(PDF) How School Libraries Improve Outcomes for Children: The New Mexico Study

How School Libraries

Improve Outcomes for Children

The New Mexico Study

Keith Curry Lance

Marcia J. Rodney

Christine Hamilton-Pennell

New Mexico State Library

Santa Fe, New Mexico

June, 2002

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table of Contents

How School Libraries ...........................................................................1

Improve Outcomes for Children.............................................................1

Table of Contents.................................................................................i

List of Tables....................................................................................ivv

Acknowledgements............................................................................vvi

Executive Summary......................................................................... vivii

School Library Development ......................................................... viivii

School & Community Differences................................................... viivii

School Librarians & Strong School Libraries....................................viiiviii

Introduction .......................................................................................1

Review of the Literature.......................................................................2

Presence of a Library Media Center with a Professional Library Media

Specialist ........................................................................................3

Learning and Teaching......................................................................4

Early Studies ................................................................................4

Instructional Role Since Information Power Guidelines ........................5

Gap Between Theory and Practice....................................................6

Level of LMS Collaboration with Teachers..........................................6

Barriers to Collaboration.................................................................7

Positive Effects of Flexible Scheduling...............................................8

Positive Effects of Collaboration .......................................................9

Positive Effects of Technology..........................................................9

Characteristics of Library Media Specialist ....................................... 11

Other Aspects of the Instructional Role........................................... 12

Information Access and Delivery....................................................... 12

Access to Print Resources ............................................................. 12

School-Public Library Relationship.................................................. 14

School Library and Equity Issues.................................................... 14

Size of Library Media Center Collection ........................................... 16

Frequency of School Library Use .................................................... 17

Role of Library Media Specialist in Program Development .................. 18

Role of Technology in Student Academic Achievement ...................... 19

Integrating Technology into the Library Media Center ....................... 20

Program Administration................................................................... 21

Maximizing Predictors of Student Achievement ................................ 22

Principal Support of the Library Media Program................................ 22

Collaborative Activities of the Library Media Specialist....................... 23

Effect of Library Media Center Staffing............................................ 24

Library Media Specialist as Provider of In-Service Training................. 25

Library Media Specialist Planning and Management Skills .................. 25

i

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Library Media Specialist Budgetary Role.......................................... 26

Summary...................................................................................... 27

Methodologies .................................................................................. 28

Sample......................................................................................... 28

Survey.......................................................................................... 28

Respondent Information ............................................................... 28

Service Hours ............................................................................. 29

School Library Staff ..................................................................... 29

Paid Staff Activities...................................................................... 29

Computers with Access to Library Resources ................................... 30

School Library Usage.................................................................... 30

School Library Collection............................................................... 30

Annual Operating Expenditures for the School Library....................... 31

Available Data................................................................................ 31

Statistical Significance .................................................................... 32

Bivariate Correlation....................................................................... 33

Factor Analysis............................................................................... 34

Multiple Regression Analysis ............................................................ 34

Findings........................................................................................... 37

Elementary School Level ................................................................. 37

School Library Predictors of Achievement Scores.............................. 37

Information Technology................................................................ 38

Information Resources ................................................................. 39

Middle School Level ........................................................................ 39

School Library Predictors of Achievement Scores.............................. 39

Print Volumes ............................................................................. 40

Database Computers.................................................................... 41

High School Level........................................................................... 41

School Library Predictors of Reading Scores .................................... 41

Library Visits............................................................................... 42

Information Technology................................................................ 43

Print Volumes ............................................................................. 44

Information Resources ................................................................. 45

Controlling for Other School and Community Factors........................... 45

Correlation Analysis for School Library Development Variables........... 46

Factor Analysis for School Library Development Variables.................. 47

Correlation Analysis for Other School and Community Variables ......... 48

Factor Analysis for Other School Variables....................................... 49

Factor Analysis of Community Socio-Economic Status (SES) Variables. 50

Regression Analysis for School Library Development, School, and

Community SES Factors ............................................................... 50

Comparison of Highest & Lowest Scoring Schools................................ 51

ii

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Conclusions...................................................................................... 59

School Library Development ............................................................ 59

School & Community Differences...................................................... 59

School Librarians & Strong School Libraries........................................ 60

Recommendations for Action .............................................................. 63

Appendices ...................................................................................... 64

Bibliography..................................................................................... 65

Participating New Mexico Elementary Schools........................................ 75

Participating New Mexico Middle Schools .............................................. 80

Participating New Mexico High Schools................................................. 85

iii

How School Libraries Improve Outcomes for Children:

The New Mexico Study

List of Tables

Table 1. Comparison of the Study Sample and the Universe of All New

Mexico Public Schools Serving Grades 4, 8 and 10, 2000................... 28

Table 2. Direct School Library Predictors of Fourth Grade Achievement

Scores in New Mexico Elementary Schools, 2000.............................. 38

Table 3. Indirect School Library Predictors of Fourth Grade Achievement

Scores via Information Technology in New Mexico Elementary School

Libraries, 2000............................................................................ 38

Table 4. Indirect School Library Predictors of Fourth Grade Achievement

Scores via Information Resources in New Mexico Elementary School

Libraries, 2000............................................................................ 39

Table 5. Direct School Library Predictors of Eighth Grade Achievement

Scores in New Mexico Middle Schools, 2000 .................................... 40

Table 6. Indirect School Library Predictors of Eighth Grade Achievement

Scores via Print Volumes in New Mexico Middle School Libraries, 2000 40

Table 7. Indirect School Library Predictors of Eighth Grade Achievement

Scores via Database Computers in New Mexico Middle School Libraries,

2000.......................................................................................... 41

Table 8. School Library Predictors of Tenth Grade Achievement Scores in

New Mexico High Schools, 2000..................................................... 42

Table 9. Indirect School Library Predictors of Tenth Grade Achievement

Scores via Visits to New Mexico High School Libraries, 2000 .............. 43

Table 10. Indirect School Library Predictors of Tenth Grade Achievement

Scores via Information Technology in New Mexico High School Libraries,

2000.......................................................................................... 44

Table 11. Indirect School Library Predictors of Tenth Grade Achievement

Scores via Print Volumes Held by New Mexico High School Libraries,

2000.......................................................................................... 44

Table 12. Indirect School Library Predictors of Tenth Grade Achievement

Scores via Information Resources in New Mexico High School Libraries,

2000.......................................................................................... 45

Table 13. Correlation Matrix for School Library Development Variables and

District Expenditures per Student for New Mexico High Schools, 2000. 47

Table 14. Factor Analysis of School Library Development Variables for New

Mexico High Schools, 2000............................................................ 48

Table 15. Correlation Matrix for School and Community Variables for New

Mexico High Schools, 2000............................................................ 49

Table 16. Factor Analysis of Other School Variables for New Mexico High

Schools, 2000 ............................................................................. 49

Table 17. Factor Analysis of Socio-Economic Status Variables for New

Mexico High Schools, 2000............................................................ 50

iv

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 18. Regression Analysis of School Library Development, School, and

Community SES Factors with Achievement Scores for New Mexico High

Schools, 2000 ............................................................................. 51

Table 19. Selected Library Media Statistics for New Mexico Elementary

Schools Scoring Highest and Lowest on the Fourth Grade Achievement

Test, 2000.................................................................................. 53

Table 20 . Selected Library Media Statistics for New Mexico Elementary

Schools Scoring Highest and Lowest on the Fourth Grade Achievement

Test, 2000.................................................................................. 54

Table 21. Selected Library Media Statistics for New Mexico Middle Schools

Scoring Highest and Lowest on the Eighth Grade Achievement Test,

2000.......................................................................................... 55

Table 22. Selected Library Media Statistics for New Mexico Middle Schools

Scoring Highest and Lowest on the Eighth Grade Achievement Test,

2000.......................................................................................... 56

Table 23. Selected Library Media Statistics for New Mexico High Schools

Scoring Highest and Lowest on the Tenth Grade Achievement Test, 2000

................................................................................................. 57

Table 24. Selected Library Media Statistics for New Mexico High Schools

Scoring Highest and Lowest on the Tenth Grade Achievement Test, 2000

................................................................................................. 58

v

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Acknowledgements

How School Libraries Improve Outcomes for Children: The New

Mexico Study was a research project directed by the New Mexico State

Library to collect information on school libraries and their relationship to

student achievement, specifically in the area of reading achievement in

grades 4, 8, and 10.

The State Library would like to thank and credit the staff of the participating

schools who made the time in their busy schedules to collect, compile, and

report data required to complete the 2001 statewide school library study.

Without their efforts, this study would not have been possible.

Very special thanks are due to the McCune Foundation for underwriting half

of the costs of this research.

Special thanks to Bill Schwarz, who did yeoman work during both the

preparation and data processing stages of this study. He assisted in

matching data across many files, mining community demographic data, and

in running the many SPSS statistical procedures that yielded the

accompanying tables.

Thanks are due to the New Mexico State Department of Education for

providing the substantial amounts of data required related to the state’s

public schools.

The review of the literature contained here updates the comprehensive

review done for the original Colorado study. Indeed, it was produced by one

of that review’s co-authors, Christine Hamilton-Pennell of Mosaic Knowledge

Works. In addition to updating her earlier effort with Lynda Welborn, she

also did an excellent job of relating previous research on this topic to the

themes of Information Power. This focus improves the organization and

readability of the literature review and makes it more relevant to the current

context of library media development.

Keith Curry Lance Marcia J. Rodney

Westminster, Colorado Louisville, Colorado

vi

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Executive Summary

New Mexico achievement test scores rise with the development of school

library programs. The relationship between school library development and

test scores is not explained away by other school or community conditions at

the high school level. There was insufficient variation in librarian staffing to

make similar claims for the elementary and middle school levels.

School Library Development

New Mexico achievement test scores tend to rise with increases in:

• school librarian and total library staff hours per 100 students;

• print volumes per student;

• periodical subscriptions, video materials, and software packages per 100

students; and

• school library expenditures per student.

Whatever the current level of development of the school library program,

these findings indicate that incremental improvements in its staffing,

collections, and funding will yield incremental increases in reading scores.

The only caveat is that school library spending cannot exert a positive

influence on academic achievement if it comes at the expense of other

school programs.

School & Community Differences

The impact of school library development on academic achievement at the

high school level cannot be explained away by:

• school differences, including:

• the percentage of classroom teachers with master’s degrees,

• teachers’ average years of experience, and

• the teacher/pupil ratio, or

• community differences, including:

• the percentage of schoolchildren living in poverty,

• the percentage of schoolchildren belonging to racial/ethnic minority

groups, and

• the percentage of adults who graduated from high school.

vii

How School Libraries Improve Outcomes for Children:

The New Mexico Study

When these other conditions are taken into account, school library

development alone accounts for 7.9 percent of variation in average

achievement scores among high schools. This figure takes into account

community socio-economic status (SES), which explains 42.0 percent of

variation in high school test scores. It also considers other school conditions

that explain no additional variation beyond the 50.0 percent attributable to

the combination of community SES and school library development factors.

Similar conclusions could not be drawn from such analyses at elementary

and middle school levels, due to a lack of variation in school library staffing

at those school levels. (Most elementary schools lacked a full-time librarian;

most middle schools had precisely one full-time librarian.)

School Librarians & Strong School Libraries

School librarians exert a complex web of effects on the school library

programs. Findings about these effects are summed up in the following

description of a strong school library.

A strong school library program is one

• that is adequately staffed, stocked and funded.

Minimally, this means

one full-time librarian and one full-time aide. The relationship, however,

is incremental; as the staffing, collections, and funding of school library

programs grow, reading scores rise.

• whose staff are actively involved leaders in their school’s teaching and

learning enterprise. A successful school librarian is one who has the ear

and support of the principal, serves with other teachers on the school’s

standards and curriculum committees, and holds regular meetings of the

library staff. Students succeed where the school librarian participates

with classroom teachers and administrators in making management

decisions that encourage higher levels of achievement by every student.

• whose staff have collegial, collaborative relationships with classroom

teachers. A successful school librarian is one who works with a classroom

teacher to identify materials that best support and enrich an instructional

unit, is a teacher of essential information literacy skills to students, and,

indeed, is a provider of in-service training opportunities to classroom

teachers. Students succeed where the school librarian is a consultant to,

a colleague with, and a teacher of other teachers.

viii

How School Libraries Improve Outcomes for Children:

The New Mexico Study

• that embraces networked information technology.

The school library of

today is no longer a destination; it is a point of departure for accessing

the information resources that are the essential raw material of teaching

and learning. Computers in classrooms, labs and other school locations

provide networked access to information resources—the library catalog,

electronic full text, licensed databases, locally mounted databases, and

the Internet. Students succeed where the school library program is not a

place to go, apart from other sites of learning in the school, but rather an

integral part of the educational enterprise that reaches out to students

and teachers where they are.

ix

How School Libraries Improve Outcomes for Children:

The New Mexico Study

10

The Impact of New Mexico School Libraries on Academic Achievement

Summary of Relationships

Achievement

Scores

Grade 4 /8 / 10

School Library

Visits

• Group visits

• Group visits for

information skills

instruction

• Individual visits

Information Resources

• Print volumes

• Periodical subscriptions

• Electronic references

• Video materials

• Software packages

• Library expenditures

Information

Technology

• Library computers

• School computers

• Catalog computers

• Licensed database computers

• Internet computers

Library Staff

• Librarian

• Total staff

Staff Activities

• Teaching cooperatively

• Providing in-service training

• Providing reading incentives

• Developing collections

• Managing computer network

• Attending faculty meetings

• Attending library staff

meetings

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Introduction

For many years, the declining condition of libraries in New Mexico public

schools has been a cause for concern to the State Library. As part of a

campaign to reverse this alarming trend, the State Librarian and others

resolved to document the impact of school libraries and librarians in New

Mexico schools and to share this information with school decision-makers—

school boards, superintendents, principals, teachers, even school librarians.

In 1993, the Library Research Service of the Colorado State Library

published a landmark study, The Impact of School Library Media

Centers on Academic Achievement. Between 1999 and 2001, successor

studies were completed in Alaska, Colorado, Iowa, Massachusetts, Oregon,

Pennsylvania, and Texas. Realizing the potential for such a study in New

Mexico, the State Library designate funds from the Library Services and

Technology Act (LSTA) for this purpose.

In 2001, the State Library contracted for this study with the team of

researchers responsible for the most of the other state studies: Keith Curry

Lance, the principal investigator for this study as well as the original

“Colorado study;” Marcia J. Rodney, a library research consultant; and

Christine Hamilton-Pennell, consulting librarian and also a co-author of the

original Colorado study.

In addition to confirming in New Mexico the findings of the first Colorado

study, this project also sought to explore several issues that were explored

in the other state studies. Those issues included:

• identifying characteristics of school librarians and school library programs

that affect academic achievement,

• assessing the contribution of collaboration between teachers and school

librarians to the effectiveness of school library programs, and

• examining the growing role of information technology in school libraries,

particularly licensed databases and the Internet.

On all three of those counts, this New Mexico study was decidedly

successful. This document reports comprehensively on the project, putting

it into perspective with past research as well as the American Association of

School Librarians’ new standards, Information Power: Building

Partnerships for Learning (1998). It also contains reports of the findings

in a variety of readily useful formats, including: an executive summary, a

brochure, and the preliminary report.

1

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Review of the Literature

This study aims to replicate and expand upon previous research showing a

link between student academic achievement and the school library media

program. With the move to standards-based education, which focuses on

what students have learned

(proficiencies or outcomes) rather than what is

taught (coverage of content), the school library media specialist is in a

unique position to help students develop the information literacy skills which

will enable them to achieve standards.

The new edition of Information Power: Building Partnerships for

Learning (ALA, 1998), reflects a change in emphasis for school library

media programs, from providing resources to students to creating a

community of lifelong learners. Three overlapping roles are identified for

school library media specialists (LMSs) in this document. The learning and

teaching role supports the instructional goals of the school in both content

(standards and subject curriculum) and process (information literacy skills).

The information access and delivery

role encompasses the more traditional

responsibilities of the LMS, those of developing the media center's collection

and services and providing access to them. A third role, program

administration, includes both management of the library media program and

larger training and advocacy functions within the school community.

This review of the research organizes the research findings under the three

roles identified for the LMS in Information Power (1998). Many of the

research studies were conducted in the context of the earlier guidelines,

Information Power: Guidelines for School Library Media Programs

(ALA, 1988). Although some of the goals in the document have changed,

the underlying mission statement remains the same:

The mission of the library media program is to ensure that students

and staff are effective users of ideas and information. This mission is

accomplished:

• by providing intellectual and physical access to materials in all formats

• by providing instruction to foster competence and stimulate interest in

reading, viewing, and using information and ideas

• by working with other educators to design learning strategies to meet the

needs of individual students. (ALA, 1998, p. 1)

2

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Presence of a Library Media Center with a Professional Library Media

Specialist

Many studies conducted before the advent of the Information Power

Guidelines dealt with the value of the mere presence of a library with a

professional librarian, reflecting the lack of centralized library service,

particularly at the elementary level. Willson (1965) showed that students

demonstrated superior gains on the Iowa Test of Basic Skills (ITBS) in

elementary schools with a centralized library and a professional librarian.

Likewise, Becker (1970) compared ITBS scores between students in

elementary schools with and without libraries and found that the presence of

a library and the guidance and function of a librarian appeared to exert

significant influence on pupil achievement in some information-gathering

skills areas.

In the study by Hale (1969), SAT scores improved among students receiving

library service from a professional. McMillen (1965) found that students in

schools with good libraries and full-time librarians performed at higher levels

in reading comprehension and in knowledge and use of reference materials

than students in schools with minimal or no library service. Didier (1982)

confirmed that student achievement in reading, study skills and use of

newspapers was significantly greater at the seventh grade level in schools

with professional library media personnel as compared to schools without

them. Student access to the library media center was also significantly

greater in schools with professional library media personnel than in schools

without them.

Yarling (1968) found that the addition of a well-equipped and managed

centralized library had a significant impact on the performance of elementary

school students in library-related skills, particularly outlining and note

taking. Students who used a new fully staffed and equipped elementary

school library also showed significant improvement in library skills test

scores in the study by Ainsworth (1969). McConnaha (1972) found that the

library skills test scores of high school students who had attended an

elementary school with both a library and a librarian who conducted a strong

library skills program were significantly higher than those of students who

did not have these advantages.

Recent statistics show that public schools have moved in the direction of

providing professionally-staffed library media programs at all levels. In a

U.S. Department of Education-funded study conducted recently (Michie and

Cheney, 2000), survey results indicate that as of 1997, 98 percent of public

3

How School Libraries Improve Outcomes for Children:

The New Mexico Study

schools have a school library media center, and 78 percent of public schools

have a state-certified library media specialist.

Learning and Teaching

Early Studies

Some research studies before the advent of Information Power

Guidelines in 1988 referred to various aspects of the LMS's teaching role.

Aaron (1975) studied a group of eighth grade students who participated in a

program in which a full-time media specialist was added to the teaching

team. In addition to showing significant improvement in language arts,

spelling, and math computation, the students in the experimental group

experienced improvement in their self-concept. Bailey (1970) studied a

group of disadvantaged first-grade students who participated in a library

resource program over a 12-week period. The experimental group showed a

significant increase in total language ability and the ability to express ideas

over the control group of disadvantaged students who received no special

treatment. DeBlauw (1973) examined the rate of cognitive growth of

students on achievement test batteries before and after implementation of a

multi-media program. Elementary students showed significant gains, but

the academic performance of high school students was unchanged by the

program. A longer-term study of twelfth grade English students by Gilliland

(1986) found that test scores on the study-locational portion of the

California Assessment Program improved during the years following

implementation of a library review program.

Gengler (1965) looked at differences in the ability to apply selected problem

solving skills between sixth grade students who were instructed by a

classroom teacher and those who received additional instruction from an

elementary school librarian. Findings showed that the mean score on a

problem solving skills examination for the librarian-teacher instructed group

was significantly higher than for the teacher instructed group. Hastings and

Tanner (1963) looked at whether improved English language skills could be

developed at the tenth-grade level through systematic library experiences

rather than the traditional emphasis on formal English grammar. The group

that eliminated all traditional emphasis on formal grammar and spelling and

instead received systematic work in the use of library references was

significantly superior to the groups that emphasized traditional work in

grammar and spelling.

4

How School Libraries Improve Outcomes for Children:

The New Mexico Study

In a study by Hutchinson (1982), English teachers gave tenth-grade

students special library skills instruction and practice over a two-week

period. Library usage among the students increased regardless of their

academic grade point averages. Hale (1970) found that an experimental

group of twelfth grade students who were given a variety of library services

and resources and the opportunity to work independently under the

supervision of the librarian showed "remarkable enthusiasm" for learning.

Barrilleaux (1965) focused on a comparison of the achievement of junior

high school students in general science classes in which textbooks were used

with students who used reference materials in the school library rather than

a textbook. Results showed that for all investigated educational outcomes,

the use of library materials without a basic textbook was the superior

method of instruction.

Instructional Role Since Information Power Guidelines

Much of the research taking place after the introduction of the Information

Power Guidelines in 1988 focuses on the instructional role of the school

library media specialist. Lance, Welborn and Hamilton-Pennell (1993) found

that students whose library media specialists played an instructional role,

either by identifying materials to be used with teacher-planned instructional

units or by collaborating with teachers in planning instructional units, tended

to achieve higher reading scores. A study by the Library Research Service in

Colorado (1998a) also found that students earned higher reading scores in

schools where the LMS played a vital instructional role, including planning

instruction with teachers, providing information literacy instruction,

providing in-service training for teachers, and evaluating students' work.

Five recent statewide studies reinforce the importance of the LMS

instructional role. A study conducted in Alaska (Lance, Hamilton-Pennell and

Rodney, 2000) revealed that students' test scores tended to rise when

library staff spent more time teaching information literacy to students and

planning instructional units with teachers. In Pennsylvania (Lance, Rodney

and Hamilton-Pennell, 2000a), test scores increased as LMSs spent more

time teaching cooperatively with classroom teachers and integrating

information literacy skills into the school's approach to standards and

curriculum. A new study in Colorado (Lance, Rodney and Hamilton-Pennell,

2000b) concluded that reading test scores rise as LMSs plan cooperatively

with teachers (at the 7th grade level), identify materials for teachers, and

teach information literacy skills to students. A Massachusetts study by

Baughman (2000) found that both elementary and middle school students

5

How School Libraries Improve Outcomes for Children:

The New Mexico Study

tended to score higher on the Massachusetts Comprehensive Assessment

System (MCAS) test when there was a library instruction program. Likewise,

in Oregon (Lance, Rodney, and Hamilton-Pennell, 2001), students tended to

score higher on reading tests when their LMSs worked with classroom

teachers to identify materials to support and enrich instructional units,

taught essential information literacy skills to students, and provided in-

service training opportunities to classroom teachers.

Gap Between Theory and Practice

Nevertheless, several researchers have identified a gap between theory and

practice. Person (1993) found a discrepancy between the real and ideal role

perceptions of LMSs. While they were aware of the roles identified in the

Guidelines, they didn't perform them as often as they would have liked.

Pickard (1993) also studied the gap between theory and practice of LMSs

performing the instructional role and found that less than 10 percent of her

sample appeared to be practicing the role to a great extent. Angelo (1994)

verified this finding in a study which showed that the majority of LMSs were

performing duties of the traditional librarian, such as student orientation and

assisting students and teachers in finding materials, while a low percentage

were performing planning and consultation roles. Kuhne (1993) concluded

that school libraries need to be more integrated into the curriculum and that

the school librarian could play a much more distinctive teacher role than he

or she does today.

McCarthy (1997) studied LMSs who were "well above average" in the New

England region and found that 58 percent of them believed that

implementation of Information Power Guidelines was only somewhat

realizable or not realizable at all. McCracken (2000), in a national survey of

more than 500 LMSs, found that they perceived all the roles in Information

Power (1988, 1998) to be more important than they were able to

implement in practice. They also perceived that they practiced the role of

information specialist (i.e., the traditional role of the librarian) to a greater

extent than that of program administrator, teacher, or instructional partner

and consultant.

Level of LMS Collaboration with Teachers

Instructional collaboration has many levels, and research indicates that LMSs

are most often involved in the less complex levels. Slygh (2000) reports on

6

How School Libraries Improve Outcomes for Children:

The New Mexico Study

the difficulty of establishing exactly what is meant by the term

“collaboration.” She uses a definition by Van Deusen and Tallman (1994)

that delineates an interdependent relationship between LMS and classroom

teacher involving a continuum of five progressively complex levels of

instructional collaboration, from gathering materials to collaborating with a

teacher to evaluate a unit. Slygh found that teachers in her sample of

Library Power schools indicated a greater frequency of collaborating with the

LMS in planning and designing instruction than in delivering it.

A study by DeGroff (1997) compared the ideal roles of the LMS as laid out in

Information Power (1988) with the actual practice of these roles. She

determined that the instructional consultant activities of LMSs were usually

limited to gathering books for instructional units and seldom involved

participating with teachers in developing, carrying out, or assessing unit

plans. Van Deusen and Tallman (1994) found that more than half of their

sample of LMSs did not assess student work at all during the study period.

Michie and Chaney (2000) found that the overall percentage of library media

centers working with teachers on curriculum development, collaboratively

teaching curriculum units with classroom teachers or collaboratively

evaluating curriculum units with classroom teachers ranged from two

percent to 21 percent, depending on the subjects taught. The greatest

amount of collaboration was with reading or English teachers.

Mosqueda (1999) studied the roles played by LMSs in 67 Florida schools

named as National Blue Ribbon Schools. While the overwhelming majority

of responding principals and LMSs agreed that the library media programs

performed well in program administration and information access and

delivery, an average of 75 percent of principals and LMSs thought the library

media programs needed to be more fully integrated into the curriculum.

Mosqueda's data does show a higher percentage of LMSs who spend time

planning instruction with teachers on a daily (40 percent) or weekly (80

percent) basis than other research studies cited, which may indicate that

LMSs in exemplary schools perform the instructional role more often than

their counterparts in less well-recognized schools.

Barriers to Collaboration

Barriers to LMSs practicing the instructional role include the attitudes of both

LMSs and teachers, as well as program limitations such as fixed scheduling,

limited resources, and lack of technology. Lai (1995) found no significant

differences between teachers' and media specialists' attitudes regarding the

LMS's role in curriculum development, instructional development and

7

How School Libraries Improve Outcomes for Children:

The New Mexico Study

technology use. Both groups believed the LMS had only a marginal role in

designing and producing materials for units. Beaird (1999) found that key

barriers to collaboration were time, desire to maintain the status quo, and

lack of resources. In a national survey of LMSs by McCracken (2000), the

biggest barriers cited to expanding their instructional role were lack of time

and resources, specifically funding and clerical help. Several researchers

(e.g., Slygh, 2000; Beaird, 1999; DeGroff, 1997) point to role confusion or

role conflict of LMSs themselves, as well as lack of understanding of the

instructional role of the LMS among the teachers and administrators they

work with, as inhibiting expanded collaboration activities.

Positive Effects of Flexible Scheduling

Giorgis (1994) discovered that the majority of elementary school teachers

perceived the LMS as a resource person and only a few as a collaborator.

Nevertheless, during the course of her study Giorgis found that these views

changed. Through flexible scheduling and collaborative planning, the

perceptions of the LMS's role by classroom teachers, administrators, children

and parents became one of teacher and collaborator.

Other researchers also attest to the positive effects of moving to flexible

scheduling. Bishop (1992) found that the most significant changes in the

role of the LMS occurred with the move to flexible scheduling and

curriculum-integrated instruction. Fedora (1993) compared two exemplary

school library media centers, one with fixed and one with flexible scheduling

and found that the LMS participated more often in planning with teachers

and as an instructional consultant in the flexibly-scheduled program. Van

Deusen (1993) and van Deusen and Tallman (1994) found that LMSs in

schools that combined both flexible scheduling and team planning had

significantly more curriculum involvement. Hughes (1998), in her study of

four library media programs in schools implementing whole language

programs, found that all four LMSs moved to flexible scheduling in the LMC

during the course of implementation and were able to provide leadership and

direction in creating a library program that promoted the active construction

of knowledge. Likewise, Beaird (1999) reported that one of the major

enhancers to increased collaboration was flexible scheduling, and McCracken

(2000) found that LMSs who use flexible scheduling perceived that they

implemented the role of instructional consultant more than those who used

fixed or combination scheduling.

8

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Positive Effects of Collaboration

Research indicates that a collaborative environment begets more

collaboration. Beaird (1999) discovered that as a result of a first experience

with collaboration with the LMS, teachers were made aware of the presence

of another professional on their campuses who could engage in collaborative

planning and teaching. LMSs became more aware of teachers’ needs as they

provide for regular students as well as those with special needs. Straessle

(2000), in her case study of a suburban junior high school, concludes that

the more teachers and administrators understand and experience the role of

the LMS as an instructional consultant, the more likely their perceptions will

change and expectations increase, thus improving teacher instruction and

student learning. Slygh (2000) reports that LMSs’ perceptions of the degree

to which their school climate was a professional community affected the

amount of instructional collaboration they performed.

Positive Effects of Technology

Technology can also support the instructional role. Everhart (1992) found

that high school library media specialists with automated circulation systems

spent significantly more time in instructional development and use of

technology than those without automated systems, although the actual time

spent in development of the educational program was quite low. Van

Deusen (1996a) found that both flexible scheduling and library automation

were positively related to the LMS performing an instructional consultation

role, as well as providing electronic support for teachers using technology,

providing individual assistance to students, and reducing the amount of time

spent on clerical duties. Jones (1994) concluded that technology expands

the teacher-librarian partnership possibilities in literature-based instruction.

E. W. Smith (1998) studied the changes occurring in media centers in

Dekalb County, Georgia after the incorporation of technology, and found

both that the media center program was a viable component of the school's

curriculum and that the media specialist's role as teacher, information

consultant, and information specialist had expanded.

Characteristics of Library Media Specialist

Perhaps the most important factor in successfully implementing the

instructional role is the characteristics and skills of the school library media

9

How School Libraries Improve Outcomes for Children:

The New Mexico Study

specialist himself or herself. Yetter (1994) found that LMSs successfully

involved in resource-based learning were energetic, healthy and

enthusiastic; showed leadership abilities; had theoretical understanding of

resource-based learning; had the ability to translate theory into effective

instructional plans; and were knowledgeable about specific learning

resources. These LMSs saw teaching as their primary function. As a result,

the colleagues of these LMSs saw them as vital participants in the

instructional process. Esser (1999) interviewed 18 female LMSs in Kentucky

to discover their initial motivations for entering the field and discovered that

a desire for autonomy was high on the list. Her data suggested that these

teacher-librarians achieve a greater degree of autonomy when they are

collaborating with teachers, and they negotiate these relationships and

create effective networks in order to carry out their work.

Farwell (1998) found that LMSs could play a pivotal role in the successful

implementation of collaborative planning if they were knowledgeable about

the curriculum, the library collection, information literacy instruction, and

instructional design and delivery. They also needed to have good

interpersonal skills and be willing to act as change agents. Van Deusen's

(1996b) case study of a school library media specialist involved in an

instructional planning process showed that she contributed as a peer and

helped to clarify, initiate, summarize and test the discussion ideas. DeGroff

(1997) concludes from her survey data that the most important factors

supporting opportunities for teachers and LMSs to work together effectively

were the librarian’s knowledge, personality, and attitudes or interests.

K. G. Alexander (1992) studied four exemplary LMSs and found that they

fulfilled most of the aspects of the instructional role. They spent large

portions of each day giving instruction, effectively managed class and

teaching time, provided instruction related to the curriculum, and used

innovative instructional methods. They also instructed different sections of

the school community, ensured that their media center had resources to

support the changing curriculum, and assisted teachers in planning

classroom instruction. Gehlken (1994) studied the school library media

programs in three blue ribbon high schools and came to similar conclusions.

In all three schools, there was a cooperative relationship between the LMS

and the faculty, with opportunities for collaborative planning and integrating

information skills into the classroom curriculum. The students in all three

schools overwhelmingly indicated that the most important service provided

by the school library media program was help from the LMS in finding and

evaluating information. Bell and Totten (1992) found that teachers

employed in academically successful schools tended to choose the library

10

How School Libraries Improve Outcomes for Children:

The New Mexico Study

media specialist significantly more for cooperation on instructional problems

than did teachers serving in academically unsuccessful elementary schools.

Other Aspects of the Instructional Role

Another aspect of the LMS's learning and teaching role identified in

Information Power (1998) is to encourage and engage students in

reading, viewing, and listening for understanding and enjoyment. Yetter

(1994) found that LMSs involved in resource-based learning were

enthusiastic about reading and books. Lai (1995) found that teachers and

LMSs both strongly agreed that the LMS should work with teachers in

helping students to develop the habit of reading.

In Australia, Todd, Lamb and McNicholas (1993) studied Year-Seven and

Year-Eleven students and found that integrating information skills into

subject content, with collaboration by classroom teachers and LMSs, had a

positive impact on student learning, including better understanding of

subject content and improved test scores. Todd (1995) analyzed the impact

of integrated information skills instruction in a Year-Seven science class.

The two treatment classes recorded significantly higher annual science

scores than the control classes.

The adoption of state content standards and the movement towards

standards-based instruction and assessment is too recent to have a

substantial research base as yet. N. A. Alexander (1998) determined that

standards policy is generally associated with improved student performance,

although there are disturbing equity issues. In the school library field, a few

research studies to date have looked at the connection between integrating

information literacy skills into the curriculum and improved student learning.

Grover and Lakin (1998) reported on the development and testing of a

Kansas model which integrates information skills into planning and assessing

learning across the curriculum. Teachers and librarians who participated in

the study indicated that the model facilitated student learning in all grade

levels studied and for units of any length. The "integrated assignment"

stage of the model was reported as a key to enhancing student learning.

In regard to standards-based education, one of the more interesting

developments of the past decade has been the elaboration and proliferation

of the value-added assessment model originated in Tennessee by Sanders,

et. al. (1997). This model is being considered in Colorado and other states

as a method for evaluating the performance of both schools and teachers.

11

How School Libraries Improve Outcomes for Children:

The New Mexico Study

In the author’s most authoritative account to date, he reports that the

strongest predictor of year-to-year improvement in students’ test scores is

teacher quality. To date, however, there have been no further reports from

Sanders or others defining precisely what “teacher quality” means in

empirical terms. Decades of library media research findings indicate that

one major factor that has demonstrated consistently a positive, strong, and

statistically significant relationship to quality teaching is a close working

relationship between the classroom teacher and the library media specialist.

Information Access and Delivery

The information access and delivery role includes providing quality resources

and technology that support the curriculum, offering convenient and flexible

access to the media center's resources and services, and providing a

welcoming environment that is conducive to learning. Recent studies show

that LMSs still perceive this role to be their most important one, and the one

they practice most often in reality (DeGroff, 1997; McCracken, 2000).

Early studies focused on service level and collection size as predictors of

academic achievement. Greve (1974) discovered that the most valuable

predictor of student test scores was the number of volumes in the school

library. Thorne (1967) examined the reading comprehension and library

skills of students using the augmented services of a Knapp Project library

versus the nominal services of a second junior high school library in a two-

year study. Findings revealed a significant difference in the mean gains of

the experimental group over the control group in reading comprehension

and library skills.

Access to Print Resources

More recent studies have focused on the connection between students'

achievement in reading and access to print resources, particularly in

libraries. The first Colorado study by Lance, Welborn and Hamilton-Pennell

(1993) concluded that the size of a media center's staff and collection is the

best school predictor of academic achievement. In that study, academic

achievement was represented by reading scores, which were highly

correlated with scores in other areas, such as writing and research skills.

Elley (1994, 1996) compared the scores of students from 32 countries on

the 1992 International Association for the Evaluation of Educational

12

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Achievements (IEA) Reading Literacy Study with data on the home

environment and school and public libraries. He concluded that access to

print, and especially the size of the school library, was the strongest

predictor of reading achievement. Froese (1997) compared the IEA reading

scores for British Columbia with variables related to school and classroom

libraries and found that students who have the opportunity to borrow books

from libraries have a considerable achievement advantage over those who

cannot.

In his meta-analysis of reading research studies, Krashen (1993) concluded

that more free voluntary reading results in better reading comprehension,

writing style, vocabulary, spelling and grammatical development. He also

determined that when books are readily available and the print environment

is rich, more reading is done. Even second-language learners will be more

successful in language acquisition when they read more in the second

language. Children get a substantial percentage of their books--from 30 to

90 percent--from school, classroom and public libraries. They also read

more when they have a comfortable, quiet place to read, such as the school

library. Ramos and Krashen (1998) concluded that simply providing

interesting books to children is a powerful incentive for reading, perhaps the

most powerful incentive possible.

McQuillan (1997) drew similar conclusions. He found that access to print via

the home, school and public library, and frequency of free reading accounted

for nearly 80 percent of the variance in fourth grade reading test scores. He

also reported a correlation between school and public library quality, library

use, and amount of reading done by school children. In McQuillan's (1998)

meta-analysis of literacy studies, there was considerable evidence that the

amount and quality of students' access to reading materials is substantively

related to the amount of reading they engage in, which in turn is the most

significant determinant of reading achievement. More reading leads to

better reading achievement.

Other researchers have also demonstrated a relationship between free

voluntary reading and academic achievement. Digiovanna (1994) found that

the amount of recreational reading was positively correlated with higher

academic achievement levels for third, fifth, and seventh graders. Halliwell

(1995) demonstrated a relationship between eighth graders' self-perceptions

of being free voluntary readers and the degree of their success on the

Missouri Writing Assessment. Lipscomb (1993) reported on the self-selected

recreational reading of first through third-graders over a nine-week period in

the summer and found that the total number of words read was a significant

predictor of students' overall reading achievement and word recognition.

13

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Access to print through public libraries has been shown to contribute to

students' academic achievement. A Library Research Service study (no.

153, 1998) reported that in Colorado school districts scoring in the highest

third on the 1997 Colorado State Assessment reading test, circulation of

children's materials per capita by public libraries was 50 percent higher than

in school districts scoring in the lowest. There were similar results for states

scoring highest on the NAEP reading test. Ramos and Krashen (1998)

reported that even one classroom trip per month to the public library had a

positive impact on students' reading. McQuillan (1997) found that SAT

scores were positively correlated with per capita public library circulation.

School-Public Library Relationship

A Library Research Service study (no. 150, 1998) reported that students are

likely to earn higher reading scores if there is a relationship between the

library media program and local public libraries. Such a relationship might

include public library staff presenting booktalks at the school library, and the

local public library providing a summer reading program. Similarly, Lance,

Hamilton-Pennell and Rodney (2000a) found that Alaska students' test

scores tended to be higher when the LMC had a cooperative relationship with

the public library. Michie and Chaney (2000) reported that 60 percent of

LMCs participated in some type of cooperative activity with a local public

library, including borrowing materials for teachers or the LMC, informing the

public library of curriculum or homework needs, and coordinating class visits

to the public library.

School Library and Equity Issues

Several researchers point to the potential importance of the school library as

a factor in equalizing access to print resources for disadvantaged children.

McQuillan (1997, 1998) found a strong negative correlation between poverty

and print resources at home. He concluded that school and public libraries

could help increase access to print for low-income communities and schools,

thus improving their students' reading achievement. The survey data

collected by Baughman (2000) suggest that children from a lower

socioeconomic stratum who have a school library program obtain a higher

mean test score than similar children from schools without such a program.

14

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Halle, Kurtz-Costes and Mahoney (1997) reported that the number of books

in the homes of low-income, African-American children was related to

children's reading scores at the end of the following year. They concluded

that providing access to children's books through libraries may be one of the

most important things disadvantaged communities and schools can do.

McQuillan (1998a) studied the public library use of language minority

students and found that Spanish-speaking households are much less likely to

have access to books, and, therefore, fewer opportunities to further literacy

development. He concludes that both public and school libraries must make

concerted efforts to reach out to language minority parents and their

children. Godina reports that Mexican-background students demonstrate

different literacy practices in their homes and communities than those

acknowledged at school, where they are viewed as having limited English

proficiency and enrolled in low academic tracks. The local public library

becomes an important locale for these students' literacy learning because it

provides culturally-relevant reading materials.

Unfortunately, school libraries often appear to reflect the economic

conditions of their communities. Krashen and O'Brian (1996) reported that

socio-economic status was the most powerful predictor of student reading

achievement in the Los Angeles Unified School District. Both Krashen

(1996) and McQuillan (1998) made the point that the low student reading

scores in California could be traced to the deplorable state of its school and

public libraries. Allington, Guice, Baker, Michelson, and Li (1995) studied

the variations in access to books in school libraries in twelve high- and low-

income neighborhoods. They discovered that high-income schools had 21.5

books per student, whereas the low-income schools had 15.4 volumes. They

also discovered disparities in number of magazine subscriptions, size of

classroom libraries and access policies. McQuillan, LeMoine, Brandlin and

O'Brian (1997) studied access to school libraries and found that students in

high-achieving schools serving largely middle-class children provided greater

access to books, more time to read in school, and more liberal circulation

policies than those from lower-achieving schools in largely low-income

neighborhoods. Smith, Constantino and Krashen (1996) found, not

surprisingly, that school libraries in high income communities such as

Beverly Hills had around three times as many books per student as school

libraries in low-income communities such as Compton and Watts. Public

libraries in high-income communities also had about twice as many books as

those in low-income communities.

Two recent studies reinforce these conclusions. Duke (2000) studied 20

first-grade classrooms from very low- and very high-socioeconomic status

(SES) districts. Data indicated substantial differences between low- and

15

How School Libraries Improve Outcomes for Children:

The New Mexico Study

high-SES classrooms in all major areas examined--amount of print

experienced, type of print experienced, and number of print-related

activities. Neuman and Celano (2001) focused on the neighborhood settings

in which literacy begins for young children, and the extent to which literacy

is a potential factor contributing to differences in achievement. What they

found was striking. Access to print resources, including children’s books

available for purchase, public areas where children might observe people

reading, and the size and quality of book collections in local childcare

centers, school and public libraries, varied dramatically between the

neighborhoods of middle- and low-income children. Children from middle-

income neighborhoods were likely to be deluged with a wide variety of

reading materials, while children from poor neighborhoods would have to

“aggressively and persistently seeks them out.” These children would have

to rely on public institutions, which provided unequal resources. School

libraries in these neighborhoods were in serious disrepair.

A case study of a Library Power elementary school in Chattanooga (Oberg,

1999) shows that the library media program can make a difference in a low-

income school. Lakeside Elementary School, largely composed of low-

income African-American children, improved LMC collections and facilities,

developed collaborative planning between teachers and librarians through

professional development; provided flexible scheduling to the LMC; and

hired a full-time LMS. The school experienced dramatic improvements in

student learning, as evidenced by scores on the TCAP (Tennessee

Comprehensive Assessment Program), that are attributable at least in part

to the Library Power Initiative.

Size of Library Media Center Collection

Size of the school library collection has been shown to be a positive predictor

of student academic achievement (Greve, 1974; Lance, Welborn and

Hamilton-Pennell, 1993; Elley, 1996). Baughman (2000) and Lance, Rodney

and Hamilton-Pennell (2000b) found that the per pupil book count was

correlated with higher test scores. Lance, et al, (2001) also found that

Oregon reading test scores rose with increases in print volumes per student

and periodical subscriptions per 100 students. Krashen (1995) found that a

significant predictor of NAEP reading comprehension test scores was the

number of books per student in school library media centers. Similarly,

McQuillan (1997) reported that SAT scores were positively correlated with

the number of books per student in the school library. Krashen and O'Brian

(1996) did not find a significant relationship between books per student and

16

How School Libraries Improve Outcomes for Children:

The New Mexico Study

student achievement in the Los Angeles Unified School District. They

postulated that the number of books in a school library has little effect on

literacy if access to them is restricted, the books are badly out of date, and

students do not know where they are, as is the case in many California

school libraries.

Frequency of School Library Use

Frequency of library use has also been positively linked to student

achievement scores. Baughman (2000) found that students at each grade

level score higher when there is increased use of the LMC and more open

hours at the library. Koga and Harada (1989) investigated the attitudes of

students in Australia, Japan, Korea and Thailand towards school libraries and

found that students with a keen attitude toward learning tended to use the

library more often and demonstrated better academic achievement. The

LMS at Lakeside Elementary School, a Library Power elementary school in

Chattanooga (Oberg, 1999), found a direct relationship between the number

of times students had been in the library and the level of their test scores in

reading comprehension and reference skills. A Library Research Service

report (no. 149, 1998) showed that states with above average reading

scores on the 1994 NAEP reading test have schools where students visit the

school library media center more frequently and borrow more books and

other materials.

Flexible scheduling appears to support more frequent library use by

individual students. Fedora (1993) found that in a flexibly scheduled library

media program, students have more frequent access individually and in

small groups than in a fixed-schedule program, where nonscheduled use is

rare. Van Deusen (1996b) reported that instances of the LMS providing

individual assistance to students was higher in flexible than fixed schedule

situations. Lance, Rodney and Hamilton-Pennell (2000b) found that middle

schools with high test scores tend to have LMCs that report a higher number

of individual visits to the LMC on a per-student basis.

The extent to which flexible scheduling occurs in library media centers varies

by type of school and educational level. According to recent survey results

(Michie and Chaney, 2000), 95 percent of public secondary school library

media centers and 60 percent of public elementary school library media

centers (70 percent total) report that they use flexible scheduling, although

it is most often used in combination with regular scheduling as well.

Mosqueda (1999) reports a similar figure, with 75 percent of the LMCs of the

17

How School Libraries Improve Outcomes for Children:

The New Mexico Study

reporting National Blue Ribbon Schools operating on a flexible schedule.

Abdoler-Shroyer (1999) reports that in her population of Missouri's

combined K-12 schools, 88 percent of the districts responding used the LMC

for teacher released time at the elementary level, with 87 percent seeing

students on a fixed schedule. At the high school level, 83 percent schedule

high school classes on a flexible basis.

Role of Library Media Specialist in Program Development

The role of the LMS in developing and providing access to the library media

program has received a fair amount of attention in the research. Pembroke

(1997) found that, when school librarians provide reading guidance or a

bibliography, reluctant fifth grade readers can be motivated. Other

motivating factors included access to the library and books; an adequate

collection of print and non-print materials; and an inviting environment.

Martin (1996) found that as library services increased (including reference,

information skills, curriculum integration, interlibrary loan, reading guidance,

and technical assistance), third grade test scores also increased. She found

a statistically significant relationship among all the library media center

variables (collection size, expenditures for the collection, staffing, and

services) and overall achievement in grades 3, 5, and 11, indicating that the

whole is greater than the sum of its parts. A Library Research Service study

(no. 150, 1998) reported that students are likely to earn higher reading

scores if there is a collection development policy for the school library media

center. Lance, Hamilton-Pennell and Rodney (2000) also found that Alaska

students' test scores tend to be higher when the LMC has a strong collection

development policy.

K. G. Alexander's (1992) study of four exemplary media specialists found

that they all provided continuous access to their media centers; assisted

individual users; designed flexible circulation policies; used innovative

methods to promote their media centers; and developed media center

collections which supported all areas of the curriculum. Gehlken (1994)

reported that in all three blue ribbon high schools studied, the LMS flexibly

scheduled classes; organized and cataloged the collection; provided an

inviting climate conducive to learning; assisted students in traditional and

electronic methods of information access; and involved faculty in the

selection of materials. The media centers in all three schools were

organized, automated, easily accessible, and provided materials in a variety

of formats across all levels and subject areas. The media center facilities

were inviting, aesthetically pleasing, safe, and user-friendly.

18

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Role of Technology in Student Academic Achievement

The role of technology in promoting student achievement has been the focus

of several recent studies. In their review of educational technology

research, Sivin-Kachala, Bialo, and Langford (1997) concluded that using

technology has a positive effect on student achievement, attitudes toward

learning, and student self-concept. Silverstein, Frechtling and Miyaoka

(2000) found that technology usage had a small but significant impact on

Illinois students' achievement, particularly at the higher grade levels,

specifically eleventh grade science. Paul, VanderZee, Rue and Swanson

(1996) reported that using the Accelerated Reader technology-based literacy

program had a positive affect on student academic performance and

attendance rates, especially for socio-economically disadvantaged children in

urban areas. Peters (1998), on the other hand, found no significant

difference in reading achievement when students used the technology-based

Electronic Bookshelf Program.

Wenglinsky (1998) found a positive correlation between computer use and

academic achievement in mathematics, but only if computers were used to

convey higher-order skills or engage in learning games. Use of computers

for drill and practice, the lower-order skills, was negatively related to

academic achievement for both fourth and eighth grade students.

Significantly, disadvantaged groups had less access to those aspects of

technology that positively affected educational outcomes. Page (1999)

reported that the presence of classroom technology had a positive effect on

the mathematics achievement of low socioeconomic elementary school

students, although reading achievement remained inconclusive.

DeFrieze (1998) found that reading achievement appears to be more

influenced than mathematics achievement by the use of computers,

particularly in an unstructured environment. She speculates that in an

unstructured environment teachers have more control over the programs

each student accesses, which may be the key to influencing students' higher

achievement in reading. A significant finding by Bohannon (1998) is that

high frequency of school computer use results in students earning

significantly higher mean scores on reading achievement tests. This is true

for males and females, as well as high and low socioeconomic groups. She

concludes that frequent use of computers requires more continuous reading

practice, regardless of the activity.

19

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Integrating Technology into the Library Media Center

There is perhaps no place where the library media specialist's role has

changed more in the last ten years than in the integration of technology.

Person (1993) reported that LMSs don't see a separate, organized

technological media role for themselves, but see technology as a means to

accomplish the goals and missions of the media center program as

expounded in Information Power (1988). A Library Research Service

report (no. 141, 1998) concluded that students earn higher reading scores if

their school library media programs incorporate the latest information

technology. Such technology includes a district-wide catalog, access to

online databases, resources available through a local-area network, and

access to the World Wide Web and the statewide library network. Lewanski

(1998) showed a statistically-significant correlation between the use of

computer-assisted library research and improvement in overall critical

thinking skills. The control group, using a traditional paper-based process

did not show such a relationship.

Lance, Rodney and Hamilton-Pennell (2000a) found that Pennsylvania

students' test scores increase as LMSs spend more time managing

information technology. Students also earn higher reading test scores in

both Pennsylvania and Colorado (Lance, Rodney and Hamilton-Pennell,

2000b) where networked computers link school libraries with classrooms,

labs and other instructional sites, enabling access to LMC resources, licensed

databases and the Internet. Alaska students' test scores tend to be higher

when the LMC is equipped to provide access to the Internet (Lance,

Hamilton-Pennell and Rodney, 2000).

An Illinois study (Silverstein, Frechtling and Miyaoka, 2000) revealed that

almost all Illinois schools are now connected to the Internet. McCracken

(2000) found that 99% of high school LMCs, 95% of middle school LMCs,

and 84% of elementary school LMCs had Internet access. Nevertheless,

access to technology in school library media centers still varies widely.

Powell's (1998) survey of 300 elementary and secondary school library

media centers in Tennessee revealed a wide variability in technology access.

McCarthy (1997) found that less than 50 percent of the New England school

library media centers in her sample had automated circulation and

cataloging systems, and these were mostly in middle and high school

libraries. McCracken’s figures (2000) show that the percentage of LMCs with

online catalogs ranged from 60% in elementary schools to 84% in high

schools. She also reported that LMSs perceive that integrating technology

20

How School Libraries Improve Outcomes for Children:

The New Mexico Study

into the LMP is more important than the extent to which they perceive they

have actually implemented it into their practice.

High-achieving schools tend to have more technological resources. Baule

(1997) found that schools with exemplary technology were also more likely

to have high-quality school library media programs. Yetter (1994) observed

that the library media centers in successful resource-based learning schools

had modern, spacious facilities designed for flexible use and access to

technology. Gehlken (1994) noted that all three blue ribbon schools studied

had library media centers which were committed to increasing student

access to technology, and which had the flexibility and electronic capabilities

to accommodate the changing needs created by new technologies. Students

identified the electronic catalog, computer printer workstations and copying

machines as some of the most important services provided by the library

media program.

As Wenglinksy (1998) and DeFrieze (1998) demonstrated, it is not the

amount of technology or computer use that counts in promoting student

achievement, but how

it is used. Many researchers (for example, McQuillan,

1996; Lance, Welborn and Hamilton-Pennell, 1993) have found no

correlation between reading achievement and amount of computer software

available. Technology must be integrated into the school library media

program to influence academic achievement. An elementary school in

California with a high Latino population (Ferguson, 2000) saw student

achievement gains that were attributed to the technological upgrade of the

LMC. Reading scores rose 8% from the previous year after the school hired

a professional LMS who refurbished the LMC’s technological resources and

added reading literacy programs such as Accelerated Reader and the

S.T.A.R. Reading testing program.

Program Administration

The program administration role involves effective management of the

human, financial and physical resources of the library media program. This

role also provides leadership within the larger learning community.

Adequate staffing, budget, and administrative support are key to the success

of this role. When all these factors are maximized, research studies show a

dramatic impact on student achievement.

21

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Maximizing Predictors of Student Achievement

Lance, Rodney and Hamilton-Pennell (2000b) found that when all LMC

predictors of students achievement (i.e., staffing, expenditures, information

resources and technology) are maximized, Colorado reading test scores tend

to run 18 percent higher in the fourth grade and 10 percent higher in the

seventh grade. In Pennsylvania (Lance, Rodney and Hamilton-Pennell

2000a), reading test scores tend to run 10 to 15 points higher when LMC

predictors (i.e., staffing expenditures, information resources and technology,

and information literacy activities of library staff), are maximized. In Alaska

elementary schools with well-developed library media programs, 86 percent

of students scored proficient or above on state reading tests compared with

73 percent of students in schools with less well-developed media programs

(Lance, Hamilton-Pennell and Rodney, 2000). In Oregon, Lance, Rodney,

and Hamilton-Pennell (2001) concluded that whatever the level of a school’s

library program, incremental improvements in its staffing, collections, and

budget will yield incremental increases in reading scores. They also found

that test scores rose when LMSs spent more time attending faculty

meetings, meeting with library media staff at the district level, and meeting

regularly with the principal.

Principal Support of the Library Media Program

Yetter (1994) found that schools that had successfully implemented

resource-based learning had a common understanding and support from the

principal, teaching faculty and library media specialist about the centrality of

the library media program in the school's instructional process. These

schools provided planning time for teachers and library media specialists to

work collaboratively, clerical support for the LMS, flexible scheduling in the

library media center, and principal support of the library media program.

Likewise, Farwell (1998) determined that in schools with successful

collaborative planning, the principal served as an advocate for collaborative

planning and information literacy instruction, and provided financial support

for the library media program, adequate clerical staff, and time during the

school day for LMSs and classroom teachers to plan together.

Mosqueda (1999) concluded that the most important finding in her study of

school library media programs in Florida blue ribbon schools was the overall

favorable perception of principals and LMSs in regard to the leadership role

they played in their schools. The reported strengths of the library media

programs in these schools supported the research on what constitutes an

22

How School Libraries Improve Outcomes for Children:

The New Mexico Study

exemplary program, including open access, administrative support, well-

equipped facilities, technology integration, reading support, and good

staffing. Gehlken (1994) reported that in all three blue ribbon high schools

she studied, the principals actively supported and promoted the library

media program. Standridge (1996) reported that student achievement in

urban elementary schools was positively impacted by greater participation of

the certified staff in school-based decision making, especially in the areas of

goals, vision, mission, and curriculum and instruction.

There appears to be a two-way relationship between administrative support

and LMSs performing the instructional role. In schools where there was

fiscal and organizational support for the library media program, including

automated systems and paid support staff, van Deusen (1996a) found that

LMSs performed the instructional consultation role to a greater extent.

Lumley (1994) concluded that instituting a curricularly integrated and

flexibly scheduled library media program required leadership on the part of

the library media specialist as well as principal support, resulting in strong

leadership roles for the LMS in curriculum, instruction and staff

development. Van Deusen (1996a) reports that the availability of support

staff and automated library systems was positively related to LMSs' doing

more consulting work with teachers and spending less time on

nonprofessional tasks. McCracken (2000) found that the factors most

frequently cited by LMSs as important to expanding their instructional role

were support of the school administration and support of other teachers at

the school.

Collaborative Activities of the Library Media Specialist

Lance, Rodney and Hamilton-Pennell (2000b) found that LMS involvement in

collaborative activities has a direct impact on test scores. Higher levels of

collaboration result from meeting with school administrators, serving on

standards and curriculum committees, working with faculty at school-wide

staff meetings, and meeting with library media staff at the building level.

Data from Pennsylvania (Lance, Rodney and Hamilton-Pennell 2000a)

reinforce the finding that test scores increase as LMSs spend more time

serving on curriculum and standards committees. Hughes (1998) reports

that LMSs who successfully participated in their schools' move to whole

language became part of their school's leadership team, attended

professional development workshops with teachers in the school, and spent

more time pulling together curriculum-related resources in the library media

center.

23

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Effect of Library Media Center Staffing

Lance, Welborn and Hamilton-Pennell (1993) found that library media

centers which have more endorsed staff tend to have staff who spend more

time identifying materials for instructional units developed by teachers and

more time collaborating with teachers in developing such units. They found

that as the LMS's instructional role increases, the size of the library media

center's staff and collection increases, which, in turn, is a direct predictor of

student reading achievement. Martin (1996) also found a significant positive

relationship between school library media center staffing and student

achievement, especially in high school reading. Schools employing more

media center staff had higher achievement test scores.

A Library Research Service study (no. 141, 1998) showed that student

reading scores were higher in schools where there is a state-endorsed library

media specialist and where the LMS is supported by an aide. This is

unfortunately not yet the norm. Michie and Chaney (2000) report that as of

1997, public school libraries had an average of 0.9 professional staff

(included non-certified professional staff) and 0.6 other paid employees,

overall. Twenty-two percent of public schools lacked a full-time or part-time

library media staff person who is state-certified. McCracken (2000) reported

that the biggest barriers cited by LMSs to expanding their instructional role

were lack of time and resources such as funding and support staff.

Studies in Alaska, Pennsylvania, Colorado, Massachusetts, and Oregon

(Lance, Hamilton-Pennell and Rodney, 2000; Lance, Rodney and Hamilton-

Pennell, 2000a; Lance, Rodney and Hamilton-Pennell, 2000b; Baughman,

2000; Lance, Rodney and Hamilton-Pennell, 2001, respectively) attest to the

value of full-time, endorsed media staff. In Alaska, at both elementary and

secondary levels, the presence of a full-time librarian was a very strong

predictor of average and above achievement. Regardless of the level of

librarian staffing, however, the more LMS staff time was devoted to

delivering library and information literacy instruction to students and

providing in-service training to teachers, the higher the test scores.

Colorado reading scores increased with increases in LMS hours per 100

students (for seventh grade) and total staff hours per 100 students.

Pennsylvania reading test scores increased with increases in LMS staff hours

and support staff hours. In Massachusetts, at elementary and high school

levels, students who were served by full-time LMSs had higher test scores

than those in schools without full-time LMSs, and non-professional staff

24

How School Libraries Improve Outcomes for Children:

The New Mexico Study

assistance also made a positive difference in average test scores. Oregon

reading scores rose with increases in total staff (both professional and

support) per 100 students.

Library Media Specialist as Provider of In-Service Training

The involvement of a school library media specialist in technology-based

staff training can support student achievement. Michie and Chaney (2000)

report that in 1997, 43 percent of public school library media programs with

access to the Internet arranged Internet training for teachers and 37 percent

for administrators. In Colorado (Lance, Rodney and Hamilton-Pennell,

2000b), reading test scores increased with the amount of time LMSs spent

as in-service trainers of other teachers. Alaska and Pennsylvania students'

test scores tended to be higher when library staff spent time providing in-

service training to teachers (Lance, Hamilton-Pennell and Rodney, 2000;

Lance, Rodney and Hamilton-Pennell). Wenglinsky (1998) found that

teacher's professional development in technology and the use of computers

to teach higher-order skills were both positively related to academic

achievement in mathematics and the social environment of the school.

Library Media Specialist Planning and Management Skills

School library media specialists in effective schools tend to have good

planning, communication and management skills. Yetter (1994) observed

that library media specialists in successful learning-based schools were

expert in developing effective library media programs which were congruent

with the state and national Information Power (1988) guidelines. The

basic library procedures and processes in their library media programs

functioned smoothly. A Library Research Service report (no. 150, 1998)

indicated that students are likely to earn higher reading scores if there is a

plan for the development of their school library media program. Gehlken

(1994) reported that in all three blue ribbon high schools the LMSs took

proactive steps to update students, teachers and administrators about new

materials, technology, and services. Lumley (1994) concluded that

instituting a curricularly integrated and flexibly scheduled library media

program in an elementary school required LMS leadership in site-based staff

development and good communication with staff and principal support.

25

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Library Media Specialist Budgetary Role

A very important administrative role for the LMS is to obtain an adequate

budget for the library media program. Angelo's (1994) study of

Massachusetts school library media programs revealed that more than 90

percent were operating at the minimum level according to state standards in

the areas of personnel, collection, and budget. Lancaster (1998) surveyed

superintendents about their attitudes toward elementary LMSs. She reports

that though there appears to be agreement on the importance of and role of

the LMS, in concrete terms, they do not adequately fund or staff the

program in many cases. Almost half of the superintendents spent less than

two percent of their total budget on the LMC.

Lance, Welborn and Hamilton-Pennell (1993) found that students at

Colorado schools with better funded library media centers tend to achieve

higher average test scores, whether their schools and communities are rich

or poor and whether adults in the community are well or poorly educated.

Bruning (1994) also reported a positive relationship between student

achievement measures and the proportion of a school district's budget spent

on library materials, for both high- and low-income districts. Studies in both

Colorado and Oregon (Lance, Rodney and Hamilton-Pennell, 2000b; Lance,

Rodney and Hamilton-Pennell, 2001, respectively) found that student

reading scores increased with increases in library media expenditures per

student. Baughman (2000) reported that in Massachusetts, at elementary

and middle school levels, average test scores were higher in schools with

larger per pupil expenditures for library materials.

These findings are particularly significant since studies seeking a relationship

between school spending as a whole and student performance have shown

mixed results. Krashen (1995) found that expenditures for education did

not affect reading comprehension scores, while Powell and Steelman (1996)

did find that school spending was positively linked to state SAT and ACT

performance. A review of over 400 studies of student achievement by

Hanushek (1997) demonstrated that there is not a strong or consistent

relationship between student performance and school resources after

variations in family input are taken into account. Hedges, et. al. (1994) in

their meta-analysis of studies of differential school inputs on student

outcomes, show that a positive relationship between resources and

educational outcomes does exist and is significant enough to be of practical

importance. While there is no clear mandate for increasing school spending

in general to support student achievement, the research does show that

increasing expenditures for school library media materials has a positive

effect.

26

How School Libraries Improve Outcomes for Children:

The New Mexico Study

A World Bank meta-analysis of funding studies (Acevedo, 1997?), indicated

that differences in aggregate education budgets do not appear to have a

tight association with learning outcomes, although class instructional time,

school library resources and textbooks, and class frequency of homework are

inputs positively associated with improved test scores. Based on this

analysis, the PARE program (Programa para Abatir el Rezago Educative)

provided increased resources (including library resources) to Mexican

schools. Data indicate that on average, PARE assistance had a significant

positive effect on learning outcomes in Spanish. Acevedo concluded that

roughly doubling the school resources allocated per student overcame a

30% deficit in test scores among rural students.

Summary

The impact of school library media programs on academic achievement is

well documented in the research literature. Studies demonstrate

consistently that well-equipped, quality school library media centers that

have professional staff involved in instruction contribute to the academic

success of their students. Likewise, both higher order uses of technology

and expenditures for library materials support student achievement. All

three roles of the school library media specialist identified in Information

Power (1998) lead to greater integration of the school library media

program into the larger learning community and promote greater student

achievement.

27

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Methodologies

Sample

In 2000, New Mexico had 759 public schools. Of the 400 elementary schools

serving fourth grade, 208—52 percent—participated in this study. Of the

177 junior high and middle schools serving eighth grade, 100—56 percent—

chose to be included. Of the 152 high schools serving tenth grade, 72—47

percent—participated in this study. (See Table 1.)

Throughout this study, the participants were treated as three distinct

samples, one for each tested grade. Table 1 reports the number in the

sample for each grade and its proportion of the universe of all schools

serving that grade.

Table 1. Comparison of the Study Sample and the Universe of All New Mexico

Public Schools Serving Grades 4, 8 and 10, 2000

Grade

Number in

sample

Number in

Universe

Sample as percent

of universe

4

th

208 400 52%

8

th

100 177 56%

10th 72 152 47%

Survey

The survey of school library programs focused on several sets of potential

predictors of academic achievement. These included: LMC hours, LM staff

and their activities, technology, LMC usage, LM resource collections, and

finances.

Respondent Information

The questionnaire began with several items identifying the responding

school—its name and address, grades served, and its school district—and

the individual respondent—his/her name and title as well as telephone and

fax numbers and e-mail address. All of this information was required to

assess and address potential deficiencies in the initial response rate to the

28

How School Libraries Improve Outcomes for Children:

The New Mexico Study

survey. The grades served were especially important as they made it

possible to determine which tested grades a school included.

Service Hours

The second part of the questionnaire contained items concerning the school

library’s hours of operation—both during and after school in a typical school

week and in a typical week during summer months. It is expected that

schools with higher test scores will be those with libraries that have longer

hours.

School Library Staff

This part of the questionnaire contained items requesting the numbers of

people and total person-hours worked by paid staff with different types of

qualifications. As noted earlier, one of the most consistent findings of

research about the impact of school libraries is the value of staffing them

with individuals who are professionally trained for the job. Another

consistent finding in past research is the importance of having support staff

who free professionals to do their job.

Paid Staff Activities

Perhaps the most fundamental question examined by this study was the

value of staffing school libraries with trained individuals who engage in

particular professional-level activities. The synergy of these activities,

proved to have considerable impact on test scores. While the original

Colorado study found strong evidence for the importance of the library

media specialist’s instructional role, those findings were based on just two

items—the number of hours library media staff spent identifying and

providing materials for instructional units developed by teachers and

planning instructional units with teachers. The 2000 New Mexico

questionnaire included a much more comprehensive list of staff activities.

Additional activities on this list included, among others, hours per typical

week staff spent: providing library/information literacy instruction to

individuals or groups; providing in-service training to teachers and other

staff; and teaching collaboratively with classroom teachers. The rationale

for asking practitioners to parse their time so many ways was to obtain more

29

How School Libraries Improve Outcomes for Children:

The New Mexico Study

specific insights about exactly what it is that library media specialists do that

makes a difference in how students perform on achievement tests. Despite

an absence of research at this level of detail, it seemed reasonable to expect

that some activities would be more effective than others and that their

effectiveness might vary by school level (elementary, middle, high school).

Computers with Access to Library Resources

A great deal of detailed information about library technology was collected

by the next section of the questionnaire. Respondents were asked to

identify numbers of school computers both in and under the jurisdiction of

the school libraries and elsewhere in the school from which networked

library resources may be accessed. Of those numbers, they were further

asked to identify numbers of computers meeting various descriptions (e.g.,

with access to the library catalog, licensed databases, and the Internet).

School Library Usage

The next part of the questionnaire solicited statistics about how often

students and staff (i.e., administrators, teachers, others) interacted with

school library staff for different purposes, including library/information

literacy instruction. This section also included items for circulation of library

materials as well as counts of materials loaned to other libraries and

obtained from outside the library (e.g., interlibrary loans, intra-district

loans). Previous research and conventional wisdom indicate that school

librarians who impact student performance are those who are most actively

engaged with teachers and students alike, particularly more direct

involvement in teaching and learning activities. Evidence from previous

research also supports the assumption that students who read more—both

for school purposes and voluntarily—do better on tests.

School Library Collection

Despite the increasingly critical role played by school library staff in the

instructional process, what most people think of first when the school library

is mentioned is its collection. This section of the questionnaire solicits an

inventory of the collection by format, including traditional print sources

(e.g., books, magazine and newspaper subscriptions), non-print items (e.g.,

30

How School Libraries Improve Outcomes for Children:

The New Mexico Study

videos, software packages, and other audio-visual materials), and the

rapidly growing “electronic” sector (e.g., CD-ROM, laser disk, and online

database subscriptions). Traditionally, conventional wisdom dictated that

the larger the collection, the better. As electronic sources of information

proliferate and the value of up-to-the-minute information increases, this

assumption becomes more questionable. Another wildcard related to this

issue is the age of library collections. A larger collection is not necessarily a

better one, if it consists increasingly of deteriorating volumes containing

obsolete information.

Annual Operating Expenditures for the School Library

Although few school libraries have budgets that include personnel costs,

many have budgets for print and non-print materials, electronic access to

information, and miscellaneous operating expenses. This section of the

questionnaire asked for the total figure in whole dollars spent for these

items.

Available Data

This study depends on demographic data that, whenever possible was

obtained at the school or neighborhood level.

The New Mexico Department of Education provided data on both the number

and the percentage of students eligible to receive free or subsidized school

lunches in each school. The percentage of the student body eligible to

receive school lunch assistance was then used as a school-specific poverty

variable.

Each school’s enrollment, subdivided by ethnicity, was provided by the New

Mexico Department of Education. Categories included were Native

American, Asian, Black, Hispanic, White and Other. Five of these variables,

Native American, Asian/Pacific Islander, Black, Hispanic, and Other were

then combined to determine the minority percentage of the school

population. The school minority percentages were utilized in correlation

analysis.

The educational attainment variable demonstrates the general level of

education in the school’s surrounding population. Educational attainment

31

How School Libraries Improve Outcomes for Children:

The New Mexico Study

data was extracted from the United States Census Bureau Web site. The

variable refers to the percentage of people age 25 and over with a high

school diploma or equivalency or higher.

Total school expenditure data on a district level were also provided by the

State Department of Education. The per student expenditure amount was

then accorded to each school within that district. Total school budget and

per student expenditures were the only factors considered. There were no

program breakdowns in such areas as Talented and Gifted or Vocational.

This study also took into consideration two key teacher characteristics, the

percentage of teachers with a master’s degree and the average years of

service. These figures were also provided by the New Mexico Department of

Education.

The test scores used as indicators of fourth and eighth grade students’

academic achievement in this study are 1999-2000 scores on the Language

Arts portion of the standards-based portion of the New Mexico Achievement

Assessment Program. The standards-based results are reported by

performance level, with proficiency settings determined by New Mexico

educators against state content standards. Achievement indicators for tenth

grade students were drawn from the 1999-2000 New Mexico High School

Competency Examination. Here the building-level composite score provided

by the state indicated the percentage of students passing all the sub-tests of

the Competency Examination, including language arts, social sciences,

reading, mathematics, science, and composition.

Statistical Significance

Statistical significance is an often-misunderstood concept. Usually, when a

statistical finding is reported, the first question someone asks is “Is that

figure significant?” In this context, the intuitive response is to question the

magnitude or size of the figure or the difference between two figures. There

are no statistical tests to determine if a difference between two groups is

“big enough,” particularly if the groups in question represent an entire

universe of subjects rather than a sample.

Statistical significance is about reliability or consistency. When a sample is

studied, instead of the entire universe (in this case, school libraries in New

Mexico public schools), a pertinent question is “Are these results truly

32

How School Libraries Improve Outcomes for Children:

The New Mexico Study

representative of the universe, or would different samples yield dramatically

different results?”

Throughout later sections of this report, statistical significance is reported as

“p,” as in “probability.” Three common milestones for statistical significance

are reported: “p < .05,” “p < .01,” and “p < .001.” Respectively, these

designations indicate that the probability of reported results being an

accident or a coincidence is less than one in 20, one in 100, or one in 1,000.

Conversely, these figures may be interpreted to indicate 95, 99, or 99.9

percent certainty that the results are representative.

Generally, the levels of statistical significance reported represent two-tailed

tests—ones indicating the probability that the reported results might be

either exaggerated or suppressed. In some cases, however, significance

levels are based on one-tailed tests—ones indicating only the likelihood that

results reported may be exaggerated.

Throughout this study, statistical significance is most often reported in

association with Pearson product-moment correlation coefficients in bivariate

correlation analyses.

Bivariate Correlation

In this study, bivariate correlation analysis served two purposes: 1) informing

decisions about eliminating or combining variables, and 2) assessing the

direction and strength of the relationship between two variables, such as the

ratio of library media specialist staff hours to students and reading test scores.

Pearson’s correlation coefficient (r) indicates the extent to which two variables

change together on a scale of –1.00 to zero to 1.00. Negative values indicate

that a decline in one variable is associated with an increase in another, while

positive values indicate that two variables increase together. For each report

of this statistic, there is a corresponding indication of its statistical significance.

(See earlier discussion about interpreting statistical significance.) In addition

to assessing the direction and strength of relationships, Pearson’s r helped to

determine if any data elements were so strongly associated as to be either

unnecessary or problematic if used together. In some cases, this statistic

provided the basis for decisions to combine variables. Such data reduction

was deemed desirable as it focused and simplified the model to be tested.

33

How School Libraries Improve Outcomes for Children:

The New Mexico Study

The significance of a bivariate correlation may be tested on a one- or two-

tailed basis. Two-tailed significance testing assumes that the direction of the

presumed relationship is unknown, thus one wishes to assess the possibility of

error in either direction. One-tailed significance testing assumes that the

direction of the presumed relationship is known, making it necessary to assess

the possibility of error in only one direction. Because previous research

consistently indicates that the effects on academic achievement of the library

media variables under study are positive, some relationships significant on a

one-tailed basis, but narrowly not on a two-tailed basis, will be reported.

Factor Analysis

While correlation analysis examines relationships between pairs of variables,

factor analysis establishes relationships among groups of related variables.

This technique was particularly useful when two or more variables needed to

be combined, but were measured on different scales (e.g., dollars and

percentages).

Instead of reporting the correlation of each variable with each other variable,

factor analysis helps to create composite factors by reporting factor loadings

that indicate how strongly and in what direction each variable is related to a

factor. A factor loading indicates how much weight is assigned to a given

factor for a given variable. Factors on which a variable loads highly are closely

related to that variable. At this stage, factor analysis was a more efficient

method of confirming--and discovering--relationships among variables than

comparing multiple relationships among pairs of variables.

Exactly how factor analysis works need not be understood. When sorted by

factor, the results are easy to interpret, since a researcher can readily identify

variables that load highly on a given factor. The researcher, however, must

interpret what a factor represents and decide what to call it. The factor

analysis technique will also generate a factor score based on a school's values

on the variables that load on a factor. In several instances, closely related

variables were replaced by a factor score.

Multiple Regression Analysis

Multiple regression was used to weigh the effects of library media variables

relative to school and community variables as predictors of academic

34

How School Libraries Improve Outcomes for Children:

The New Mexico Study

achievement. This technique is especially useful in assessing complex

relationships among several potential predictors, because it weighs the

importance of each predictor variable while ruling out the effects of the others.

This application of multiple regression techniques is a path analysis because

both research and practice suggest a certain cause-and-effect order among

the variables. In this model, community variables precede school variables,

and school variables precede library media variables. All three sets of

predictors precede—and may affect directly and/or indirectly—academic

achievement. Multiple regression is used to assess the strength and direction

of each separate path from variable to variable. These relationships are

reported as path coefficients (i.e., betas or standardized regression

coefficients).

Correlation and factor analyses of the original data elements helped to refine

the model. They provided the basis for decisions to eliminate redundant

variables and combine those that were so closely related as to produce

statistical "static." In a path analysis via multiple regression, such "noise"

complicates a model unnecessarily and suppresses the effects of other

predictors statistically.

It is very important to note that this type of analysis makes two kinds of

assumptions. It assumes causal order. The presumed cause-and-effect order

in this model is suggested by previous research and practical experience. It is

intuitively obvious that the status of library media centers may depend on

more general school circumstances, just as they, in turn, may be driven by

community conditions. It is equally apparent, however, that each of these

sets of variables may affect academic achievement either directly or indirectly

via some other variable not represented in this model.

An assumption of causal closure supposes that no critical variables are omitted

from the model. This assumption is addressed as fully as was practically

possible by this study. Without apology, its focus is on assessing the impact of

school library media centers on academic achievement. The community and

school variables included represent major antecedent conditions that might

explain away that impact. For instance, the possibility that a correlation

between the level of library media (LM) staffing and test scores might be a

spurious result of generally high levels of staffing in a school was addressed by

including the teacher-pupil ratio. Similarly, the possibility that a correlation

between time spent by LM staff on library/information literacy instruction and

test scores might be a spurious result of community affluence or socio-

economic advantages was addressed by considering several alternative

measures of those variables. Likely predictors of academic achievement for

35

How School Libraries Improve Outcomes for Children:

The New Mexico Study

which data are not available include parental involvement in a student’s

education, extra-curricular activities of students, characteristics of school

curricula, and pedagogical techniques of teachers. Nonetheless, no

relationships between such likely predictors and the level of development of

LM programs are anticipated.

Because the original number of variables was large, it is assumed that an

acceptable degree of causal closure was established. Nonetheless, Multiple R

Square (R

2

) is taken as a sufficient statistical indicator of the extent to which

the model may not be causally closed. This statistic indicates the percentage

of variation in test scores which is explained by a given group of predictors.

Separate analyses were conducted for elementary, middle, and high school

levels. In each case, multiple regression was used to generate initial path

coefficients. Variables whose path coefficients were less than .10 and which

were not statistically significant at at least the .05 level (generally accepted

standards) were automatically eliminated from the analysis.

36

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Findings

Utilizing the combination of statistical techniques described above, both

direct and indirect relationships between school libraries and academic

achievement were explored. The findings concerning both types of

relationships are presented by school level. Then, the impact of school

libraries on academic achievement is examined more closely, taking into

account other school and community conditions that make up the often

complex environment in which New Mexico’s school libraries operate.

Elementary School Level

At the elementary school level, library variables demonstrate positive and

statistically significant relationships to reading scores, some of which are

direct and some of which are indirect.

School Library Predictors of Achievement Scores

In New Mexico elementary schools, fourth grade achievement scores tend to

rise with increases in

• the total number of library staff;

• the percentage of their time school librarians spend managing computer

networks;

• numbers of computers in the school that provide access to the library

catalog, licensed databases, and the Internet;

• the number of print volumes in the school library; and

• the number of electronic reference sources relative to students. (See

Table 2.)

•

37

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 2. Direct School Library Predictors of Fourth Grade Achievement Scores in

New Mexico Elementary Schools, 2000

Variable

Correlation

coefficient (r)

Statistical

significance (p)

Library Staffing

Total number of staff

.167

.017

Staff Activities (Percent of Weekly Hours)

Managing computer network

.158

.029

Information Technology

Internet computers in school

.185

.014

Catalog computers in school .159 .031

Database computers in school .222 .003

Information Resources

Print volumes

.332

.000

Electronic references per 100 students

.163 .028

Information Technology

Fourth grade test scores rise with the level of library-related information

technology in New Mexico elementary schools. In turn, elementary schools

with more such technology tend to have:

• more school librarians who work more hours per week; and

• school librarians who spend more time each week meeting with their

principals, providing in-service training to teachers, and providing reading

incentive programs to students. (See Table 3.)

Table 3. Indirect School Library Predictors of Fourth Grade Achievement Scores

via Information Technology in New Mexico Elementary School Libraries, 2000

Information Technology

Variable

Internet

computers in

school

Catalog

computers in

school

Database

computers in

school

Library Staffing

Number of school librarians

.157*

Weekly librarian hours

.182* .185*

Staff Activities (Weekly Hours)

Meeting with principal

.181*

Providing in-service training .276* .367* .406*

Providing reading incentive programs .177*

* p < .05

38

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Information Resources

The achievement test scores of New Mexico fourth graders improve with the

size of library collections in elementary schools. Elementary schools with

larger print collections and access to more electronic resources tend to have:

• more school librarians who work more hours per week, and

• higher levels of staffing generally, including more support staff who work

more hours per week. (See Table 4.)

•

Table 4. Indirect School Library Predictors of Fourth Grade Achievement Scores

via Information Resources in New Mexico Elementary School Libraries, 2000

Variable

Print

volumes

Electronic

references per

100 students

Library Staffing

Number of school librarians

.238**

Weekly librarian hours .324**

Total staff per 100 students .190*

Total staff hours per 100 students .292**

* p < .05, ** p < .01

Middle School Level

At the middle school level, library variables demonstrate positive and

statistically significant relationships to achievement scores--some direct,

some indirect.

School Library Predictors of Achievement Scores

In New Mexico middle schools, eighth grade achievement scores tend to

improve with increases in

• the number of weekly hours the school library is open;

• extra time spent by library staff on selected activities, including:

planning with teachers and providing them with in-service training,

developing library collections and creating incentives for students to read,

and managing school computer networks;

• the number of print volumes in the library’s collection; and

39

How School Libraries Improve Outcomes for Children:

The New Mexico Study

• the number of library computers that provide access to licensed

databases. (See Table 5.)

Table 5. Direct School Library Predictors of Eighth Grade Achievement Scores in

New Mexico Middle Schools, 2000

Variable

Correlation

coefficient (r)

Statistical

significance (p)

Library Hours

Weekly hours school library open

.196

.052

Staff Activities (Weekly After-School Hours)

Planning with teachers

.295

.023

Providing in-service training to teachers .252 .055

Providing reading incentive programs .251 .055

Developing library collections .255 .052

Managing computer networks .283 .030

Information Resources

Print volumes

.298

.006

Information Technology

Database computers in school library

.251

.013

Print Volumes

Eighth grade achievement scores increase with the number of print volumes

in middle school library collections. Middle schools with larger print

collections tend to have:

• longer weekly hours of operation,

• more librarian and total weekly staff hours, and

• librarians who spend more hours per week teaching cooperatively with

classroom teachers. (See Table 6.)

Table 6. Indirect School Library Predictors of Eighth Grade Achievement Scores

via Print Volumes in New Mexico Middle School Libraries, 2000

Variable

Correlation

coefficient (r)

Statistical

significance (p)

Library Hours

Weekly hours open

.210

.050

Library Staffing

Weekly librarian hours

.306

.004

Weekly total staff hours .301 .005

Staff Activities (Weekly Hours )

Teaching cooperatively

.271

.012

40

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Database Computers

Eighth grade achievement scores also increase with the number of library

computers that can be utilized to access licensed databases. Middle schools

with more library computers that can access licensed databases tend to

have:

• more school librarians who work more weekly hours, and

• who spend more weekly hours teaching cooperatively with classroom

teachers. (See Table 7.)

Table 7. Indirect School Library Predictors of Eighth Grade Achievement Scores

via Database Computers in New Mexico Middle School Libraries, 2000

Variable

Correlation

coefficient (r)

Statistical

significance (p)

Library Staffing

Number of school librarians

.264

.009

Weekly librarian hours .313 .002

Staff Activities (Weekly Hours )

Teaching cooperatively

.323

.001

High School Level

At the high school level, library variables demonstrate positive and

statistically significant relationships to achievement scores, both direct and

indirect.

School Library Predictors of Reading Scores

In New Mexico high schools, tenth grade achievement scores tend to

improve with:

• the general level of library staffing;

• the extent to which classroom teachers and librarians teach

cooperatively;

• the frequency with which both individuals and groups visit the school

library, especially—in the case of groups—for information literacy

instruction;

• the availability of library computers to students, particularly computers

that provide access to the library catalog and the Internet; and

41

How School Libraries Improve Outcomes for Children:

The New Mexico Study

• the size of the library’s collection, including its numbers of books, videos,

and software packages, and spending on such items. (See Table 8.)

Table 8. School Library Predictors of Tenth Grade Achievement Scores in New

Mexico High Schools, 2000

Variable

Correlation

coefficient (r)

Two-tailed

statistical

significance (p)

Library Staffing

Total weekly staff hours per 100 students

.239

.046

Staff Activities (Percent of Weekly Hours)

Teaching cooperatively

.265

.032

Library Visits

Group visits per 100 students

.229

.058 *

Group information skills instruction per 100 students .292 .014

Individual library visits per student .225 .059 *

Information Technology

Library computers per 100 students

.297

.012

Library Internet computers per 100 students .318 .007

Library catalog computers per 100 students .255 .036

Information Resources

Print volumes

.214

.092

Videos per 100 students .340 .004

Software packages per 100 students .312 .012

Library expenditures per student .232 .057 *

*Achieves one-tailed statistical significance at p < .05.

Library Visits

Tenth grade achievement scores improve with numbers of individual and

group visits to school libraries, especially group visits for information literacy

instruction. High school libraries tend to be visited more often when they

have:

• more librarian hours and more staff who work longer weekly hours,

• librarians who spend more of their time attending faculty meetings and

serving on key school committees,

• more library computers that provide access to the library catalog and the

Internet, and

• larger video and software collections and higher spending on collections

generally. (See Table 9.)

42

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 9. Indirect School Library Predictors of Tenth Grade Achievement Scores via

Visits to New Mexico High School Libraries, 2000

Library Visits

Variable

Group library

visits per 100

students

Group ISI

visits per 100

students

Individual

library visits

per student

Library Staffing

Librarian hours per 100 students

.307*

Total staff per 100 students

.621** .533** .239*

Total staff hours per 100 students .446** .491** .542**

Staff Activities

(Percent of Weekly Hours)

Attending faculty meetings

.276*

Serving on standards/curriculum committees .271*

Information Technology

Library computers per 100 students

.371**

.462**

.321**

Library Internet computers per 100 students .340** .407** .319**

Library catalog computers per 100 students .350** .479**

Information Resources

Videos per 100 students

.508**

.462**

.330**

Software packages per 100 students .364** .417**

Library expenditures per 100 students .353** .319** .431**

* p < .05, ** p < .01

Information Technology

Tenth grade test scores tend to rise with the availability of information

technology in the library. School libraries that have more computers,

especially ones that provide access to the library catalog and the Internet,

tend to have:

• more library staff working more hours per week, and

• librarians who spend a greater share of their time meeting with their

principals, providing in-service training to teachers, and providing reading

incentive programs for students. (See Table 10.)

43

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 10. Indirect School Library Predictors of Tenth Grade Achievement Scores

via Information Technology in New Mexico High School Libraries, 2000

Information Technology

Variable

Library

computers per

100 students

Library Internet

computers per

100 students

Library catalog

computers per

100 students

Library Staffing

Total staff per 100 students

.559**

.484**

.397**

Total staff hours per 100 students .702** .660** .467**

Staff Activities

(Percent of Weekly Hours)

Meeting with principal

.258*

Providing in-service training .327**

Providing reading incentive programs .265* .252*

* p < .05, ** p < .01

Print Volumes

Tenth grade achievement scores tend to improve with the size of a high

school library’s print collection. High schools that have larger print

collections in their libraries tend to have:

• more weekly hours of librarian and total library staffing, and

• librarians who spend more weekly hours planning and teaching

cooperatively with classroom teachers, meeting with other library staff,

and attending faculty meetings. (See Table 11.)

Table 11. Indirect School Library Predictors of Tenth Grade Achievement Scores

via Print Volumes Held by New Mexico High School Libraries, 2000

Variable Print volumes

Library Staffing

Weekly librarian hours

.266*

Total staff hours

.490**

Staff Activities (Weekly Hours)

Planning with teachers

.296*

Teaching cooperatively

.584**

Meeting with other library staff

.302*

Attending faculty meetings

.252*

* p < .05, ** p < .01

44

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Information Resources

Tenth grade achievement scores tend to increase with the size of video and

software collections in high school libraries and overall library spending per

student.

High school libraries with larger video and software collections and higher

library spending tend to be those with:

• higher levels of staffing (especially larger total numbers of staff and total

weekly staff hours), and

• librarians who spend more weekly hours providing reading incentive

programs for students. (See Table 12.)

Additionally, larger software collections are found in high school libraries

with:

• more librarian staff hours, and

• librarians who spend more weekly hours providing in-service training to

classroom teachers.

Table 12. Indirect School Library Predictors of Tenth Grade Achievement Scores

via Information Resources in New Mexico High School Libraries, 2000

Information Resources

Variable

Videos per

100 students

Software

packages per

100 students

Library

expenditures

per student

Library Staffing

Total librarian staff hours

.415**

Total staff per 100 students

.529** .476** .492**

Total staff hours per 100 students

.608** .594** .597**

Staff Activities

(Percent of Weekly Hours)

Providing in-service training

.292*

Providing reading incentive programs

.391** .440** .370**

* p < .05, ** p < .01

Controlling for Other School and Community Factors

To assert that the previously described relationships between school library

characteristics and academic achievement are cause and effect, it is

45

How School Libraries Improve Outcomes for Children:

The New Mexico Study

essential to control for other school and community factors. Before such an

analysis can be conducted, however, relationships among school library,

other school, and community variables must be explored to eliminate

redundant ones and to combine strongly related ones.

Correlation Analysis for School Library Development Variables

In the vast majority of recent school library impact studies, a critical factor

has been the level of development of the school library—its staffing, its

collections, and its funding. For this analysis, a somewhat more elaborate

School Library Development Factor was developed. Of the many details

collected about school library infrastructure in New Mexico elementary

schools, positive and highly statistically significant relationships with each

other were exhibited by six ratios: total staff hours per 100 students,

volumes per student, subscriptions per 100 students, video materials per

100 students, software packages per 100 students, and school library

expenditures per student. (See Table 13.)

An additional ratio, overall district expenditures per student, also correlated

very strongly with volumes per student as well as subscriptions, video

materials, and software packages per 100 students. Somewhat curiously,

this overall expenditures ratio did not correlate excessively strongly with the

school library expenditures ratio, despite the other very strong correlations

just identified.

The need to include overall district expenditures per student in the School

Library Development Factor is unusual, but reflects a reality worth

recognizing. Strong school libraries cannot exert an impact on academic

achievement in isolation. As this correlation indicates, a school library can

only be effective if it is integrated fully into the daily operation of its school.

District spending per pupil was not correlated so strongly with school library

characteristics as to explain away the importance of the latter. These

relationships were sufficiently strong, however, to indicate that higher levels

of school funding and stronger school library programs combine to exert a

positive influence on student performance. In other words, a school cannot

strengthen its library at the expense of other programs and expect positive

results.

46

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 13. Correlation Matrix for School Library Development Variables and District

Expenditures per Student for New Mexico High Schools, 2000

School

Library

Development

Variable

Librarian hours per 100

students

Total staff hours per

100 students

Volumes per student

Electronic references

per 100 students

Subscriptions per 100

students

Audio materials per 100

students

Video materials per 100

students

Software packages per

100 students

Library expenditures

per student

District expenditures

per student

Librarian hours per

100 students

1.00

Total staff hours per

100 students

1.00

Volumes per student

.590

**

1.00

Electronic references

per 100 students

1.00

Subscriptions per 100

students

.607

**

.732

**

.326

**

1.00

Audio materials per

100 students

.803

**

.540

**

1.00

Video materials per

100 students

.608

**

.518

**

.502

**

.331

**

1.00

Software packages per

100 students

.594

**

.510

**

.418

**

.288

*

.612

**

1.00

Library expenditures

per student

.597

**

.597

**

.300

*

.668

**

.327

**

.644

**

.844

**

1.00

District expenditures

per student

.613

**

.140 .640

**

.397

**

.619

**

.409

**

.530

**

1.00

* p < .05, ** p < .01

Italicized variable names are those for which either excessively strong relationships or an absence of

relationships with other variables justified eliminating them.

Factor Analysis for School Library Development Variables

To achieve data reduction, other data items in the school library

development category were discarded, and these six items were combined

into a single School Library Development Factor based on the results of a

factor analysis. All six of the items in question load highly on the resulting

factor, which explains over 68 percent of the variation among the

constituent statistics. These high factor loadings indicate that each of these

statistics makes a substantial contribution to the composite variable—the

School Library Development Factor—thereby achieving the goal of data

reduction without substantial loss of information. (See Table 14.)

47

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 14. Factor Analysis of School Library Development Variables for New Mexico

High Schools, 2000

School Library Development Variable Factor Loading

Total staff hours per 100 students .809

Volumes per student .796

Subscriptions per 100 students .865

Videos per 100 students .825

Software packages per 100 students .759

Library expenditures per student .885

District expenditures per student .841

Initial eigenvalue = 4.784, percentage of variation explained = 68.336, principa.841l component

analysis

Correlation Analysis for Other School and Community Variables

Previously reported bivariate correlations between school library

characteristics and reading scores are insufficient alone to claim cause and

effect relationships. A variety of school and community conditions might

create antecedent conditions that would explain both school library

characteristics and reading scores. For instance, it might be that rich

schools or communities with well-educated adults have both high reading

scores and strong school libraries. If that were the case, a strong school

library could not be credited as even a partial cause of high reading scores.

It would simply be another fortuitous consequence of favorable school or

community conditions. (See Table 15.)

Before such relationships can be explored, however, relationships among

school and community variables must be examined to ensure that there are

no confounding relationships between any of them. Examining these

relationships also sheds light on how they might be combined to achieve

data reduction, thereby simplifying the model to be tested.

The correlation coefficients for school and community variables indicate that,

indeed, they do fall into two groups. The three inter-related school variables

are the percentage of teachers with master’s degrees, teachers’ average

years of experience, and the teacher-pupil ratio. The three inter-related

community variables are the percentages of poor and minority students and

the percentage of adults in the community (including parents) who

graduated from high school. Each trio of variables in these two groups is

closely enough related to each other, both substantively and statistically, to

48

How School Libraries Improve Outcomes for Children:

The New Mexico Study

justify exploring via factor analysis the possibility of combining them into

school and community factors.

Table 15. Correlation Matrix for School and Community Variables for New Mexico

High Schools, 2000

School/

community

variable

Percent of teachers

with master’s degrees

Teachers’ average

years of experience

Teacher-pupil ratio

Percent of poor

students

Percent of minority

students

Percent of adults

high school

graduates

Percent of teachers with master’s

degrees

1.00

Teachers’ average years of experience .577** 1.00

Teacher-pupil ratio .381 ** .374** 1.00

Percent of poor students -.270* -.278* 1.00

Percent of minority students .287* .569** 1.00

Percent of adults high school

graduates

-651**

1.00

Italicized variables and correlation coefficients indicate related teacher variables.

Bold variables and correlation coefficients indicate related community variables.

Factor Analysis for Other School Variables

Not surprisingly, a factor analysis of the three school variables generated

one strong factor that combines teachers’ levels of education and experience

and teacher-pupil ratios conducive to smaller class sizes. Thus, a single

factor score can be utilized to summarize those three data elements. The

resulting School Factor explains almost two-thirds of the variation in the

constituent variables. (See Table 16.)

Table 16. Factor Analysis of Other School Variables for New Mexico High Schools,

2000

School Variable Factor Loading

Percent of teachers with master’s degrees .838

Teachers’ average years of experience .835

Teacher-pupil ratio .705

Initial eigenvalue = 1.897, percentage of variation explained = 63.220, principal component analysis

49

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Factor Analysis of Community Socio-Economic Status (SES) Variables

Similarly, a factor analysis of the three community variables generated a

single factor explaining more than two-thirds of the variation among

communities in terms of poverty, race/ethnicity, and adult educational

attainment. Notably, this Community Socio-Economic Status (SES) Factor is

a negative one, as a community’s score on it rises as the percentages of

poor and minority students increase and the percentage of adult high school

graduates decreases. (See Table 17.)

Table 17. Factor Analysis of Socio-Economic Status Variables for New Mexico High

Schools, 2000

Community SES Variable Factor Loading

Percent of poor students .930

Percent of minority students .729

Percent of adults high school graduates -.779

Initial eigenvalue = 2.004, percentage of variation explained = 66.793, principal component analysis

Regression Analysis for School Library Development, School, and Community SES

Factors

The preceding correlation and factor analyses generated a single factor score

each representing School Library Development, School, and Community

Socio-Economic Status (SES), setting the stage for a very simple,

straightforward regression analysis. That statistical technique makes it

possible to weigh the impact of School Library Development on achievement

scores, while controlling for School and Community SES differences that

might explain away any impact as something other than cause and effect.

(See Table 18.)

50

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 18. Regression Analysis of School Library Development, School, and

Community SES Factors with Achievement Scores for New Mexico High Schools,

2000

Factor

R

Square

R

Square

Change

Sig.

F

Change

Standardized

Beta

Coefficients

Sig. of

t for

Beta

Community SES .648 .404 .420 .000 -.663 .000

School library development .706 .471 .079 .025 .281 .023

School factor excluded.

For New Mexico high schools, the results indicate that the Community SES

Factor exerts by far the stronger influence on academic achievement. The

combined effects of poverty, racial/ethnic minority status, and low

educational attainment explained a whopping 42.0 percent of the variation in

students’ reading scores. An additional 7.9 percent of such variation,

however, is explained by the School Library Development Factor. After

these two factors are taken into account, the remaining School Factor was

excluded, as it explained no additional variation in student performance.

Such regression analysis results were not reproducible for elementary and

middle schools, not because school librarians and libraries are unimportant

at those levels, but because there was too little variation in librarian staffing

levels from school to school. At the elementary level, an overwhelming 68

percent of participating schools did not have a full-time librarian; conversely,

at the middle school level almost 65 percent reported precisely one full-time

librarian—no more, no less. With the key variable, librarian staffing, simply

not varying much from one school to another, it was not possible to isolate

the impact of school libraries from other school or community conditions at

these school levels.

Comparison of Highest & Lowest Scoring Schools

The foregoing analyses indicate that improvements in school library

programs can help to change one of the state’s lowest scoring schools on

state achievement tests to one of the highest scoring schools. The following

tables compare the 25 highest scoring schools with the 25 lowest scoring

ones.

51

How School Libraries Improve Outcomes for Children:

The New Mexico Study

With only a few exceptions, schools with higher average test scores

• have demonstrably better developed school library programs—i.e., higher

levels of staffing, collection development, and funding;

• have library staff who spend more time engaged in activities that

contribute to collaborative teaching and learning; and

• have more extensive and sophisticated computer networks extending the

reach of the school library. (See Tables 19 through 24.)

52

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 19. Selected Library Media Statistics for New Mexico Elementary Schools

Scoring Highest and Lowest on the Fourth Grade Achievement Test, 2000

Elementary school averages

(4

th

grade)

Variable

25

highest

scoring

schools

25

lowest

scoring

schools

Percent

difference

(lowest to

highest)

4

th

Grade Reading Scores

(Percent proficient & above) 85.78 30.82 178%

Staffing (Weekly hours)

LM Specialist hours 16.28 14.13 15%

per 100 students 4.27 3.78 13%

Total LM staff hours 38.03 41.51 (8%)

per 100 students 13.31 13.05 2%

Staff Activities (Weekly hours)

Offering reading incentive activities

6.22 6.58 (5%)

Percent 17.77 14.85 20%

Meeting with library staff 0.57 0.51 12%

Percent 1.59 1.22 30%

Meeting w/standards, curriculum committees 0.54 0.44 23%

Percent 1.39 1.01 38%

Managing computers, network 4.90 3.93 25%

Percent 13.91 7.67 81%

Weekly LMC Usage

Individual visits for information skills instruction 228.29 231.70 (1%)

per student 0.60 0.52 15%

Group visits for information skills instruction 15.65 15.43 1%

per 100 students 4.47 3.78 18%

Technology

Number of LMC computers

4.24 4.47 (5%)

per 100 students 1.77 1.19 49%

Number of LMC computers with:

- Internet connection 3.62 3.36 8%

per 100 students 1.15 0.85 35%

- Access to school library catalog 3.68 3.64 1%

per 100 students 1.07 0.86 24%

- Access to library databases 2.59 2.49 4%

per 100 students 0.83 0.63 32%

School computers not in LMC with access to LMC 39.25 34.08 15%

per 100 students 9.11 9.20 (1%)

School computers not in LMC w/access to LMC

with:

- Internet connection 37.59 24.24 55%

per 100 students 8.80 7.72 14%

- Access to school library catalog 21.49 10.72 100%

per 100 students 4.52 1.79 153%

- Access to library databases 20.07 6.22 223%

per 100 students 4.36 1.27 243%

53

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 20 . Selected Library Media Statistics for New Mexico Elementary Schools

Scoring Highest and Lowest on the Fourth Grade Achievement Test, 2000

Elementary school averages

(4

th

grade)

Variable

25

highest

scoring

schools

25

lowest

scoring

schools

Percent

difference

(lowest to

highest

4

th

Grade Reading Scores

(Percent proficient & above)

85.78 30.82 178%

School Library Collection

Number of print volumes

11,038.39 7,215.06 53%

per student 31.40 26.74 17%

Number of electronic reference titles 10.45 8.68 20%

per 100 students 3.00 1.11 170%

Current magazine & newspaper subscriptions 13.10 14.45 (9%)

per 100 students 4.88 4.83 1%

Audio materials - tapes, CDs, LPs 94.07 115.40 (18%)

per 100 students 33.21 28.16 18%

Video materials – tapes, disks 221.43 215.40 3%

per 100 students 62.35 57.24 9%

Computer software packages for in-school use 7.95 11.35 (30%)

per 100 students 4.68 3.36 39%

54

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 21. Selected Library Media Statistics for New Mexico Middle Schools

Scoring Highest and Lowest on the Eighth Grade Achievement Test, 2000

Middle school averages (8

th

grade)

Variable

25

highest

scoring

schools

25

lowest

scoring

schools

Percent

difference

(lowest to

highest)

8

th

Grade Reading Scores

(Percent proficient & above) 53.51 17.43 207%

Staffing (Weekly hours)

LM Specialist hours

29.21 23.80 23%

per 100 students 6.44 6.62 (3%)

Staff Activities (Weekly hours)

Planning with teachers

2.60 2.32 12%

Percent 5.65 5.23 8%

Teaching with teachers 11.25 9.57 18%

Percent 25.27 21.67 17%

Providing in-service training 1.75 1.23 42%

Percent 3.42 2.34 46%

Offering reading incentive activities 2.73 2.59 5%

Percent 5.79 6.19 (6%)

Doing collection development 2.99 2.74 9%

Percent 6.30 5.76 9%

Meeting with library staff 0.78 0.62 26%

Percent 1.51 1.57 (4%)

Meeting with principal, other administrators 0.56 0.68 (18%)

Percent 1.17 1.47 (20%)

Attending faculty or staff meetings 0.83 0.81 2%

Percent 1.75 1.87 (6%)

Meeting w/standards, curriculum committees,

teams

0.53 0.51

4%

Percent 1.08 1.11 (3%)

Managing computers, network 6.36 6.79 (6%)

Percent 14.81 13.85 7%

Weekly LMC Usage

Individual visits to LMC 356.14 330.53 8%

per student 0.82 1.43 (43%)

Group visits to LMC 15.76 13.95 13%

per 100 students 5.11 4.77 7%

Individual visits for information skills instruction 135.35 110.84 22%

per student 0.35 0.75 (53%)

Group visits for information skills instruction 10.00 8.26 21%

per 100 students 2.97 2.67 11%

55

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 22. Selected Library Media Statistics for New Mexico Middle Schools Scoring

Highest and Lowest on the Eighth Grade Achievement Test, 2000

Middle school averages (8

th

grade)

Variable

25

highest

scoring

schools

25

lowest

scoring

schools

Percent

difference

(lowest to

highest)

8

th

Grade Reading Scores

(Percent proficient & above) 53.51 17.43 207%

Technology

Number of LMC computers 10.97 9.44 16%

per 100 students 3.19 2.78 15%

Number of LMC computers with:

- Internet connection 9.74 8.68 12%

per 100 students 2.91 2.45 19%

- Access to school library catalog 7.46 5.05 48%

per 100 students 2.21 1.37 61%

- Access to library databases 7.29 4.11 77%

per 100 students 2.15 1.38 56%

School computers not in LMC w/access to LMC

with:

- Internet connection 64.36 63.94 1%

per 100 students 19.01 27.32 (30%)

- Access to school library catalog 23.06 16.84 37%

per 100 students 3.97 6.01 (34%)

- Access to library databases 44.84 20.06 124%

per 100 students 16.26 13.13 24%

School Library Collection

Number of print volumes

10,864.24 8,193.75 33%

per student 33.97 26.57 28%

Number of electronic reference titles 12.13 10.20 19%

per 100 students 3.65 5.25 (30%)

Current magazine & newspaper subscriptions 23.89 23.19 3%

per 100 students 6.57 11.58 (43%)

Audio materials - tapes, CDs, LPs 85.94 81.83 5%

per 100 students 19.69 28.65 (31%)

Computer software packages for in-school use 16.70 11.83 41%

per 100 students 4.88 9.20 (47%)

56

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 23. Selected Library Media Statistics for New Mexico High Schools Scoring

Highest and Lowest on the Tenth Grade Achievement Test, 2000

High school averages (10

th

grade)

Variable

25

highest

scoring

schools

25

lowest

scoring

schools

Percent

difference

(lowest to

highest)

10

th

Grade Reading Scores

(Percent proficient & above) 91.32 71.73 27%

Staffing (Weekly hours)

LM Specialist hours

29.80 31.93 (7%)

per 100 students 5.65 4.83 17%

Total LM staff hours 63.61 65.84 (3%)

per 100 students 19.71 14.57 35%

Staff Activities (Weekly hours)

Planning with teachers 4.28 3.98 8%

Percent 6.43 6.00 6%

Teaching with teachers 14.59 12.22 19%

Percent 21.75 16.58 31%

Doing collection development 4.41 4.18 6%

Percent 7.28 5.93 23%

Meeting with library staff 1.28 1.08 19%

Percent 1.81 1.92 (6%)

Meeting with principal, other administrators 1.16 0.99 17%

Percent 2.01 1.61 25%

Attending faculty or staff meetings 0.80 0.59 36%

Percent 1.33 1.06 25%

Meeting w/standards, curriculum committees,

teams 1.06 0.76 39%

Percent 1.88 1.37 37%

Weekly LMC Usage

Individual visits to LMC 356.45 319.24 12%

per student 1.15 0.78 47%

Group visits to LMC 16.68 12.96 29%

per 100 students 3.64 2.18 67%

Individual visits for information skills instruction 157.13 157.81 0%

per student 0.36 0.21 71%

Group visits for information skills instruction 13.42 9.54 41%

per 100 students 2.98 1.62 84%

57

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Table 24. Selected Library Media Statistics for New Mexico High Schools Scoring

Highest and Lowest on the Tenth Grade Achievement Test, 2000

High school averages (10

th

grade)

Variable

25

highest

scoring

schools

25

lowest

scoring

schools

Percent

difference

(lowest to

highest)

10

th

Grade Reading Scores

(Percent proficient & above) 91.32 71.73 27%

Technology

Number of LMC computers 13.26 11.68 14%

per 100 students 3.41 2.22 54%

Number of LMC computers with:

- Internet connection 12.62 10.67 18%

per 100 students 3.27 1.92 70%

- Access to school library catalog 10.69 8.37 28%

per 100 students 2.21 1.25 77%

- Access to library databases 11.70 9.81 19%

per 100 students 2.87 1.64 75%

School computers not in LMC with access to LMC 75.50 77.48 (3%)

per 100 students 27.34 26.13 5%

School computers not in LMC w/access to LMC

with:

- Internet connection 76.39 2%

per 100 students 26.93 25.67 5%

- Access to school library catalog 44.67 36.67 22%

per 100 students 11.79 8.86 33%

- Access to library databases 70.83 64.54 10%

per 100 students 24.90 19.85 25%

School Library Collection

Number of print volumes

14,502.57 12,380.88 17%

per student 31.43 31.06 1%

Number of electronic reference titles 17.70 38.89 (54%)

per 100 students 19.81 16.47 20%

Current magazine & newspaper subscriptions 37.82 36.88 3%

per 100 students 8.68 9.56 (9%)

Audio materials - tapes, CDs, LPs 117.50 67.04 75%

per 100 students 27.73 19.90 39%

Video materials – tapes, disks 437.22 475.39 (8%)

per 100 students 146.96 96.59 52%

Computer software packages for in-school use 16.18 11.22 44%

per 100 students 9.83 4.47 120%

Annual Operating Expenditures

Total operating expenditures

$12,003.97

$11,362.57 6%

per student $25.49 $22.08 15%

78.27

58

How School Libraries Improve Outcomes for Children:

The New Mexico Study

Conclusions

New Mexico achievement test scores rise with the development of school

library programs. The relationship between school library development and

test scores is not explained away by other school or community conditions at

the high school level. There was insufficient variation in librarian staffing to

make similar claims for the elementary and middle school levels.

School Library Development

New Mexico achievement test scores tend to rise with increases in:

• school librarian and total library staff hours per 100 students;

• print volumes per student;

• periodical subscriptions, video materials, and software packages per 100

students; and

• school library expenditures per student.

Whatever the current level of development of the school library program,

these findings indicate that incremental improvements in its staffing,

collections, and funding will yield incremental increases in achievement test

scores. The only caveat is that school library spending cannot exert a

positive influence on academic achievement if it comes at the expense of

other school programs.

School & Community Differences

The impact of school library development on academic achievement at the

high school level cannot be explained away by:

• school differences, including:

• the percentage of classroom teachers with master’s degrees,

• teachers’ average years of experience, and

• the teacher/pupil ratio, or

• community differences, including:

• the percentage of schoolchildren living in poverty,

• the percentage of schoolchildren belonging to racial/ethnic minority

groups, and

• the percentage of adults who graduated from high school.

59