaction be­ cause computers were the only reference source used or because a different struc­ tural relationship exists between com­ puter use and patron satisfaction versus print reference use and patron satisfac- For library skills in general, the better students feel their skills are, the more satisfied they will be with the library. tion. The library instruction sessions were all conducted in an electronic classroom, and English composition faculty pro­ moted the importance of electronic data­ bases, while in effect disregarding print sources. For a variety of reasons, students may have been predisposed to favor or place higher value on electronic informa­ tion sources. Their hands-on experience in the instruction sessions was almost exclusively centered on electronic sources and searching for periodical citations. Even when they used the catalog, it was mostly to see if the library subscribed to particular journals. Handouts, with the exception of the catalog handout, also focused on journals. For library skills in general, the better students feel their skills are, the more sat­ isfied they will be with the library. The university is located on a busy, nontradi­ tional, urban campus, where students place a high value on time and efficiency. It was surprising, therefore, to find that the time necessary to obtain books and conduct library research was not a factor in student satisfaction with the library. It also is interesting to note that wait time for articles is only slightly correlated with satisfaction and that wait time for books is not correlated with patron satis­ faction at all. This result is somewhat sur­ prising because the literature has indi­ cated that long waits are associated with dissatisfaction. However, the results of pre­ vious research and the results obtained in this study are not necessarily incongruous. It is possible that for some service aspects, a two-factor conceptualization is appropri­ ate. Two-factor theory states that certain variables have the power only to dissat­ isfy people whereas others have the power only to satisfy. A dissatisfying factor has the power to dissatisfy patrons if perfor­ mance is below some threshold level. However, after performance reaches the threshold, dissatisfaction disappears, but further improvements in the factor do not increase satisfaction. After the threshold levels of performance have been reached for all dissatisfying factors, satisfaction increases as performance on satisfying factors increases.27 Two-factor theory was originally de­ veloped to explain job satisfaction; there is some evidence that a two-factor conceptualization is relevant to consumer satisfaction situations as well. If two-fac­ tor theory is relevant in library services, it is possible that wait time perceived to be excessive may dissatisfy patrons, but wait times perceived as reasonable do not have the power to raise satisfaction rat­ ings. If this is true, the measured effect of library instruction regarding wait time— that is, the observed effect of increasing expected wait time—may be seen as hav­ ing a positive effect on satisfaction, not because satisfaction increases but, rather, because the library instruction may have prevented patrons from becoming dissat­ isfied. The relevance of two-factor theory to library services in general, and with regard to wait time in particular, is an area requiring additional research effort. In summary, of the variables measured in this study, the only effects of instruc­ tion that are directly associated with http:increases.27 332 College & Research Libraries July 1998 longer-term patron satisfaction appear to be related to patron self-assessed skill lev­ els. In this regard, the study findings are consistent with those of Pearson and Tiefel—increasing patrons’ confidence in their skill at using libraries in itself has a positive impact on their perception of the library. One major opportunity uncovered during this project is that of improving pa­ tron perception about the library staff, es­ pecially with regard to helpfulness. Actual perceived helpfulness was significantly lower than expected, yet staff helpfulness is more strongly correlated with patron sat­ isfaction than is any other single variable. Expectations of helpfulness did not change as a result of the instruction, which in ret­ rospect probably is a good thing: If expec­ tations had risen following the instruction, and perceived staff helpfulness remained unchanged, it is likely that overall satis­ faction among this group would have been lower. The managerial implication is that either patron expectations of staff helpfulness ought to be lowered (a dan­ gerous strategy) or perceptions of staff helpfulness ought to be raised. Conclusions Development of a comprehensive “driv­ ers of satisfaction” model will have sig­ nificant implications for assessment, ac­ countability, and library administration. The explicit recognition that patron sat­ isfaction is one component of library qual­ ity means that its measurement is a logi­ cal component of assessment. Libraries frequently assess individual aspects of library service—for example, reference, instruction, circulation, document deliv­ ery, collection development. In addition, they would be well served by looking at the degree to which such aspects, indi­ vidually as well as collectively, contrib­ ute to overall satisfaction. Within this framework, understanding what drives satisfaction at a specific li­ brary requires an ongoing measurement effort. It is unlikely that any two libraries will have the same “drivers” weighted in the same manner, so although the evalu­ ation process described here may be gen­ eralizable, the specific findings probably are not. The first step is for each library to identify what variables are correlated with satisfaction and then to determine which activities under the control of the library staff influence those variables. The ultimate goal of such an effort is to un­ derstand and manage those activities that determine user satisfaction with the li­ brary. Notes 1. Youjae Yi, “A Critical Review of Consumer Satisfaction,” in Review of Marketing 1990, ed. Valarie A. Zeithaml (Chicago: American Marketing Association, 1990); Robert A. Peterson and William R. Wilson, “Measuring Customer Satisfaction: Fact and Artifact,” Journal of the Academy of Marketing Science 20 (winter 1992): 61–71. 2. David K. Tse and Peter C. Wilton, “Models of Consumer Satisfaction Formation: An Ex­ tension,” Journal of Marketing Research 25 (May 1988): 204–12; Richard A. Spreng, Scott B. MacKenzie, and Richard W. Olshavsky, “A Reexamination of the Determinants of Consumer Satisfaction,” Journal of Marketing 60 (July 1996): 15–32. 3. Yi, “A Critical Review of Consumer Satisfaction”; Richard L. Oliver, “A Cognitive Model of the Antecedents and Consequences of Satisfaction Decisions,” Journal of Marketing Research 17 (Nov. 1980): 460–69. 4. Jill Coupe, “Undergraduate Library Skills: Two Surveys at Johns Hopkins University,” Research Strategies 11, no. 4 (fall 1993): 187–201. 5. David N. King and John C. Ory, “Effects of Library Instruction on Student Research: A Case Study,” College & Research Libraries 42 (Jan. 1981): 31–41. 6. Virginia Tiefel, “Evaluating a Library User Education Program: A Decade of Experience,” College & Research Libraries 50 (Mar. 1989): 249–59. 7. Patricia Daragan and Gwendolyn Stevens, “Developing Lifelong Learners: An Integra­ tive and Developmental Approach to Information Literacy,” Research Strategies 14, no. 2 (spring Effects of Library Instruction 333 1996): 68–81. 8. Dorothy F. Davis, “A Comparison of Library Bibliographic Instruction Methods on CD­ ROM Databases,” Research Strategies 11, no. 3 (1993): 156–63. 9. Patricia F. Vander Meer and Galen E. Rike, “Multimedia: Meeting the Demand for User Education with a Self-Instructional Tutorial,” Research Strategies 14, no. 3 (summer 1996): 145–58. 10. Trudi E. Jacobson and Janice G. Newkirk, “The Effect of CD-ROM Instruction on Search Operator Use,” College & Research Libraries 57 (Jan. 1996): 68–76. 11. Vicki Coleman, Yi Xiao, Linda Bair, and Bill Chollett, “Toward a TQM Paradigm: Using SERVQUAL to Measure Library Service Quality,” College & Research Libraries 58 (May 1997): 237. 12. Jacobson and Newkirk, “The Effect of CD-ROM Instruction on Search Operator Use,” 76. 13. Roland Person, “Long-Term Evaluation of Library Bibliographic Instruction: Lasting En­ couragement,” College & Research Libraries 42 (Jan. 1981): 19–25. 14. Penelope Pearson and Virginia Tiefel, “Evaluating Undergraduate Library Instruction at The Ohio State University,” Journal of Academic Librarianship 7, no. 6 (Jan. 1982): 351–57. 15. Tiefel, “Evaluating a Library User Education Program.” 16. Sonia Bodi, “Teaching Effectiveness and Library Bibliographic Instruction: The Relevance of Learning Styles,” College & Research Libraries 51 (Mar. 1990): 113–19. 17. Carroll H. Varner, Vanette M. Schwartz, and Jessica George, “Library Instruction and Tech­ nology in a General Education ‘Gateway’ Course: The Student’s View,” Journal of Academic Li­ brarianship 22, no. 5 (Sept. 1996): 355–59. 18. Coupe, “Undergraduate Library Skills.” 19. King and Ory, “Effects of Library Instruction on Student Research.” 20. Arlene Greer, Lee Weston, and Mary Alm, “Assessment of Learning Outcomes: A Mea­ sure of Progress in Library Literacy,” College & Research Libraries 52 (Nov. 1991): 549–57. 21. Diane Prorak, Tania Gottschalk, and Mike Pollastro, “Teaching Method and Psychological Type in Library Bibliographic Instruction: Effect on Student Learning and Confidence,” RQ 33, no. 4 (summer 1994): 484–95. 22. Connie J. Ury and Terry L. King, “Reinforcement of Library Orientation Instruction for Freshman Seminar Students,” Research Strategies 13, no. 3 (summer 1995): 153–64. 23. Vander Meer and Rike, “Multimedia.” 24. Danuta A. Nitecki, “Changing the Concept and Measure of Service Quality in Academic Libraries,” Journal of Academic Librarianship 22, no. 3 (May 1996): 181–90; Coleman, Xiao, Bair, and Chollett, “Toward a TQM Paradigm,” 237–51. 25. Danuta A. Nitecki, “Assessment of Service Quality in Academic Libraries: Focus on the Applicability of the SERVQUAL,” in Proceedings of the 2nd Northumbria International Conference on Performance Measurement in Libraries and Information Services, September 7–11, 1997 (Newcastle upon Tyne: Information North for the Department of Information and Library Management, University of Northumbria at Newcastle, 1998), 181–96. 26. Questions are available from the authors. 27. F. Herzberg, B. Mausner, and B. Snyderman, The Motivation to Work, 2nd ed. (New York: Wiley, 1959).