College and Research Libraries Research Notes Factors Related to Frequency of Use of CD-ROM: A Study of ERIC in an Academic Library Anne-Marie Belanger and Sandra D. Hoffman This study investigated some of the factors as- sociated with frequency of use of ERIC CD- ROM, including gender, age, level of familiar- ity with computers, and level of study. During the winter semester of 1988, 231 education stu- dents from Concordia University were sur- veyed. The instrument used was a mailed ques- tionnaire, which elicited a 63.6% rate of return. The analyses carried out included fre- quencies of variables and cross-tabulations, us- ing chi-square to test for independence between variables. All factors were found to be statisti- cally significant with the exception of level of study. Since the results obtained for gender were strong, further analysis was carried out controlling for gender. INTRODUCTION With the emergence of new information technology, library users are increasingly involved in the use of computers. Al- though computers were originally used primarily in technical services, online cir- culation systems, online catalogs, and CD-ROMs are now increasingly available for use by the public. Among the first CD-ROM products was the ERIC database, an index of educa- tional resources published in paper as Current Index to Journals in Education and Resources in Education. ERIC CD-ROM was introduced by SilverPlatter in 1986, and Concordia University Libraries acquired it in February 1987. Through our work at the reference desk and through the response of education students to the ''library expe- rience,'' a library exercise many education teachers require students to complete as part of their course, we became aware that students required various levels of train- ing. Since this instruction would be added to the work load of existing reference staff, it seemed important to determine if there were identifiable groups for whom in- struction could be designed. To begin our research, we decided to investigate some of the factors that might be correlated to frequency of use of CD-ROM. This paper describes a pilot project conducted with education students at Concordia Univer- sity and reports the results of our prelimi- nary analysis. REVIEW OF THE LITERATURE Studies on user reaction to CD-ROM products are beginning to appear in the lit- Anne-Marie Belanger, now Cataloguing Librarian, was Acting Education Librarian at Norris Library, at the time of this study. Sandra D. Hoffman is Commerce Librarian at Norris Library, Concordia University, Montreal, Canada QC H3G 1MB. 153 154 College & Research Libraries erature1; however, none has addressed specifically frequency of use of CD-ROM. Studies of online catalogs are numerous, however, and since online catalogs are a related technology, some of these findings were of interest to us. A landmark study by Matthews and Lawrence surveyed sixteen different on- line systems in twenty-nine libraries and more than 8,000 users and 4,000 nonus- ers.2 A combination of research methods consisting of surveys, online monitoring, focus group interviews, and feature analy- sis were used. This study found that ''age, sex, education level, academic discipline and academic affiliation appear to be asso- ciated with frequency of library, online and other catalog use.' '3 Prior experience with computers was not found to be a sig- nificant variable for users of online cata- logs. However, Matthews also found that "non-users have had much less experi- ence with other computer systems when compared to users of the online catalog. " 4 In Lolley's study of university students' use of automated and card catalogs, he concluded that perhaps ''a significant fac- tor in the acceptance and use of the on-line catalog is previous experience with com- puters either in the work or home environ- ment. Almost 60% of the students re- sponding to the survey indicated such previous experience.' ' 5 Because opinion is divided and available research is on an- other technology (online catalogs), it seemed appropriate to investigate the in- fluence of previous computer experience or computer-related skills on frequency of use of CD-ROM. Among other factors, student status, age, and sex were studied. Student sta- tus, identified by Matthews as a signifi- cant factor affecting online catalog use, had also been examined in an academic setting by Pease and Gouke. 6 Although neither study compared frequency of use of online catalogs by graduates and un- dergraduates, this was identified as an important factor for the purposes of our study. Age was a variable looked at by both Lolley and Matthews. Lolley found that students under thirty preferred the online catalog over the card catalog, and Mat- March 1990 thews noted that "non-users are slightly older than users of the online catalog.''7 Collis examined sex differences in atti- tudes toward computers of 3,000 young adults and found that "males are consis- tently more positive about using com- puters than are females, and more likely to express interest and pleasure in using a computer.' ' 8 Therefore, females may be less frequent users of ERIC CD-ROM than males. PROBLEM TO BE STUDIED The purpose of the investigation was to determine if the frequency of student use of ERIC CD-ROM could be correlated to: (1) gender, (2) age, (3) range (extent) of ex- perience with computers, and ( 4) student status. Since ERIC is an index of educa- tional resources, our investigation was limited to use by education students. In order to determine if the above factors were related to the use of ERIC CD-ROM, the following hypotheses were tested: 1. There is a statistically significant rela- tionship between frequency of use of ERIC CD-ROM and gender. 2. There is a statistically significant rela- tionship between frequency of use of ERIC CD-ROM and age. 3. There is a statistically significant rela- tionship between frequency of use of ERIC CD-ROM and level of familiarity with computers. 4. There is a statistically significant rela- tionship between frequency of use of ERIC CD-ROM and level of study. DEFINITIONS Some of the variables identified are de- scribed in further detail. The level of famil- iarity with computers was defined in terms of the following nine factors: (a) use of a computer at home; (b) use of a com- puter at work; (c) use of a computer in a place other than home or work, e.g., a friend's place, the university, a public li- brary, etc.; (d) use of a computer for pro- gramming; (e) use of a computer for word processing; (f) use of educational soft- ware; (g) use of computer games; (h) hav- ing taken a computer course; (i) familiarity (level of ease/unease) with the computer keyboard. The level of familiarity with computers was measured in terms of an aggregate score. Student status was defined as under- graduate, master's, and doctoral. Within the undergraduate level were included all students registered in a program leading to a B.A. in Education and all students en- rolled in the program leading to the Con- cordia University Certificate in Education. The graduate level was broken down into two groups, master's and doctoral candi- dates. The master's category included di- ploma students. SIGNIFICANCE OF THE PROBLEM TO BE STUDIED To date, few research studies have been published analyzing use of CD-ROM tech- nology. End-user training is often men- tioned as an area of concern because it is so time-consuming. 9 Since most libraries have limited human and material re- sources, the results of this study may en- able libraries to design workshops that tar- get groups of students (e.g., older, female, undergraduate). If groups can be identified, it would be possible to tailor content and level of instruction to their needs. For example, if undergraduate stu- dents constitute a significant group of in- frequent users, it may be advisable to in- clude instruction in core undergraduate education courses. If females or mature students are infrequent users, workshops limited to these sex/ age groups may be tai- lored to their needs. If familiarity with computers is an important factor in fre- quency of use, workshops may need to be designed and advertised as including a ba- sic introduction to computers. SAMPLING AND METHODOLOGY The methodology used was a survey Research Notes 155 consisting of a mailed questionnaire deal- ing with frequency of use of ERIC CD- ROM. The population studied was stu- dents from the Education Department of Concordia University. For purposes of this study, students were grouped into three categories as outlined in table 1. The size of each of our populations was determined by adding the number of part- time and full-time students within each group. Our sample size was determined by using the table published by Krejcie and Morgan. 10 When we selected our ran- dom sample, we chose an equal number of part-time and full-time students at the un- dergraduate level, given that numbers of part-time and full-time students were al- most equal at this level of study. All doctoral-level students were included in the sample. To ensure an adequate return of the questionnaire, two mailings were planned and carried out. Of the 361 questionnaires mailed, 231, or 63.6%, were returned. As an incentive for the students to return the questionnaire, we offered to those willing to include their names, addresses and telephone numbers an opportunity to par- ticipate in a drawing for two gift certifi- cates to be used at a local bookstore. 11 A pretest of the survey instrument was carried out using a sample of twenty-five students. The wording of several ques- tions was changed to clarify the meaning, based on the suggestions of two experi- enced researchers. 12 The revised question- naire was pretested again using a sample of fourteen staff members. DATA ANALYSIS The SPSS/PC + statistical package was used to analyze the data collected in our study. Missing cases were accounted for in all the statistical procedures. Frequen- TABLE 1 STUDENT ENROLLMENT, SAMPLE SIZE, AND QUESTIONNAIRES RETURNED Student Enrollment Sample Size Questionnaires Returned Level of Study Number % Number % Number % B.A. and Certificate 264 47 160 44 88 38 Diploma and Masters 277 50 184 51 128 55 Ph.D. 17 3 17 5 13 6 No status given 2 1 Total 558 100 361 100 231 100 156 College & Research Libraries des, the first analysis carried out on the data, tabulated the number of responses in any one category. These were reported in percentages. The second analysis carried out on the data was cross-tabulation, using chi- square to test for independence between variables. The variable frequency of use of ERIC CD-ROM was tested for possible as- sociation with the following variables: age, gender, level of familiarity with com- puters, and level of study. The level of sig- nificance chosen for this study was .05. In addition to the chi-square analysis, the strength of the relationships was tested in the following manners: hypothe- sis 3 was tested using Somers' D; hypoth- eses 1 and 4 were tested using Cramer's V; and hypothesis 2 was testing using Ken- dall's Taub. 13 RESULTS When frequency of ERIC CD-ROM use was cross-tabulated with gender, age, and, level of familiarity with computers respectively, the results of the chi-square tests were all significant at a level of at least .05. The hypotheses were accepted at this level of significance. When frequency of ERIC CD-ROM use was cross-tabulated with level of study, the result of the chi.:square test was signifi- cant at a level of .0565. Consequently, the hypothesis was rejected; i.e., no signifi- cant relationship was found between level of study and frequency of ERIC CD-ROM use (see table 7). March 1990 DISCUSSION The value of Cramers' V for the fre- quency of CD-ROM use by gender in table 2 is .25735. This value, in conjunction with a chi-square significance level of .0016, suggests that a fairly strong relationship exists between a student's gender and fre- quency of CD-ROM use. The correlation between gender and frequency of use is evident at all levels (see table 2), but is most marked at the level of greatest use. As expected from the literature, men were more likely to use and continue to use ERIC CD-ROM. Matthews found that ''the typical user of the online catalog is male."14 Given the strong correlation be- tween gender and frequency of use, when analyzing the relationship between fre- quency of use and other factors (e.g., age, status, etc.), we analyzed males and fe- males independently as well, in order to see whether apparent correlations with other factors might actually be indirect gender effects. The data for these analyses are reported in tables 4, 6 and 8. From table 3 we see that there is a statis- tically significant relationship between age and frequency of CD-ROM use in the total population. However, when age and frequency of use are analyzed controlling for gender, results are statistically signifi- cant only for females (table 4). That is, when age and frequency of use are tabu- lated separately for males and females, only the relationship between females and frequency of use is significant at the .05 level. Although the results for males were TABLE2 CROSS-TABULATIONS FREQUENCY OF CD-ROM USE BY GENDER Gender Count Row CD-ROM Use Column% F M Total Never 94 20 114 54.0% 35.1% 49.4% Once 19 4 23 10.9% 7.0% 10.0% 2-5 times 42 15 57 24.1% 26.3% 24.7% 6 or more times 19 18 37 10.9% 31.6% 16.0% Column Total 174 57 231 75.3% 24.7% 100.0% Number of Missing Observations = 0 Chi-Square = 15.29926 df = 3 Significance = .0016 Cramer's V = .25735 Research Notes 157 TABLE 3 CROSS-TABULATIONS FREQUENCY OF CD-ROM USE BY AGE Count Age 45and Row CD-ROM Use Column% to24 25-34 35-44 older Total Never 22 34 43 14 113 43.1% 41 .5% 66.2% 43.8% 49 .1% Once 6 6 5 6 23 11.8% 7.3% 7.7% 18.8% 10.0% ยท 2-5 times 17 24 9 7 57 33.3% 29.3% 13.8% 21.9% 24.8% 6 or more times 6 18 8 5 37 11.8% 22.0% 12.3% 15.6% 16.1% Column total 51 82 65 32 230 22.2% 35.7% 28.3% 13.9% 100.0% Number of Missing Observations = 1 Chi-Square = 17.26492 df = 9 Significance = .0447 Kendall's Tau B = - .08783 TABLE4 CROSS-TABULATIONS FREQUENCY OF CD-ROM USE BY AGE CONTROLLING FOR GENDER Count Age 45and Row CD-ROM Use Column% to24 25- 34 35-44 older Total Never Female 21 29 31 12 93 44.7% 48.3% 77.5% 46.2% 53.8% Male 1 5 12 2 20 25.0% 22.7% 48.0% 33.3% 35.1% Once Female 5 4 5 5 19 10.6% 6.7% 12.5% 19.2% 11.0% Male 1 2 0 1 4 25.0% 9.1% 0.0% 16.7% 7.0% 2-5 times Female 16 18 3 5 42 34.0% 30.0% 7.5% 19.2% 24.3% Male 1 6 6 2 15 25.0% 27.3% 24.0% 33.3% 26.3% 6 or more times Female 5 9 1 4 19 10.6% 15.0% 2.5% 15.4% 11.0% Male 1 9 7 1 18 25.0% 40.9% 28.0% 16.7% 31.6% Column total Female 47 60 40 26 173 27.2% 34.7% 23.1% 15.0% 100.0% Male 4 22 25 6 57 7.0% 38.6% 43.9% 10.5% 100.0% Female: Chi-Square= 19.77204 Male: Chi-Square = 8.15426 Significance = .0194 Significance = .5187 Kendall's Tau B = - .12299 Kendall's Tau B = - .14104 NOTE : Column percentages are calculated separately for female and male respondents . not statistically significant, we examined the data to identify trends. The strength of the relationship be- tween frequency of use and age of females is a fairly weak, negative association, as indicated by the value of Kendall's Tau B of- .12299. That is, frequency of use is in- versely associated with age in females. For men, if we combine the frequency- of-use categories of 2-5 times and 6 or more times across the age groupings, we see that use of CD-ROM two or more times holds fairly constant at approxi- mately 50% across the table, with the ex- ception of the 25-to-34 age grouping. Here the percentage increases to 68.2%. The men in this age group may be full-time students who have more discretionary 158 College & Research Libraries time than working students who study part-time, or they may be enrolled in computer-related programs (e.g., educa- tional technology or computer-assisted learning). In future surveys it would be desirable to ask students to identify their program of study or specialization. If we combine the same frequency-of- use categories for women as we did for men (i.e., 2-5 times and 6 or more times), we see that the percentage of women us- ing CD-ROM two or more times holds at approximately 45% up to age 34, drops to 10% in the 35-to-44 age group, and rises again to 34.6% in the 45+ age group. This lower use by females in the 35-to-44 age group may be attributed to a period in life occupied with child rearing and career ad- vancement, when family and/ or work commitments may allow little time for ex- perimenting with new technology. Fur- ther investigation is needed to determine if lower use in this age group is typical for all women and if so, why. It is also inter- esting to note that 48% of the males in this age group have never used CD-ROM. Per- haps the same factors are at work for both men and women in this age group. In the age category up to 24 years, it is difficult to draw conclusions concerning the effects of gender. Although the fe- males are fairly evenly divided between never/once (55.3%) and two or more times (44.6%), there are too few males (four) to report trends. Again, in students over 45, it is not pos- March 1990 sible to see trends for male users, as the group includes only six men. The women in this age group are concentrated in the never category of use (46.2%). In summary, patterns of frequency of use differ for men and women in the 25-to- 34 and 35-to-45 age groups. For men, an increase in use was noted in the 25-to-34 age group, with more or less consistent levels of use in the remaining age groups. For women, there was a noticeable de- crease in the frequency of use in the 35-to- 44 age group. These preliminary findings seem to indicate that patterns of use vary for males and females in different age groups, but further investigation is needed. Table 5 shows that there is a statistically significant relationship between level of familiarity with computers and frequency of CD-ROM use in the total population, as evidenced by a significance level of .0171. However, a Somers' D value of .17321 in- dicates that the strength of the association is relatively weak. When familiarity with computers and frequency of use are ana- lyzed controlling for gender, there is no statistically significant relationship (table 6). That is, when familiarity and frequency of use are tabulated separately for males and females, neither relationship is statis- tically significant at the .05 level. How- ever, we analyzed the results to identify trends . . In table 6, the column totals show the distribution of females and males across TABLES CROSS-TABULATIONS FREQUENCY OF CD-ROM USE BY LEVEL OF FAMILIARITY WITH COMPUTERS Count Score Score Aggr~~:e Score Score Score Row CD-ROM Use Column% 9-13 14-18 19-24 25-30 31-36 Total Never 21 28 25 22 10 106 58.3% 56.0% 50.0% 40.7% 34.5% 48.4% Once 6 3 7 5 1 22 16.7% 6.0% 14.0% 9.3% 3.4% 10.0% 2-5 times 6 14 11 18 6 55 16.7% 28.0% 22.0% 33.3% 20.7% 25.1% 6 or more times 3 5 7 9 12 36 8.3% 10.0% 14.0% 16.7% 41.4% 16.4% Column total 36 50 50 54 29 219 16.4% 22.8% 22.8% 24.7% 13.2% 100.0% Number of Missing Observations = 12 Chi-Square = 24.54212 df=12 Significance = .0171 Somers' D = .17321 Research Notes 159 TABLE6 CROSS-TABULATIONS FREQUENCY OF CD-ROM USE BY LEVEL OF FAMILIARITY WITH COMPUTERS CONTROLLING FOR GENDER Count Score Score Aggr~~:e Score Score Score Row CD-ROM Use Column% 9-13 14-18 19-24 25-30 31-36 Total Never Female 21 26 17 18 4 86 63.6% 59.1% 47.2% 46.2% 36.4% 52.8% Male 0 2 8 4 6 20 0.0% 33.3% 57.1% 26.7% 33.3% 35.7% Once Female 5 3 5 4 1 18 15.2% 6.8% 13.9% 10.3% 9.1% 11.0% Male 1 0 2 1 0 4 33.3% 0.0% 14.3% 6.7% 0.0% 7.1% 2-'5 times Female 5 11 9 13 3 41 15.2% 25.0% 25.0% 33.3% 27.3% 25.2% Male 1 3 2 5 3 14 33.3% 50.0% 14.3% 33.3% 16.7% 25 .0% 6 or more times Female 2 4 5 4 3 18 6.1% 9.1% 13.9% 10.3% 27.3% 11.0% Male 1 1 2 5 9 18 33.3% 16.7% 14.3% 33.3% 50.0% 32.1% Column total Female 33 44 36 39 11 163 20.2% 27.0% 22.1% 23.9% 6.7% 100.0% Male 3 6 14 15 18 56 5.4% 10.7% 25.0% 26.8% 32.1% 100.0% Female: Chi-Square = 9.87957 Male : Chi-Square = 15.62582 Significance = .6265 Significance = .2090 Somers' D = .13947 Somers' D = .15838 NOTE : Column percentages are calculated separately for female and male respondents . the scores. Women are fairly evenly dis- tributed across the scores from 9 to 30, varying from a low of 20.2% to a high of 27. 0%. At the highest score level, there are only 6.7% of the women. On the other hand, the percentage of men increases as the score levels increase. In fact, only 16.1% of the men have scores lower than 19. When we look at levels of frequency of use, the numbers of men are so low in score categories 9-13 and 14-18 that no trends can be seen. For men, their com- puter experience does not appear to be as- sociated with their frequency of CD-ROM use except at the level of 6 or more times in the score categories of 19-24 and higher. In the never category of use, the per- centage of females decreases across the scores from a high of 63.6% in the 9-13 cat- egory to a low of 36.4% in the 31-36 score category. Conversely, at the level of use of 6 or more times, the percentage of women increases as the scores increase, with a slight decrease in percentage in the 25-30 score category. That is, the percentage ranges from a low of 6.1% in the lowest score category to a high of 27.3% in the highest score category. These trends indi- cate an association between frequency of CD-ROM use and familiarity with com- puters for women. This agrees in part with Matthews' study, which found a correla- tion between nonuse of online public ac- cess catalogs and less experience with computers. 15 We found a corresponding correlation in nonuse of ERIC CD-ROM for women. The correlation between level of study and frequency of use was not found to be statistically significant at the .05 level (ta- ble 7). However, at the level of use of 6 or more times, there was a correlation in that there were 7. 9% of the undergraduates and 21.1% of the graduates. When level of study and frequency of use are analyzed controlling for gender, neither relation- ship is statistically significant at the .05 level (table 8), but the trend found in the 160 College & Research Libraries March 1990 TABLE7 CROSS-TABULATIONS FREQUENCY OF CD-ROM USE BY LEVEL OF STUDY Count Level of Study Row CD-ROM Use Column% Under~aduates Graduates Total Never 48 66 114 53.9% 46.5% 49.4% Once 11 12 23 12.4% 8.5% 10.0% 2-5 times 23 34 57 25.8% 23.9% 24.7% 6 or more times 7 30 37 7.9% 21.1% 16.0% Column total 89 142 231 38.5% 61.5% 100.0% Number of Missing Observations = 0 Chi-Square = 7.54257 df = 3 Significance = .0565 Cramer's V = .18070 TABLE 8 CROSS-TABULATIONS FREQUENCY OF CD-ROM USE BY LEVEL OF STUDY CONTROLLING FOR GENDER Count Level of Study Row CD-ROM Use Column% UnderS!aduates Graduates Total Never Female 45 49 94 57.0% 51.6% 54.0% Male 3 17 20 30.0% 36.2% 35.1% Once Female 9 10 19 11.4% 10.5% 10.9% Male 2 2 4 20.0% 4.3% 7.0% 2-5 times Female 20 22 42 25.3% 23.2% 24.1% Male 3 12 15 30.0% 25.5% 26.3% 6 or more times Female 5 14 19 6.3% 14.7% 10.9% Male 2 16 18 20.0% 34.0% 31.6% Column total Female 79 95 174 45.4% 54.6% 100.0% Male 10 47 57 17.5% 82.5% 100.0% Female: Chi-Square = 3.13650 Male : Chi-Square = 3.57%8 Significance = .3711 Significance = .3106 Cramer's V = .13426 Cramer' s V = .25060 NOTE: Column percentages are calculated separately for female and male respondents. total population is again apparent. That is, at the level of use of 6 or more times, the percentage of graduates is higher for both females and males (6.3% of the under- graduates and 14.7% of the graduates for females, and 20% of the undergraduates and 34.0% of the graduates for males). This trend was apparent for both males and females. This higher level of use on the part of graduate students is not unex- pected, given that at this level of study more research is required. In addition to the questions testing the hypotheses, several questions were asked for administrative purposes. In order to determine if further publicity was needed, we asked students when they first learned the library had acquired ERIC CD-ROM. Although 43.5% of our respondents re- ported learning of ERIC CD-ROM be- tween February and September 1987, there were still 24.3% who learned of it only through our questionnaire. We asked students who had used ERIC CD-ROM if they planned to use it again in order to determine if this was an area re- quiring further study. Only 7% of the stu- dents who had tried ERIC CD-ROM re- ported that they would not use it again. Of those who had not used it, 82% planned to do so. It would appear that students are receptive to this new technology and, hav- ing tried it, are planning to use it again. Respondents were asked if they would be interested in instruction classes, if available, and 76.9% replied affirmatively. Since the time of our survey, SilverPlatter has added a tutorial to the ERIC software that may partially meet this need. Further investigation will be required to ascertain its effectiveness. SUMMARY AND CONCLUSIONS The results of this study may enable li- braries to concentrate limited resources and personnel in areas where the greatest need has been identified. Some of these areas are discussed below. Since a correlation was found between frequency of use of ERIC CD-ROM and level of familiarity with computers for women, training sessions for people with little or no previous computer experience might include a very basic introduction to computers and the computer keyboard, as well as instruction in elementary search- ing techniques. Ideally, hands-on experi- ence would be an invaluable part of these workshops. Since the results of our study indicate that frequency of use is not significantly related to level of study, it may not be es- . sential to hold separate workshops for un- dergraduate and graduate students. Results associated with age may indicate the advisability of holding workshops for mature students so that presentations can be geared to their needs. Furthermore, it may be advisable to publicize to mature students the advantages of learning to use CD-ROM, e.g., eventual time savings and Research Notes 161 the ability to combine concepts. As noted earlier, many education stu- dents have had an extensive introduction to the library because they must complete as part of a course requirement a very de- tailed library experience which requires them to use education indexes. In fact, 78.8% use the library at least once a month, and only 4.3% never use the li- brary. Therefore, increased general library instruction and increased library publicity would have minimal impact on this popu- lation. A surprising finding was the fact that 24.3% of our respondents learned that the library had acquired ERIC CD-ROM through our survey, although it was ex- tensively publicized when acquired. This finding may indicate the need for continu- ing and/or specialized publicity, perhaps through such channels as the registration process. Of our respondents, almost half had never used ERIC CD-ROM. This can par- tially be accounted for by the fact that half of the nonusers were unaware of its exis- tence prior to our survey. However, fur- ther investigation may determine other factors that may be influencing nonuse. For example, do students not have a need to use indexes within certain programs of study? Do some working students have access to library materials through their school board library? Since ERIC is an index of educational re- sources, it seemed appropriate to limit our study to education students. Future stud- ies might incorporate more academic dis- ciplines and appropriate CD-ROM prod- ucts. It would also be useful to compare the response of students in several univer- sities. It is interesting to note that our pop- ulation is 75% female and 25% male. When we look at level of study, the women are almost evenly divided be- tween undergraduates (45.5%) and grad- uates (54.6%), while the men are primarily graduate students (82.5% graduates as compared with 17.5% undergraduates) . This profile may be typical of most educa- tion departments, but further investiga- tion would be necessary to verify this. In conclusion, this preliminary research has identified some factors associated 162 College & Research Libraries March 1990 with frequency of use of ERIC CD-ROM in education students at Concordia Univer- sity. However, further studies are needed to determine whether these findings are valid for students in other universities and in other disciplines. REFERENCES 1. N. A. Mullan and A. R. Blick, "Initial Experiences of Untrained End-users with a Life Sciences CD-ROM Database: A Salutary Experience,'' Journal of Infonnation Science 13:139-41 (1987); Tim Miller, "Early User Reaction to CD-ROM and Videodisc-based Optical Information Products in the Library Market," Optical Infonnation Systems 7:205-9 (May/June 1987); and Deborah Bezanson, "Integrating CD-ROM with Printed and Online Services: A Silver Platter End-User Perspective," Optical Infonnation Systems 7:387-90 (Nov./Dec. 1987). 2. Joseph R. Matthews, Gary S. Lawrence, and Douglas K. Ferguson, eds., Using Online Catalogs: A Nationwide Survey: A Report of a Study Sponsored by the Council of Library Resources (New York: Neal-Schuman, 1983); and Joseph R. Matthews and Gary S. Lawrence, "Further Analysis of the CLR Online Catalog Project," Infonnation Technology and Libraries 3:354-76 (Dec. 1984). 3. Matthews and Lawrence, "Further Analysis," p.359. 4. Matthews, Lawrence, and Ferguson, Using Online Catalogs, p.93. 5. John Lolley, "University Night Students and the Online Library Catalogs," Research Strategies 2, no.4:177 (1984). 6. Sue Pease and Mary Noel Gouke, "Patterns of Use in an Online Catalog and a Card Catalog," College & Research Libraries 43:286 Ouly 1982). 7. Lolley, "University Night Students," p.174; and Matthews, Lawrence, and Ferguson, Using Online Catalogs, p.95. 8. Betty Collis, "Psychosocial Implications of Se~ Differences in Attitudes toward Computers: Results of a Survey," International Journal of Women's Studies 8:209 (May/June 1985). 9. Jean Reese, "CD-ROM Technology at Vanderbilt University: Impact on Library Staff and the Educational Community," Optical Infonnation Systems 9:38-43 Oan./Feb. 1989); Stephen P. Harter and Susan M. Jacques, "Optical Disc Systems in Libraries: Problems and Issues," RQ 27:516-27 (Summer 1988). 10. Robert V. Krejcie and Daryle W. Morgan, "Determining Sample Size for Research Activities," Educational and Psychological Measurement 30:608 (Autumn 1970). 11. Carol Walton, Susan Williamson, and Howard D. White, "Resistance to Online Catalogs: A Comparative Study at Bryn Mawr and Swarthmore Colleges," Library Resources & Technical Services 30:390 (Oct./Dec. 1986). 12. Diane Mittermeyer, Assistant Professor, McGill University, Graduate School of Library and Information Science, and John Leide, Associate Professor, McGill University, Graduate School of Library and Information Science. 13. Frank P. Besag and Peter L. Besag, Statistics for the Helping Professions (Beverly Hills: Sage, 1985), p.305-7, 286-93. Sidney Siegel and N. John Castellan, Nonparametric Statistics for the Behavioral Sciences, 2d ed. (New York: McGraw-Hill, 1988), p.303-10, 225-32; and Frank M. Andrews and others, A Guide for Selecting Statistical Techniques for Analyzing Social Science Data, 2d ed. (Ann Arbor, Mich.: Institute for Social Research, University of Michigan, 1981), p.8-9. 14. Matthews, Lawrence, and Ferguson, Using Online Catalogs, p.87. 15. Matthews, Lawrence, and Ferguson, Using Online Catalogs, p.93.