Understanding the gap between an AmericaReads program and the tutoring sessions: The nesting of challenges.

Abstract. Increasingly, students receive tutoring for reading assistance. While some information exists about the features of various programs, little is known about the overlay between the planned tutoring program and what a tutor actually does. This investigation fills that void by exploring whether tutors in an AmericaReads program implement the planned program. To make this determination, artifacts stemming from the tutoring sessions, audiotapes of representative lessons, interviews of supervisors, and questionnaires completed by tutors were examined. The findings unveiled an inconsistent fidelity to the tutoring expectations. To further understand why these mismatches occurred, specific features of this AmericaReads program were examined. Here, organizational difficulties were detected that further explained the challenge of a program like AmericaReads to ensure that tutor actions reflect the planned program.

Within the field of education, as well as the community at large, students' reading attainment continues to be a key concern. Recent performances on national and state tests do little to assuage that concern. Since many experts believe that one-to-one tutoring can affect the reading achievement of young children (Ainsworth, 1995; Gaustad, 1992; Topping, 1998; Wasik, 1998), schools and service agencies often turn to tutoring for assistance. In doing so, many questions arise about who should tutor and how the tutoring should proceed. In this article, we describe an AmericaReads program instituted, in part, to provide the basis for examining existing tutoring programs, as well as the issues and research linked to them.

Tutoring Issues

Defining Tutoring

Despite variations in the components of tutoring programs and disagreements among researchers on the definition, tutoring, in general, is characterized by an interaction between a tutor who assists a tutee. The information exchanged during the tutoring session focuses on curriculum content, particularly on an area of weakness for the tutee. Typically, the purpose of tutoring is one of the following: 1) preventing academic problems, 2) providing remediation for those having difficulty, 3) maintaining students' current academic status, or 4) enriching children's academic abilities. Generally, the components of effective tutoring programs involve trained tutors, structured materials, and specific procedures for implementation. Descriptions of tutoring programs show a wide variety of differences in characteristics such as tutee age, materials used, time and place of tutoring session, amount and quality of training, evaluation, frequency and duration of sessions, and types of tutors (Topping, 1998). This study exa mines tutoring designed to promote reading.

Selecting and Training Tutors

In selecting tutors, some programs use teachers. Being a teacher, however, does not excuse the tutor from subsequent training regarding a program's specific features. For example, in Reading Recovery, a one-to-one 1st-grade program developed by Marie Clay that uses teachers to tutor, tutors complete intensive training conducted by teacher trainers. These trained teachers/tutors then work one-to-one with 1st-grade students (DeFord, Lyons, & Pinnell, 1991).

Success for All, an intervention program for 1st-graders having difficulty in reading, also uses certified teachers as tutors. In addition to their general professional background, teachers intending to tutor using Success for All strategies generally receive three days of inservice training and are regularly observed by the program facilitator (Slavin, Madden, Dolan, & Wasik, 1996; Slavin, Madden, Karweit, Livermon, & Dolan, 1990).

Finally, Early Steps, as implemented by Santa (Santa & Hoien, 1999), is a one-to-one, pull-out tutoring program. Like the previously described programs, it taps an educationally experienced group--selecting from 1st-grade teachers, Title I teachers, and a Language Arts Coordinator. Tutors in the Early Steps tutoring program receive intensive training in the form of a graduate class that meets weekly for one month and then monthly thereafter. The Language Arts Coordinator serves as the on-site teacher trainer.

Other programs depend upon community volunteers. In part, this decision stems from a need to serve more students in and out of school. The chance to serve more students does not eliminate the need to serve them well. Simply stated, the best of volunteer training does not compensate for the rich background that teachers and reading specialists bring to a tutoring session. As Wasik (1998) cautions, "Without adequate training, untrained volunteers could be more of a hindrance than a help to a struggling child" (p. 284). Perhaps this is what led Snow, Burns, and Griffin (1998) to suggest that volunteer tutors be used to provide practice and motivational support to struggling readers only.

Those programs that do use volunteer tutors to teach children take several steps to compensate for the tutor's lack of formal understanding about reading. As recommended by Shannon (1998) and Wasik (1997), quality volunteer tutoring programs provide supervision, feedback, and training in reading. We briefly review a few of these programs below.

The Howard Street Tutoring Program (Morris, Shaw, & Perney, 1990) is an example of a program that demonstrates Shannon's and Wasik's recommendations. This program is an after-school program for children in grades 2 and 3 who are having problems with reading. This program features community unpaid volunteer tutors, ranging from college students to retirees. A teacher trains the tutors and subsequently becomes their tutoring supervisor, creating lesson plans for each tutoring session. A professional teacher designs the lessons delivered by the volunteers. During training, the tutors spend three or four days observing tutoring sessions and discussing techniques. Then they conduct tutoring sessions and receive feedback from the supervisor. Finally, they begin the tutoring sessions. Every child is tutored twice a week for one hour.

Some volunteer tutoring programs use volunteers to deliver a Reading Recovery-type program. In these instances, tutors receive extensive training (often about 150 hours of instruction on children's development as readers, as well as appropriate instructional strategies) before beginning their work with children. In addition, the future tutors observe teacher-leaders in tutoring sessions. During their first week of tutoring, the volunteers are observed by a teacher-leader, who provides feedback and training as needed. Finally, throughout the school year, volunteers and teacher-leaders spend two hours per week together discussing students and strategies. Such programs clearly recognize the importance of initial and ongoing training.

HOSTS is a well-known nationwide volunteer tutoring program. In HOSTS programs, volunteer tutors are trained and observed by a teacher, usually the school's reading specialist or Title I teacher, who also organizes the materials, places volunteers with children, and develops a diagnostic plan for each child. This teacher is the program coordinator. The coordinator receives three days of initial training to review HOSTS materials and to learn how to use the materials to make a diagnostic plan for each child/tutee. The coordinators attend three more days of training throughout the year and have yearlong access to the HOSTS hotline for answers to their questions. At each site, the coordinators provide the ongoing training and feedback for the tutors who actually deliver services to the children. A recent study (Kellner, 1999) suggests that implementation of the HOSTS program is very dependent on the program coordinator. Kellner's close examination of three HOSTS sites in a single district revealed three very different programs. At one site, the teacher coordinator implemented a program modeled on Reading Recovery, with extensive tutor training and ongoing supervision. At a second site, the program coordinator wrote a single, "one lesson fits all" lesson plan on the board for all tutors to follow with all tutees. The tutors received one hour of training on the classroom's tutoring procedures. At a third site, the program coordinator prepared folders with skill activities, which tutors collected for use with their tutees as they entered the tutoring room. These tutors were provided with a three-hour training session exploring the range of activities to be implemented with the tutees. Which of the three reveals fidelity to the HOSTS program?

While literature describing various programs' tutor selection and training procedures is available, little information is available on how the selected tutors use the training information in their actual tutoring of young children. Juel's (1996) study of college students' tutoring of at-risk 1st-graders comes close to doing what Adler (1999) suggests is greatly needed: a description of how the tutoring sessions reflect the training the tutors have received. The goal is to answer the question: Do tutors do what they have been instructed to do? While Juel closely examined the instructional exchanges between the tutor and the young children, she did not link her observations of the tutors' tutoring behaviors to their training to be tutors. Adler contends that this information is important for researchers and teachers to know.

Concerned that tutors, while trained, may not implement the recommended lessons as presented to them in their training and through the materials provided, Adler set out to study what college-age tutors do during tutoring sessions. Studying a university's AmericaReads tutoring program for a year, she learned that although the tutors generally implement the broad components of the program--that is, they had the tutees reread a previously read book and read a new book--they do not consistently implement all components as directed. For example, 21% of the tutees never experienced word study, a "required" component from this university's perspective. Furthermore, Adler discovered that tutors often followed their own intuition, rather than seeking professionals' advice. Adler concludes with the plea for more research on

how . . . tutors understand their training and carry out what they learn through tutoring sessions. . . . Clearly, more studies are needed to better understand the nature of programs, like the AmericaReads Challenge Program, which rely on relatively inexperienced tutors to work with struggling readers. (p. 22)

The Heart of the Tutoring Program: The Tutoring Session

As Adler and Kellner suggest, the true nature of a tutoring session can be learned from what happens during each tutoring session, not from the program's name. Several programs describe themselves as following the Reading Recovery model. The Reading Recovery model consists of reading familiar stories, engaging in daily writing activities, and introducing a new book. The approach is holistic, with a heavy emphasis on reading connected text. Tutees work on word recognition strategies or other needs that surface during a text's reading. The onus is on the tutor to use a student's performance to direct subsequent instructional actions, which are termed "teaching points" by Reading Recovery teacher trainers and tutors (DeFord, Lyons, & Pinnell, 1991).

Success for All is one tutoring program that claims to model itself on Reading Recovery. However, success for all tutors focus on mastery of specific skills taught during classroom reading instruction. Therefore, coordination between the tutoring and the classroom program assumes importance.

In addition, many volunteer tutor programs model their tutoring sessions after Reading Recovery even though, as previously explained, this program originally used teachers as tutors. The basic lesson components shared by these programs include rereading a familiar book, completing a writing activity, engaging in a word analysis/letter-sound activity, and reading a new book. Some programs have added other components, such as phonemic awareness activities, word recognition strategies, mastery of letters and words, and phonics. For example, tutors in Juel's program are encouraged to build their lessons by selecting four out of the following seven components: 1) reading children's literature (either the child reads or the tutor reads); 2) a writing activity, involving a story, a message, or a topic chosen by the child; 3) reading "build-up" readers, which are books with high-frequency or frequently repeated, words; 4) journal writing, in which the tutor and child participate; 5) alphabet books, with pictures and words for each letter; 6) phonemic awareness activities, such as reading books with rhyming words and recognizing patterns in words; and 7) letter-sound activities, such as spelling and recognizing patterns in words (Juel, 1996).

Of course, not all programs follow the Reading Recovery model. HOSTS, for example, provides a different set of components and structure. HOSTS tutors follow computer-based lesson plans that focus on a child's weak areas. This structured lesson plan designates the materials for the session. Within this plan, tutors receive worksheets that focus on isolated skills for their tutees to complete. Tutees spend little time reading books.

Variable features of volunteer tutoring programs include time and place of tutoring sessions, type of tutor, age of tutee, eligibility for tutoring services, and number of tutoring sessions. While most of the programs take place in schools, some take place in community buildings such as community centers, child care centers, and university buildings. For example, the Howard Street program takes place in a community building in a disadvantaged neighborhood. Also, some programs do not pull children out of the classroom. Instead, they tutor in the classroom (e.g., Success for All) or after school.

Acknowledging Variations Across Programs

The ages of tutees range from kindergarten to 6th grade. Children selected for tutoring programs tend to experience difficulty in reading, and they live in high poverty areas. Eligibility is determined by test scores, classroom performance in reading, teacher recommendation, and/or principal recommendation. Most of the children are considered at-risk for reading failure.

Most tutoring sessions last 30 minutes. A few last 45 minutes to an hour; one tutoring program, Early LD, offers tutoring for 10 minutes per day. However, tutoring programs seem to be divided on the number of tutoring sessions provided. Some of them provide tutoring two times per week, some are two to four times per week, and some provide tutoring every day of the week. Reading One-to-One and Book Buddies claim that the number of tutoring sessions a child receives is related to achievement. Juel (1996) claims that how the time is spent (i.e., the choice of activities) matters more than how long the session lasts.

While there is documentation that tutoring programs by professionals have yielded positive results, whether or not volunteer tutoring programs actually affect students' reading remains unclear. While some programs document encouraging trends, few have received sophisticated attention to determine whether the tutoring accounts for observed reading gains. In addition, studies comparing the efficacy of various tutoring pools have not received enough attention. Therefore, many challenges remain: addressing the issues and implications surrounding the role of volunteers in the schools as well as appropriate training; identifying the components of effective tutoring programs; and, of utmost importance, determining whether or not children benefit.

To add to what is known and to explore these questions of ongoing importance, the authors investigated how the selected tutors use the training information in their actual tutoring of young children in an AmericaReads program. Two questions guided this inquiry: 1) Do tutors carry out recommended lessons as presented to them in their training and materials provided? and 2) If tutors stray from their trainers' expectations, how do the staff assistants and others respond? Understanding these questions required consideration of the challenges inherent in establishing a high-quality volunteer program aimed at benefiting its tutees' reading achievement--a piece not addressed in the accounts of existing programs. Consequently, the following questions assumed importance: What categories of challenges and dilemmas emerge when a university accepts responsibility for an AmericaReads volunteer tutoring program? How does the staff respond to these challenges? and Do the solutions correct or exacerbate each challenge? In combination, these questions presage a wider study that will address the reading achievement of the young children in this university's AmericaReads program.

Method

This AmericaReads Program

The program involved a partnership between two mid-Atlantic school districts, service agencies within those districts, and a university. The participating school districts and agencies identified tutees in grades K-3 and provided suitable tutoring locations. University professors and staff trained the tutors, made tutor-tutee matches, scheduled the tutoring sessions, made resources available to tutors, and evaluated the project.

The total number of tutoring sessions per tutee ranged from nine to 32. Each session lasted for 30 minutes. Due to tutee schedules, some tutees were tutored once each week, while other tutees were tutored three times a week. Each tutor received training about the general format for each session and had access to materials. Tutors also received a training manual, designed by two participants in this program for the state, for further explanation and future reference. Specifically, tutors learned general literacy tutoring strategies, the attributes of a quality tutoring session, and specific tutoring strategies for use with emergent and novice readers. Sessions with emerging readers had the following components: shared reading of a familiar text, read aloud of a new book followed by discussion, a phonemic awareness or other strategy activity, and a writing activity. The following components framed a session with a novice reader: rereading of a text, reviewing previously read content or strategies, reading a ne w book and discussing it, participating in a reading strategy activity, and completing a writing activity related to the text. Central to the sessions for novice and emerging readers was engagement with a good book.

The goals of this program were to train and place undergraduates to serve as reading tutors for at-risk children in kindergarten through grade 3, to explore the effectiveness of the state's recently issued reading tutoring manual, and to assess the impact of the reading tutoring program on young children's reading achievement. This study focused on the quality and consistency of the prescribed tutoring sessions.

Participants

A total of 340 children were served by this program. Over the course of the year, 55 children moved from the participating school or discontinued their participation in the service agency program and thus the tutoring program. Of the remaining 285 children, complete data were available on 151. These children are the participants in this study. The children were in kindergarten through grade 3. They came from districts that provided approximately 37 percent of the students a free or reduced lunch, and that have a 39 percent minority population. The tutees were selected by school district personnel who considered them atrisk of reading failure and, therefore, most in need of additional reading support.

The 76 tutors (64 females and 12 males) came from a Federal Work-Study Program. Therefore, they were all enrolled in an undergraduate program at the local university and were eligible for participation in the Federal Work-Study Program. Before tutoring began, they received seven hours of mandatory training. This included information about general literacy tutoring strategies, the attributes of a quality tutoring session, and specific tutoring strategies for use with emergent and novice readers. In addition, they were provided information on how to administer the pretest assessment instruments and how to use the results to inform their selections of books and choices of activities to use with the children. Following this initial training, the tutors received ongoing, on-site assistance from the Project Assistant and supervisors.

Supervisors

Initially, a Vista volunteer coordinated the program by herself. When her work load proved unmanageable, two coordinators were employed. One was a doctoral student at the University who was assigned this task as part of her public service assistantship. The other, who had previously worked as a teacher and principal, was hired with grant money.

University Personnel

Two university faculty members and one staff member oversaw the program. As previously explained, the university faculty wrote the training manuals and conducted the initial training sessions. In addition, they designed the research project linked to this program. The staff member assisted with various managerial features of the program and attended the staff meetings.

Data Collection and Analysis Procedures

The data collected and analyzed informed the authors' central intention to focus on the integrity of the tutoring sessions. They also informed the related questions that linked to this purpose--the challenges of implementing a volunteer tutoring program.

Three sources offered insights as to what transpired during AmericaReads tutoring sessions: 1) the tutor-maintained Record Keeping Forms, whereby tutors detailed their plans for each tutoring session (see Appendix A); 2) tutor audiotapes of a representative lesson; and 3) questionnaires completed by all tutors that included Likert-scaled and open response items (see Appendices B and C). A final source--individual interviews with the supervisors--afforded tutors insights regarding the program's strengths, needs, and shifts. (Refer to Appendix D for a listing of the interview questions.)

Record Keeping Forms. A form was randomly selected from those completed by 30 tutors and analyzed in light of the components tutors were expected to implement during each tutoring session. The sequence of implementation of the components was labeled as either consistent or inconsistent with the prescribed order of activities. Percentages of consistency were calculated. This descriptive statistical account of individual tutoring sessions, as well as an aggregate of them, indicated the overlap or disconnection between the sessions as intended and implemented (see Table 1).

Tutor Audiotapes. All tutors audiotaped one tutoring session. From the tapes submitted, the coordinator and supervisors randomly selected 10 tapes for transcription, which were read by two people. First, the activities were "chunked" to understand the tutoring session components and to compare the components against that which tutors had been instructed to use. Then, within each chunk, tutor behaviors were analyzed to understand how the tutor scaffolded the child's pronunciation of unknown words and comprehension of the book(s) read. This coincides with an overall expectation of conversation analysis to "understand the nature of roles, relationships and social norms in settings and thereby discern the underlying structure of social reality" (Adler & Adler, 1994 p. 387).

Questionnaires. Responses to the Likertscaled questionnaires were analyzed quantitatively. This resulted in descriptive statistics for each statement (see Table 2). The responses to the open-ended questions were analyzed qualitatively. Specifically, the responses for each question were listed by ideas. When appropriate, these ideas were collapsed under a representative label. A simple tally tracked the resurfacing of ideas. The coding and tallying of ideas continued until the responses for each item were completely considered.

Supervisor Interviews. Each individually conducted interview was audiotaped and transcribed. Analysis of their responses began with each question. Each idea was numbered and collapsed with comparable statements. Then, these grouped responses were labeled. Once completed, this listing and categorizing of responses captured the range and convergence of ideas offered across questions and supervisors.

Following the analysis of each data source, the researchers collapsed information across sources in order to confirm or disconfirm a finding. This offered a form of triangulation (Pitman & Maxwell, 1992). Upon completion, these discrete and linked analyses framed an understanding of the research questions.

Results and Discussion

Transcriptions of the audiotapes of 10 tutoring sessions and the tutors' RecordKeeping Form descriptions of 30 tutoring sessions provide the primary sources of information for the following findings. Survey data on the frequency with which all tutors reported on the implementation of the instructional components of their tutoring sessions also are included. (Table 2 provides information from the 80 survey respondents.) Together, these data provide important insights into tutors' understandings of tutoring young children in reading, and they highlight the tutoring-related nested challenges. Data are presented by the prescribed tutoring instructional components.

Reread. Tutors were directed to begin each tutoring session with the tutee rereading the leveled book that was read during the tutee's previous tutoring session. Presumably the tutee had successfully read the book in the previous session and, therefore, would begin the session feeling comfortable, and be able to read with ease.

Tutors in this study did not consistently do as they were directed. Tutors self-reported beginning at least 67.4% of their tutoring sessions with the tutee rereading the previous session's new book. The RecordKeeping Form data suggest a higher percentage (86.6%) and the transcriptions of the 10 tutoring sessions suggest a lower percentage (40%). Transcription data indicate that an additional 30 percent of the tutors did have their tutee reread the previous session's book eventually, after the tutee completed a skill activity (e.g., tutee read words from flash cards, tutee copied words from a list created by the tutor that ended in "ly," tutee read names of the alphabet from letter cards). Thirty percent of the tutors did not provide tutees with the opportunity to reread the previous session's book; one tutor began with the reading of a new book, one tutor read the previous session's book to the tutee, and one tutor moved from the tutees' reading of words on flash cards to the reading of a new book.

The tutoring session transcription data also provide insights into the tutors' skill at selecting leveled books appropriate for their tutees. Recall the researchers' expectations that the tutees would reread these books with ease. Again, the expectations were not met. The data suggest the tutees struggled with the rereadings, requiring the tutors to repeatedly scaffold the tutees' decoding of words by directing them to sound the word out; by giving clues ("If there is no 'e,'it's 'her'. If there is an 'e,'it's 'here'."); telling the word; reading along with the tutee; reading the book to the tutee, with the tutee reading the predictable phrases or alternate sentences; providing the letter sounds and asking the tutee to blend the sounds into the words; telling the tutee to use classroom resources to help them read the word; or telling the tutee to reread the sentence. The tutors provided a range of 2 to 14 decoding scaffolds (mean = 10.25). The frequency of supports needed led to the conclusion that, with one exception, the books were too difficult for the tutees. Of course, when the tutor, not the tutee, read the book, it was not possible to reach any conclusion regarding the tutee's ability to read. Furthermore, the data suggest that tutors expected tutees to read every word correctly; tutors directed tutees' attention to each and every misread word and required that the word be read correctly before they permitted the tutee to read on. One tutor did recognize the challenge the book presented to a tutee. "Do you want to try a different one? Some of these are graded accordingly. So, we'll try this one."

Reading a New Book. Tutors were directed to follow the rereading of the tutee's last tutoring session book with the reading of a new book. Before reading the book, tutors were to introduce the book to the tutee. Survey data suggest that 82.5 percent of the tutors engaged the tutee's attention in the book before asking the tutee to read at least most of the time. Transcription data suggest a smaller percentage (60%) of these tutors introduced the book to the tutee; those tutors who did direct the tutee's attention to the book read the book's title and asked the tutee to predict what the book would be about, connected the book with other books the tutee had read with the same character, talked about the book's vocabulary that might be unfamiliar to the tutee and was necessary to understand the story, or did a "picture walk" through the book. The RecordKeeping Form did not specifically ask tutors to provide information about the book introduction activities.

Following the introduction, tutors were to support the tutee's reading of the book. Tutors were to have selected new books that the tutees would be able to read with reasonable ease and minimal tutor support. To guide their book choices, tutors were trained to use pretest data and data gathered from each tutoring session. The three data sources provide confirmation that, in fact, tutors did provide a new book for their tutees to read; survey data (90%), Record Keeping Form data (86.7%), and transcriptions (100%) all indicate that tutees read a new book during most tutoring sessions.

Unfortunately, again, the transcription data suggest that, with one exception, the tutors' new book selections greatly challenged the tutees. The mean number of decoding scaffolds offered ranged from 2 to 15 (mean = 7.82). The decoding strategies the tutors used in the book rereading instruction component were used again to support the tutees' reading of the new book, and again the tutors generally required the tutees to read each word correctly: "Try again." "Keep guessing." "Wait! Look at that word." "What does that say? What does it start with?" "Wait; use the picture to help you." Tutors often praised tutees when they were able to decode challenging words, sometimes with specific praise (e.g., "You used the picture to help you") and other times with general praise (e.g., "Good job"). In two instances, the book selected proved so challenging the tutors shifted away from the tutees reading, to the tutor talking the tutee "through" the book, or to the tutor reading the book with the child reading the refrain. The tutees complained (e.g., "Is this the last page?") and engaged in avoidance behavior ("Hey, I can't read the big letters." "I have a question." "Wait! I have a question.").

Talk About the Book. Tutors were directed to follow the tutees' reading with talk about the book. Transcription (60%), Record Keeping Form (63.3%), and survey (79% at least most of the time) data combine to describe the tutors' activity following the reading of the book. Transcription data provide details of the talk. Tutors connected the book with the tutee's personal life, or asked recall questions, or asked sequence questions. Those tutors who did not follow the reading with talk moved immediately to the next component.

Teaching a Reading Strategy. Survey data suggest that at least most of the time, 66.2% of the tutors followed talk about the book with an activity that focused the tutees' attention on the print. Record Keeping Form and transcription data describe the activities tutors implemented with their tutees and noted under Teaching a Reading Strategy. Tutors engaged tutees in one or more of the following strategies: sequencing pictures from the book, playing hangman, playing word bingo, making a book by using the book as a model, drawing a picture and then writing a sentence, making a graphic organizer like a Venn diagram to compare the child with the character, finding words in the book that begin with a particular letter, finding specific letters in the book, drawing pictures of objects that begin with a letter, making an alphabet book with pictures of objects from the book that begin with each letter, writing sentences and having the tutor correct incorrectly spelled words, and writing two sentences with words tha t the tutee had mispronounced during the reading. These strategies, at best, provide practice around a reading skill. However, they stray from the intention to offer instruction about a reading strategy. Of more concern (as Table 1 indicates), tutors often omitted any attention to this tutoring expectation.

New and Old Challenges. Lack of tutor reliability resurfaced here. Fidelity to the model, as prescribed in the training materials, was inconsistently noted. Then, new challenges were nested within each of the tutoring instructional components. The selection of books appropriate to each tutee proved to be a challenge to tutors and tutees alike. Tutees struggled to read the texts, and tutors struggled to use their limited repertoire of strategies to help the tutees read. In the end, tutors shifted from the prescribed to the recalled.

Understanding the mismatch between the tutoring on site and the program on paper unveiled the nesting of these challenges within the AmericaReads organizational structure. Clarifying the organizational challenges further explains the tutoring outcomes.

As previously explained, this program generated much enthusiasm from schools and service providers. This success triggered a need for additional staff and activated a challenge--establishing lines of communication and creating a sense of community and teamwork with an expanded number of supervisors, tutors, students, public schools, and agencies. As one supervisor lamented, "I don't think that anyone realized the immensity of the program." The original organizational features grew to be inadequate, and new ways of behaving became critically important. One supervisor's words framed a shared concern: "My biggest challenge was I didn't know the rules.... That was really frustrating. I think when it really bothered me the most was when there was a possibility that it was going to impact the kids." For these supervisors, determining and implementing a more helpful package of organizational arrangements required them to work as a team and function as a cohesive whole. They describe this progression as moving from working as individuals to acting as a "well-rounded professional team" that "shared all of the responsibilities" and worked at "establishing organization and structure." Tutors appreciated the resulting changes, as indicated by their response to the question: "Describe how changes adopted in the AmericaReads Program second semester have been most helpful to you." One hundred percent of the respondents cited the increased organization of the program and their weekly meetings.

The tightening of the program's infrastructure revealed specific concerns. For example, the plan that existed for establishing contacts with schools and agencies proved inadequate. Principals or agency heads did not always get back to the supervisors, and site coordinators often failed to sufficiently explain the program to the teachers. These oversights, in turn, affected the tutors who, as indicated by questionnaire responses (see Appendix B, Part II, item 6) and responses to open-ended items (see Appendix C, item 6), initially worked at sites where those in charge were not clear on what was expected of them. Over time, this situation improved. The supervisors worked more directly with site coordinators and teachers, becoming more concrete in explaining the program and the tutors' responsibilities. These lines of "professional communication" became straightforward and concise. Supervisors' interview comments capture the awareness and sensitivity that directed their shifts: 1) "The teachers have 100 differe nt things going on. They don't think we're the most important thing in the world, so we need to make it as easy as we can for them"; and 2) "This is a job working with people, not numbers. And if you want success with your numbers, you have to work with the people first."

Conclusion

In the best of worlds, tutoring programs would focus on the important aspects of tutor competence and student achievement. This AmericaReads project began with that intention. However, myriad circumstances quickly obscured the major features of this vision. These frustrating problems--nested challenges--highlight the need to remain mindful of the complex context of volunteer tutoring programs. We must reflect on the whole--accepting the mundane and more trivial pieces alongside larger and more meaningful expectations. Now, the researchers are less flexible in what they consider relevant to implementing a tutoring program. The items underscored here are far from exhaustive.

First, a volunteer program must embrace a unified staff and tutoring pool committed to fulfilling the program's guidelines. This program struggled with competing agendas. School personnel, overworked and pressured, tried to usurp the tutors for their perceived needs. Too little was done to bring these teachers along as partners in implementing the tutoring program. Although the tutors received training and support, their numbers and the scant availability of field supervision left them undersupported and overwhelmed. In the end, tutors chose to "go their own way." The inability to unify the commitment of each participant in the AmericaReads program to understand and implement the features of its tutoring component precluded an ability to examine the efficacy of the tutoring package.

Second, school-university/program partnerships must be nourished. This necessity goes beyond the previously mentioned communication with teachers to include administrators, support personnel, and community members. Without synergy among the tutoring program, the hosting organization, and the array of stakeholders who support its intention, the hope of having tutoring events overlay with the designed tutoring features lessens.

Third, the tasks expected of tutors must match what they can easily and willingly accomplish. For the tutors in this program, strategy instruction was elusive. When not omitted, it became practice. This finding becomes instructive for future programs: Provide training in components that tutors can and will accomplish, and that hold promise for improving a student's reading achievement.

Finally, and linked to the previous challenge, volunteer programs must show evidence that they help students read better. Importing a program that lacks a strong research base is unwise. Importing an expensive program is not always feasible. Regardless, not having local evidence of success is simply reckless. Time and resources are simply too precious.

Obtaining coherence between a tutoring plan and what a tutor does is not impossible. However, it requires a supportive organizational foundation that goes beyond training the tutors. Once that coherence exists, then the stage is set to explore the question of ultimate significance: What tutor characteristics, and which tutoring events, contribute to improving a child's ability to read?

References

Adler, P. (1999). The AmericaReads challenge: An analysis of college students' tutoring. Ann Arbor, MI: CIERA.

Adler, P A., & Adler, P (1994). Observational techniques. In N. K. Denzin & S. Lincoln (Eds.), Handbook of qualitative research (pp. 377-392).

Ainsworth R. (1995). Turning potential school dropouts into graduates: The case for school-based one-to-one tutoring. (Report No. UD 030633). Washington, DC: National Commission for Employment Policy. (ERIC Document Reproduction Service No. ED 387 568).

DeFord, D. E., Lyons, C. A., & Pinnell, G. S. (Eds.). (1991). Bridges to literacy: Learning from Reading Recovery. Portsmouth, NH: Heinemann.

Gaustad, J. (1992). Tutoring for at-risk students. (Rep. No. EA 024556). Eugene, OR: Oregon School Study Council. (ERIC Doc. Rep. Serv. No. ED 353 642)

Juel, C. (1996). What makes literacy tutoring effective? Reading Research Quarterly, 31, 268-288.

Kellner, M. B. (1999). A study of three HOSTS sites in the Red Clay Consolidated School District. Exec. Position Papers Completed in Fulfillment of an Ed.D. Newark, DE: University of Delaware.

Morris, D., Shaw, B., & Perney, J. (1990). Helping low readers in grades two and three: An after-school volunteer tutoring program. Elementary School Journal, 91, 133-150.

Pitman, M.A., & Maxwell, J.A. (1992). Qualitative approaches to evaluation. In M. D. LeCompte, W. L. Millroy & J. Preissle (Eds.), The handbook of qualitative research in education (pp. 729-770). New York: Academic Press.

Santa, C. M., & Hoien, T. (1999). An assessment of Early Steps: A program for early intervention of reading problems. Reading Research Quarterly, 34, 54-79.

Shannon, T. (1998). On the effectiveness and limitations of tutoring in reading. Review of Research in Education, 23, 217-234.

Slavin, R., Madden, N., Karweit, N, Livermon, B., & Dolan, L. (1990). Success for All: First-year outcomes of a comprehensive plan for reforming urban education. American Educational Research Journal, 27, 255-278.

Slavin, R. E., Madden, N. A., Dolan, L. J., & Wasik, B. A. (1996). Every child, every school. Success for All. Thousand Oaks, CA: Corwin Press.

Snow, C. E., Burns, M. S., & Griffin, P. (Eds.). (1998). Preventing reading difficulties in young children. Washington, DC: National Academy Press.

Topping, K. (1998). Effective tutoring in AmericaReads: A reply to Wasik. The Reading Teacher, 52, 42-50.

Wasik, B. (1997). Volunteer tutoring programs: Do we know what works? Phi Delta Kappan, 79, 282-287.

Wasik, B. (1998). Volunteer tutoring programs in reading. A review. Reading Research Quarterly, 33, 266-292.

Table 1

The Consistency Between the Planned Tutoring Events and the Events the
Tutor Implemented

Tutoring Component      Consistency Percentage

Review previous lesson           43%
Reread previous book             57%
Read new book                    80%
Talk about book                  63%
Strategy instruction             17%
Reflection                       87%
Table 2

Tutor Responses to Questionnaire Items About the Structure of the
Tutoring Sessions

Item 1: began the session by having my tutee read the book from the last
session.


B:  1(8, 10%)   2(32, 40%)  3(22, 28%)  4(18, 22%)
E:  1(8, 10%)   2(17, 22%)  3(26, 32%)  4(29, 36%)

Item 2: I talked about the book with my tutee before the reading.


B:  1(10, 9%)   2(21, 26%)  3(19, 24%)  4(30, 38%)
E:  1(3, 4%)    2(11, 14%)  3(27, 34%)  4(39, 49%)

Item 3: I had my tutee read one new book.


B:  1(4, 5%)    2(6, 7%)    3(23, 29%)  4(47, 59%)
E:  1(3, 4%)    2(5, 6%)    3(14, 18%)  4(58, 73%)

Item 4: I talked about the book with my tutee after the reading.


B:  1(3, 4%)    2(2, 3%)    3(18, 22%)  4(57, 71%)
E:  1(0, 0%)    2(1, 1%)    3(10, 13%)  4(69, 86%)

Item 5: I had my tutee write a sentence about the book.


B:  1(20, 25%)  2(27, 34%)  3(17, 21%)  4(16, 20%)
E:  1(5, 6%)    2(22, 28%)  3(28, 35%)  4(25, 31%)

Item 6: I taught a skill lesson.


B:  1(16, 20%)  2(37, 46%)  3(17, 21%)  4(10, 13%)
E:  1(3, 4%)    2(24, 30%)  3(25, 31%)  4(28, 35%)

Item 7: I talked with my tutee about how to be abetter reader.


B:  1(23, 29%)  2(28, 35%)  3(20, 25%)  4(9, 11%)
E:  1(8, 10%)   2(24, 30%)  3(29, 36%)  4(19, 24%)

Item 8: I read a book to my tutee.


B:  1(10, 13%)  2(34, 43%)  3(19, 24%)  4(17, 21%)
E:  1(9, 11%)   2(41, 51%)  3(10, 13%)  4(20, 25%)

Item 9: Overall, the structure of my sessions matched the description in
the AmericaReads Challenge Packet.


B:  1(4, 5%)    2(21, 26%)  3(35, 44%)  4(20, 25%)
E:  1(0, 0%)    2(8, 10%)   3(29, 36%)  4(43, 54%)

Note: Judgments were made on a 4-point scale (1=applies to none of the
tutoring sessions, 2=applies to some of the tutoring sessions, 3=applies
to most of the tutoring sessions, and 4=applies to all of the tutoring
sessions). B represents responses at the beginning of the tutoring. E
represents responses at the conclusion of the tutoring.
Appendix B: Tutor Questionnaire

As part of the AmericaReads grant that funded your tutoring, we need to
collect information about the program. The following questionnaire is
designed to provide that information. It is divided into two sections.
Part I explores the format of your lessons. Part II seeks your input
about the challeges you encountered as a tutor: Your responses are
confidential and will be aggregated with those of other tutors.
Therefore, your indentity will remain anonymous. We appreciate your
honest responses to the questions and look forward to learning from
them. Complete a questionnaire for each student you tutored.

Part I: The following statement apply to the features of your tutoring.
For each statemetn, use the following scale to circle the number that
best represents the beginning (B) and ending (E) sessions with your
student. Circle a 1 if the statement applies to some of the tutoring
sessions. Circle a 3 if the statement applies to most of your tutoring
sessions. Circle a 4 if the statement applies to all of your tutor
sessions.



1. I began the session by having my tutee read the  B: 1 2 3 4
    book from the last session.                     B: 1 2 3 4

2. I talked about the book with my tutee before     B: 1 2 3 4
    the reading.                                    B: 1 2 3 4

3. I had my tutee read one new book.                B: 1 2 3 4
                                                    B: 1 2 3 4

4. I talked about the book with my tutee after the  B: 1 2 3 4
    reading.                                        B: 1 2 3 4

5. I had my tutee write sentence about the book     B: 1 2 3 4
                                                    B: 1 2 3 4

6. I taught a skill lesson.                         B: 1 2 3 4
                                                    B: 1 2 3 4

7. I read a book to my there.                       B: 1 2 3 4
                                                    B: 1 2 3 4

8. Overall, the structure of my sessions matched    B: 1 2 3 4
    the description in the AmericaReads Packet.     B: 1 2 3 4
Part II: The following statements represent challenges that you might
have encountered during your tutoring. Use the following scale to
indicate whether you faced these challenges at the beginning and/or end
of your sesions. Circle an N if you never faced this challenge. Circle
an S if you sometimes faced this challenge. Circle an A if you always
faced this challenge.


 1. It took too long to receive my assignment  B: N S A
                                               B: N S A
 2. There was too much time between initial    B: N S A
     training and first tutoring session       B: N S A
      (one month).
 3. There was lack of transportation. Getting  B: N S A
     to and from my site was difficult.        B: N S A
 4. I wasn't allowed to follow the lesson pla  B: N S A
     in the AR Challenge Packet.               B: N S A
 5. The program didn't provide the materials   B: N S A
     when I needed them.                       B: N S A
 6. My site was not clear on what I was        B: N S A
     supposed to do.                           B: N S A
 7. The teachers weren't receptive to my       B: N S A
     presence.                                 B: N S A
 8. I was confused about the folder system.    B: N S A
                                               B: N S A
 9. I lest my tutee's folder.                  B: N S A
                                               B: N S A
10. The tutoring was too short.                B: N S A
                                               B: N S A
11. Paying for my travel costs was a           B: N S A
     hardship.                                 B: N S A
12. Keeping the university and sites aware     B: N S A
     that I wouldn't be there was too          B: N S A
      difficult.
13. My coordinator would say she contacted     B: N S A
     me, didn't receive the message.           B: N S A
14. I never knew when the group meetings       B: N S A
     were taking place.                        B: N S A
15. I didn't understand how to administer      B: N S A
     the test.                                 B: N S A
16. I don't know enough about teaching         B: N S A
     reading to understand how to help         B: N S A
      my tutee.

Appendix C: Tutor Evaluation

1. Describe how your tutees have changed from when you began working with them. (You may want to consider attitude as well as achievement.)

2. Describe how your tutoring style has changed from the beginning of the year until now.

3. Describe which changes adopted in the second semester of the AmericaReads program have been most helpful to you.

4. What materials have you found to be most useful? Are there any materials that you would like to have provided that might assist tutors in the future?

5. What changes would you like to see instituted in the program for the next school year?

6. In what ways, if any, would you like to have school-based personnel (school coordinator/ teachers, etc.) provide more assistance to you?

Appendix D: Interview Questions

As coordinators for AmericaReads, we are interested in better understanding the challenges you encountered. With your permission, we would like to record your responses, but what you say will remain confidential.

1. How would you describe your role at the beginning and the end of the program? What caused that role to change (or remain the same)?

2. When serving in a new role, everyone has things they would do again and things they would avoid. What were those things for you?

3. What were your biggest challenges in working with school personnel? Tutors? The other coordinators? The overall program? How did you meet these challenges?

4. What were your biggest successes in working with school personnel? Tutors? The other coordinators? The overall program?

5. If you were to advise someone assuming this role next year, what would you say?

6. Are there some areas our questions didn't cover that would better help us understand the challenges you encountered?