This chapter describes research on effective instructional practices to develop the literacy of adolescents and adults and identifies needed research. Individuals needing to improve their literacy have diverse characteristics, literacy development needs, learning goals, and challenges to learning. Settings of instruction are wide-ranging and include local education agencies, community organizations, community colleges, prisons, and workplaces. Across these programs and often within a single program, the instruction has diverse aims to help adults attain employment or work skills, career advancement, a general educational development (GED) credential, a college degree, the ability to assist children with school, or other practical life goals. Thus, the first part of the chapter describes the population and the contexts of literacy instruction. Because formal literacy instruction in the United States occurs mainly in adult education programs and developmental education courses in college, we organize the discussion around these two learning contexts.
The second part of the chapter characterizes the state of research on instructional practices for adults. As explained in Chapter 1, adult is defined in this volume as individuals ages 16 and older not enrolled in K-12 school, consistent with the eligibility for participation in federally funded adult literacy education. A recent systematic review of research on instructional approaches for adult literacy populations has been funded by the National Institute for Literacy in partnership with the U.S. Department of Education and the National Institute for Child Health and Human Development (Kruidenier, MacArthur, and Wrigley, 2010). In synthesizing the evidence on instruction, we draw on this review, which we then augmented with
additional searches of quantitative and qualitative research. We include English language learners and adults with disabilities in describing the population of adults with literacy development needs but discuss the research on instruction with these populations in subsequent chapters. The chapter concludes with a summary of the extent of current knowledge of effective practices in adult literacy instruction and directions for future research.
There are many reasons why individuals seek to develop their literacy skills as adults. Some study to obtain a high school equivalency diploma; others seek to help their children and families with education, health, and other practical life matters; and others seek to learn English or enhance skills for new job responsibilities. Others may have a higher level of literacy but have not yet developed the reading and writing skills needed in college. Adults who wish to develop their literacy receive instruction in two main types of settings: adult education programs and developmental courses in college, especially in community colleges. Two types of adult education are found in college settings: (1) adult literacy programs for individuals who wish to complete their secondary education and (2) developmental education1 for students formally enrolled in college programs.
The U.S. Department of Education reports that nearly 2.6 million adults enrolled in federally supported adult education programs during the 2006-2007 fiscal year, the most recent year for which complete data are available. Adult education programs are largely supported by federal and state funding, which together provides about two-thirds of the funding for adult literacy programs, according to a national survey of adult education programs (Tamassia et al., 2007). Other sources of funding are local governments, private donations, and, to a small degree, fees and tuition paid by the participants. The U.S. Department of Education’s Office of Vocational and Adult Education administers the federal funds, which are appropriated to designated state agencies in a competitive granting process, consistent with the Workforce Investment Act, Title II, Adult Education and Family Literacy Act (AEFLA). Each state must provide matching funds to qualify for this allocation.
The Adult Education Program Survey (AEPS; Tamassia et al., 2007) provides information on a nationally representative sample of adult edu-
1 We use the term developmental education (also called remedial instruction) to refer to the broad array of services and specific courses provided to college students with weak skills.
cation programs and enrolled learners during the 12-month period 2001-2002.2 At the time of the survey, 3,108 adult education programs were offered in 29,424 learning sites. More than 1,200 adult education programs funded under the Adult Education and Family Literacy Act participated in the survey. During this period, the median budget for a program was $199,000; with a median enrollment of 318 learners per program, the median expenditure per learner was $626.
According to the survey, adult education programs offer three main types of literacy instruction:
1. Adult basic education (ABE) provides instruction to adults who lack “competence in reading, writing, speaking, problem solving or computation at a level necessary to function in society, on a job or in the family” (National Reporting System for Adult Education, 2001, p. 25).
2. Adult secondary education (ASE) is “designed to help adults who have some literacy skills and can function in everyday life,3 but are not proficient or do not have a certificate of graduation or its equivalent from a secondary school” (National Reporting System for Adult Education, 2001, p. 25). Adults usually attend ASE classes to obtain a GED or adult high school credential.
3. English as a second language (ESL) instruction is “designed to help adults who are limited English proficient achieve competence in the English language” (National Reporting System for Adult Education, 2001, p. 25).
English as a second language serves the largest number of students, followed closely by adult basic education: 43 percent of adult learners receive ESL instruction, 40 percent receive ABE instruction, and 19 percent participate in ASE instruction. Most English language learners (85 percent) who attend a program attend ESL programs. Of native language learners, two-thirds attend ABE and one-third attend ASE programs.
Instruction is offered in many different places and programs that vary widely in size and number of learning sites. According to the AEPS, local education agencies are the major providers of adult education, offering 54 percent of the programs surveyed, followed by community-based organizations (25 percent), community colleges (17 percent), and correctional
2 The AEPS, funded by the U.S. Department of Education, was designed and conducted by the Educational Testing Service and Westat, Inc., with the involvement of staff of the Office of Vocational and Adult Education and the National Center for Education Statistics.
3Since these definitions for adult basic education and adult secondary education were produced, there has been a trend for jobs that pay above a poverty wage to require higher levels of literacy.
institutions (2 percent). And 3 percent of programs were offered by “other” entities, such as libraries, departments of human services, institutions for people with disabilities, and coalitions made up of the various provider types. Community colleges offer the largest programs in terms of the median number of students enrolled.4Table 3-1 shows the percentage of program types (ABE, ASE, ESL) offered by each type of provider.
There is not a simple alignment of learning goals with program type or location. For example, English language learners may be taught reading and writing skills in ESL classes in a workplace education setting or in a community college ABE program. Although the major goal of students in both settings may be to increase English language proficiency, the instructional aims will differ, with one focused on meeting specific job requirements and the other on developing more general literacy practices. Similarly, the goal of earning a GED certificate may be addressed in settings as diverse as prisons and volunteer library literacy programs.
Most participants (80 percent) in adult education programs surveyed in 2001-2002 were adolescents and young adults ages 44 and younger pursuing goals related to education, family, and work: 34 percent were ages 16 to 24; 46 percent were ages 25 to 44; 16 percent were ages 45 to 59; and 2 percent were ages 60 and older. Although originally designed for adults, the programs are increasingly attended by youth ages 16 to 20 (Hayes, 2000; Perin, Flugman, and Spiegel, 2006). Nonnative adults participating in ESL programs (those not born in the United States) were somewhat older than native adult learners in ABE and ASE programs, with 60 percent between the ages of 25 and 44 (versus 46 percent for native adults).
The diversity of languages spoken by English language learners points to a need to understand the factors that influence the development of literacy in English for speakers of different languages and respond to the practical challenge of delivering instruction effectively to linguistically diverse learners. According to the AEPS, 57 percent of adults in adult education programs were native to (born in) the United States. English was the home language
4Community colleges are defined in the AEPS as institutions of higher education (e.g., junior colleges without residential facilities) that offer degrees below a bachelor’s degree or technical degrees or certificates, such as in mechanical or industrial arts and applied sciences (e.g., technical colleges). Community colleges also provide continuing education, apart from the college programs, which are the site of ABE programs; college degrees or certificates are not awarded as part of these programs.
Community-based organizations are religious and social service groups, libraries, volunteer literacy organizations, literacy coalitions, community action groups, and other kinds of public or private nonprofit groups. Local education agencies are typically public schools or school districts, which in addition to providing K-12 education offer adult education classes open to all members of the community. Correctional institutions are prisons and jails funded by the state to provide adult basic education services to incarcerated adults.
TABLE 3-1 Instructional Program Types Offered by Each Type of Provider (in percentage)
|Adult basic education||36||35||42||52|
|Adult secondary education||20||11||17||18|
|English as a second language||44||55||42||31|
SOURCE: Data from the Adult Education Program Survey (Tamassia et al., 2007). Data are from a nationally representative sample of 3,108 programs during 2001-2002.
for 94.7 percent of these adults; Spanish was the home language for 4.5 percent.5 Almost 43 percent of adults were nonnative to the United States (versus 14 percent in the general population in 2002, the year of the survey). Of these adults, 3 percent spoke English as the home language, 62 percent spoke Spanish, 15.8 percent spoke an Asian language, 3.8 percent spoke a European language, and 14.7 percent spoke a language categorized in the survey as “other.”
Most native-born adults in adult education have completed ninth to eleventh grade (68 percent); about 14 percent had less education than that, and 20 percent had more (16 percent completed high school or received a GED credential, and 4 percent reported having “some college”). Nonnative learners show a broader range of educational attainment compared with native-born adults; that is, they appear in larger numbers at both the highest and lowest levels of education. More nonnative learners had completed some college (28 percent) and more had completed high school (22 percent), but more also reported having an education lower than ninth grade (28 percent); 17 percent completed ninth to eleventh grade. This variation within and across populations presents an additional challenge to programs that must design instruction for adults with such diverse educational backgrounds and degrees of proficiency in a first and second language.
5Home language was defined as the first language learned at home in childhood and still understood as an adult.
A portion of adults participating in adult basic literacy studies can be expected to have some form of learning disability that would require differentiated instruction and the provision of appropriate accommodations. There is no consensus, however, on the estimated numbers of adult learners who may have such a disability. The estimates range from one-tenth to more than half (Patterson, 2008). There are no program reporting requirements regarding the prevalence of learning disabilities among participants in federally supported literacy programs. According to the AEPS, only 34 percent of programs reported screening for learning disabilities, and of these, only 4 percent reported using cognitive or clinical instruments. Most—62 percent—relied on self-reports. Thus, it is likely that many adults may have gone unrecognized as having a learning disability, especially older students. Others may have been mislabeled, may not remember or have known that they were identified as having a learning disability, or may be uncomfortable disclosing their learning disability. With this caveat, 89 percent of programs reported providing services to at least one adult with learning disabilities. There is a need for more reliable information about students with learning disabilities in programs and for research on instructional effectiveness to clearly define these samples and identify the practices that promote their progress.
As described in Chapter 2, reading is generally understood to be comprised of the fluent reading of words and sentences and the comprehension of text. One source of information about the component skills of low-literate adults (third to eighth grade reading-level equivalent) comes from a research initiative funded by the U.S. Department of Education, the National Institute of Child Health and Human Development, and the National Institute for Literacy to develop instructional interventions for low-literate adults in adult education programs and to evaluate their effectiveness (see Appendix D for details about these studies). Findings from these studies and other research (see Kruidenier, MacArthur, and Wrigley, 2010) show that adults can have difficulties with any or all of the crucial aspects of reading: alphabetics (phonemic awareness and word analysis), fluency, vocabulary, or comprehension. Thus, it is important to comprehensively assess adults’ profile of starting skills to plan the appropriate instruction.
According to these studies, lack of fluent decoding is a source of reading difficulty for a significant number of low-literate adults, especially below the eighth grade reading-level equivalent (Alamprese et al., 2011; Greenberg et al., 2011; Hock and Mellard, 2011; Sabatini et al., 2011). Decoding dif-
ficulties are observed among adults performing at each of the six levels of the National Reporting System, the system used to assess the literacy performance of adults in federally funded adult education programs (Mellard, Fall, and Mark, 2008; Mellard, Woods, and Fall, 2011). Thus, even at higher levels in the National Reporting System (NRS), adults can differ greatly in their word-level reading skills.
Three studies have tested whether the reading component patterns of adults match similar models of reading developed with children (MacArthur et al., 2010a; Mellard, Fall, and Woods, 2010; Nanda, Greenberg, and Morris, 2010). These studies suggest that for adults with low literacy, the reading models were not similar. Specifically, low-literate adults appear to lack the fluent integration of word reading, language, and comprehension skills shown by young children who learned to read on a normative time-table. The comprehension skills of the low-literate adults were more similar to those of children with low reading skills than to typically developing child readers, in that they did not generate an integrated representation of the meaning of a passage by connecting words, phrases, sentences, and paragraphs and making inferences using information provided in the text and background knowledge (see the discussion of comprehension in Chapter 2).
The measurement of reading comprehension for either research or practice remains a challenge. As mentioned in Chapter 2, a more integrated approach needs to be taken to the study and assessment of reading comprehension. Depending on the assessment chosen, different subskills of reading comprehension are tapped or assessed to a greater or lesser degree (Cutting and Scarborough, 2006; Hock and Mellard, 2005). Some reading comprehension tests relate more strongly to word recognition skills, others relate more strongly to oral language ability, and the tests have only low-to-moderate correlations with one another (Keenan, Betjemann, and Olson, 2008). Furthermore, the format of the reading comprehension assessment appears to affect test performance (Eason and Cutting, 2009; Francis et al., 2005; Spear-Swerling, 2004). Reading comprehension measures for research and practice are needed with adult norms and that comprehensively assess components of reading comprehension in the context of valued everyday literacy activities.
Despite the capacity of writing to facilitate reading development and the need for adults to be able to write for work, education, and other purposes, writing has not been included in major surveys of adult learners, nor have writing skills been a focus of adult literacy research (Gillespie, 2001). It is known, however, that low-literate adults spell less accurately, their spellings are inconsistent (Dietrich and Brady, 2001), and their errors show more nonphonetic and morphological errors in comparison to the spelling of reading-matched adults (Greenberg, Ehri, and Perin, 1997, 2002; Worthy and Viise, 1996). Adult literacy students also have been reported
to have great difficulty with descriptive and argumentative writing (Berry and Mason, in press; MacArthur and Lembo, 2009). Few standard tests of writing achievement are available to assess progress over time with norms for adults, much less adults with basic literacy development needs. The time required to score written compositions can present a challenge to the valid assessment of writing in research and for instruction.
Information about the instructional practices used in adult education programs is not available from the Adult Education Program Survey, although general characteristics are provided, such as whether the instruction was classroom-based or one-on-one instruction. On average, learners participated in adult education programs for less than 100 hours over the course of a program year, according to the Adult Education Program Survey. Only about one-third of adults made reading gains equivalent to a grade level during the program year. These findings are consistent with the levels of participation and progress reported in the few published studies of interventions designed to develop the literacy of adults with low-to-intermediate skills (see Appendix C) and other information gathered from individual researchers and practitioners working in the field. Reading is a complex skill, and research on the development of complex skills and expertise suggests that about 3,000 hours are required for mastery (Chi, Glaser, and Farr, 1988); 100 hours represent 3 percent of that amount, and so it is likely to be insufficient for learning for many adults, even if the goal is not expert mastery. Thus, one primary reason for limited progress may be that adults lack sufficient amounts of instruction and practice for improving skills.
It is not clear why some adults persist with literacy instruction and others do not. Sabatini et al. (2011) reported that those who persisted with a literacy intervention tended to be older, on average, with poorer basic reading skills. This finding is consistent with the higher dropout rates reported for younger adult education students (Flugman, Perin, and Spiegal, 2003). Younger students who have lower reading scores when entering ABE and GED programs are more likely to drop out of the programs than older, higher skilled students (Dirkx and Jha, 1994). Adults report a wide range of factors that positively or negatively affect persistence in adult education, which include transportation, competing life demands, supportive relationships, and self-determination (Comings, 2009). Reasons reported for dropping out of adult education include family problems, the pace of instruction (either too fast or two slow), health issues, dislike of classwork,
and inconvenient class location or schedule (Perin and Greenberg, 1994). About one-third of adult education programs report that they provide noninstructional support services (transportation, child care, psychological counseling) in an attempt to ease some of the barriers that adults experience, paid for with in-kind services contributed by the community (Tamassia et al., 2007).
For all providers, instruction was delivered mainly by part-time staff members and volunteers, with larger percentages of individuals in these categories (versus full-time staff) filling an instructional role (see Table 3-2). The expertise of instructors in adult education programs is highly variable (see Table 3-3 and Box 3-1). According to the Adult Education Program Survey, across provider types, instructional staff is the largest program expenditure; professional development is the smallest. Volunteers deliver a significant portion of the instruction in adult basic literacy programs, and the most commonly reported educational requirement for volunteers was a high school diploma or equivalent. The most commonly reported education requirement for full-time and part-time instructors was a bachelor’s degree, followed by K-12 certification. Table 3-3 shows instructor credentials as reported by ABE, ASE, and ESL programs in 2001-2002. It appears that the bulk of instructors have inadequate or no specific training in best methods for teaching in adult literacy programs (see also Box 3-1).
When special needs are considered, the situation is even more extreme (Tamassia et al., 2007). It is vital to use reliable methods to diagnose learning and reading disabilities and to adjust instruction accordingly. Across ABE, ASE, and ESL instruction, about 2 percent or fewer of programs required their full-time, part-time, or volunteer instructors to have special education certification. This problem is compounded by the fact that special education degree programs rarely focus on the needs of adult literacy students.
ESL instructors and the learners they serve face the dual challenge of
TABLE 3-2 Percentage of Staff in an Instructional Role by Role and Staff Type
SOURCE: Data from the Adult Education Program Survey (Tamassia et al., 2007).
TABLE 3-3 Credentials of Instructors in Adult Education Programs by Staff Type and Type of Instruction (percentage of each staff type with the credential)
Type of Instruction
K-12 teaching certificate
Adult education certificate
K-12 teaching certificate
Adult education certificate
NOTE: The table includes the three most common instructor credentials reported by programs in a nationally representative survey of adult education programs.
ABE = adult basic education; ASE = adult secondary education; ESL = English as a second language; TESOL = teachers of English to speakers of other languages.
SOURCE: Data from the Adult Education Program Survey (Tamassia et al., 2007).
Characteristics of Adult Literacy Instructors
Adult basic education teachers
• work mostly part time.
• may leave the field more often than K-12 teachers.
• are often required to teach in multiple subject areas.
• have scant formal education related to teaching adults, although many are qualified and have taught in K-12.
• have in-service preparation as their primary form of professional development.
• are not consistently funded to participate in in-service professional development.
• have access mostly to short-term training and conferences.
• are hindered by systemic constraints from participating in professional development.
SOURCE: Adapted from Smith and Gillespie (2007).
improving both spoken language and literacy skills in English, and, as mentioned earlier, their students speak a variety of languages. This challenge to instructors is expected to grow: U.S. Census Bureau projections show net international migration is likely to account for more than half of the nation’s population growth between 2000 and 2015 (Kirsch et al., 2007).
Although some part-time and full-time adult literacy instructors have K-12 teaching certifications and have taught in K-12 schools, evidence suggests that many teachers of grades 1 through 12 do not feel confident in teaching reading and writing and are likely to lack the requisite knowledge and skills. To illustrate, results from a survey published in 1994 on the phonics knowledge of experienced reading teachers showed that only 10-20 percent of the teachers could accurately identify consonant blends in written words, only 21 percent knew what an inflected verb was, and only 27 percent could identify morphemes in a word (Moats, 1994, 2004). Teachers with limited knowledge of language structure will be less able to teach effectively to learners at any age. Furthermore, in one survey, only 32 percent of K-12 teachers whose classes included students with disabilities felt well prepared to address their academic needs (National Center for Education Statistics, 2010a).
With respect to writing, one-third of primary grade teachers have reported that they were poorly prepared to teach writing by their college teacher preparation program (Cutler and Graham, 2008). The number increased to 66 percent in grades 4 to 6 (Gilbert and Graham, 2010), dropped to 47 percent in middle school (Graham et al., 2010) but appears most problematic among high school teachers (Cutler and Graham, 2008; Graham and Gilbert, 2010), with 71 percent reporting that they were inadequately prepared (Kiuhara, Graham, and Hawken, 2009).
Although no data were identified on the preparation of instructors of adults specific to reading and writing, it is reasonable to assume from the information available that the knowledge and skills of the instructors are highly uneven. Many instructors also are likely to have a view of the trajectory for adult literacy instruction that fits better with the world of formal K-12 schooling developed prior to the information age than to adult learners and the levels and forms of literacy needed today.
Most programs in the AEPS reported having access to educational technologies, although it is not clear how appropriate the technologies were for literacy practice and instruction. Most programs reported having computers, audiovisual equipment, and Internet connectivity; however, it is not evident what access learners have to computers during each classroom session, the supports that would be needed to secure access outside class,
and the supports needed by learners and instructors to use technology tools effectively.
In the AEPS, programs reported that adult learners were assessed on a regular basis, although the assessments that programs reported using most often were measures to meet federal accountability requirements. The NRS is the system through which all federally supported adult education programs report their annual program data, which must include assessments of learners’ progress. Currently, although not at the time of the survey, states must use one or more assessments that have been determined to be valid and reliable measures and programs must administer pre- and posttests in accordance with the test publishers’ guidelines. The U.S. Department of Education uses a panel of experts to review the standardized tests annually as part of its process for approving assessments submitted by the states.6 These measures are for accountability purposes, however, and reliable information is not available about the range of assessments and assessment practices that instructors and programs use to plan the appropriate instruction. A sound approach to assessment to support and monitor learning at the individual, program, and systems levels is systematic, with linkages among the various purposes of assessment and extensive professional training and supports needed to implement the assessments reliably. More information is needed about the methods used for diagnostic, placement, and formative assessment to ascertain adults’ skill development needs in order to plan instruction and track progress in component reading and writing skills and functional literacy related to broader learning goals.
The precise number of academically underprepared college students is not known: estimates for community college entrants range widely, from 40 to 90 percent (Perin and Charron, 2006). National data have not been reported on the specific limitations in college students’ reading and writing skills. Wang (2009) reported in a study of first-year college students enrolled in a developmental reading course that only 55 percent could identify explicitly stated main ideas in text, only 42 percent could comprehend implicit main ideas, and only 11 percent were aware of a global main idea in text. Similarly, Perin, Keselman, and Monopoli (2003) found in a study of community college students that many students attending the highest level
6A current list of approved assessments may be found at http://www.nrsweb.org/foundations/implementation_guidelines.aspx [Jan. 2012].
of developmental education had great difficulty identifying the main ideas in text in order to write summaries.
At present there is not a universally accepted definition of college readiness. The policies and regulations that govern eligibility for enrollment in credit-bearing courses, as well as student assessment and placement, pedagogy, staffing, and completion, vary from state to state, college to college, and program to program. There is also considerable variability across types of higher education institutions about the level of writing and reading proficiency that necessitates remediation.
Conventionally, community colleges and other open-enrollment colleges give placement tests to all incoming students and consider anyone above a cut point to be prepared for postsecondary learning. Other colleges may use placement measures for students admitted with lower grades or SAT scores. Placement measures vary across colleges and, among colleges using the same measures, cut scores vary and are adjusted from time to time within colleges for reasons that are not easy to determine (Perin, 2006). Furthermore, it is not clear from research that the placement scores in use or the literacy skills they assess are valid predictors of college academic performance (Hughes and Scott-Clayton, 2011). In research, readiness for postsecondary learning has not been assessed using measures derived from research on reading and writing. In practice, states and test services companies write descriptions of reading and writing capabilities for twelfth graders that currently serve as default standards but have no empirical grounding or predictive validity (e.g., ACT, undated; Grigg, Donahue, and Dion, 2007; Salahu-Din, Perskey, and Miller, 2008; University of the State of New York, 2005). A recent national effort to develop K-12 Common Core Standards includes literacy standards for twelfth grade and may inform future definitions (National Governors Association and Council of Chief State School Officers, 2009).
Developmental education courses are the primary mechanism used to increase students’ skills in colleges (Kozeracki and Brooks, 2006).7 More than half of community college students enroll in at least one developmental education course during their college tenure (Bailey, Jeong, and Cho, 2010). One study of 250,000 students from 57 colleges in 7 states found that, among students enrolled for the first time in fall 2003 to fall 2004, 59 percent were referred for remedial instruction and 33 percent of the referrals were specifically for reading (Bailey, Jeong, and Cho, 2010). Remedial reading and writing instruction in college is widely reputed among education researchers to focus on drill and practice on small subskills
7Funding for developmental education varies by state. Sources of funding may include state and local appropriations, tuition, and federal funds to the extent that students use federal financial aid to pay tuition (Education Commission of the States, 2000).
without strong linkages to the literacy activities that are part of the college curriculum (Grubb, 2010). Although there are many descriptive reports on instructional practices used in single classrooms and colleges, few quantitative data are available on the outcomes for students.
Alternate or complementary approaches to addressing the skill needs of underprepared college students include “college success” courses, college learning centers, and the incorporation of literacy skill development into disciplinary coursework. College success courses, which are increasingly required for incoming students, do not explicitly teach reading and writing skills but rather college study and research strategies that require the use of reading and writing (Derby, 2007; Pan et al., 2008; Zeidenberg, Jenkins, and Calcagno, 2007). College learning centers provide assistance from peer or professional tutors in a variety of areas that include reading and writing (Brittenham et al., 2003; Gordon, 2008; Hock, Deshler, and Schumaker, 1999; Hodges and White, 2001; Perin, 2004). They also offer legally mandated supports for students with disabilities, which can involve classroom accommodations (Gordon, 2004) or specialized tutoring (Hock, Deshler, and Schumaker, 1999; Mull, Sitlington, and Alper, 2001).
Some college instructors choose to teach basic skills to underprepared students who do not attend developmental education courses to enable them to comprehend and write about what is being taught in a discipline. These instructors intentionally incorporate literacy skills into disciplinary coursework (Juchniewicz, 2007) similar to content area literacy in secondary education (Moje and Speyer, 2008). This type of basic skills instruction is not formally recognized, and it has been referred to as “remediation in disguise,” “hidden remediation,” and “submerged remediation” (Grubb, 1999, pp. 194-195).
ESL courses are offered in colleges to teach language skills to students with low English language proficiency. These courses tend to be administered separately from developmental education, although they may integrate written and oral language instruction (Kaspar, 1996; Scordaras, 2009; Song, 2006). Between 1979 and 2008, the number of school-age children ages 5 to 17 who spoke a language other than English at home increased from 3.8 to 10.9 million, or from 9 to 21 percent of the population in this age range (National Center for Education Statistics, 2010b). Thus, the proportion of English language learners in higher education is increasing, especially in community colleges (Cohen and Brawer, 2003; Smith, 2010a). College students who are not fully proficient in English include “Generation 1.5” students: these students have a primary language other than English, have attended school in the United States for some period of time, and are fluent in informal but not academic English. They tend not to self-identify, however, as needing to take ESL courses (Blumenthal, 2002; DiGennaro, 2008; Goldschmidt, Notzold, and Miller, 2003; Matsuda, 2003). Those
who do complete ESL courses can still require additional reading or writing instruction in college.
We identified no source of information about the qualifications (training, credentials, skills) of the nation’s developmental education instructors to teach reading and writing, despite increasing concerns about the quality of developmental education (e.g., Grubb, 2010) and the need to better support the academic progress of community college students (Sperling, 2009; Zachry and Schneider, 2010). In one qualitative study (Kozeracki, 2005), 36 developmental English instructors who responded to structured interviews pointed to challenges that include a lack of maturity and motivation of students to do college work, language differences that may be best addressed in ESL classes, possible learning disabilities that may never have been diagnosed, socioeconomic conditions that make it very difficult for students to progress academically, and expressed student anger over being placed in developmental classes. The faculty from colleges in two states with an enrollment greater than 15,000 students and varied organizational structures for their developmental education programs report that they do not feel competent to address the needs of their developmental education students. The knowledge faculty gain from their own graduate training is significantly different from the knowledge they need to teach developmental classes. Although they hold advanced degrees in their discipline (e.g., English), the instructors may not be familiar with evidence-based techniques for teaching low-skilled readers and writers.
As the committee examined the research literature on instructional practices, we made certain assumptions. First, our central concern is to understand the state of the research on effective practices to develop reading and writing skills among low-literate adults and college students, including students who are proficient speakers of English and those who are learning English. Although other populations may need assistance to develop literacy or compensate for declines in their literacy, we focus on research with these populations because they represent the overwhelming majority of participants in adult education programs and developmental education courses who experience particular difficulty in achieving the literacy levels needed for economic, educational, social, and personal success in U.S. society. We do include, however, studies on the out-of-school literacy practices of disaffected youth who are still in K-12 education because these students are at risk for dropout and may eventually attend adult literacy programs.
Second, we recognize that different types of research questions call for
different methodological approaches. Questions about effectiveness are best answered with well-designed randomized controlled trials and other controlled experiments, which yield the most interpretable findings. We also reviewed correlational data that controlled for extraneous factors and that were analyzed with such methods as hierarchical linear regression to yield insights about hypotheses to pursue with experimental methods. We included studies that had at least one quantitative outcome pointing to an association between an instructional practice and the learning of reading and/or writing skills. Appendix D describes more fully the procedures used to conduct the research reviews and describes the studies retained for further consideration by the committee. Quantitative studies were excluded if they did not describe specific instructional practices or curricula (e.g., they assessed program attendance on a literacy outcome) or if the outcome was derived from self-report and not a direct measure of skills.
Many quantitative studies of the effectiveness of adult literacy instruction have serious methodological flaws that limit the ability to determine best practice. However, we adopted a pragmatic approach and assumed that, although the research was not of optimal quality for this purpose, it would be useful to examine for themes that suggest directions and hypotheses for future research.
We examined descriptive and qualitative research to reveal the variety of goals, techniques, and materials that are being used and studied in relation to reading and writing instruction. We assume that qualitative research makes the strongest contribution to knowledge when it follows established procedures for qualitative research (e.g., Denzin and Lincoln, 2005) and also systematically (1) states explicit goals for literacy instruction, (2) describes the practices used to achieve stated goals, and (3) analyzes links between observed practices and well-described literacy outcomes. Such findings can provide information for generating hypotheses to test in effectiveness research. We focused our search on identifying qualitative research studies with these features. When used in conjunction with quantitative experiments, qualitative research can provide rich descriptive information about learners and the instructional context, such as how the instructional practices were implemented and the provision of other supports for learning. This information helps to interpret experimental research findings and identify the conditions that may facilitate or hinder instructional effectiveness.
Sources of information gathered include a recent comprehensive review of literature on adult literacy instruction (Kruidenier, MacArthur, and Wrigley, 2010), augmented with targeted literature searches as needed to draw conclusions about the state of the research base and needs for development. These reviews focused on studies of practices to develop the reading and writing skills of adults in basic and secondary education and academically underprepared students in college (see Appendix D; findings
in the appendix for adults learning English are discussed in Chapter 8). An additional search was conducted of practices used in programs for adults with low literacy in other countries to identify practices to study with adults in basic and secondary education in the United States. We also explored the literature available on the effectiveness of practices used in programs for disengaged youth.
Although there is a large literature on adult literacy, the committee found a striking lack of useful, high-quality research for identifying the features of effective instructional practice. There are at least four reasons for this state of affairs:
1. Progress in adult literacy research has been hampered by the high attrition of research participants.
2. The research has lacked systematic focus on the development of reading and writing skills.
3. The research, whether quantitative or qualitative, does not include methods for systematically identifying associations or cause-effect relations between an instructional practice and outcomes.
4. Research funders and thus researchers of literacy have chosen to focus mainly on preschool and K-12 populations, a situation that has constrained the amount of research on how to further develop the literacy of adults outside school.
Despite such shortcomings in the research base, it is important to examine the existing corpus of research to try to understand the variety of instructional practices in use and to identify specific needs for future research on instructional effectiveness.
Our search terms and the other resources from which we draw directly targeted the many disparate types and locations of literacy instruction. We organize our discussion here into the two general categories of instruction for adults in education programs and instruction for academically underprepared college students. We discuss findings on literacy instruction for English language learners and those with learning disabilities in later chapters.
The National Institute of Child Health and Human Development, the Office of Vocational and Adult Education, and the National Institute for Literacy invested $18.5 million from 2002 to 2006 in large-scale research to
develop and evaluate effective approaches to literacy instruction for adults with low literacy (i.e., third to eighth grade reading-level equivalent). The research applied knowledge of the components of reading and writing and effective practices validated with children and adolescents in K-12 settings. Appendix D reports details of these studies.
A main finding from this body of work is that the interventions tested with low-literate adults did not differ from “business as usual” in adult education programs, despite being more systematic and structured in their approach. A second finding is that both the interventions and business as usual had small effects or no effects on various component skills. Notably, although the adults in these interventions did show decoding problems, as described earlier, the interventions with a strong decoding component were no more effective in remediating componential or functional skills than interventions without a strong decoding component or business as usual in programs. One exception was a structured decoding curriculum that included an emphasis on spelling and showed gains on some decoding and word recognition measures (Alamprese et al., 2011). Instruction that targeted fluency and comprehension also either produced no effect or gains that did not differ from the gains experienced with the less systematic and structured approaches used in programs.
Thus, it is not clear from this set of studies what range of approaches might be effective in developing skills sufficiently for fluent reading with comprehension. The instruction may need to be more explicit than what was offered in these interventions, with more opportunities for extensive practice. The instruction also may need to target particular areas of decoding difficulty and develop vocabulary to a greater extent, while providing more opportunities to practice and integrate skills in the context of reading actual text with scaffolding and feedback. Low-literate adults show difficulty understanding the meaning of text beyond the word or sentence level (e.g., Mellard, Fall, and Woods, 2010). Thus, they may need to develop the knowledge and skills for making connections across text elements, drawing inferences, and generating an overall representation of the meaning of a text. These adults may benefit from explicit instruction in comprehension strategies and development of vocabulary and background knowledge relevant to the text. All of these hypotheses, which remain to be tested, are consistent with principles of effective instruction for struggling readers from K-12 research (see Chapter 2).
The intervention studies displayed several limitations and constraints that may have affected the results. The researchers reported that instructional procedures were difficult to implement as intended in the context of adult education, given that many participants did not persist to the end of the studies. To combat the high attrition rates throughout adult education programs, some studies tested shortened versions of interventions that are
effective in K-12 settings. Furthermore, a sizeable percentage of the participants reported having learning disabilities, consistent with what would be expected from other studies with the population (Mellard and Patterson, 2008). Although several of the interventions were adapted from those that have been effective with children and adolescents with learning disabilities, the interventions or the placement procedures that were used may not have addressed underlying skill deficits. In addition, most of the outcome measures used in this research were developed and normed for children, and patterns of observed adult skills do not fit into the literacy levels available (Greenberg et al., 2009).
Beyond these intervention studies, a coherent and sustained base of research does not exist on the effectiveness of adult literacy instruction. There are only a handful of quantitative research experiments that include outcomes measured with standardized tests or researcher-developed measures of the components of reading (alphabetics, decoding and word recognition, fluency, vocabulary, and reading comprehension) spelling, and writing. One large study compared gains from a variety of instructional approaches in 130 ABE classrooms and found that the greatest gain was found for structured instruction in alphabetics with effect sizes of .37 to .42 (Alamprese, 2009). Most studies use designs, however, that are not adequate for concluding that the instructional approach caused the observed results.
Some studies have reported pre-post gains in terms of grade levels, but the reasons for the gains are not clear, and the amount of gain varies substantially (Gold and Horn, 1992; Gold and Johnson, 1982; Maclay and Askov, 1988; Messemer and Valentine, 2004; Shippen, 2008). There are major problems in using grade equivalents to denote adult literacy levels or progress in the literacy learning of children or adults. Grade equivalent scores do not represent an absolute standard, nor do they represent equal units at different levels of development.8 For adult learners, some assessment instruments are calibrated to important everyday literacy demands, and the scales from such instruments may be a far better indicator of adult literacy progress than grade equivalents.
8The misconceptions about what grade equivalent scores mean have been widely noted (e.g., Airasian, 1994; American Educational Research Association, American Psychological Association, and National Council on Measurement in Education, 1985; Miller, Linn, and Gronlund, 2009; Stiggins, 1997). The grade equivalent scale is not an equal-interval scale, although grade equivalent scores are often treated as if they represent equal units. This leads to the common misperception that someone who moves from 2.5 to 2.9, for example, has “grown” the same amount as someone who moves the same number of grade equivalents at a different level on the scale (e.g., from 8.5 to 8.9). Yet the amount of growth in ability needed to move from 2.5 to 2.9 is much greater than the amount required to move from 8.5 to 8.9. Furthermore, because grade equivalent units are not an equal-interval scale, they should not be used in mathematical calculations, such as determining the mean.
All of the gains in research to date are small relative to the amount of gain that would be needed for someone to achieve levels of literacy required for functional literacy (e.g., obtaining a high school diploma or postsecondary certificate or degree). The degree to which literacy gains may be accelerated is not clear nor the rate of gain to expect with engagement in instruction that has been demonstrated to be effective. A priority for research is to experiment with a variety of ways to more fully engage learners for longer periods of time to determine how to maximize literacy gains depending on the particular skills to be developed, the characteristics of the learner, and the features and intensity of the instruction. An additional priority is to develop more valid ways of measuring adults’ literacy gains than grade level equivalents with assessments normed for the population and designed to show progress in the specific component skills targeted and related improvements in valued literacy capabilities. One example of this approach is the Degrees of Reading Power (DRP) measure, which is a criterion-referenced assessment for use in grades 1 through 12 designed to measure reading facility needed for valued everyday activities (a specific level on the DRP implies the ability to read a job application, another level implies the ability to read a driving license test, etc.).
A large literature is available on practices used in literacy programs for adults in other countries. Few studies have tested whether particular curricula or pedagogies used in the programs result in better literacy skills or include quantitative assessments of the cognitive and literacy skills of learners. There are notable exceptions, however, that provide insight into practices that may be effective with low-literate adults in the United States (Abadzi, 2003; Baynham et al., 2007; Brooks et al., 2001, 2007). These include an evaluation of a research-based, functional adult literacy program developed and implemented for women in Turkey and evaluations of a program implemented in England (Skills for Life) and Northern Ireland (Essential Skills) as part of a national effort to increase adult literacy and numeracy skills. Many of the findings from these studies are consistent with K-12 research on effective practices for teaching reading and writing. An important feature of the research studies is that they include descriptive information about learners, learning contexts, and available supports to help interpret experimental research findings and understand the conditions that influence instructional effectiveness.
The functional literacy program developed in Turkey aims to develop dimensions of literacy (e.g., word recognition, listening and reading comprehension, writing), cognitive skills (e.g., critical thinking), and functional skills (e.g., performing everyday tasks) (Durgunoğlu, 2000; Durgunoğlu, Oney, and Kuscul, 2003; Kagitcibasi et al., 2005). It also includes components to promote the confidence and empowerment of women in society (e.g., discussion of legal rights). Volunteer tutors who
implement the curriculum receive intensive training in literacy and numeracy and how to teach adults and communicate with them effectively. Compared with existing courses, the new curriculum was more effective in meeting literacy skill goals (e.g., word recognition, spelling, reading comprehension), affective goals (e.g., increased self-confidence), functional goals (e.g., being able to find the right bus) and sociocultural goals (e.g., knowing about rights and voting). Affective and societal outcomes were evident 1 year later. Sustaining cognitive and literacy skills depended on the starting levels of skill, with students at higher levels of skill and continued self-study showing more sustained benefits.
In other research, between 2003 and 2006 the National Research and Development Centre (see http://www.nrdc.org.uk [Jan. 2012]) conducted a large-scale pre-post examination of a total of 1,649 adults who participated in programs in England to identify the effectiveness of the literacy, numeracy, and ESL practices that were implemented (Rhys-Warner and Vorhaus, 2007). A study of 298 participants at varied reading levels showed that gains in reading comprehension were modest and highly variable (Brooks et al., 2007). Predictors of progress included starting levels of literacy, self-study outside class, and having time to engage in pair and group work in class.
With respect to writing, for 199 learners studied who received between 40 and 79 hours of instruction (average 51 hours), progress in writing was slow, but several factors distinguished the classes with the greatest increases in writing scores: (1) learners spent time on the composition of texts of different kinds; (2) writing skills, such as spelling, grammatical correctness, and punctuation, were developed in the context of meaningful writing tasks; (3) there was time to discuss the writing process and the writing task; and (4) individual feedback and support were provided as learners drafted, revised, and proofed their work (Grief, Meyer, and Burgess, 2007).
Embedded case studies allowed for a deeper examination of effective instructional practices and point to several predictors of progress that warrant future attention in research with low-literate adults. These include (1) clear planning by the teachers, both strategic and on the spot (i.e., using the opportunities of the moment to teach); (2) explicit framing for the learner to provide a rationale for what is to be learned, the activities to be completed, and how these will help the learner; (3) focusing attention on how language is structured while encouraging and supporting talk in the language to be developed; (4) repeated reviewing and reworking of linguistic items (e.g., new words or structures) in different contexts; (5) professional vision and an understanding of objectives about language teaching and the ability to use and combine materials and activities creatively to work toward these objectives; (6) learning spoken and written language for practical purposes; (7) collaborative group work; (8) safe and fun learning to create a motivating environment that avoids labels and feelings of failure if one’s written
and spoken language is not consistent with certain standards; (9) avoidance of practices associated with decreased sustained engagement with literacy (irrelevant content, inappropriate teaching methods, inadequate teacher training, failure to take account of students’ expectations and needs, poor initial learning, top-down, didactic programs, and discrete skill instruction removed from content); and (10) skilled teachers who have time for professional development.
For some adults in these studies, shorter term deliberate instruction on fluency and phonological processing helped reading comprehension (Abadzi, 2003; Burton et al., 2010; Durgunoğlu, Oney, and Kuscul, 2003). Other adults showed little or no improvement, however, consistent with findings from the large-scale interventions for low-literate adults discussed earlier. These results point to the need to study in detail why progress in developing these skills is slow for many adults and why certain interventions are effective for some adults but not others. Learners reported several factors they perceived to help their progress: peer support, trusting the teacher, and explicit feedback, especially validation of their efforts and progress (Hannon et al., 2006; Ward and Edwards, 2002). Such findings indicate a need to develop various methods of assessment so that learners can continually assess themselves and each other to monitor progress toward learning goals (Dymock and Billett, 2008; Prins, 2010; Ward and Edwards, 2002).
The provision of professional development and support for educators affected program effectiveness (Balatti, Black, and Falk, 2007; Durgunoğlu, Oney, and Kuscul, 2003; McNeil and Smith, 2004). Fostering persistence with learning was a challenge in these studies that was met with efforts to provide programs in communities that are easily accessible by learners (Brooks et al., 2001; Guenther, 2002; McNeil and Smith, 2004). Practices associated with sustained effects on persistence over time include developing learners’ confidence and integrating literacy into their everyday lives, so that skills are used in meaningful and relevant ways and continue to be practiced (Aoki, 2005; Brooks et al., 2007; Dardour, 2000; Durgunoğlu, Oney, and Kuscul, 2003; McNeil and Smith, 2004; Prins, 2010; Puchner, 2003; see also Hurry et al., 2010; Thompson, 2002).
Beyond the intervention research we have described, most research in adult literacy education has been descriptive and qualitative (e.g., Kruidenier, 2002; see Appendix D). As mentioned earlier, the research is limited in its ability to identify practices and other influences on reading, writing, and literacy and, in fact, often did not set out to meet such goals. Yet an examination of this research as well as research on practices used with disengaged youth reveals topics that are important to pursue in future research to identify effective approaches to adult literacy instruction. The research often converges with findings from K-12 research on reading and writing and with research on learning.
Several themes from the available research about adult literacy warrant particular attention as topics for future research on adult literacy instruction: collaborative learning; contextualized instruction; instructional materials; writing instruction; funds of knowledge and authentic learning experiences; and social, psychological, and functional outcomes of literacy instruction. Before describing these topics, we note our examination of the literature reveals a number of popular theoretical frameworks used to guide the development of instructional programs for adults. The primary approaches include andragogy (Knowles, Holton, and Swanson, 2005), transformational learning (Mezirow, 1981, 1998), theories of self-directed learning or autonomy (Garrison, 1997; Tough, 1978), learning styles (see http://www.c-pal.net/course/module4/m4_learning_styles.html [Jan. 2012]), and multiple intelligences theory (Gardner, 1999, 2004). All of these approaches make assumptions about the learning preferences and needs of adults that have not been adequately tested.
Many of these approaches have not been informed by theory substantiated with empirical findings in cognitive science, motivation, developmental science, or neuroscience. For example, there is scant evidence that instruction matched to self-reported learning styles (visual, verbal, auditory) or distinct intelligences (linguistic, logical mathematical, musical, intrapersonal) improves instructional outcomes. In some of the approaches, the concepts are not defined well enough to measure, and findings (and the theories themselves) are underdeveloped even when measurement is plausible. Tailoring instruction to build on a student’s strongest skills, as a by-product, also decreases opportunity to build up weaker areas of skill. In general, there has been a rush to apply these approaches in adult education and literacy instruction without empirical examination of their core principles. This is not to say that all of the claims embedded in these approaches are inaccurate. Some claims (e.g., that autonomy and self-direction are important for learning and that collaborative learning and group work are beneficial) are supported by research in various disciplines and thus need further specification and evaluation in the context of adult literacy instruction.
Collaborative learning has been assumed to facilitate learning for several possible reasons. It has the potential to create a sense of community and connection that supports engagement with learning (Sissel, 1996; Soifer, Young, and Irwin, 1989), and it presents authentic opportunities
to engage adults in literacy tasks in communities of practice (Street, 2005; Taylor et al., 2007). Collaboration is also hypothesized to develop independence and familiarity with each learner’s strengths and challenges that can shape modeling and coaching (Taylor et al., 2007) and provide social support (Tett and Maclachlan, 2008).
It is uncertain, however, whether collaboration works in the ways hypothesized to develop valued literacy outcomes. Findings from K-12 research on reading and writing suggest that collaborative learning activities may facilitate learning under some conditions (see Chapter 2). The conditions that enable adults to benefit from collaboration need to be determined in future research. As others have noted (e.g., Bryan, 1996; Fingeret and Drennon, 1997; Hofer and Larson, 1997; Street, 2005; Taylor et al., 2007), such research must pay attention to setting explicit goals, the structure of the instruction and how groups are established, the literacy tasks used, and the quality of interpersonal interactions in groups.
Contextualized instruction is of particular interest to adult literacy practitioners both in the United States and internationally (Aoki, 2005; Casey et al., 2008; Guenther, 2002; McNeil and Smith, 2004; Thompson, 2002). The contextualization of skills is defined here as an instructional approach that creates explicit connections between the teaching of reading and writing and instruction in an academic discipline or content area (e.g., science, history, financial management, health, parenting, civics and government, engineering, mechanics). Many terms have been used to refer to contextualization, including contextual teaching and learning (Baker, Hope, and Karandjeff, 2009; Johnson, 2002), contextualized instruction (Parr, Edwards, and Leising, 2008; Wisely, 2009), content-area literacy (McKenna and Robinson, 2009), embedded instruction (Simpson et al., 1997), writing-to-learn (Klein, 1999), integrative curriculum (Dowden, 2007), situated cognition (Stone et al., 2006), theme-based instruction (Dirkx and Prenger, 1997), anchored instruction (Bottge et al., 2007), curriculum integration (Badway and Grubb, 1997), academic-occupation integration (Grubb and Kraskouskas, 1992; Perin, 2001), infused instruction (Badway and Grubb, 1997; Perin, 2001), developmental education learning communities (Weiss, Visher, and Wathington, 2010), workplace literacy (Mikulecky and Lloyd, 1997), and functional context education (Sticht, 2005).
Whatever term is used, the work tends to converge on the themes of (1) teaching skills with direct reference to real-world events and practices (Berns and Erickson, 2001; Carrigan, n.d.; Dirkx and Prenger, 1997; Fuchs and Fuchs, 2001; Goldman and Hasselbring, 1997; Johnson, 2002;
Jurmo, 2004; Karweit, 1998; Orpwood et al., 2010; Sticht, 2005; Stone et al., 2006; Weinbaum and Rogers, 1995) and (2) instruction in the basic skills needed in content courses (Boroch et al., 2007; Martino, Norris, and Hoffman, 2001; Perin et al., 2010; Snyder, 2002; Wisely, 2009). In some cases, contextualization occurs through the merging of basic skills and subject-matter instruction (Grubb, 1996; Guthrie et al., 1999; Paquette and Kaufman, 2008). Furthermore, the connection between basic skills and disciplinary learning is also seen in the newly developed national literacy standards for career and college readiness, which specify competencies for reading and writing in history, social studies, and science (National Governors Association and Council of Chief State School Officers, 2009).
The effectiveness of contextualized instruction has not been sufficiently evaluated for any population, including adult literacy students. Research is needed to identify the features of various contextual approaches that lead to both development of literacy skills and achievement of broader learning goals. A recent review yielded a small body of descriptive and experimental research with adolescents and adults that linked specific instructional practices to reading, writing, and mathematics outcomes, suggesting the value of pursuing this approach (Perin, 2011).
Much of the available research on adult literacy describes the use of authentic texts gathered from actual contexts in which adults used these materials (e.g., a workplace, a restaurant) or ways of reading fiction and nonfiction (Beaverstock, Bhaskaran, and Brinkley, 2009; Castleton, 2002; Fallon, 1995; Fingeret and Drennon, 1997; Forell, 2006; Pinsent-Johnson, 2007; Rhoder and French, 1994) and descriptions of how to match a learner with text and “debugging” it to bring it into the “learners’ instructional zone” (Rogers and Kramer, 2008). Notably absent from the literature are uses of instructional texts and materials systematically developed to match adults’ skill development needs, that connect with the interests of adult learners, and that draw from knowledge of other populations about the importance of reading varied forms of text for development of reading comprehension. The only quantitative study (nonexperimental) of the effects of authentic literature on the development of adults’ component skills was designed to study the effects of extensive reading and combined extensive silent reading of authentic literature chosen by students with teacher read-alouds and group discussion to teach beginning readers (reading at grade equivalent 1-3). This approach showed mixed findings. It was associated with increases in expressive vocabulary and fluency but not word analysis, receptive vocabulary, or reading comprehension (Greenberg et al.,
2009). A priority for research is the development of instructional materials and texts and effective practices for their use in developing both adults’ componential literacy skills and functional literacy outcomes.
A very small body of research focuses on writing in basic education students. Descriptions of instructional practices across studies are consistent with many classrooms adopting versions of the writing process approach first made popular by researchers of children’s learning to write, such as Graves (1983) and Calkins (1994) and, at the middle school level, Atwell (1987) (see Chapter 2). The process taught is meant to model what good writers do—brainstorm, draft, get feedback, revise, and edit (not necessarily in a rigid order)—(Beaverstock and McIntyre, 2008; Fiore and Elsasser, 1987; Padak and Baradine, 2004; Weibel, 1994). In most cases, students have the choice of what to write about in process approaches but are encouraged to draw on life experiences for topics and to write in the narrative form (Carter, 2006; Gaber-Katz and Watson, 1991; Moni, Jobling, and van Kraayenoord, 2007; Pharness, 2001; Shor, 1987; Siegel, 2007; Street, 2005; Woodin, 2008). This approach is believed to create feelings of ownership and help students be less reluctant to write (Street, 2005). The approaches sometimes include mini-lessons to teach specific technical aspects of writing (Fuller, 2009) and various forms of feedback (student to student or teacher to student).
The research we identified does not tend to focus, however, on practices to develop adults’ writing skill. The emphasis is more on documenting how teachers might help adults feel comfortable with writing, find their voice, develop an identity as a writer, understand how writers write, and use writing to bring about social change than on documenting how teachers engaged students in improving the technical aspects of writing for practical purposes. The instruction stresses writing for self-expression and communication; the process is assumed to be as important as the product. Thus, whether these various forms of writing instruction develop the component skills needed to perform literacy tasks for practical purposes, such as GED attainment, career success, financial management, health maintenance, and fulfillment of parental responsibilities, is not systematically studied and requires further research. Such research needs to consider findings from K-12 (see Chapter 2), which indicate that the process approach to teaching writing works best with professional development, that it may be more effective when combined with explicit instruction to develop specific skills, and that it may not be as effective in developing writing skill for those adults who struggle with writing.
Research on youth literacy practices suggests several approaches used with youth out of school that might inform the development of instructional practices for adolescents and adults in basic and secondary education programs. The approaches include a “funds of knowledge” framework, disciplinary literacy, cultural modeling, inquiry-based instruction, and anchored instruction. All of these approaches assume that people bring knowledge and experiences as well as literacy practices to learning that educators should understand and use to build new knowledge, support engagement, and establish shared expectations for learning. Curricular interventions that draw from community, family, and peer group funds of knowledge have been developed for elementary school children (e.g., Au and Mason, 1983; Heath, 1983; Moll, 1992; Moll and Greenberg, 1990; Moll and Whitmore, 1993), as well as adolescents (Gutiérrez, Rymes, and Larson, 1995). For example, teachers have used language and concepts drawn from students’ lives as a bridge to support their development of deep understandings of academic language (see Gutiérrez et al., 1999) and to build disciplinary knowledge and language (e.g., Lee, 1993, 1995, 2001; Moje et al., 2001a, 2001b, 2004a, 2004b; Morrell, 2002, 2004).
There has been a long tradition of community-based and after-school programs of media-intensive and arts-based instruction, especially for marginalized youth (e.g., Buckingham, 2003; Eccles and Gootman, 2002; Kafai, Peppler and Chiu, 2007; Peppler and Kafai, 2007; Soep and Chavez, 2005). Often drawing on popular cultural forms, including music and film and digital media, such programs include literacy-related skills and practices by immersing participants in language-rich and multimodal activities to reengage youth with learning. Although such programs do not typically measure success via academic literacy gains, research that has compared students who participate in these programs with nonaffiliated youth has suggested superior academic and social performance (Heath, Soep, and Roach, 1998; see Hull et al., 2006).
The qualitative research on adult literacy (see Appendix D) suggests an array of psychological, social, and functional factors that may result from or influence effective instruction to develop literacy skills. Similarly, the ultimate purposes of adult literacy programs in other countries are broad and studies of their effectiveness have included psychological outcomes (e.g., self-confidence, achievement of personal goals), functional outcomes (e.g., better performance at work), economic outcomes (e.g., employment), and social outcomes (e.g., positive engagement with family
or society) (Andersen and Kooij, 2007; Aoki, 2005; Balatti, Black and Falk, 2007; Casey et al., 2008; Durgunoğlu, Oney, and Kuscul, 2003; Dymock, 2007; Guenther, 2002; Hannon et al., 2006; Hua and Burchfield, 2003; Hurry et al., 2010; Kagitcibasi, Goksen, and Gulgoz, 2005; Prins, 2010; Prins, Toso, and Schafft, 2009; Puchner, 2003; Thompson, 2002). Far from being tangential, assessments of such broader social, economic, and functional outcomes can help to reveal both the conditions that support effective learning and instruction and the full impact of a literacy program that is measured not only in terms of literacy skill outcomes but greater and more effective involvement in family, work, and society. There is a need, however, to develop more reliable assessments of the full range of social, psychological, instrumental, and functional outcomes associated with effective adult literacy instruction (Dymock and Billett, 2008; Prins, 2010).
As in adult education, research has not focused on evaluating instructional approaches to improve the literacy skills of underprepared college students; for example, the committee identified only seven small studies from 1990 to 2009 (see Appendix D; Caverly, Nicholson, and Radcliffe, 2004; Friend, 2001; Hart and Speece, 1998; Martino, Norris, and Hoffman, 2001; Rochford, 2003; Scrivener et al., 2008; Snyder, 2002). Most reported small gains in various aspects of literacy, but problems with the study designs prevent drawing conclusions about effectiveness. Only one study included a randomized design; it tested the effects of a learning community approach that produced small gains (e.g., higher pass rates for college placement reading and writing test) (Scrivener et al., 2008). None of the studies compared teaching methods. The number of teaching methods researched was approximately equal to the number of studies; thus, a sustained program of research is not available for understanding which approaches are likely to work well for which students if implemented on a large scale, how to implement the approaches, and the conditions that support effectiveness. Progress in the reading and writing skills that were taught in these studies was not commonly or directly measured.
Similarly, descriptive studies with the population lack sustained and programmatic research on instructional approaches (see Appendix D). As for adult education, descriptive studies of practices used to develop reading and writing skills did not usually describe outcomes or analyze links between the practices and change in the outcomes of students.
A body of work on writing with low-skilled postsecondary students, especially studies focused on text-based analyses and cognitive process approaches, converge with findings from the K-12 literature and warrant
further mention. This research has consisted of quantitative experiments, quasi-experiments, and longitudinal correlational studies, as well as content analysis, discourse analysis, and case studies.
Text-based analyses of the writing of college students and English language learners in college have focused on error correction, sentence length and variation, audience awareness, and proficiency with specific genres. These studies report the nature, timing, and modality of feedback on elements of writing (Duijnhouwer, Prins, and Stokking, 2010; Hassel and Giordano, 2009; Morra and Assis, 2009; Sheen, Wright, and Moldawa, 2009; Yeh, Gregory, and Ritter, 2010). For example, the modality of instructors’ comments (i.e., written or audio-recorded) (Morra and Assis, 2009) and the type of feedback instructors provide (Duijnhouwer, Prins, and Stokking, 2010) can affect students’ abilities to cope with increasing difficulty in assignments (Hassel and Giordano, 2009) and increase their self-efficacy and motivation to continue tasks with difficult writing prompts, although the feedback on progress did not affect students’ actual writing performance (Duijnhouwer, Prins, and Stokking, 2010). Several approaches are associated with students’ ability to self-correct errors in sentences. Explicitly correcting their errors for them appears to be less effective than explicitly teaching types of error patterns (Shaughnessy, 1979) or teaching students to identify errors in their own writing, using such strategies as reading aloud (Bartholomae, 1980), proofing their own papers with explicit instruction in error labeling (Morra and Assis, 2009), or using online error correction and analysis feedback systems (Yeh, Gregory, and Ritter, 2010).
In the early 1980s, research proliferated on the cognitive processes of students’ writing and the problems experienced by students referred to as at-risk, underprepared, basic, or remedial college writers (Hull and Bartholomae, 1984). Often single-subject designs and case studies document writers’ cognitive processes using think-aloud protocol analysis and qualitative observations of students’ writing. Cognitive process approaches for basic writers have focused on helping students to be attentive to the needs of audiences (Flower, 1979) and become aware of times when they might be prone to writer’s block (Rose, 1984) or to overedit their writing (McCutchen, Hull, and Smith, 1987). Students also become aware of “rigid rules and inflexible plans” (Rose, 1984) that limit their abilities to produce the drafts needed to successfully complete assignments. With explicit instruction that targets these barriers to writing, students have produced longer and more detailed first drafts (Eves-Bowden, 2001). With explicit instruction to develop self-regulated learning, students demonstrate a wider range of metacognitive abilities to guide their writing processes (Nuckles, Hubner, and Renkl, 2009). Specific types of mini-lessons also have emerged from this research and warrant further study of their effectiveness with
adults. These lessons include drafting to move students from writer-based prose to audience awareness, error identification, and the use of software and approaches to help them develop text and manage their writing process. Consistent with K-12 studies showing the benefit of peer assistance with writing and the positive effect of writing on comprehension and the learning of content, college students who have opportunities to receive feedback from peers about their writing show increased learning of subject matter (Cho and Schunn, 2007, 2010; Cho, Schunn, and Kwon, 2007). These were not students in developmental education courses, however, and thus the approaches need to be evaluated with college students who need to develop their literacy skills.
There is a severe shortage of research on effective reading and writing instruction for adults, despite the large population of U.S. adults needing to develop their literacy skills (Baer, Kutner, and Sabatini, 2009; Kutner et al., 2007) and the fact that adult literacy instruction has been offered for many years (Sticht, 1988). The shortage exists for several reasons, including the high attrition rate of research participants, a lack of attention to reading and writing skills as an outcome of literacy instruction, and the use of methods that do not allow for identification of cause-effect relations between an instructional practice and outcomes. More broadly, the field has lacked a comprehensive, sustained, and systematic agenda to produce curricula, practices, texts, and other tools that meet the skill development needs of adult learners. Research funders and thus researchers of literacy have chosen to focus mainly on preschool and K-12 populations, a situation that has constrained the amount of research with adults outside school.
The research that does exist has several limitations, consistent with Beder’s (1999) observations more than a decade ago. The high rates of attrition and lack of well-controlled experiments have led to a body of research with small sample sizes and results that are difficult to interpret. The research also suffers from assessments that lack validity for the population; inadequate descriptions of the subgroups of adults being studied; over-reliance on self-report; vague or incomplete descriptions of instructional practices, outcomes, and study procedures; and a lack of standards against which to judge the utility and significance of findings (e.g., few agreed-on curricula or standard practices that can be tested, varying learning objectives not linked to standard measures, or expected effect sizes).
Box 3-2 summarizes priorities for research given the limited knowledge of the effectiveness of instructional approaches and adults’ learning contexts. The chapter has highlighted additional specific topics that warrant
Needs for Research on Adult Literacy
Development and Instruction
• Development and testing of motivating instructional practices and materials and noninstructional supports for effective instruction and sustained engagement and persistence with learning.
• Controlled experiments to evaluate curricula and instructional practices with a focus on explicit and systematic instruction, opportunities for extensive practice, and well-designed texts to build language and literacy skills related to functional learning goals.
• Valid methods of measuring component skill and functional literacy gains based on adult norms.
• Qualitative research using rigorous methods to obtain rich description and analysis to point to possible links between instructional practices and learning outcomes, help to interpret findings from effectiveness research and establish the boundary conditions of an instructional effect, and assess implementation fidelity (how practices were actually implemented).
• Establishment of standard research protocols for defining subgroups of learners and generating more complete and comparable information across studies related to the characteristics of learners, instructional practices, the quality of implementation, and instructional outcomes.
• Alignment of standards for literacy instruction, with empirical linkages among literacy activities, goals, and standards, across the programs and systems that provide literacy instruction (K-12, adult education programs, postsecondary education).
• Ongoing collection of educationally meaningful data across the systems that provide literacy instruction on learners’ skills, learners’ characteristics, the quality of instruction, and learning environments to enable effective instructional planning and delivery.
• Studies to identify (1) learning progressions for diverse subgroups of adults in the context of instruction and (2) how instructional approaches might need to differ at various points in the lifespan and according to the needs of particular subgroups of adults.
• Development and testing of professional development systems to ensure that teachers have the knowledge and skills required to implement instruction effectively and differentiate instruction to meet learners’ needs.
future study, such as collaborative learning and contextualized approaches to teaching reading and writing.
To elaborate on these priorities, only a handful of interventions have been tested to develop the skills of low-literate adults in adult basic education, adult secondary education, or colleges. Although gains have been reported, they are not substantial for this population either in terms of the size of intervention effects or gains observed against the amount of gain
needed to be functionally literate. More needs to be known about the features of instruction and the intensity and duration required to maximize gains for adults who vary widely in their literacy skills.
Except for a few studies, research on component literacy skills has not been a priority and has not drawn on findings about effective literacy instruction with K-12 students. Although research with young children yields information about the targets of instruction and effective practices, the instruction may not always work as well for adults. Research is required to validate, identify the boundaries of, and extend this knowledge for the adult population. For example, it is clear that the adults who read at the eighth grade equivalent level and lower lack sufficient reading fluency to support optimal comprehension. That is, their word recognition processes are slow and divert cognitive capacity from making sense of the text. A substantial number of adult learners lack some word recognition (decoding) skills. Although research shows that direct decoding instruction is effective in developing word reading for most young students, the conflicting findings obtained thus far for adult learners suggest that the approaches used with children may not be as effective. They may not be sufficiently motivating or may not be implemented with the intensity or duration needed to be effective for some learners. Differences in cognitive function at different ages (e.g., size of any short-term phonological store, attentional capacity that might allow the use of strategies that are not possible with children) also may call for different phonemic awareness and word analysis strategies to accelerate progress. Longitudinal studies will be valuable for discovering how the processes of reading and writing might change with age and how instruction to develop reading and writing skills might need to differ at various points in the life span.
Consistent with K-12 research, it is likely that multiple approaches, if designed following principles of learning and instruction reviewed in this volume, may prove to be effective. Regardless of the approach, it can be assumed that the instruction should create a positive climate for adults that draws on their knowledge and life experiences, uses materials and learning activities that develop valued knowledge and skills, and supports adults as much as possible in regulating their own learning. It is also important to ensure that instructional activities to develop such skills as word recognition and decoding are provided when specific diagnostic evidence suggests that they are needed.
Research needs to identify the approaches that are effective for identified subgroups of adults, unless there is evidence that a particular approach works for all. At some level, it is obvious that instruction needs to be differentiated, just as in K-12 research, depending on the particular skills and other characteristics of learners and larger learning goals. Research is required to understand the constraints on generalizing findings across
subgroups and how to effectively differentiate instruction. A significant portion of adult learners will have multiple disabilities, including learning disabilities. This fact is inevitable, since K-12 students with diagnosed disabilities have lower literacy levels than students without disabilities (National Assessment of Educational Progress, 2010). This fact suggests both the need for careful assessment to drive choices in reading and writing instruction for adults and the likely need for instruction to overcoming specific reading and writing disabilities. As described in Chapter 2, research with K-12 students does not suggest that reading and writing instruction for students with and without learning disabilities is qualitatively different; rather younger populations with disabilities have benefitted from more intensive, explicit, and systematic instruction that targets specific reading and writing difficulties and the transfer of learned skills, and they require more opportunities to practice.
An important area for research is the development and evaluation of texts and other tools for learning experiences that are linked to well-defined instructional objectives and to adults’ skill development needs. Instructors now select and adapt texts and materials with little guidance from research and in an ad hoc fashion. An interdisciplinary effort involving researchers, practitioners, and curriculum developers would help to address the lack of supportive reading materials for instruction and practice. These materials need to (a) be appropriately matched to assessed proficiencies and scaffold the practice needed to become facile in applying component reading skills, (b) present topics of interest to the reader and valued content related to broader learning goals, and (c) draw from research on what is known about practice with varied forms of texts to facilitate reading comprehension.
Any effort to apply what is known from K-12 research should be informed by research on learning with adults (see Chapter 4). For the extrapolations to work, however, and to engender confidence in the proposed approaches, they must be grounded in a theory of learning that is supported and understood well enough to provide a basis for the extrapolation. To date, popular frameworks in the adult education field about how adults learn lack sufficient empirical support, and the theories of adult cognition and learning from psychology and cognitive science have been developed with homogeneous populations and, most often, samples of convenience. Substantial conceptual and empirical work is needed to validate and further develop theories of learning in research studies that include broader populations of adults, such as those needing to develop their literacy skills. Some of this work could be accomplished in the context of research designed to identify effective approaches to reading and writing instruction.
It is difficult to interpret the available literature because the research does not include standard descriptions of subgroups of the population or account for possible group differences in the findings. The adult literacy
learner population is diverse in terms of age, culture, languages spoken, learning disabilities, literacy attained to date, educational background, and other life experiences that require systematic attention. In addition, reading ability is confounded with many psychological and social variables that may influence the effectiveness of instruction. For example, reading ability, major depression, and conduct disorders, all significantly predict dropping out of high school, and reading ability in high school is related to minority status and lower socioeconomic status (Daniel et al., 2006). Poor adolescent readers self-report higher levels of depression, trait anxiety, and somatic complaints than typical readers (Arnold et al., 2005). Enhanced descriptions of learners would help to identify constraints on learning that need attention and help to interpret the results of effectiveness research. Similarly, research reports lack critical information about the components of the interventions or how they were implemented. As a result, it is difficult to ascertain the precise nature of the instructional practices that have led to (or failed to produce) gains. A more coordinated approach with established research protocols is needed to accumulate useful information about well-defined subgroups of learners and to produce a more coherent and interpretable body of information about effective instruction.
Qualitative research in adult literacy has suffered from thin descriptions and inadequate analysis of linkages among instructional goals, practices and outcomes. It also lacks attention to component literacy skills. There is an unfortunate alignment of research methods with the instructional goals and practices studied. Qualitative methods are used in studies grounded in sociocultural theories of learning and literacy. These studies focus mainly on social and psychological goals for literacy instruction and not on the teaching of reading and writing skills. Quantitative methods are used in research on reading and writing skills grounded in cognitive theories of learning. These studies focus on explicit and systematic instruction to facilitate acquisition of component skills (e.g., decoding, vocabulary, and comprehension).
If established standards for analytic methods are followed, qualitative research, which ranges from ethnography to basic observation checklists, has the potential to contribute in at least three ways to the identification of effective literacy instruction for adult learners: (1) it points to possible links between instructional practices and learning outcomes; (2) it yields rich descriptions of individuals and environments that help to establish the boundary conditions of an instructional effect; and (3) it helps to assess implementation fidelity (how practices were actually implemented). The research, whether quantitative or qualitative, needs to be designed to establish a clear relationship between an instructional practice and literacy outcomes that incorporate both component literacy skills and functional literacy tasks related to learning goals.
For many adults, the enhancement of component reading and writing skills itself is not the ultimate objective, but the attainment of larger life goals related to career and educational advancement and improvement in the lives of their families. The instructional practices studied in research must have clearly stated learning goals and objectives. These should take into account both the need to develop component skills of reading and writing and the literacy facility needed for education, work, parenting, and other purposes. It is also important to empirically document the particular constellations of component literacy skills needed to perform important literacy tasks associated with larger learning goals (e.g., GED preparation, college entry and completion, fulfillment of parental responsibilities, performance of workplace skills, participation in civic responsibilities).
Although learners across programs share literacy development needs and learning goals, the current system of instruction is a loose mix of programs in many places that lack coordination and coherence with respect to what is taught and how. There also is a lack of alignment to be addressed in the learning objectives for literacy development across adult education, colleges, and K-12 instruction. Adult literacy research is hampered by the lack of a coherent system and established curricula with materials and standard practices that can be tested. An empirical mapping of component skills to literacy tasks and learning goals would offer a basis for aligning literacy instruction across places and systems of instruction and for developing standard instructional curricula and practices to meet the needs of diverse learners across learning contexts.
At present, information is limited from adult education programs and colleges about the specific reading and writing development needs of the adults they serve; the instruction that is used and whether it is implemented effectively; and whether the instruction facilitates development of reading and writing skills needed to achieve broader learning goals. There is a need for ongoing collection across the systems that provide literacy instruction of data on learners’ skills, the quality of instruction they experience, and other characteristics of learners and learning environments to enable planning and implementing instruction effectively and the tracking of progress. A sound assessment system is needed to support and monitor learning at the individual, program, and systems levels to plan instruction and track progress in the component reading and writing skills and functional literacy skills related to broader learning goals.
A primary problem to resolve is how to engage adults in the amount and intensity of instruction and practice that is required to develop literacy skills and conduct the needed research. High attrition rates, which are typical of both adult literacy and college developmental education programs (Alamprese, 2009; Comings, 2009; Goldrick-Rab, 2007), compromise the integrity of research findings and can be a disincentive to research. Several
factors that are known to affect the amount and intensity of instruction and sustained engagement with learning need attention in future research. As discussed in more detail in Chapter 5, there are several ways in which an instructional approach or environment can affect motivation to persist, among them inappropriate focus or inadequate quality of the instruction, lack of clear learning objectives, failure to be explicit about or to set appropriate expectations about progress, lack of awareness of the progress that has been made, and unwanted identity as a remedial student or low-literate adult. As discussed in this chapter, studies of low-literate adults in other countries also show possible reasons for the lack of sustained engagement with literacy.
Time for learning is usually constrained for adults because of limited program funds and locations (a few hours of instruction are offered a few days per week), participants’ work schedules, transportation difficulties, child care responsibilities, and other life demands. Even when personal motivation is high and instruction is appropriately motivating, some subgroups are unlikely to persist, such as those with jobs who need several hours to get to and from a learning site. Some low-literate adults have social service needs associated with poverty (Alamprese, 2009; Tamassia et al., 2007)—teenage pregnancy, physical disability, illness, alcoholism, drug addiction, or domestic violence—(Sandlin and Clark, 2009) that need to be addressed. These various barriers and problems may lead teachers and programs to offer services and advocacy in conjunction with literacy instruction, which, as covariates in impact studies, can be hard to control or measure. Although some amount of attrition may be handled with more effective instruction, expanding the scope of instructional research to systematically account for these other factors and reduce barriers to learning appears necessary if reading and writing instruction is to be effective—and effectively studied—with this population.
Another clear impediment to instructional effectiveness and to conducting the needed research is the highly variable knowledge and expertise of adult literacy instructors (Smith and Gillespie, 2007). Instructors vary in their knowledge of reading and writing development, assessment, curriculum development, and pedagogy. The training instructors receive is generally limited and professional development is constrained by lack of funding, inflexible locations, work, and other life demands. Nonetheless, the instructors must reliably assess learners’ skills, plan and differentiate instruction, and select and adapt materials and learning activities to meet the skill development needs of learners who differ greatly in their neurobiological, psychosocial, cultural, and linguistic characteristics. To be effective, teachers will need to have the requisite tools for instruction, the technical knowledge and expertise, professional development, and ongoing supports.