Selection or Censorship? School Librarians and LGBTQ Resources

Abstract

All students, including those who identify as lesbian, gay, bisexual, transgender or queer (LGBTQ), benefit from quality library books that reflect their experiences. This mixed-method research project examines whether public high school library professionals self-censor their library collections when it comes to materials with LGBTQ themes. Quantitative data were collected from 120 Ohio public high school libraries and 12 school librarians were interviewed. The results suggest that school libraries tended to have significantly fewer LGBTQ-themed books than titles with other kinds of controversial content, and that certain school-based factors such as high enrolment, racial diversity, liberal-leaning community locations; and presence of certified school librarians, were present in schools with more inclusive LGBTQ collections. Qualitative findings supported the quantitative analyses in that participants stressed the importance of a supportive community and administration when developing a quality, inclusive library collection.

Headnote

All students, including those who identify as lesbian, gay, bisexual, transgender or queer (LGBTQ), benefit from quality library books that reflect their experiences. This mixed-method research project examines whether public high school library professionals self-censor their library collections when it comes to materials with LGBTQ themes. Quantitative data were collected from 120 Ohio public high school libraries and 12 school librarians were interviewed. The results suggest that school libraries tended to have significantly fewer LGBTQ-themed books than titles with other kinds of controversial content, and that certain school-based factors such as high enrolment, racial diversity, liberal-leaning community locations; and presence of certified school librarians, were present in schools with more inclusive LGBTQ collections. Qualitative findings supported the quantitative analyses in that participants stressed the importance of a supportive community and administration when developing a quality, inclusive library collection.

Introduction

In July 2010, a 15-year-old student named Brent created a stir among school librarians when a recent posting from his blog The Naughty Book Kitties was published in School Library Journal (SLJ). In this posting, Brent explained just how important books were to him and to other young adults in helping them come to terms with their sexuality. Brent and his friends combed bookstores and public libraries looking for gay and lesbian characters, but when he turned to his own school library, his school librarian informed him that, "This is a school library. If you're looking to read inappropriate titles, go to a bookstore" (Limited Shelf Life, 2010). Librarians around the world were aghast, and sent letters to the editor of SLJ and posted comments on Brent's blog that reflected concern for his school library. Brent's blog post's closing statement was a wake-up call to many school librarians:

The world needs more librarians who are devoted to finding the right book to put in the right person's lap, not librarians who think they can decide what's "inappropriate" and what's not based on their personal prejudices. There are tons of gay teens struggling to find a group to fit into. LGBT YA lit helps us realize that no, we aren't alone, and no, we aren't worthless. It helps us discover that we are part of the LGBT group, which includes tons of brilliant people, doing brilliant things. (Limited Shelf Life, 2010)

Most school librarians can attest to the thrill of pairing students with the perfect book, introducing teens to characters that will inspire, comfort, challenge, and excite them as well as help them to move beyond their limited experiences. Students who are struggling to find a place to belong so often take solace in books; it is vital that all students have access to stories that validate their feelings and experiences. It is this desire that all students have access to high-quality, inclusive school library collections that motivated this study.

While there have been a variety of small studies conducted intended to determine whether school librarians self-censored books, including titles that specifically addressed LGBTQ topics, to date, no one has yet undertaken a large scale research project that incorporates both quantitative and qualitative methods. There is a need to ascertain to what degree librarians may be self- censoring their collections, as well as to understand librarians' philosophies about library collection development and justifications for their selection decisions. This project seeks to provide that knowledge.

Research Hypotheses

Phase One. More than half of the sampled schools will show evidence that suggests that school librarians are censoring their collections.

Phase Two. There will be certain variables about the librarian, school, and/or community (e.g., the age of the librarian, student enrollment in a school, Ohio Department of Education (ODE) ranking or the local political climate) that will correlate with higher or lower instances of self-censorship.

Phase Three. Some factors that may cause librarians to self-censor books with LGBTQ themes include past experiences with material challenges, personal belief systems, and/or pressure from administrators to avoid controversy.

Research Questions

Phase One RQ:

What is the evidence that suggests school librarians may be practicing self-censorship when it comes to including library materials with gay, lesbian, bisexual, or transgender content?

Phase Two RQ:

What is the evidence that suggests certain attributes of the librarian, school, or community may correlate with higher or lower instances of self-censorship?

Phase Three RQ:

What actual professional and personal experiences affect how school librarians approach collection development and materials selection? What internal or external factors influence their decisions during the selection process?

LGBTQ Literature is Vital to School Library Collections

The need for welcoming, informative and inclusive library collections is particularly pertinent for lesbian, gay, bisexual, transgendered and questioning (LGBTQ) students who often find school libraries to be safe havens and use libraries to locate information, despite the fact that information can easily be accessed elsewhere (Whelan, 2006). In fact, Whelan (2006) argued in "Out and Ignored" that consequences for excluding LGBTQ literature from the library can be severe for students. Whelan (2006) asserted that a strong gay literature collection could discourage teasing and bullying in schools and foster a climate of tolerance and acceptance. LGBTQ literature in schools may helps all teen patrons because such titles "offer guidance or support to teens that might otherwise choose to remain silent about the prevalence of LBGTQ-related issues within their own respective lives" (Manfredi, 2009, p. 28). Manfredi (2009) further stressed that teens who do not identify as also benefit from reading these works as they may become "heterosexuals who are committed to ideals of equality and dedicated to fighting homophobia" (p. 28). Martin also recommended providing LGBTQ collections to create "safe, welcoming environments" and to inspire the formation of gay-straight alliances (GSAs) that create dialogue and combat homophobia (2006, p. 39). Kevin Jennings, founder of the Gay, Lesbian and Straight Education Network (GLSEN), described his own experience with libraries: "The library was also the first place I found any information about gay people that was objective and gave me a sense that I might have a future. I probably would not be exaggerating to say libraries saved my life" (2006, p. 22).

School Climate and Safety

Maintaining a quality LGBTQ collection may literally save lives by supporting an environment of tolerance and acceptance. A recent study (Toomey, McGuire & Russell, 2012) revealed that students who attended schools that supported LGBTQ students with resources such as Gay Straight Alliances (GSAs) perceived their schools to be safer than those who attended schools without such clubs. The study's participants reported witnessing fewer incidences of harassment and were more likely to state that students with non-normative gender identity were safe at school. Students who attended schools without supportive policies and organizations claimed to witness acts of bullying and intimidation, particularly towards males who were perceived to be gay, a type of aggression well documented in schools. The results of 2013 survey conducted by the GLSEN (Kosciw, Greytak, Palmer, & Boesen, 2014) suggested that schools can be very hostile environment for LGBTQ teens. The GLSEN survey researchers found that more than half of LGBTQ students felt unsafe at school due to their sexual identity; approximately 65% heard homophobic comments frequently; and that more than half heard such comments from teachers and staff! LGBTQ students also reported that they face a significant amount of harassment, assault, and discrimination due to their sexuality or gender expression. On a positive note, the GLSEN survey (Kosciw, Greytak, Palmer, & Boesen, 2014) echoed the Toomey, McGuire and Russell (2012) survey results: schools with LGBTQ resources and supports, including the presence of GSAs, were significantly less likely as schools without to have those serious bullying and harassment problems in their schools.

They Ask for Them by Name

Many of the important reasons that school libraries need to include LGBTQ books in their collection is that students are demanding these titles. Like the aforementioned young blogger Brent, all teens, regardless of their sexuality, are actively searching for stories to help them understand their often chaotic and confusing realities. In a study of gay, lesbian, and bisexual students in American schools, Sanelli and Perreault (2001) compiled a list of education reforms that students would like to see implemented, and high on the list was a request for materials with gay characters. According to one of the student survey participants, "[W]e read love stories all the time in English, but there's never anything on lesbian love or there's not like gay love stories" (p. 74). Alexander and Miselis (2007) found that libraries were heavily used by LGBTQ teens, and "that this group was seeking information about understanding their gay identities, coming out, learning gay social 'rules,' and where to connect with others like them" (p. 45). For LBGTQ students, such novels not only can act as a mirror, but also as a window through straight students can view the struggles and confusion LGBTQ teens face (Rauch, 2011). "Friends and family of the [LGBTQ] community, as well as teachers, counselors, clergy, and in fact, any professional [who] works with the public, all have a need for accurate information about this group of people" (Alexander & Miselis, 2007, p. 44).

Are Librarians Censoring LGBTQ Resources?

With such compelling reasons to build strong, quality LGBTQ collections, one must wonder why some school librarians seem to resist including books with homosexual themes. According to a 2009 survey on self-censorship by School Library Journal, 70% of the library professional claimed that concern about possible parent reactions factored into their decisions to acquire controversial titles (Whelan, 2009b). Library survey respondents were also concerned about objections from administrators, community members, and students. Librarians who had previously faced a censorship challenge, one of the most stressful and harrowing experiences a school librarian can face, were significantly more apprehensive (Whelan, 2009b). Some librarians were so adamant about avoiding controversy that they reported resorting to subversive ploys, advising other librarians to "spend your entire budget on regular (which is to say, non-controversial) materials. Then when someone comes asking why you don't have And Tango Makes Three, you can just say you're out of money" (Maycock, 2011, p. 11). Librarians also reported limiting access to young adult books with LGBTQ themes by shelving them in the adult section or placing them on restricted shelving where students cannot access them without special permission. Other librarians reported that they "hid" controversial titles by removing references to "homosexuality" and related identifiers in the catalog records, to make the titles all but invisible to young patrons and potential censors (Clyde & Lobban, 2001).

The Ethics of the Profession. While there are valid and understandable concerns, librarians are bound by professional ethics to include materials for all patrons; rejecting books because their content might cause problems violates those ethics. In many ways, pre-emptive censorship is inconsistent with readers' right assured in the United States Constitution. Several federal rulings, including the landmark 1982 Board of Education, Island Trees Union Free School District v. Pico, have ruled that students have First Amendment rights, and that barring students from accessing age- appropriate books because the content might be objectionable violates students' rights and is a form of censorship (Whelan, 2009a).

Previous Studies on Self-Censorship. Self-censorship, the practice of not acquiring materials because of their potential to generate controversy, is nearly impossible to document. Librarians who engage in self-censorship are generally unwilling to admit it, no one is monitoring it, and the profession lacks open discussion on the topic (Whelan, 2009a). Burke (2008) examined public opinion trends over time related to removal of gay-themed materials from libraries and found that overall, the public has become less inclined to have objectionable material removed from libraries. Moody (2004) studied Australian librarians and their self-censorship habits via questionnaire, and found that while the majority of librarians professed to fight censorship, they often avoided controversial materials. Freedman and Johnson (2001) studied the attitudes of teachers and classroom censorship when using young adult literature, and found that teachers were extremely likely to censor their classroom collections if they knew that some of the books were controversial. Alexander and Miselis (2007) surveyed librarians to inquire about LGBTQ programming in their libraries and to speculate as to why their LGBTQ materials are not challenged by community members. In 2009, Whelan reported that SLJ survey of librarians about self-censorship, were the most specific sources of information regarding the censoring practices of library professionals.

The aforementioned studies all relied on self-reported information via surveys; another way to study potential censorship may be to examine the holdings of libraries for LGBTQ materials to check if they are being purchased and circulated. Researchers have conducted quantitative studies using online public access catalog records to objectively gauge whether libraries circulated certain controversial titles (Bellows, 2005; Coley, 2002; Cook 2004). In these studies, the researchers searched library catalogs for specific titles, and found their presence or absence to be evidence of librarian self-censorship.

Method

The researcher employed a sequential mixed-methods research design for this study. This mixed approach allows the researcher to collect and analyze of quantitative data because these data tend to produce results accepted by policymakers. The researcher then used qualitative to explain and expand on the quantitative results. In this study, quantitative analyses delivered empirical data that alone did not tell a complete story; participant narratives were vital to more completely understand reasons behind self-censorship and selection decisions. Both sets of data were necessary for this research to effect change.

Phase One

The objective of the first phase was to collect and analyze quantitative library collection data to determine which libraries, if any, showed signs of self-censorship by the school librarian. The collections of 120 public high school libraries across the state of Ohio were evaluated using an instrument created by the researcher.

Sample. The sample for Phase 1 included 120 Ohio public high school libraries. Using the ODE database to sort the schools by typology (based on size and income of districts), 20 high school libraries (15 for the sample and five alternates) were selected via stratified random sample that represented urban, suburban, small town, and rural locales; a range socioeconomic circumstances; and all geographic sectors of Ohio. An additional requirement for inclusion in the sample was web-based access to the school library's Online Public Access Catalog (OPAC).

Sample. The sample for Phase 1 included 120 Ohio public high school libraries. Using the Ohio Department of Education (ODE) database which sorts the schools into eight typologies (two each of urban, suburban, small town, and rural, subdivided based on size and income of districts), 20 high school libraries (15 for the sample and five alternates) were selected from each typology via stratified random sample that represented all typologies; a range socioeconomic circumstances; and all geographic sectors of Ohio. An additional requirement for inclusion in the sample was web-based access to the school library's OPAC.

The Instrument. For this project, the library collections were measured with an instrument that included a list of quality LGBTQ literature. Created by the researcher, a certified librarian with extensive experience in young adult literature, the instrument included titles of high quality drawn from personal knowledge, published research and revised and a variety of websites dedicated to LGBTQ literature. The researcher also revised the list based on feedback from other Ohio librarians shared via the Ohio Educational Library Media Association (OELMA) email list. The final instrument consisted of two lists of young adult titles, one lists of only titles with LGBTQ content and a "control" comparison list which that did not include such books but did include titles known to be controversial for other reasons. Both lists contained the same number of titles and were limited to contemporary young adult fiction. The books on both lists were similar in length and reading level, were intended for high school students, had mostly been published in the past two decades, and were still in print and commercially available through the major book distributors. The significant difference between the two lists was the inclusion of LGBTQ content in all of the books on the experimental list and the absence of LGBTQ content in the control list titles.

The LGBTQ List. The experimental list consisted of 28 titles, all of which contain LGBTQ content. Each of these books is considered a young adult title, meaning that it is targeted to young adults and has been positively reviewed in library journals as appropriate for high school students. In order to qualify for this list, the book must contain at least one significant character who identifies in some manner as LGBT or Q, and issues relating to that character's sexuality or gender difference must be germane to the plot. In these books, the LGBTQ character is generally the teenage main character or a character close to the main character (e.g. parent, sibling, or friend). Each of the books included in the instrument has been formally recognized as an LGBTQ book in one of three ways: by having been a recipient of ALA's Stonewall Award; been included on ALA's Rainbow List; and/or by having been formally challenged in a school due to LGBTQ content. The LGBTQ List is detailed in Appendix A.

The Control List. The researcher created a second list of controversial titles as a "control," or comparison list. This list was designed to determine whether a librarian was not selecting titles with LGBTQ content while still purchasing books with other types of controversial content. The 28 titles on the control list were all considered young adult fiction; positively reviewed by library journals as appropriate for high school students; and had been formally challenged in school or public libraries. Some of the books were subsequently banned, while others were retained and returned to the library. None of the books on the control list were LGBTQ titles, which means that the main character does not identify as LGBTQ and that there are not any mentions or inclusions of characters or events in the story relating to LGBTQ issues. Because all of the books had been challenged, they contained a number of other controversial elements, including violence, drug use, heterosexual sexual content, occult activities, profanity, and/or religious issues. Appendix B includes the full control list.

Measurement. In order to assess a collection's signs of librarian self-censorship, each library's collection was measured against the two lists of titles. Findings that a library possessed a large number of books on the Control List but very few of the books on the LGBTQ List were taken to indicate that self-censorship may be occurring. Specifically, libraries that possessed less than a quarter of the titles on the LGBTQ list were seen as quite possible experiencing some degree of librarian self-censorship of LGBTQ materials, especially if the same library contained a large number of the Control List titles.

In order to sort the titles by "adoptability," a Rasch model test was run for each list. This statistical model, a commonly used method in item response theory, is often used in educational research in part to determine the difficulty of test items, allowing researchers to sort items from easiest to most difficult. The measures will vary in ranges with a midpoint of 0, with very wide ranges indicating larger differences of degrees of difficulty (Keeves & Alagumalai, 2005). Using this model, a score was calculated for each title, with higher scores indicating books that were more difficult to adopt, or least often found in the school library catalogs. It is important to note that the Rasch measures for each list are calculated using only the various adoption rates of that particular list and the scores from one list cannot be compared to those of another. In other words, an LGBTQ title with a Rasch measure of 1 should not be considered to have the same ease or difficulty of adoption as a Control title with a measure of 1 - it is not an "apples to apples" comparison.

After each title was assigned a Rasch score, another Rasch score was generated for each library according to the specific titles each held, creating scaled scores that could then be used to more accurately compare the libraries. These new library Rasch scores were based not only on how many titles the collection contained, but also on each of those titles' adoptability score according to the Rasch model. Two scores were generated for each library, one for the LGBTQ titles and a separate one for the Control titles. Rasch measures range from negative numbers to positive numbers, with 0 as the midpoint, so in order to make the scores easier to use, for this project the Rasch scores were adjusted to make all the scores a positive number. Libraries with high scores for both lists are considered to have diverse collections containing a considerable number of titles that other schools are hesitant to adopt. Libraries with low scores on the LGBTQ list but higher than average scores on the Control titles show signs of censorship of LGBTQ titles. Libraries with low scores on both lists are not of interest - their lack of current titles in general more likely represents a lack of funding than potential censorship.

Phase Two

The objective of Phase 2 was to collect and analyze demographic and other data about the librarian, the school, and the community in order to look for variables that appeared to relate to self- censorship. Using publicly available information resources, pertinent data were recorded for the participating librarians in the schools identified in Phase 1.

Sample. For the next phase, the sample of 120 school libraries was pared down to 60. The 30 school libraries with the least evidence of censorship and the 30 schools with the most evidence of censorship were selected to become the sample for Phase 2, while the school libraries in the middle were removed from further study. The 120 school libraries studied in Phase 1 were ranked according to the findings from the most likely to be censored to the least. The top and bottom quartiles of this list (N=60) comprised the sample for Phase 2, specifically the quarter of schools with outcomes suggesting high instances of self-censorship (n=30) and the quarter of schools that showed the least evidence of possible self-censorship (n=30).

Measurement. There are countless variables that will influence a librarian in book selection. After considering the availability of comparable current data sets, , the researcher selected to examine specific variables about the librarian (i.e. certification type and age), the district (i.e. enrollment and l rating) and the community (diversity, income and ideology) to determine if any of the variables correlated with the possible evidence of self-censorship.. The variable list was not meant to be exhaustive, but rather, should be seen as an initial attempt to establish external variables that might have a relationship with self-censorship.

The maintains a comprehensive web site that disseminates a wide variety of information about Ohio Schools (see http://reportcard.education.ohio.gov). Demographic information about each school district studied in this project was collected earlier in the project during the sampling process, as it was provided along with the typology information for each school. Ohio issues each school and district an annual State Report Card letter grade, which provides publically available information about the school and district including its standardized test scores; ODE designation; and specific demographic information for the school including enrollment, student diversity by ethnicity, and average socio-economic status of the district's families. Information about each school's librarian was also located on the ODE's website. The Educator Search feature allows interested parties to access publically available information about educators, including their Ohio professional certifications and their date of birth. Finally, the Ohio Secretary of State's website archives extensive election data. The political ideology of the community served by each school was determined by the county's 2012 U.S. Presidential election results. Current American political trends suggest that social conservatives tend to vote for Republican candidates and social liberals tend to vote for Democratic candidates (Quinn, 2015). The 2012 Presidential election results by county serves to distinguish communities that are socially conservative from those that are more liberal.

Procedure. Using online resources, information about the variables listed above was collected and recorded. A Pearson product-moment correlation coefficient bivariate analysis was run including each of the listed variables to search for significant correlations between the data and the measures of potential self-censorship. Correlation coefficients were computed using Statistical Package for the Social Sciences (SPSS) software among the scales for Number of LGBTQ Titles, Political Lean of County of School, Age of Librarian, Certification of Librarian, School Student Enrollment, School Percentage of Minority Students, School Percentage of Economically Disadvantaged Students, and the ODE Designation of the school. Using the Bonferroni approach to control for Type I error across the seven correlations, a p value of less than .007 (.05/7 = .007) was required for significance

Phase Three

Phase 1 identified several libraries in which librarians may be self-censoring their collection. Phase 2 examined other variables regarding librarians, schools, or communities might influence self-censorship. However, numbers alone do not tell the whole story, so the objective of Phase 3 was to explore the practices of some of the librarians whose libraries were scrutinized in this study.

Sample. Twelve school librarians were selected from the Phase 2 schools, six each from two groups created in Phase 2. The librarians were selected for a variety of reasons, including interesting information regarding their libraries gleaned during the prior phases (e.g. a library which labeled books as LGBT in the OPAC and a librarian who owned more titles on the experimental list than the control list) or their unusually high or low placement on the Rasch model. Of the 19 total candidates, 14 responded to the request to be interviewed, and 12 librarians completed the process. The characteristics of the sample schools and librarians are pseudonymously detailed in Appendix C.

Interviews. The purpose of the interviews was to explore how librarians make selection decisions, particularly when it comes to controversial topics. The interviews were mostly conversational, but each included specific open-ended questions. The master list of interview questions is reflected in the tables in Appendix D.

Procedure. The selected librarians were contacted by email to inform them of the study. The librarians who expressed interest in the study were then contacted again to arrange a time for a telephone interview, projected to last for approximately 30 minutes. Those librarians were also sent an email attachment with the list of questions in advance, so that they could begin thinking about selection and censorship and be better prepared to discuss these topics in the interview. To ensure that participants were fully informed of their rights and consent to be interviewed, they were also sent the Research Consent Form. The interviews were conducted in person, over the phone, or via email conversations. Detailed notes were taken during the interview, and a full written account of the conversation was written immediately following the interview. All participants agreed to be contacted later if necessary in order to verify or expand on any information gleaned during the interview.

Phase One Results and Analysis

Over a two-week period in January 2014, the researcher searched the OPACs of the 120 selected school libraries for the 58 titles on the LGBTQ list and the control list. For each school, the researcher noted whether the library owned the title in any format (e.g., book, paperback, electronic book) or if the library catalog had no record of the title. The researcher then totalled the scores for each list for each library. Overall, the 120 libraries combined owned 465 of the LGBTQ titles and 1,236 of the control titles, or on average, just over one LGBTQ title for every three control titles on the list. Almost 70% of the libraries held less than a quarter of the titles on the LGBTQ list, while only 2% contained less than a quarter of the control list titles. Conversely, 80% of school libraries owned more than half of the control list titles, while only 8% of them shelved more than half of the LGBTQ titles. On the surface, it certainly appeared that libraries are selecting significantly more controversial titles that are devoid of LGBTQ-themed content.

However, each book is unique, and raw scores do not distinguish between more or less controversial titles on the two lists. It was readily apparent upon examination of the totals that certain titles seemed to be adopted by many schools, while others were included in very few. In other words, some seemed "easy" for schools to adopt, while others seemed more "difficult."

Phase Two Results

Phase 2 involved the study of additional variables that may have an impact on whether a library collection will have a diverse young adult collection. Additional data were collected for the 60 schools selected at the end of Phase 1.

The results of the correlational analyses presented in Table 1 show that four of the seven correlations were statistically significant and had Pearson correlation coefficients greater than .35.

There were statistically positive correlations between the number of LGBTQ titles owned by a school library and the political lean of the county in which the school resides (specifically the tendency of the citizens to vote for the Democratic presidential candidate in 2012); the type of certification held by the librarian (specifically those certified in library media); the number of students enrolled in the school (specifically larger school enrollments); and the percentage of minority students enrolled in the school (specifically large minority populations).

Some other factors that I expected to have in impact on selection decisions did not have a statistically significant effect on whether or not librarians tended to purchase LGBTQ-themed materials. There were no significant correlations between the age of the librarian, percentage of economically disadvantaged students, or ODE designation and the tendency to purchase or reject materials with LGBTQ themes.

Results: Phase Three

In Phase 3, twelve librarians were interviewed between mid-April and early June 2014. These discussions centered on how selection decisions were made in their school library. Four librarians were interviewed in person during a visit to their school library, five were interviewed over the telephone, and three, due to various circumstances, were only able to communicate via email. The interviews were conversational in nature, and no recording devices were used. For the face-to- face and telephone interviews, I took detailed notes during the interview and wrote a full report from those notes immediately after.

While I initially contacted equal numbers of librarians in the two groups of libraries selected from the collections that appeared not to censor LGBTQ titles and those who did appear to censor LBGTQ titles, I received significantly more responses from the group with the diverse LGBTQ collections. The first eight interviews described were from the first group, while the final four librarians manage collections that show some degree of censorship.

Demographic data about the schools and the librarians for the eight schools in the first group and the four schools in the second group are listed in Appendix C. Responses to interview questions are reflected in Appendix D.

Common Traits of Libraries

Each of the 12 librarians had unique circumstances, experiences, and perspectives that affected how selection decisions were made in their school libraries. The quantitative data analysis from the second phase of this study suggested that four variables strongly correlate with the tendency of school to own a diverse LGBTQ collection. Those four factors (having a certified school librarian in the library, being located in in a politically "blue" county, having a diverse student body, and having a large student enrollment) were reflected in nearly all the library programs studied in the third phase of the study.

Library Media Certification. Statistical analyses during the second phase showed a strong, positive correlation between the presences of a certified library professional and a larger LGBTQ book collection. During phase three, it was confirmed that in seven of the eight libraries considered to have diverse LGBTQ collections, the professional in charge of the high school library was certified in library media, while only one of the four of the libraries with few or no LGBTQ- themed novels employed a full time, certified high school librarian.

Community values. Nearly every librarian interviewed discussed the community their libraries served, and those community values seemed to directly impact the degree to which the librarian felt comfortable providing potentially controversial materials to the students. John Green High School and Nancy Garden High School both had librarians that spoke very highly of their communities and how much those communities valued open mindedness and resisted censorship. They both felt extremely supported by their communities and felt no trepidation about providing a diverse collection of materials for their students. The librarian at Crutcher High School, while acknowledging that his district was rather conservative, still mentioned the pride the community had in its small, independent school district, and that the community valued the library. Although his library saw far more reconsideration cases than any other in the study, he expressed faith in the selection policy and reconsideration committee, and pointed out that the people that challenged materials represented only a small fraction of the community, which was still largely supportive of the library. Librarians with sparse LGBTQ collections also pointed to community values and support to explain the low numbers of LGBTQ materials. The library aide at Pullman High School lamented the complete lack of support for the libraries in her school, while the librarians at Bradbury and Hopkins High Schools mentioned that the strong conservative values in their small rural location created a rather hostile climate for LGBTQ-themed materials.

Minority Population. The statistical analyses performed earlier in the study showed a strong positive correlation between diverse student bodies and more LGBTQ-friendly library collections. The library demographics for the libraries studied in the third phase definitely supported that finding, as all of the schools who held strong LGBTQ collections had a significant minority populations (with one exception), while three of the four schools with sparse LGBTQ collections had almost no minority populations. Perhaps it stands to reason that schools with students representing a variety of ethnicities would need a variety of books; Ms. Montag of Bradbury School actually justified her lack of LGBTQ diversity in her collection due to her perception that there were no LGBTQ students in her school, pointing out that she also doesn't buy books with African American characters because they don't have any black students, either. Conversely, Mr. Bethune from Crutcher High School maintains a very diverse collection, even though his students are almost 90% white. He points out that not all students are the same as the majority, and that their needs are important, too. He also made the very valid point that sometimes we read beyond our own experiences, and that is how we learn.

Student Enrollment. The strongest statistical correlation during the second phase was the positive relationship between large student bodies and diverse LGBTQ collections. This link was strongly supported by the demographics of the schools studied in the third phase, as all eight schools with strong LGBTQ collections had at least 650 students, while three of the four schools with weaker LGBTQ collections had less than 330 students in the entire high school. Those three schools were also K-12 schools, meaning that their school was generally the only school in the district, and while there was some distinction between the elementary and high school collections, in many cases the district maintained a single library for all the students from kindergarten through twelfth grade. It is likely that the constant presence of younger students caused some trepidation on the part of the librarian when it came to purchasing books targeted only to more mature students.

Funding. Student poverty did not relate to either the presence or lack of LGBTQ-themed books in the library. However, the level of student poverty does not always reflect a school district's financial situation and how much money is available to fund the libraries. I was surprised to find that the level of funding that the librarians received in their annual budget did not seem to play a role in whether or not the librarian purchased LGBTQ-themed materials. I expected that librarians with limited funds might be so judicious with their meager dollars that they might avoid controversial titles, while libraries with money to spare would put those extra dollars to use creating a large, diverse, comprehensive collection. That expectation was not supported by the data gathered in the interviews. For instance, the most generously funded library (i.e., Bradbury High) employed a librarian who proudly censored the collection, while many sparsely funded school libraries (e.g. Rowling High, Garden High, Sanchez High) still managed to have some diversity in the shelves. And in some cases, libraries with diverse LGBTQ collections had generous budgets while their less diverse counterparts had very small ones. The development of a diverse collection seemed to have more to do with the librarian's commitment to building one and the community's decision to support the library than how much or how little funding the library receives.

Conceptualizing Censorship

Each of the 12 librarians had a slightly different philosophy about collection. Some librarians were extremely liberal and felt that students should have access to pretty much anything. The school librarians at Green and Garden High schools willingly purchased pretty much anything their students asked for, with full confidence in both their community and administrative support and belief in their selection policies. The former public librarian who oversaw the library at Myracle High school based her beliefs about student access to information on her experience in public libraries. Others admittedly walked the line, such as Kyla Meeks at Sanchez High School who expressed concern with books containing profanity and other potentially objectionable material, and promoted use of the public library to avoid purchasing such books herself. However, the Sanchez library still owned a great deal of controversial books, which suggests that Ms. Meeks still has a pretty liberal professional opinion about what is age appropriate for high school students. Ms. Streit is aware of her tendency to censorship in the Hopkins High library, which she attributes partly to having to ask individually for everything she wants to buy, but primarily due to her principal's specific command that she not purchase LGBTQ materials for the library. Ms. Streit does recognize that her unwillingness to purchase the items amounts to censorship, but justifies her actions as necessary to keep her job, believing that from within can she slowly effect change. Finally, Ms. Montag shared the most conservative perspective toward collection development, believing that only materials that reflected her community's very limited experience were appropriate for her students. She considered her choices to be part of selection, and not that of censorship.

Discussion

The first phase of the study revealed a wide disparity between library collections. While some school libraries were clearly providing a significant number of materials with LGBTQ content, the majority of the libraries possessed very few, and some none at all. This phase produced quantitative evidence that the bulk of Ohio librarians who are purchasing controversial books are still hesitant to buy materials with LGBTQ content.

Of the factors that were found during the second phase to correlate with increased inclusion of LGBTQ materials (large student population, diverse student population, liberal communities and library staff certified in library media), school districts have control over only one factor: a certified librarian. However, the qualitative data showed that even in small, rural, conservative districts, a certified librarian is sometimes the only apparent difference between a well balanced, diverse library collection and a limited one. Certified librarians are educated and grounded in principles of equity, inclusion and access, and those principles are critical for collection development.

Nevertheless, all is not lost for school libraries located in small buildings, conservative areas or racially homogenous communities. It became apparent from both the quantitative and qualitative data that the influence of the "community" strongly impacted the library program in both positive and negative ways. However, it is truly the perception of the librarian as to the extent of community support for diverse and inclusive collections that seemed to make the difference. While it may be easy to feel confident about selection choices when you teach in a largely liberal, privileged community, as Ms. Spiegelman and Mr. Kenyon did, Mr. Bethune of Crutcher High School also showed that even in a conservative county and in a community where some tend to challenge library holdings frequently, a strong belief in the ethics of librarianship, including equity of access, as well as a strong selection policy, can bolster a librarian's courage to continue to provide diverse materials for all students. Mr. Bethune reminded me during our conversation that the vocal minority who object to library materials do not speak for the entire community. It is important for school leaders to be aware of the voices within the community, and to make sure that supportive community members are also heard.

Finally, 12 librarians shared their experiences and philosophies regarding selection and censorship. Their narratives revealed a variety of interpretations regarding selection, age- appropriateness, censorship and access. The most salient variable with the most impact on selection across the 12 librarians seemed to be administrative and community support; those who felt supported seemed more comfortable with potentially controversial materials than did those who felt stifled. All were very aware of their community's collective values, although some were willing to challenge the status quo, either overtly or subversively, while others acquiesced. The existence and use of a quality district selection and reconsideration policy was also found to be important. While more restrictive librarians seemed to view it as a mandate to exclude certain titles, the librarians with more inclusive collections tended to regard the selection policy as a safeguard against censorship, knowing that a procedure is in place to protect controversial books from arbitrary removal.

Limitations

As with all studies, this project does have some limitations. The instrument, two carefully selected lists of titles, is limited. Whenever choices are made to include some books and exclude others, some quality titles will be overlooked. It is possible that libraries could have LGBTQ-friendly holdings that are not acknowledged by this study. Also, these lists measure Young Adult fiction only. School libraries with quality LGBTQ nonfiction resources would not be recognized. Although there was an attempt to control for other variables, including budget issues, it is possible that the controls did not effectively neutralize them.

Conclusion

This study has uncovered fascinating and useful knowledge about how librarians are choosing whether or not to include LGBTQ materials in school libraries and the reasons behind their decisions. By unveiling discriminatory attitudes and practices, as well as highlighting inclusive and beneficial collection development, librarians may be able to effect meaningful change

Librarians and other school leaders committed to creating a more inclusive and inviting environment for all of their students, but especially for those who identify as LGBTQ, would serve their students well to develop a diverse school library collection that includes high quality materials with LGBTQ content. This study has shown that the presence of a school librarian certified in library media, paired with a supportive administration, encourages such welcoming library collections and environments. It has also shown the impact of the local community on library collections, and how the librarian's perception of the political and ideological bent of the local area can color selection decisions. School leaders, working within the community, should make certain that positive, supportive voices are heard, while librarians must ensure that the entire school community can be served by the library collection. School- and community-based support groups, such as GSA clubs, also help LGBTQ students feel valued and supported. The literature shows that supportive school environments discourage bullying and encourage students to become allies, and that LGBTQ-themed literature can be a lifesaver to young LGBTQ students.

Young adult book selection can be a daunting task for high school librarians, and there are a number of complicated factors that impact how diverse the YA collection can be. Realities vary for every school librarian, as do local community priorities and individual librarian philosophies. This study reveals the overwhelming need to improve Ohio's school library collections. There are already some schools with outstanding library collections and library professionals in place. Now is the time for school leaders to emulate those schools and commit to a strong and inclusive library media collection.

References

References

Alexander, L. B., & Miselis, S. D. (2007). Barriers to LGBTQ collection development and strategies for overcoming them. Young Adult Library Services, 5(3), 43-49.

Bellows, N. K. (2005). Measuring self-censorship in school media centers. Unpublished Master's thesis, University of North Carolina, Chapel Hill.

Burke, S. K. (2008). Removal of gay-themed materials from public libraries: Public opinion trends, 1973-2006. Public Library Quarterly, 27(3), 247-264.

Clyde, L. A., & Lobban, M. (2001). A door half open: Young people''s access to fiction related to homosexuality. School Libraries Worldwide, 7(2), 17-30.

Coley, K. (2002). Moving towards a method to test for self-censorship by school library media specialists. School Library Media Research 5.

Cook, J. C. (2004). LGBTQ teen literature: Is it out there in Indiana? Indiana Libraries. 23(2) 25-28.

Freedman, L. & Johnson, H. (2001) Who's protecting whom? I hadn't meant to tell you this, a case in point in confronting self-censorship in the choice of young adult literature. Journal of Adolescent & Adult Literacy. 44( 3) 56-69.

Jennings, K. (2006). Librarians make a difference. Knowledge Quest, 34(5), 22-22.

Keeves, J. P., & Alagumalai, S. (2005). Applied Rasch measurement: A book of exemplars: papers in honour of John P. Keeves. (Springer e-books.) Dordrect: Springer.

Kosciw, J. G., Greytak, E. A., Palmer, N. A., & Boesen, M. J. (2014). The 2013 National School Climate Survey: The experiences of lesbian, gay, bisexual and transgender youth in our nation's schools. New York: GLSEN.

Limited shelf life. (2010). School Library Journal, 56(7), 15-15.

Manfredi, A. (2009). Accept the universal freak show. Young Adult Library Services, 7(4), 26-31.

Martin, H. J. (2006). A library outing: Serving queer and questioning teens. Young Adult Library Services, 4(4), 38-39.

Maycock, A. (2011). Issues and trends in intellectual freedom for teacher librarians: Where we've come from and where we're heading. Teacher Librarian 39(1) 8-12.

Moody, K. (2004) Censorship by Queensland public librarians: Philosophy and practice. Australian Public Libraries and Information Services. 17(1) 168-85.

Quinn, J. (2015). Social Conservatism - Social Conservatives. Retrieved from http://usconservatives.about.com/od/typesofconservatives/a/SocialCons.htm

Rauch, E. W. (2011). GLBTQ collections are for every library serving teens! Teacher Librarian 39 (1) 13-16.

Sanelli, M. (2001). ''I could be anybody'': Gay, lesbian, and bisexual students in U.S. schools. NASSP Bulletin, 85(622), 69.

Toomey, R. B., McGuire, J. K., & Russell, S. T., (2012). Heteronormativity, school climates, and perceived safety for gender nonconforming peers. Journal of Adolescence, 35187-196.

Whelan, D. L. (2006). OUT and ignored. School Library Journal, 52(1), 46-50.

Whelan, D. L. (2009a). A dirty little secret. (cover story). School Library Journal, 55(2), 26-30.

Whelan, D. L. (2009b). SLJ self-censorship survey. School Library Journal. Retrieved from http://www.schoollibraryjournal.com/article/CA6633729.

Appendix

(ProQuest: Appendix omitted.)