Comparing gifts to purchased materials: a usage study

Library Collections, Acquisitions, & Technical Services 24 (2000) 351 359 Comparing gifts to purchased materials: a usage study Rob Kairis* Kent State University, Stark Campus, 6000 Frank Ave. NW, Canton, OH 44720, USA Abstract The aim of the study was to compare the value of gift materials added to the general collection of a small to medium-sized academic library with materials selected and purchased by the library. The measure employed to determine the relative value of the gifts vs. non-gifts was how often the materials were used within a 12-month period. Two groups of 77 books (gifts and purchased items) were tracked over the course of 1 year to determine if the gifts were used significantly less. Results indicate that donated books were used on average 0.87 times whereas non-gifts were used 1.38 times. A t-test was also applied demonstrating that the two groups were significantly different at the 0.05 level. This provides empirical data libraries can use to re-evaluate existing policies and procedures for handling gifts. 2000 Elsevier Science Ltd. All rights reserved. Keywords: Book gifts; Donations; Usage; Use studies; Collection development 1. Introduction By definition, small to medium-sized academic libraries need to be selective when deciding on acquiring the materials that are added to their collections. Budgets, staff, and shelf space are factors that guide collection development efforts in such libraries to purchase materials that support the programs offered by the college or university. An important criterion to be considered for any item added to the library s collection is its usefulness. It can be argued that donations to small to medium-sized academic libraries add materials that ignore this important principle. The lure of obtaining something for free often results in the relaxing of normally applied selection criteria. Not only does this lead to the addition of materials that would not typically be added, it can be detrimental to the collection as a whole, * Corresponding author. Tel: 1-330-499-9600; fax: 1-330-494-6212. E-mail address: rkairis@stark.kent.edu (R. Kairis). 1464-9055/00/$ see front matter 2000 Elsevier Science Ltd. All rights reserved. PII: S1464-9055(00)00091-9

352 R. Kairis / Libr. Coll. Acq. & Tech. Serv. 24 (2000) 351 359 because important collection development initiatives are ignored. Dennis Dickinson has suggested that academic libraries, especially small college libraries, seek to develop working collections of materials that are useful and will be used. Most gifts-in-kind emphatically do not contribute to this latter goal; often they actually detract from it [1]. An important attribute used to determine the value of an item, after it has been added to the library collection, is the number of times it is used. Therefore, anticipating the level of usage for a potential addition to the collection should pay a major role in deciding whether or not it gets added. As Liz Chapman has written, the use to which a particular book will be put is an integral part of its value. Equally, in most normal libraries, there is no point in having a valuable book which will never be used [2]. Libraries that are highly selective when deciding what items to purchase desire an efficient and effective collection small, but useful. Anticipating the potential use of items selected for purchase is an important consideration. A question arises as to whether this same practice is included in the process used to add donated materials to a library s collection. When the library is particularly selective in deciding what materials are selected, purchased, and added to the collection (as is the case for small to medium-sized academic libraries), anticipated use should be considered an important factor. Central to this discussion is the suggestion that anticipated use may not be an important consideration when gifts are selected for addition to the library collection. This study was to determine whether or not gift materials added to the general collection of a small to medium-sized academic library are as valuable as materials selected and purchased by the library. The measure employed to determine the relative value of the gifts and non-gifts is how often the materials were used. 2. Review of the literature The literature on gift materials in libraries is limited and provides little data on the use, usefulness, and cost of the gift materials in libraries [3]. This is particularly true for usage studies that compare the use of gift materials to non-gift materials. The literature does include some anecdotal data regarding librarians impressions concerning how useful gifts are. Robinson has reported that out of 47 Tennessee academic libraries surveyed, 71% felt that most gift books are not useful [3]. Dole has suggested that gifts should be considered a benefit if the value of the material acquired exceeds the costs involved in acquiring them, and if the materials fit the library s collection development policy [4]. Quantitative methods can establish the relative costs of processing purchased materials and gifts. Because, by definition, gifts do not incur a purchase price, under most circumstances the cost to a library for adding gifts should be less than purchased materials. In addition, libraries sometimes receive out-of-print materials as gifts that would otherwise be very costly for the library to purchase. Still, sometimes processing gifts has added costs (putting nameplates on books, maintaining files on donations, corresponding with donors, itemizing and evaluating donated items, etc.). The most difficult aspect of comparing gifts and purchased materials is defining value. As previously noted, the value of any item added to a library s general collection needs to take into account the actual usage of the item. In deciding which gifts are added to

R. Kairis / Libr. Coll. Acq. & Tech. Serv. 24 (2000) 351 359 353 a collection, this emphasizes the need (at least to some extent) to consider the anticipated use of the item. Although there are plenty of opinions available that suggest that gifts added to the collection are, for the most part, little used when compared to materials that are selected and purchased by the library, little empirical evidence has been generated affirming this position [1]. Diodato and Diodato proved this assertion over 15 years ago by comparing circulation statistics of gift and non-gift materials shelved within two separate call number ranges. were identified by the presence of a plate identifying the donor. Usage was determined by dates stamped on the due date slips [5]. Diodato s analysis employed two models. The first model was used to determine the relative cost per circulation for gifts and non-gifts (a concept that is similar to Dole s first condition, describing a gift s value). The results showed that non-gifts circulated on average four times more than gifts. This suggested that the cost for processing gifts should be one quarter (or less) than that for processing non-gifts. The second model measured the number of books that had circulated at least once. Forty percent of gifts circulated at least once, whereas two-thirds of non-gifts had at least one circulation [5]. Although these results do seem to confirm the opinions of librarians regarding the lower usage of gifts compared to non-gifts, some feel that libraries need more information on the degree to which gift materials are actually used and are useful in libraries [3]. Further examination of the topic should be conducted for the following reasons: Y More than 15 years have passed since an empirical study comparing the usage of gifts to non-gifts was conducted. Y Automated library systems make data collection easier and arguably more accurate. Y Usage studies should include all types of uses (in-house use, renewals, along with basic circulation figures). Y A study comparing groups containing an equal number of items over a specific period of time may produce more statistically valid results. 3. Methodology The Stark Campus Library is one of seven small academic regional campus libraries of Kent State University, containing close to 80,000 volumes. It also serves as the library for the Stark State College of Technology, which shares the campus with Kent State. The combined student FTE for both institutions is around 4,400. There is a collection development policy that provides guidelines for Kent s regional campus libraries. The section relating to materials selection states in evaluating materials for selection or renewal, librarians shall consider the purpose, authority, scope, intended audience, relevance to the curriculum, and cost [6]. Although it may be assumed that most librarians will somehow factor into the selection process the concept of anticipated use, this criterion is not specifically included. The section related to gifts simply states that the same criteria used for purchased materials should be applied to gifts. On average, the library spent about $70,000 on monographs during each of 2 years

354 R. Kairis / Libr. Coll. Acq. & Tech. Serv. 24 (2000) 351 359 Table 1 Processing of donated materials for 1996/1997 Number % Donated items 528 100 Initially selected 126 24 No. of selected items owned 30 24 (of initially selected) No. added to collection 96 18 No. tested for usage 77 15 (1996/1997 1997/1998). As is consistent with similarly sized libraries (in terms of volumes held and budget), the selection process is quite focused. In addition, each year the library receives several hundred gift items, primarily books. The basic procedure for handling donated items is as follows: 1. Donors are informed that the library may not add all (or any, for that matter) of the gifts to the collection. The library disposes of items that are not added to the collection; some materials are passed on to other libraries, whereas many are simply thrown out. 2. If requested, the library will provide a letter of thanks to the donor, although most donations are anonymous. 3. An evaluation of the gifts is conducted based upon such factors as the condition, age, and subject content of the material. This is very similar to the process used by the library to weed existing material from its collection. 4. All initially selected items are searched in the library s online catalog. If an existing copy is found, the gift is added only if the usage of the existing copy is considered high. During the 1996/97 academic year, 528 items were donated. The materials consisted mainly of books, but also included some periodicals and videocassettes. The amount of donated materials during 1996/1997 was consistent with the amount of donated materials during previous years. After applying the procedure listed above, 77 books were added to the library s general collection. Table 1 illustrates that, of all the items donated, 96 (18%) were added to the library and 77 (15%) were added to the general circulating collection for testing (the other 19 items were either non-book materials or books added to other specific collections in the library). The 77 books added to the general collection were coded for later identification in the library s Innovative Interfaces online system and put into circulation in July 1997. The fact that only 18% of the donated items were actually added to the collection indicated that the existing procedure for identifying gift materials for addition was already quite selective, although predicting usage was not a specific quantifiable measure applied to determine which gifts were added to the collection. The plan was to compare the usage of these 77 gift books with the first 77 purchased books added to the general collection during July 1997 (the control group). Because the library purchased a total of 1,250 books during 1997/1998, this sample represented 6% of the total purchased during the year. Still, 77 books were selected to make the sample sizes identical. The first 77 purchased books were chosen so that each sample could be placed into

R. Kairis / Libr. Coll. Acq. & Tech. Serv. 24 (2000) 351 359 355 Table 2 Comparison of the content of gift and non-gift samples Social sciences 6 37 Sciences 68 3 Humanities 3 37 Total 77 77 circulation at the same time. It should be noted that no attempt was made to compare the content of the gift books with the content of the purchased books. A later analysis separated the samples into three categories, Social Sciences, Sciences, and Humanities. Table 2 shows that the content of the two groups of books was very different. This could be perceived as a weakness to the study. Still, the intent of the study was to compare gifts added to the library to an equally sized group of purchased books. If content-similar purchased books were selected, this would have delayed placing the non-gift books into circulation and the gift books would have been available for circulation for a longer period of time than the non-gift books. So emphasis was placed on having same-sized groups and an equal period of time for each group to be used rather than on the content of each group. In this study, a use was not qualified. Any indication that a patron handled a book was counted as one use. The library considered a book to be used if it was checked out, it was renewed, or it was used in-house. All three types of uses were tracked separately. In-house use was counted every time a book that was not checked out to a patron needed to be re-shelved. Without documenting the extent to which an item was used by a patron, there was no quantifiable way to justify weighting one use more than another. Just because one book was checked out, and another was used in-house, did not imply that the checked out item was used more than the one used in-house. Indeed, one inherent disadvantage in usage studies has been determining how useful any particular item was to the patron. Therefore, the only reasonable measurement standard was to count all types of use equally. In July 1998, usage statistics of the 77 coded gift items were compared to usage statistics of the control items (the first 77 purchased items added to the general collection during July 1997). The library s online system was able to isolate both groups of 77 books by codes and dates in the book item records. In addition, separate fields in those items records were used to count the three types of uses. The online system also had the capability of reporting those statistics in a meaningful way. After gathering the usage statistics for both groups, some standard statistical measures were applied to determine the average usage of gifts and non-gifts. As an added measure, to discover whether the difference between the groups was within the range of variance between statistically similar groups, a t-test was employed to determine whether or not the difference was significant at the 0.05 level. A comparison of the totals for each use type and the average (or mean) use per book are good indicators for describing differences between the groups. The t-test is a more detailed examination of the difference of the means of the groups that determines if this difference was significant enough to validate the findings of the study.

356 R. Kairis / Libr. Coll. Acq. & Tech. Serv. 24 (2000) 351 359 Table 3 Usage comparison of non-gifts and gifts for 1 year Books 77 77 No. used at least once 42 33 % used at least once 55% 43% No. checked out 73 44 No. renewed 30 13 No. used in-house 3 10 Total usage 106 67 Average use per book (mean) 1.38.87 4. Results After accumulating usage information for a year, in July 1998 reports were generated from the library s online system comparing the usage of the 77 gifts to the 77 non-gifts. Table 3 displays basic usage statistics for these two groups of books. These statistics clearly showed that the non-gift books were used more times than the gift books over the course of 1 year. Less than half of the gifts were used one or more times (43%, strikingly similar to Diodato s results), whereas more than half of the non-gifts were used at least once (55%). On average, each gift was used less than once (0.87 times), whereas each non-gift was used 1.38 times. An interesting aspect of this comparison of usage between gifts and non-gifts is that the one area where gifts out-performed non-gifts is in in-house use (gifts were used in-house more than three times as much as non-gifts). Although all types of uses are equally weighted, one could make the assumption that after retrieving these gifts from the shelves, patrons found the books not useful enough to check out. Although this would further illustrate that gifts are used less than purchased items, without examining exactly how each book (gift and non-gift) was used, it would not be appropriate to assume that in-house use is not as valuable as other methods of use. The statistics do seem to indicate that gifts are less used than purchased items. However, a further statistical measure needs to be applied to see if this difference is statistically significant. Table 4 shows how the two equal-sized groups were broken down so that a t-test could measure the difference between the means (0.87 for gifts and 1.38 for non-gifts). An important piece of information for calculating the t-value is the square of the distance from the mean for each book in both groups. Table 4 shows the sum of these squares (248.08 for non-gifts and 126.7 for gifts). From the data in Table 4, information needed to calculate the t-value is derived (Table 5). Figure 1 shows the formula for calculating the t-value. Substituting the values in Table 5 into this formula a t-value of 2.00 is established. With 76 degrees of freedom (k 1), the value demanded for significance at the 0.05 level is 1.99. Because the t-value in this study is 2.00, it is determined that there is a significant difference at the 0.05 level. Therefore, it is accurate to state that gift books were used significantly less than non-gift books in this study.

R. Kairis / Libr. Coll. Acq. & Tech. Serv. 24 (2000) 351 359 357 Table 4 Usage distribution and statistical measures for non-gifts and gifts k Number of times used (X) Distance from the mean (x) Square of distance from the mean (x 2 ) Number of times used (X) Distance from the mean (x) Square of distance from the mean (x 2 ) 1 8 6.62 43.87 6 5.13 26.32 2 7 5.62 31.62 4 3.13 9.80 3 7 5.62 31.62 4 3.13 9.80 4 5 3.62 13.13 4 3.13 9.80 5 5 3.62 13.13 3 2.13 4.54 6 4 2.62 6.88 3 2.13 4.54 7 4 2.62 6.88 3 2.13 4.54 8 4 2.62 6.88 3 2.13 4.54 9 3 1.62 2.64 3 2.13 4.54 10 3 1.62 2.64 3 2.13 4.54 11 3 1.62 2.64 2 1.13 1.28 12 3 1.62 2.64 2 1.13 1.28 13 3 1.62 2.64 2 1.13 1.28 14 3 1.62 2.64 2 1.13 1.28 15 3 1.62 2.64 2 1.13 1.28 16 3 1.62 2.64 2 1.13 1.28 17 2 0.62 0.39 2 1.13 1.28 18 2 0.62 0.39 2 1.13 1.28 19 2 0.62 0.39 1 0.13 0.02 20 2 0.62 0.39 1 0.13 0.02 21 2 0.62 0.39 1 0.13 0.02 22 2 0.62 0.39 1 0.13 0.02 23 2 0.62 0.39 1 0.13 0.02 24 2 0.62 0.39 1 0.13 0.02 25 2 0.62 0.39 1 0.13 0.02 26 2 0.62 0.39 1 0.13 0.02 27 2 0.62 0.39 1 0.13 0.02 28 2 0.62 0.39 1 0.13 0.02 29 1 0.38 0.14 1 0.13 0.02 30 1 0.38 0.14 1 0.13 0.02 31 1 0.38 0.14 1 0.13 0.02 32 1 0.38 0.14 1 0.13 0.02 33 1 0.38 0.14 1 0.13 0.02 34 1 0.38 0.14 0 0.87 0.76 35 1 0.38 0.14 0 0.87 0.76 36 1 0.38 0.14 0 0.87 0.76 37 1 0.38 0.14 0 0.87 0.76 38 1 0.38 0.14 0 0.87 0.76 39 1 0.38 0.14 0 0.87 0.76 40 1 0.38 0.14 0 0.87 0.76 41 1 0.38 0.14 0 0.87 0.76 42 1 0.38 0.14 0 0.87 0.76 43 0 1.38 1.90 0 0.87 0.76 44 0 1.38 1.90 0 0.87 0.76 45 0 1.38 1.90 0 0.87 0.76 46 0 1.38 1.90 0 0.87 0.76 47 0 1.38 1.90 0 0.87 0.76 48 0 1.38 1.90 0 0.87 0.76 (continued on next page)

358 R. Kairis / Libr. Coll. Acq. & Tech. Serv. 24 (2000) 351 359 Table 4 (continued) k Number of times used (X) Distance from the mean (x) Square of distance from the mean (x 2 ) Number of times used (X) Distance from the mean (x) Square of distance from the mean (x 2 ) 49 0 1.38 1.90 0 0.87 0.76 50 0 1.38 1.90 0 0.87 0.76 51 0 1.38 1.90 0 0.87 0.76 52 0 1.38 1.90 0 0.87 0.76 53 0 1.38 1.90 0 0.87 0.76 54 0 1.38 1.90 0 0.87 0.76 55 0 1.38 1.90 0 0.87 0.76 56 0 1.38 1.90 0 0.87 0.76 57 0 1.38 1.90 0 0.87 0.76 58 0 1.38 1.90 0 0.87 0.76 59 0 1.38 1.90 0 0.87 0.76 60 0 1.38 1.90 0 0.87 0.76 61 0 1.38 1.90 0 0.87 0.76 62 0 1.38 1.90 0 0.87 0.76 63 0 1.38 1.90 0 0.87 0.76 64 0 1.38 1.90 0 0.87 0.76 65 0 1.38 1.90 0 0.87 0.76 66 0 1.38 1.90 0 0.87 0.76 67 0 1.38 1.90 0 0.87 0.76 68 0 1.38 1.90 0 0.87 0.76 69 0 1.38 1.90 0 0.87 0.76 70 0 1.38 1.90 0 0.87 0.76 71 0 1.38 1.90 0 0.87 0.76 72 0 1.38 1.90 0 0.87 0.76 73 0 1.38 1.90 0 0.87 0.76 74 0 1.38 1.90 0 0.87 0.76 75 0 1.38 1.90 0 0.87 0.76 76 0 1.38 1.90 0 0.87 0.76 77 0 1.38 1.90 0 0.87 0.76 Total 106 248.08 67 126.70 5. Conclusions This study indicates that gift materials are used less by library patrons than materials selected and purchased by the library. The fact that the library used in the study was a small academic library with highly selective procedures for adding donated materials (only 18% of Fig. 1. Formula for calculating the t-value.

R. Kairis / Libr. Coll. Acq. & Tech. Serv. 24 (2000) 351 359 359 Table 5 Statistical measures needed for t-test calculation Books (k) 77 77 Total usage ( X) 106 67 Mean usage (M) 1.38 0.87 Sum of squares of distance from the mean ( x 2 ) 248.08 126.70 Population variance estimate (s 2 x 2 k 1) 3.26 1.67 Squares of the SEM s 2 x (s 2 k) 0.04 0.02 the materials given to the library during the 1996/1997 academic year were added to its collection) lends credibility to the contention that gifts are used less than purchased items. Still, although the study shows that gifts are significantly less used than purchased materials, the study does show that gifts are used by library patrons. However, librarians may want to revise current gift policies or procedures to consider placing more emphasis on judging the anticipated use of gifts before deciding to add them to the collection. With the ability of integrated online library systems to track and report usage, this type of study could be replicated by other types and sizes of libraries. Factors such as library size and type may yield significantly different results. This study suggests questions on how individual libraries should go about their management of donated materials: Y Should the library accept gifts? Y Do gift policies and procedures use the same selection criteria as purchased materials? Y Does the process of accepting gifts cost more than the value of the material added? Y How important is the aspect of usage for considering the value of library materials? This study could be enhanced by using larger groups of records or by tracking usage over a longer time period. Also, comparing groups of materials with similar content may improve the quality of such a study. Other studies using added or expanded measures might produce different results. This paper does not suggest that libraries abandon the practice of accepting gifts. It does provide information that can be used by librarians to re-evaluate their current policies and procedures for treating gifts. Finally, it also provides some empirical evidence supporting the belief that many librarians have held regarding donated materials that they are not as useful (or as valuable) as purchased materials. References [1] Dickinson DW. Free books: are they worth what they cost? Library Issues 1997;17:1 4. [2] Chapman, L. Donations: the unacceptable face of acquisitions. Library Association Record 1986;88:503 504. [3] Robinson WC. Gift materials in Tennessee academic and public libraries. Tennessee Librarian 1990;42:12 24. [4] Dole WV. and block purchases: are they profitable. Library Acquisitions: Practice and Theory 1983;7:247 254. [5] Diodato, LW, Diodato VP. The use of gifts in a medium sized academic library. Collection Management 1983;5:53 69. [6] Collection Development Policy, Kent State University Regional Campus Libraries, latest revision by the Regional Campus Library Directors Council, Kent State University, 1998.