Can Community Colleges Protect Both Access and Standards? The Problem of Remediation
by Dolores Perin - 2006
A large number of community college students have difficulty with postsecondary-level reading, writing, and math demands, necessitating remedial education. A qualitative case study was conducted to investigate state and institutional practices for remediation in 15 community colleges selected for region, size, and urbanicity. The six states in which the colleges were located varied on the level of regulation of institutional remedial policy and were placed on a spectrum ranging from laissez-faire to micromanagement. Most of the states and all the institutions in the study required the assessment of students' academic skills, and the institutions mandated assessment even when the states did not require it. The types of assessment instruments varied, and subjective measures such as institutional tests, course grades, and student self-report played an important role in placement decisions. The colleges tended to require that low-scoring students attend remedial courses even in the absence of a state mandate. A wide variety of practices were used to determine student readiness to advance in or exit from remediation. Many of the institutions had procedures designed to require remediation early in the student's program, but both assessment and placement mandates appeared to be softened either at the state or institutional level, with the effect of reducing the number of students who were required to enroll in remedial courses. This trend is discussed as a struggle between the access mission of the traditionally open-door community college, and the drive to protect educational standards.
Although they have completed secondary education, a large number of college students lack the literacy and mathematics skills needed to learn at the postsecondary level (Spann, 2000). Many of these academically under-prepared students attend community colleges. There they receive remediation (also called developmental education), which has been defined as a class or activity intended to meet the needs of students who initially do not have the skills, experience or orientation necessary to perform at a level that the institutions or instructors recognize as 'regular' for those students (Grubb et al., 1999, p. 174). Developmental education programs typically provide reading, writing, and math courses, and tutoring and counseling, all designed to prepare students to participate fully in a postsecondary-level learning experience (Casazza, 1999).
All publicly funded community colleges offer developmental education programs (National Center for Education Statistics, 2003), and almost half provide contracted remedial courses to business and industry (Shults, 2000). Developmental education enrollments are often used to measure the extent of academic difficulty. At least half of community college students need such courses (Jenkins & Boswell, 2002), up to 80% enroll in at least one remedial course, and in urban institutions, over one quarter of students enroll in remedial courses (Shults). Because developmental education enrollments appear to underestimate students' difficulty with college-level reading, writing, and math (Perin & Charron, in press), the number of academically underprepared students may be even higher.
Developmental education is central to the community college mission (Howard & Obetz, 1996) and has been seen not only as an illustration of a commitment to educational access (Grubb et al., 1999) but also as a benefit to democratic society in general (McCabe, 2000). Nevertheless, community colleges have been criticized for being overly involved in remediation at the expense of baccalaureate transfer (Nora, n.d.; Rhoads & Valadez, 1996) and for duplicating K-12 education (see Oudenhoven, 2002). A survey conducted by Immerwahr (1999) found that college, business, and government personnel felt that the largest problem facing colleges is that too many new students need remediation (p. 28). However, remediation is integral to the open admission policy of community colleges (McGrath & Spear, 1991) and is a necessary service if the community college door is to be kept open (Zeitlin & Markus, 1996).
Providing remediation may create tension between access and standards goals, placing community colleges in a bind. Access goals are achieved if all applicants with a secondary education credential are admitted to postsecondary programs.1 Along with the commitment to access, community colleges also wish to maintain high standards,2 a goal that is threatened by the presence of large numbers of low-skilled entrants. A necessary, if not sufficient, condition for meeting standards goals would be to teach all under-prepared students the reading, writing, and math skills they need for college-level work. According to an ideal model of remediation, entering students are assessed for academic difficulty and then placed in remedial courses if their skills are found to be below the college level (Boylan, Bliss, & Bonham, 1997; Roueche & Roueche, 1999).
In fact, considerable variation exists in remedial assessment and placement policy, which is formulated by both states and local institutions. Although local governance can eclipse state regulation of community college policy (Shaw, 1997), the states often influence institutional activities through their roles in funding, regulation, consumer advocacy, and guidance (Richardson, Bracco, Callan, & Finney, 1999). The types and strictness of state policy for remedial education have been traced to variations in the level of higher education coordination (Shaw, 2000), differences in ideology (Gum-port & Bastedo, 2000), and the level of interest in raising educational standards (Mazzeo, 2002). Shults (2000) and Jenkins and Boswell (2002) summarized the remedial policies of states and community colleges, but neither study aimed to study the practices of specific institutions in relation to the policy of the states in which they were located. However, signs of inconsistency between state and local remedial policy were suggested in a study by Shaw (1997) who, in the course of investigating the impact of ideology on policy, found differences in assessment and placement policy within two community colleges in different states that had similar state remedial policy. A mismatch between state and local policy may compromise states' abilities to accomplish some of their primary goals (Richardson et al.).
One possible reason for state-local inconsistency specifically in the area of remediation is the ambiguity inherent in the term college-level; there is little consensus on what students need to know or be able to do to learn from the postsecondary curriculum (Merisotis & Phipps, 2000; Oudenhoven, 2002; Phipps, 1998). Whatever the explanation for the inconsistency, there has been little in-depth investigation of how postsecondary remedial policies are operationalized at state and local levels, how state and institutional remedial policies are related, or how these policies bear on access and standards goals. A case study of remedial assessment and placement policy was conducted within the National Field Study, a multitopic study of current issues in community colleges of the Community College Research Center at Teachers College, Columbia University. The following questions were asked: (1) What are the state and institutional policies for remediation? (2) To what extent do the National Field Study states vary concerning the regulation of remediation in the community colleges? (3) To what extent are remedial practices at the National Field Study sites consistent with state policy? (4) How do specific remedial practices function regarding access and standards goals?
The case study sites were 15 community colleges located in six states representing the major regions of the United States: Northwest, West, Southwest, Midwest, South, and Northeast. All the states had large, well-developed community college systems. Although generalizations cannot easily be made from qualitative case study research, the current findings have relevance beyond the sample because four of the six states together accounted for almost half of community college enrollments in the United States. Further, the community colleges within each state were selected to vary in size, urbanicity, amount of autonomy versus state regulation,3 community college boards, state policy roles (including regulatory and steering), means of financing the college operation, and level of state funding for community colleges (a full description the selection of states and sites may be found in Bailey & Morest, in press). A pool of candidate colleges was generated, and the final selection was made based on the willingness of the institution to participate. The sites, states, urbanicity, enrollment, and proportion of minority participation are shown in Table 1. All figures are from the Integrated Postsecondary Education Data System (IPEDS; 2000) except where indicated. It can be seen that the sample consisted of five urban, five suburban, one mixed (urban and suburban), and four rural community colleges in which enrollment ranged from 1,854 to 28,862, with 5% to 96% minority participation.
This qualitative case study was instrumental (Stake, 1995) in its attempt understand state and institutional remedial policy, and interpretive (Merriam 1988) in its construction of categories to describe remedial practices. The study was approved by the Institutional Review Board at Teachers College, Columbia University, and participants provided signed consent in advance of the data collection. Confidentiality of both institutions and individuals was promised, and fictitious names are used in reporting the findings.
Each site was visited by a team of researchers from the Community College Research Center, including at least two senior researchers and three research assistants who were knowledgeable about community college education. An interview protocol containing questions and probes for all the topics being studied was developed jointly by the project leaders, with sets of questions designated for the four categories of interviewee: administrator, faculty, counselor, and student.
A total of 630 people were interviewed individually or in groups for approximately 1 hour each. The interviews were open ended, beginning with the questions on the protocol. On the assumption that each interviewee had unique experience (Stake, 1995), the interviewers made tactical decisions about how to use the time allotted for the interview. In the interests of capturing as broad a range of information as possible both within and across sites, not all questions were asked of all interviewees or in all colleges, and the time allotted to each question varied depending on the interviewees' roles and backgrounds. Although the findings cannot be generalized, an advantage of the methodology is that it affords "particularization" (Stake) and "specification" (Patton, 1990)that is, an in-depth examination of issues concerning remediation in community colleges. Each interview was transcribed; the transcripts constituted the main source of data for this study and were supplemented by review of materials such as college catalogs, curriculum materials, institutional reports, and state policy documents.
From the total of 458 interview transcripts generated in the project, 290 (63%) were selected for coding for the remediation study. This set of transcripts was selected through a word search using 20 different terms relating to developmental education. The transcripts were analyzed using QSR-Nud*Ist version N5 software (http://www.qsrinternational.com). An initial set of 76 inductive free nodes was created, which was transformed to a set of deductive tree nodes. In a second pass through the data, the 76 codes were arranged in 12 larger categories that were then used to recode the transcripts (codes available from the author.) A coding report was produced for each of the 12 codes, from which salient information was selected for reporting. In identifying this information, a single mention of a policy or practice was considered as important as several mentions because limitations in the time that interviewees were available for participation and the researchers' planned flexibility in interviewing did not allow questions to be distributed equally across interviewees and sites.
The current report concerns the segment of the data relating to assessment and placement policy. (Findings on other aspects of the remediation study are reported in Perin, 2004, and Perin & Charron, in press). Information about assessment and placement policy found in the college and state documents was categorized into the same areas as the QSR codes and used to augment the findings from the interviews.
The author initiated the coding and then collaborated with a research assistant who subsequently conducted the bulk of the coding under the author's supervision. The assistant in turn trained and supervised another individual who performed some of the coding. The author checked samples of the coding, and disagreements were resolved through discussion. The document review was conducted by the research assistant under the author's supervision.
STATE AND INSTITUTIONAL POLICY FOR REMEDIATION
Five different categories pertaining to assessment and placement emerged: (1) assessment: mandatory or voluntary; (2) type of assessment measure used; (3) setting of cut scores: state or institution; (4) remedial placement: mandatory or voluntary; and (5) timing of remediation: when in the student's program the assessment and remediation occur. The information was organized in terms of stated policy, adaptations of policy, and implementation of policy, as shown in Table 2.
Skills assessment: mandatory or voluntary?
Skills assessment is important because it is the first step toward placing students in remedial classes. The assessment identifies the student's level of skill, which in turn determines whether remedial placement is needed. All but one state (Washington) had policy mandating the assessment of the reading, writing, and math skills of incoming students. Illinois policy contained a loophole in allowing the colleges to establish their own assessment procedures while not enforcing the mandate.
Nationally, the majority of community colleges mandate skills assessment (National Center for Education Statistics, 2003; Shults, 2000) and all the institutions in the current sample did so. Although not required by the state, the two National Field Study sites in the state of Washington either mandated assessment (NWRCC) or encouraged it so strongly (NWSCC) that it was classified in this research as an institutional mandate. An interviewee at NWRCC stated the importance of assessing students' skills.
If you do not have a placement test, if the students self-select their math classes, then when you walk into that math class, you have three types of students. You have the students: this class is perfect for them, their math skills are right here, this is where they need to be. You have the student that is under-selected, they really belong in a high math class, they are going to be so bored to death, they're going to be sliding under the table, snoring. And the third student is the student who has no business being in this class because their skills are not high enough. (Developmental Education Faculty, NWRCC)
In most community colleges, students who enter the college with low scores on a previously administered achievement measure must take a locally selected skills assessment test. Although Jenkins and Boswell (2002) reported that no state uses high school exit exams to determine placement in college remedial courses (p. 3), the current study found that in Texas, students who had passed a state high school graduation test were exempt from the skills testing. However, not all students who passed the test were ready for the college curriculum, and at SWSCC, there was some trial and error placement at the beginning of each semester:
[At the] very beginning of college level of math courses the drop rate is so high and it's because these are the students that still can't do it. They didn't have to take our assessment test, so no one even suggested take a remedial course first, to brush up. So students sometimes put themselves back within the first week I'm over my head. Can I get into your class? This goes on the first week of classes a lot. (Academic Faculty, SWSCC)
Although all the National Field Study colleges required skills assessment, in several instances, the specific ways in which this policy was implemented had the effect of softening the mandate, thereby reducing the number of students who could be placed in remediation. First, despite state policy, several institutions did not assess all skill areas: SWSCC did not assess writing skills, and MWSCC claimed that it did not require writing or math assessment for all students. Parenthetically, a closer look at MWSCC showed that writing and math assessments were necessary for the large part because any student who wished to enroll in college-level English and math (which were needed to graduate from many programs) was required to take the writing and math assessment tests. MWSCC also required that students who needed reading remediation obtain an advisor's permission to enroll in these courses. Both of these practices can be interpreted as attempts on the part of MWSCC to protect the standards of its college-level English and math courses. Further, limiting the number of assessment areas while requiring full assessment for degree aspirants suggests a certain amount of policy confusion.
A second adaptation of a state assessment mandate was that in some institutions (e.g., SMCC), assessment was not required for most of the students in career programs. This policy reduced by a substantial amount the number of students who might need remediation, because many community college students intend to obtain career degrees (Dougherty, 1994). A third adaptation was found at MWUCC and MWSCC, where instructors of the college-level English and math courses could override the assessment requirement altogether by signing into their courses students who had not taken the placement exams; in fact, this adaptation may have been rare:
And it's still always up to a faculty member to override if they choose to. But . . . most faculty members won't . . . because they're not trying to keep them out, they're trying to help them, and it doesn't do them any good to get in before . . . they're ready. (Developmental Education Faculty, MWSCC)
In summary, most of the states and all the institutions in the study mandated the assessment of students. If the state did not mandate assessment (Washington), the colleges did it anyway. Three local assessment practices appeared to weaken the state or institutional mandates for assessment: Not all skills were assessed, not all students were assessed, and instructors could override the assessment requirement.
Selection of assessment instruments
Reading, writing, and math skills were assessed at the National Field Study sites using state-developed tests, commercially available measures, and locally developed informal instruments. The state and commercial measures were standardized and normed (i.e., objective measures of performance). In contrast, the local tests were informal measures without norms and whose validity and reliability were unknown (Salvia & Ysseldyke, 2004). However, although they are subject to bias, informal measures are appealing to educators because they are directly related to curriculum (Lipson & Wixson, 2003).
The selection of assessment instruments was determined by state policy in only three of the National Field Study states: Florida, Texas, and New York (municipal system). New York (municipal) had the strictest mandate in requiring that a single commercially available measure be used by all colleges in the system. Florida and Texas4 both developed their own tests. Florida mandated the use of its instrument in all the state's higher education institutions but softened this policy by allowing grades in certain specified courses to substitute for the test. Texas permitted the colleges to choose between the state test and any other measure. The remaining statesWashington, California, and Illinois, and the state university system in New Yorkdid not have policy for assessment measures.
The institutions varied considerably regarding their selection of assessment instruments. Eight of the 15 sites used a single measure rather than multiple measures. Three of these eight did so as a result of a state mandate (SUCC, SMCC, and NEUCC), and the others by local choice (NWSCC, WRCC, MWRCC, NESCC, and NERCC). The seven remaining sites used varying combinations of state, commercial, and local measures. In Texas, SWUCC used the state test and a commercially developed measure, but SWSCC used a combination of state, commercial, and institutional tests. Three institutions, MWSCC, MWUCC, and NWRCC, located in Illinois and Washington, used a combination of commercial and institutional tests.
A number of the sites had changed their assessment strategies in recent years. For example, at SWSCC, an earlier writing measure had been abandoned because the results offended the high schools. When high school performance was substituted for the college test, there was a negative effect on the standards of the college English course:
English was using up until [a few years ago] a holistic writing sample, where the English faculty would grade this writing sample by students and make a holistic decision whether they could go into college English or needed to be remediated. That got very controversial with the high schools, and so [the college decided that] anybody who finishes high school English can go right into college English and that was fine, the high schools are happy but the students aren't successful. They're not passing. So we have had just a radical drop in the number of students in developmental English and a rise in the number of students in credit English and a huge drop in success rate in credit English. (Administrator, SWSCC)
The California institutions both used single institutional measures, which differed considerably from each other. WUCC was dissatisfied with a previously administered commercial measure, and at the time of the site visit was using its own locally developed multiple-choice writing and math tests. There were different versions of the test for native-English speakers and students with limited English proficiency.
WSCC had also abandoned the use of a commercial measure after finding a low correlation with subsequent remedial course grades5 and at the time of data collection was using a locally developed self-assessment survey in which students answered questions about their writing and math skills. Because this institution was unique in the sample in giving students free reign to decide on their remedial needs, it is worth exploring this policy in some detail. Several interviewees felt that the policy was effective. Students appreciated the autonomy, and the outcomes seemed the same as when tests are used.
The student response is that, Hey, you're treating us like adults. We are responsible. We can make this decision. We like it. It seems to be working. (Administrator, WSCC)
[English and math instructors are] claiming that whether a student is given an assessment or whether the student places him or herself, grades at the end of the semester seem to come out the same way, and ability levels sort of find where they belong. (Academic Faculty, WSCC)
Another interviewee stated that the self-assessment policy had not affected the standards of the degree-credit math courses because they have strict prerequisites. However, the college had removed the prerequisites on many other courses, so higher numbers of underprepared students began entering discipline-area courses. In other words, the self-assessment policy only had a chance of being useful if the college-level courses applied prerequisites. An accounting instructor stated,
Several years ago, basically we were told that we would need to eliminate all prerequisites for our courses, other than if it was a sequential course . . . the reasoning was . . . the idea of free access and that anybody should be able to enroll in a class they thought they were ready to handle and so . . .. We have students without the math skills to do accounting. We're supposed to institute writing across the curriculum, but we have a double problem there because not only do many kids come out of high school and come here with very poor writing skills, we have a very large ESL population . . .. While their technical skills may be good in terms of basic accounting procedures, they are very deficient in their ability to write explanations. (Academic Faculty, WSCC)
In addition, the self-placement policy was somewhat inefficient because students realized that they were in the wrong class and had to switch, losing instructional time:
You get that shakedown at the beginning of a semester, and so I've had students at the beginning of the semester, [who] have . . . said, I think this class is too easy for me, I think I should move up to the higher level class. (Academic Faculty, WSCC)
And one of the things we come up against is students who are unprepared, and we have to spend the first month of the semester sort of filtering them out, and answering a lot of questions which are maybe not germane to what we're trying to teach. They're lower level questions. And, you know, once they get out, the pace picks up a little bit in the classes. (Developmental Education Faculty, WSCC)
It can be seen that the institutions varied both between and within states regarding the selection of test instruments. If there was state policy, the institutions followed it. Given a choice between the state test and other alternatives, the Texas institutions opted for a combination of measures. Six of the 15 institutions included a locally developed informal measure in the mix, and 10 included a commercially available measure in their menu of assessment options.
In several instances, informal measures were used either as alternatives to the norm-referenced measures or to check the accuracy of such measures. Florida state policy allowed grades in certain courses to substitute for state test scores. In the developmental writing classes at WUCC and SWUCC, instructors administered a locally developed writing sample to confirm or change students' placements. Another reason for the administration of the local measure was that some students cheated on the state test.
Administration of the local measure, along with the policy that students could retake the state test at any time, could result in a precipitous drop in enrollment in specific classes:
The goal is to find out . . . is that an accurate score and do we really have good placement? . . . . We give another form of what we consider the exit level criteria, and if they score above the exit level criteria the first day of school, then obviously those people could be better served some place else. Now we don't just take that and jump that day. We do a couple of other things just to make sure that score was good. And then at that point we advise them to do whatever. And I moved out of my classes 16 and I ended up with nine. Then out of the class of nine, two passed [the state test] during the course of the semester . . . . So those two dropped, so I was to seven. One student was in an automobile accident and had to drop out of school completely because it impaired him. And so then I was at six. And then another student, I am not quite sure why, but she disappeared on me, even though I called her and said, Get yourself in here. Then she didn't and so I had to drop her. So I am at five. (Developmental Education Faculty, SWUCC)
At MWUCC and SMCC, students who were not satisfied with their remedial placements could have them changed either through discussion with an advisor or through taking a locally developed test. Standardized math scores were checked at SWSCC, where students who passed the state test but who had not had any math instruction for 3 years were required to pass an institutional math test. In all these cases, informal measures were used to override scores on objective tests.
In summary, there was wide variation in assessment instruments, both in whether single or multiple measures were used and in the use of formal or informal tests. The variety of measures matches Shults's (2000) findings. None of the National Field Study sites was entirely satisfied with its assessment approach, and overall, this topic was an ongoing concern. Informal tests played an important role in the evaluation of skill, either as a sole means of assessment or as a way to check the accuracy of the standardized measures. Shults found that just under one quarter of his community college respondents used measures of this type. This practice suggests a lack of trust in the accuracy of the state and commercially available measures in evaluating students' academic ability, a current theme in K-12 high-stakes testing.
Setting cut scores
The setting of cut scores is key to placement and has direct effects on the size of remedial enrollments. Among the three states that had policy on test instruments, two (New York, municipal; and Texas) determined the cut scores. Texas's cut score applied only to the state test. Florida recommended rather than mandated the cut score on its test. The other states had no policy relating to cut scores, which followed from the absence of regulation of test selection.
Thus, most of the National Field Study institutions set their own cut scores. The exceptions were in line with state policy: NEUCC implemented the scores set by the New York municipal system, and the two Texas institutions, SWUCC and SWSCC, used the cut scores required by the state for the statewide skills test. However, the two Texas institutions also used commercially available tests for which they set their own cut points. The two Florida colleges, SUCC and SMCC, opted to set their own cut scores rather than follow the state recommendations.
Some of the colleges grappled with the problem of whether the cut points selected were resulting in satisfactory placement decisions. This was a direct expression of the ambiguity in the concept of college-level work noted above. At MWUCC, an interviewee suggested that the cutoff was not high enough in the math area. Some students who were permitted to enter calculus courses in fact found them too difficult because, although they passed the placement test, their high school preparation was inadequate:
We have kids that come from . . . high schools where they have good math programs, or other colleges where they have good programs. But if they've forgotten . . . the level is not the right level. Because they think . . . I took algebra in high school, therefore I qualify for calculus. Well, then they sign up for the calculus course here, but we require a higher level of algebra ability than [the high school] and then . . . they wind up not doing very well. (Developmental Education Faculty, MWUCC)
Institutions adjusted cut scores to maintain standards in college-level courses or to make more precise the assignment to remedial levels. At the time of the site visit, faculty at NERCC thought that the cut score for writing was set too high. SUCC raised its cut score for math because too many underprepared students were entering college-credit math courses, and SMCC lowered its cut score for the higher level of remedial math because too many students were testing into the lower level. NESCC changed its cut scores in response to demographic shifts:
They will adjust if there is a serious report from faculty that students are being grossly misplaced. Then it's time to go back and look. And that happens when there's a shift in population. The more we get into a change in population, the more we have to get into looking at that placement test and readjusting the cut ranges. (Administrator, NESCC)
Strong effects of a change in cut scores were seen at SWUCC, where there was a drastic reduction in the number of sections of developmental writing (from 21 to 6 sections) after the state lowered the cutoff.
To summarize, most of the states in the sample left the setting of cut points to the institutions. This finding is consistent with Shults's (2000) report that 77% of a national sample of community colleges set their own cut points. In the current sample, when given a choice, the colleges set their own cut scores rather than follow state recommendations. In addition, there was a certain amount of experimentation in finding the specific cut points that reflected the needs of the individual institutions.
Remedial placement: mandatory or voluntary?
The majority of community colleges mandated placement in remedial classes of low-scoring students, but, as anticipated from Jenkins and Boswell's (2002) study, placement was a controversial issue. Tension between access and standards goals seemed to be played out most noticeably in remedial placement policy. Mandatory placement and, as some have suggested, requiring completion of remedial requirements prior to college-level study (Roueche & Roueche, 1999), protects the standards of degree-credit courses. However, remedial placement appears to result in student dropout (Boylan & Saxon, 2001; Hoyt, 1999), which threatens the access mission. Another issue concerns the reputation of community colleges. Given the extent of the need for basic reading, writing, and math skills, if all students who needed remediation were actually required to enroll in developmental education classes, the community college could acquire the reputation of a remedial institution. This is not a pleasant prospect to the large number of community college faculty who hold advanced graduate degrees and do not see themselves as remedial instructors.
At both the state and institutional levels, policy for skills assessment and remedial placement were separate, so requiring assessment did not necessarily entail mandatory placement in remedial courses. In practical terms, this meant that students who earned low scores on the skills assessment measures were not necessarily required to attend developmental education courses. As reported above, skills assessment was mandatory in five of the six National Field Study states and at all study sites. However, remedial placement was required in only four statesTexas, Illinois, Florida, and New York municipalcovering eight of the 15 National Field Study sites. As with assessment policy, Illinois softened its state mandate by leaving placement procedures entirely to the colleges.
Institutional placement policy matched state policy, with a few exceptions. The institutions in the states that mandated placement did in fact require placement. In addition, four institutions in states that did not mandate placement opted to do so: NWRCC, WUCC, WRCC, and NESCC, located in Washington, California, and New York (state system). Only three of the seven colleges that were located in states that did not mandate placement also did not require it (NWSCC, WSCC, and NERCC). Thus, regardless of whether they were directed to do so by state policy, 12 of the 15 National Field Study sites mandated placement.
As with the assessment policies described above, there was wide variation in placement policy across the sites that mandated placement, and in some places, the policy was rather complicated. Some of the colleges that mandated placement had found a number of ingenious ways to soften the policy, in effect releasing large numbers of academically underprepared students from the requirement to attend developmental education classes. For example, NESCC in New York (state system) mandated remediation but released students from the obligation if they signed a waiver. Advisors could urge students to enroll in developmental education but did not wish to apply too much pressure.
What I try to tell the students is that if it's recommended that you need remedial math or something to take it. It's a lot better catching those deficiencies early on in your college career as opposed to waiting, to all of a sudden say, Hey, I think I needed some help writing, or something like that, so I try to encourage them. But . . . students just don't want to hear that. They want to hear that they need to take [a credit-bearing English course] . . . I know the advisors really push it and if the student chooses not to take it they usually make them sign something. So they have it in the file if . . . [the counselor] advises the student to take XYZ course, student declined or whatever and just have them sign something. Because . . . we really don't push it, I mean they try to push it but if the student doesn't want to I don't think they push. I think they just let them sign off and take what they feel they wanted to take. (Administrator, NESCC)
All the sites followed conventional practice in providing developmental education separately for the three skill areas of reading, writing, and math. However, 3 of the 12 sites where placement was mandatory placed a maximum on the number of areas for which remediation was required. Again at NESCC, remediation in only one area was required; the area was of the student's own choosing and did not necessarily have to be the weakest area. At SWSCC in Texas, students who tested low in reading or math were required to receive remediation in only one of these areas, with the stipulation that if math was the area selected, the student would have to complete the whole math sequence. At SUCC in Florida, if students tested low in two areas, they were only required to enroll in one, and if they tested low in three areas, they were only required to enroll in two. The reason for not requiring remediation of all weak areas may have been that the students disliked having to take these classes: We have given them every chance to come back and get into remediation, and they refuse . . .. Some students are angry about being in the classes. They don't want to be there (Developmental Education Faculty, SWUCC).
The state of Florida mandated remediation, but at the time of the site visit, SUCC could not meet the demand for remedial sections and was issuing waivers to low-scoring students, enabling them to substitute college-level courses for remedial classes. The difficulty came from not being able to project enrollments in enough time to plan for the number of sections needed. The issuing of waivers was problematic because students who took the substituted courses often had difficulty reading the textbooks.
So then we get waivers. And in the English area, that was particularly detrimental because mostly it meant that people couldn't read the textbooks of the courses they were in because their reading level wasn't up . . . a waiver basically said, Well, you should be in [developmental education] but since we don't have a [remedial] class to offer you, go ahead and take US Government, or go ahead and take psychology. And then very often it's a recipe for failure. So we're working on precision scheduling. (Developmental Education Faculty, SUCC)
Institutions that did not mandate remediation relied on students' decisions to enroll. It was not uncommon for students in subject matter courses to have difficulty reading a course text. An instructor who observed difficulties in academic skills may urge the student to consider remediation.
I will meet with that student and . . . give that student some very serious advice and other options. . . . You can take the F from me. You can withdraw now without any stigma. You can stick it out and take an F, because I am telling you right now that is the way it's going to end up because you can't even read the book. Or I can put you right now, today, at this time in a developmental course that going to help you to be successful. Or maybe there's some GED stuff that you need to do. Or maybe you are just in the wrong place at the wrong time. (Academic Faculty, NESCC)
In summary, the major National Field Study findings concerning remedial placement are, first, that the institutions tended to require placement even in the absence of a state mandate, and second, although they required placement, many institutions seemed to be finding ways to limit the number of students who were actually placed.
Remedial advance and exit
There was no state policy within the National Field Study states concerning completion of remedial levels or exit from remediation. The institutions used three kinds of information to make these determinations: test scores, course grades, and instructors' judgment. Test scores were either from the standardized or informal measures described above. As part of the data collection for this study, information regarding remedial advance and exit was gathered for 10 of the 15 sites. At four of themSWUCC, SWSCC, SMCC, and NEUCCstudents were required to pass the assessment test to advance levels or leave remediation. However, both of the Texas colleges, SWUCC and SWSCC, tried to help students exit writing remediation as quickly as possible by exempting from the exit test students who either earned a B in the highest level remedial course or who passed the highest level remedial course and earned a B in a college composition course. The term policy confusion used earlier may also apply to the approaches described in the following two quotations:
It's a very complicated system, but what they have to do is if they exit remediation, the next semester they are required by law to retake [the state test], they don't have to pass it, just take it, then that same semester they can take the take the B or better which is [freshman composition]. But if the student enrolls in a paired course, which is [higher level developmental writing] and [freshman composition], and passes [freshman composition] with a B and leaves [the writing skills course], then they never have to take [the state test] again. But if they just exit remediation, they are still required to take [the state test], and I guess because the state wants to track the results. I am not sure of the logic behind it, but I am sure there is some. (Academic Faculty, SWUCC)
You have to exit our highest level. Exit [the remedial] course with a C or go take the state test and pass that. There are two ways out. It's either the exit course with a C, the highest level course in that area that you failed, or pass the state test . . . if they score a C or better on [first-level reading] they can go to [second-level reading]. Once they make a C or better in [second-level reading], they have achieved our exit level. But they still have to retake the [state test] and either pass it or if they don't they can take a government course and make a B or better and then that's an out. So there are some alternative outs other than passing the [state test]. (Developmental Education Faculty, SWUCC)
It appears from the above description that the subject matter course in government served as a test run for the application of college-level literacy skills. In this case, a grade in a college-credit course was used to determine readiness to exit remediation.
At SUCC and NESCC, instructors used their own judgment to decide whether a student was ready to move up or leave remediation. WUCC and NWSCC used grades in developmental education courses. Students who were at risk for failing sometimes were able to pass with a less demanding instructor: I think if the student ended up taking a summer session with an easy teacher they'd pass them on with a C. And that's usually how they end up . . . they know how to work the system (Developmental Education Faculty, WUCC).
At other institutions, mixed methods were used that may or may not have incorporated the same measures used for the initial skills assessment described previously. MWUCC students exited reading and math remediation based on a score on a commercially available test and from writing based on a teacher-designed writing sample. A combination of test scores and course grades was used at SWSCC, in which students either had to pass the state test or earn a grade of B or better in certain designated courses. NWSCC used a course grade for math and a writing sample for advancement in or exit from developmental writing. Each instructor applied his or her own guidelines for evaluating the writing sample. There had been talk of creating a common scoring rubric, but faculty were not able to agree on what skills students needed to enter college English, mirroring the ambiguity in defining college-level work alluded to in the introduction to this article.
Adding to the variability of institutional procedures, the colleges offered opportunities for appeal and override of placement decisions. At NWSCC (in Washington), where a commercially available test was used for assessment, placement was voluntary, students exited remediation based on a course grade, and students could appeal exit from developmental math by taking an institutional exam. At the same institution, placement in any area could be changed based on instructors' observations. At WUCC (in California), where placement was mandatory, students were permitted to appeal a placement by retaking the assessment test. At both WUCC and NWSCC, an advisor could override an assessment score if a student stated that he or she had taken a remedial course at another institution.
An academic instructor at WUCC suggested that the teachers and counselors may have had different ideas about what students need. From an instructor's point of view, a student could need remediation, but the counselor, acting as the student's advocate, could try to help the student avoid developmental education.
Student services are concerned with students' scholarships running out, and their whole life picture . . . . It's very important to stay in real close contact with them [the counselors]. . . . Because if you don't, they start thinking your program is a waste of time. And they start figuring out ways to counsel students out of it. Or figuring out ways to, in the computer, override prerequisites . . . they may have a new student who says, I took this course at another college. The counselors have to take their word for it and then they would override their placement scores . . . and at the same time, we need to hear what the counselors' concerns are. So that communication is very important. (Academic Faculty, WUCC)
At SMCC (Florida) and SWSCC (Texas), students were allowed to retake the skills tests any number of times until they passed. This practice opened the possibility of practice effects whereby low-scoring test takers could raise their scores on a second administration (Perin, 2002). Further, permission to retake the test was at faculty discretion, which meant that it was subject to bias. However, one instructor described some criteria that could preserve objectivity:
We have a little flexibility. I mean, you have to use your own judgment. Somebody that's been making forties the entire term, and they fail that exam with a forty, I'm not going to give them a retake. Somebody that's been making eighties the entire time, you know they've been coming to class the entire time, but they are very nervous every time. I would give them a retake. Because I would feel like it was test anxiety and it was the pressure. So it's really a judgment call. Somebody that's the C borderline kid, I would probably give him the benefit of the doubt. Say okay here let's try it. We'll give you one more chance. If you don't make it this is it. And some of them do, some of them don't. They're better off if they don't really just to do it over. (Developmental Education Faculty, SMCC)
In summary, the study states did not set policy for the advance and exit from remediation. In the colleges, as with the assessment measures, a wide variety of practices were followed to determine remedial progress and exit. Some procedures were straightforward, and others were a complex mix of test scores and course grades. In a number of cases, course grades, informal institutional tests, and instructors' judgment were used exclusively or as a possible substitute for a formal standardized test. Thus, as with the assessment measures above, there was strong presence of subjective measures in determining student readiness for the college curriculum.
Timing of remediation
As mentioned previously, although some feel that remediation should be completed before college work begins (Roueche & Roueche, 1999) remedial placement may cause attrition (Hoyt, 1999).6 To investigate the feasibility of requiring remediation as preparation for college study in the National Field Study sites, we examined the timing of remediation, defined as the point in the student's program at which remedial courses are taken. As a backdrop, it is noted that Shults (2000) found that 99% of community colleges allowed their students to take remedial courses and at least some college-credit courses simultaneously.
Texas, the only state in the sample that had policy on this issue, came fairly close to the Roueches' (1999) recommendation in requiring students to complete remediation before enrolling in upper-level college courses. The state also mandated continual remediation for any student who needed it; students who enrolled in college-credit courses at any level were required to attend developmental education concurrently if assessment scores were below the cut point. By the same token, students who dropped out of remedial courses were required to withdraw from all college-credit courses, in effect meaning that they had to leave the college.
Both Texas sites followed the state policy, but the presence or absence of a state mandate seemed to be relatively unimportant in accounting for institutional policy on timing among the National Field Study sites overall. In fact, almost all the sites had a way of influencing the timing of remediation, usually through imposing course prerequisites and stipulating when the skills assessment needed to take place. SWUCC, in Texas, went beyond the state mandate in applying remedial prerequisites to the large majority of courses, in effect barring students from them if they had not completed their remedial courses. This practice was also seen at most of the other institutions, although brakes on this control are described next. The imposition of prerequisites limited the selection of college-credit courses from which remedial students can choose. For example, at MWUCC, only a few courses in art and social sciences did not impose remedial prerequisites. Although the quality of courses was reported to be improving at the college, many students were excluded.
A result of the [mandatory skills] assessment is the slow raising of standards. There is a whole pool of students who are not going to meet these requirements. So what happens to those students? They are going to be caught in this vast wasteland of . . . expansion of . . . a real education. (Academic Faculty, MWUCC)
In a softening of the timing policy, MWUCC allowed instructors to override the prerequisites. Although this was intended to help students gain access to courses, sometimes sign-off was provided inappropriately.
We try to make [remedial placement] mandatory, but sometimes people try to go around us and try to get somebody else to [sign]. And if they get some other staff member from another department to [sign] them in, then we're kind of stuck with them . . . we can't hire a policeman to pick them up and carry them out of the room. I mean, once they're registered for the course, they're pretty much there . . . it happens every once in a while. (Developmental Education Faculty, MWUCC)
At times, students were able to bypass prerequisites because the registration system was technically unable to block them from enrolling. At NWSCC and WRCC, a student could pass all the classes for which college English was a prerequisite and then later fail the English placement exam despite having passed other classes that might have required mastery of English skills. At NWSCC, some of the advisors steered students with poor writing skills away from college courses with high writing demands, such as economics.
An exception to the use of prerequisites to enforce remediation was found at MWSCC, where writing and math prerequisites did not apply to many vocational degrees because they required technical writing and business math, which were less demanding options. In addition, although the trend was to impose prerequisites, some institutions had been removing them in order to accelerate the move through academic programs of academically underprepared students. This practice was occurring at WSCC and SMCC, although at the latter institution, the prerequisites had been removed only from certain non-transfer-level courses.
Institutions could also influence the timing of remediation through manipulating the point in the program at which the student's skills must be assessed. Although in principle, all the institutions mandated skills assessment for all incoming students, five allowed students to delay the assessment considerably. SUCC and NESCC allowed students to wait until the 12-credit point to undergo assessment. At SMCC, low-scoring students were required to enroll in developmental reading only if they declared a major, which many students did only after completing almost all their course work. Students at this institution could remain in an "undecided" status, taking many non-transfer-level courses that, as indicated above, did not have prerequisites, and then taking any necessary reading remediation at the end of the degree when they declared a major.
Four of the 15 colleges did not require assessment immediately upon college entry. MWSCC allowed the delay of reading assessment, and therefore remediation, until the 8-credit mark, and students at SUCC and NESCC could wait to take the reading test until they had taken 12 credits. Difficulties with basic academic skills could emerge in the courses taken prior to the assessment:
When I was in the classroom teaching psychology, my biggest frustration was I had students who couldn't read. They took psych because they could take X number of courses before they had to take placement testing, and they thought this would be a fun thing to take and didn't think they would have to read. But an abnormal psych textbook is a pretty complicated type of textbook. (Administrator, NESCC)
Similarly, at NERCC, where remediation was not mandated and there were few course prerequisites, students in the college-credit classes could have difficulty with reading comprehension.
We get a lot of students who come into [a college-level] course before they have done or are currently doing remedial reading. And that's a big problem because they cannot read the textbook. So they cannot understand what in the heck we're talking about, and it really puts them at a disadvantage. And they were put in there just to kind of fill the slots. And we have been saying this to [a college administrator], we've got to have remediation first.(Academic Faculty, NERCC)
Further, MWSCC and NWSCC only required writing and math assessment when students wished to enroll in college-level English or math courses. In fact, many students delayed taking these courses, along with any remedial prerequisites, until the end of their programs.
You can take all history courses and avoid taking any reading or math placement tests. . . . there's a number of courses, you can really zigzag through this institution and take language . . . take humanities courses, take virtually every occupational course. And you can zigzag and . . . you can put off English, math until the very end. (Academic Faculty, MWSCC)
In summary, although only one state had a policy regulating the timing of remediation, many of the institutions tried to induce students to take remedial courses early in the program through the use of course prerequisites. However, some colleges were removing prerequisites in order to provide a greater selection of courses for low-skilled students. If there were many remedial prerequisites, the prerequisite policy was softened in numerous ways. It was unintentionally softened when instructors could override the prerequisites or when students learned to outwit the registration system. Intentional softening was accomplished through permitting students to delay assessment or remediation. The result of these brakes on the control of the timing of remediation in some cases meant that developmental education became a graduation requirement rather than an entry condition.
State control of remedial practicelaissez-faire to micromanagement
Policy implementation studies have found that state policies tend to focus on large issues, leaving the details to local agencies (Anderson, 2003). Examination of state variation in the regulation of remediation suggested that the National Field Study states could be situated on a spectrum from laissez-faire to micromanagement. Because the extent of state regulation can only be considered in the context of other variables, such as approaches to leadership and policy unification across education sectors (Richardson et al., 1999), the amount of state control of remediation cannot be taken as an indicator of effectiveness. However, understanding the relation between state and local policy may shed light on issues concerning postsecondary access and standards.
Among the National Field Study states, Washington was characterized as laissez-faire because it mandated neither assessment nor placement and issued no regulations concerning tests to be used, cut points, or limits on remedial course taking. At the other end of the spectrum, Texas and Florida can be called remedial micromanagers. They mandated both assessment and placement, prescribed the assessment measure, and placed limits on the number of remedial courses that students can take using financial aid. New York (municipal) tended toward the micromanagement end in mandating both assessment and placement and in prescribing the test. Illinois was placed in the middle of the spectrum: Although it mandated both assessment and placement, it left specific procedures to the colleges and did not aim to enforce its policy. The style of this state was unique in the sample in its "steering" of institutional policy (Richardson et al., 1999). California and New York (state system) both tended toward the laissez-faire end of the spectrum in that they mandated only assessment and left the colleges to determine their own placement policies. In this scheme, California was placed slightly toward the micromanagement end because of its policy on developmental math: Instructors were required to hold at least a master's degree, and math courses were not permitted to take a workplace perspective.
There is some overlap between the current highly specific remediation spectrum and Richardson et al.'s (1999) categorization of states regarding higher education regulation in general. For example, Florida and Texas, micromanagers in the current study, were found to be generally regulation-heavy in the Richardson study. New York was identified by Richardson et al. as a heavy regulator; this was found for remediation in the current study in the municipal system but not the state community college system. Finally, the placement of Illinois in the middle of the remediation spectrum is consistent with Richardson et al.'s report that the state coordinates rather than regulates local higher education policy. The current findings are also consistent with K-12 accountability research based on scores on high-stakes math tests (Carnoy & Loeb, 2002) that found Florida, Texas, and New York, remedial micromanagers in the current study, to have strong K-12 accountability systems. California, showing moderate control in the current study, had slightly lower levels of K-12 accountability compared with Florida, Texas (at the time of the study), and New York. Further, Illinois, moderate in remedial regulation, fell in the center of Carnoy and Loeb's accountability metric. Finally, Washington, laissez-faire in the current study, fell at the lowest end of the K-12 accountability scale. The overlaps with the findings of both Richardson et al. and Carnoy and Loeb provide some measure of validation for the currently proposed remediation spectrum.
In trying to understand why the states might vary on the level of control of remediation, we can examine the level of academic preparedness of students in the states, as reflected in high school graduation rates and college remedial enrollments. Data reported by the National Center for Education Statistics cited by the United Health Foundation (2004) suggest that high school students are academically less prepared in the micro-management states as compared with the laissez-faire states.7 The two micromanagement states, Florida and Texas, showed rates of high school graduation within 4 years of 56% and 64%, respectively. Washington, extremely laissez-faire, had a higher 4-year high school graduation rate, 69%. California, relatively laissez-faire, also had a fairly high graduation rate, 70%. Illinois, which was somewhat laissez-faire in not enforcing its remedial mandates, also showed a relatively high graduation rate, 72%.
Moving to remedial enrollments, the type of information and the years for which it was reported varied across the six National Field Study states, making precise comparisons difficult. However, to provide an estimate of developmental education enrollment, information for the most recent year reported is presented in Table 3 (states are grouped according to their levels of control of remediation). It can be seen that the developmental education enrollments are not systematically related to state policy control. Washington, the most laissez-faire of the six states, showed a level of remedial enrollment somewhat similar to that of the micromanagement states. However, the two moderate-control states for which information was available showed a lower level of remedial enrollment than the micromanagement states. Thus, the high school graduation and college remedial enrollment rates provide some direction in explaining states' interest in controlling institutional remedial practice. Future research on the relationship between state and local policy via academic preparedness might take into account Carnoy and Loeb's (2002) report that state accountability requirements correlate with the proportion of the minority population and the size of the population overall.
Consistency of institutional practice with the policy
The National Field Study institutions followed the broad outlines of state policy but adapted policy to shape, and in some cases soften, the state mandates. The main thrust of these mandates was that if students need remediation, they should receive it. This policy seemed designed to maintain educational standards, but an unintended consequence (Pressman & Wildavsky, 1973) could be to slow the pace of students' access to and completion of degree programs (Perin & Charron, in press). McLaughlin (1987) proposed that policies are most likely to be implemented in the ways intended when they meet local needs; otherwise, local agencies will adapt them. In the current sample, numerous adaptations were found that illustrate the conflict between colleges wanting to maintain standards but also protect educational access.
Effects of specific remedial practices on colleges and students
An examination of the institutional practices summarized in the right column in Table 2 reveals several different functions: The practices can influence the number of remedial enrollments (increase or decrease), make assessment more precise, and promote student retention in the college. The practices were categorized into these functions:8
PRACTICES THAT DECREASE THE NUMBER OF STUDENTS IN REMEDIATION (D)
1. Placement or exit decisions based on subjective measures rather than objective tests: NWSCC, NESCC, WUCC, SWUCC, SWSCC, SUCC
2. Override of assessment or placement requirements: WUCC, MWSCC
3. Override, nonenforcement, or removal of remedial prerequisites: NWSCC, WSCC, WRCC, MWUCC, MWSCC, SMCC
4. Placement not required in all weak areas: NESCC, SWSCC, SUCC
5. Career students exempted from assessment: SMCC
6. Cut score lowered: SMCC
7. Assessment test readministered: SWSCC, SMCC
8. Students pass remedial courses with less demanding instructors: WUCC
9. Substitution of college-level courses for remedial courses (waivers): SUCC
10. Institution limits remedial course repetitions in absence of state mandate: WUCC, MWUCC
11. Low-skilled students steered away from high-skills college courses: NWSCC
PRACTICES THAT INCREASE THE NUMBER OF STUDENTS IN REMEDIATION (I)
1. Institution requires assessment in absence of state mandate: NWSCC, NWRCC
2. Cut score raised: SUCC
3. Students must pass standardized test to advance or exit: SWUCC, SMCC, NEUCC, MWUCC
4. Remedial prerequisites for college courses: WUCC, SWUCC, SMCC, NEUCC
PRACTICES THAT INCREASE THE PRECISION OF ASSESSMENT OR PLACEMENT (P)
1. Separate tests for native-English and ESL students: WUCC
2. Use of institutional assessment instruments to confirm or change placements previously established by standardized tests: WUCC, SWUCC, SWSCC
3. Cut scores adjusted as student demographics change: NESCC
4. Placement changed based on student appeal, readministration of test, use of institutional measure, advisor, or instructor judgment: NWSCC, WUCC, MWUCC, SMCC
PRACTICES THAT PROMOTE RETENTION OF STUDENTS (R)
1. Assessment not required until student well into college degree program: NWSCC, MWSCC, SUCC, SMCC, NESCC
2. Transcript does not show grade for failed remedial courses: SWUCC
3. College credit awarded for remedial courses: NERCC
A total of 22 local institutional practices were identified pertaining to the implementation of assessment and placement policy. Eleven of these practices (50%) appeared to decrease the number of students in remediation, along with four practices (18%) that increase remedial enrollments, four practices (18%) that increase the precision of the assessment or placement procedures, and three practices (14%) that function to prevent attrition. Additional research is needed to confirm the possibility, raised by the current findings, that community colleges are implementing assessment and placement practices that are contradictory to state mandates, and even mandates instituted by the colleges themselves, in an attempt to limit the amount of remediation provided.
As in other areas of state policy (Anderson, 2003), the National Field Study states formulated the skeleton of remedial policy and left the details to the colleges, but the states varied regarding how much remedial policy they left to the local community colleges, ranging from the laissez-faire state of Washington to the highly controlling micromanagement state of Florida. Ideally, there should be mutual accommodation between government policy and local interests (McLaughlin, 1987). In the current study, remedial assessment and placement policies seemed to reflect unsuccessful attempts to protect standards and access goals at the same time. Overall, where there was state policy, the colleges followed it in broad outline but tended to soften it, including breaking their own mandates.
Several years ago, two headlines appearing within months of each other in the nation's leading postsecondary education trade paper seemed to portray the tussle between access and standards goals within community colleges. Many 2-year colleges impose tougher academic standards (Ha-worth, 1999, p. A33), and As need for remediation grows, 2-year colleges must fill the gap, officials agree (Hebel, 1999). To state the contradiction in its most extreme form, standards goals are achieved when all students who need it receive remediation as preparation for college-level study, and access goals are met when all students are allowed to participate in the college curriculum irrespective of skills. Expressed in this way, it is impossible to protect both goals simultaneously; one must be favored over the other. In surveying local practices, strategies that force students into remediation, and thus result in an increase in remedial enrollments, can be characterized as promoting standards goals. In contrast, practices that permit the bypass of assessment and placement, resulting in a decrease in remedial enrollments, may be interpreted as promoting access goals. The current findings of 11 practices that seemed to decrease remedial enrollments, compared with only four that increased them, suggest that community colleges were sacrificing standards goals in the interest of protecting access.
An issue that complicates an understanding of the conflict between standards and access is the lack of conclusive information concerning the effectiveness of remediation. It has been claimed that remediation is effective (Boylan et al., 1997; Meristotis & Phipps, 2000) and that it is not (Johnson, 1996; Perin & Charron, in press). Of course, if developmental education outcomes are poor, requiring remediation will not protect standards at all. Unfortunately, there has never been a large-scale, methodologically sound evaluation of community college remediation (Weissman, Bulakowski, & Jumisko, 1997). A controlled evaluation would be an important precursor to the kind of research needed to understand more fully the effects of the softening of assessment and placement mandates found in the current qualitative case study on state, institutional, and student goals. A comprehensive research effort would provide direction for the policy reforms necessary for the simultaneous protection of standards and access goals in community colleges.
The research reported in this article was supported by a grant from the Alfred P. Sloan Foundation to the Community College Research Center, Teachers College, Columbia University. Special thanks to the 15 community colleges that served as case study sites; to Vanessa Smith-Morest and Thomas Bailey for their support and guidance of the project; to Kerry Charron for her assistance in data collection, coding and interpretation; and to Heather Nanni for her help with coding.
Anderson, J. E. (2003). Public policymaking (5th ed.). Boston: Houghton Mifflin.
Bailey, T. R., & Morest, V. S. (Eds.). (in press). Missions accomplished? Multiple perspectives on access and equity at the community college. Baltimore: Johns Hopkins University Press.
Boylan, H., Bliss, L., & Bonham, B. (1997). Program components and their relationship to student performance. Journal of Developmental Education, 20(3), 2-9.
Boylan, H., & Saxon, D. P. (2001, September). An evaluation of developmental education in Texas public colleges and universities (Technical Report). Austin: Texas Higher Education Board. Retrieved March 5, 2003, from http://www.thecb.state.tx.us/reports.
Carnoy, M., & Loeb, S. (2002). Does external accountability affect student outcomes? A cross-state analysis. Educational Evaluation and Policy Analysis, 24, 305-332.
Casazza, M. E. (1999). Who are we and where did we come from? Journal of Developmental Education, 23(1), 6-8, 10, 12.
Dougherty, K. (1994). The contradictory college: The conflicting origins, impacts, and futures of the community college. Albany: State University of New York Press.
Grubb, W N., Worthen, H., Byrd, B., Webb, E., Badway, N., Case, C., et al. (1999). Honored but invisible: An inside look at teaching in community colleges. New York: Routledge.
Gumport, P., & Bastedo, M. (2000). The remedial education controversy at CUNY: Perspectives on system design and endemic conflict. Paper presented to the American Educational Research Association, New Orleans, LA.
Haworth, K. (1999, January 22). Many 2-year colleges impose tougher academic standards. The Chronicle of Higher Education, 45, p. A33.
Hebel, S. (1999, April 9). As need for remediation grows, 2-year colleges must fill the gap, officials agree. The Chronicle of Higher Education, 45. Retrieved April 9, 1999, from http://chronicle.com/daily/99/04/99040902n.htm
Howard, J., & Obetz, W. S. (1996). Using the NALS to characterize the literacy of community college graduates. Journal of Adolescent and Adult Literacy, 39, 462-467.
Hoyt, J. E. (1999). Remedial education and student attrition. Community College Review, 27, 51-72.
Immerwahr, J. (1999). Taking responsibility: Leaders' expectations of higher education. Washington, DC: National Center for Public Policy and Higher Education.
Integrated Postsecondary Education Data System (IPEDS). (2000). Fall enrollment, 2000 [Data file]. Available from the National Center for Education Statistics Web site, http://nces.ed.gov/ipeds
Jenkins, D., & Boswell, K. (2002). State policies on community college remedial education: Findings from a national survey (Technical Report No. CC-0201). Denver, CO: Education Commission of the States, Center for Community College Policy.
Johnson, L. F. (1996). Developmental performance as a predictor of academic success in entry-level mathematics. Community College Journal of Research and Practice, 20, 333-344.
Lipson, M. Y., & Wixson, K. K. (2003). Assessment and instruction of reading and writing disability (3rd ed.). Boston: Allyn and Bacon.
Mazzeo, C. (2002). Stakes for students: Agenda setting and remedial education. Review of Higher Education, 26, 19-39.
McCabe, R. H. (2000). No one to waste: A report to public decision-makers and community college leaders. Washington, DC: Community College Press.
McGrath, D., & Spear, M. B. (1991). The academic crisis of the community college. Albany: State University of New York Press.
McLaughlin, M. (1987). Learning from experience: Lessons from policy implementation. Educational Evaluation and Policy Analysis, 9, 171-178.
Merisotis, J. P., & Phipps, R. A. (2000). Remediation education in colleges and universities: What's really going on? Review of Higher Education, 24, 67-85.
Merriam, S. B. (1988). Case study research in education: A qualitative approach. San Francisco: Jossey-Bass.
Miles, M., & Huberman, M. (1994). Qualitative data analysis (2nd ed.). Thousand Oaks, CA: Sage.
National Center for Education Statistics (NCES). (1996). Remedial education at higher education institutions in Fall 1995 (Technical Report, NCES 97-584). Washington, DC: Author. (Post-secondary Education Quick Information System [PEQIS])
National Center for Education Statistics (NCES). (2003). Remedial education at degree-granting postsecondary institutions in Fall 2000: Statistical analysis report (Technical Report, NCES 2004-0101). Washington, DC: U.S. Department of Education Institute of Education Science.
Nora, A. (n.d.). Community colleges in the 21st century: Revisiting and reexamining their mission. Washington, DC: American Association of Community Colleges. Retrieved December 14, 1998, from http://www.aacc.nche.edu
Oudenhoven, B. (2002). Remediation at the community college: Pressing issues, uncertain solutions. New Directions for Community Colleges, 117, 35-44.
Patton, M. Q. (1990). Qualitative evaluation and research methods (2n ed.). Newbury Park, CA: Sage.
Perin, D. (2002). The location of developmental education in community colleges: A discussion of the merits of mainstreaming vs. centralization. Community College Review, 30, 27-44.
Perin, D. (2004). Remediation beyond developmental education: The use of learning assistance centers to increase academic preparedness in community colleges. Community College Journal of Research and Practice, 28, 559-582.
Perin, D., & Charron, K. (in press). Lights just click on every day: Academic preparedness and remediation in community colleges. In T Bailey & V. Morest (Eds.), Defending the community college equity agenda. Baltimore: Johns Hopkins University Press.
Phipps, R. (1998). College remediation: What it is, what it costs, what's at stake (Technical Report). Washington, DC: The Institute for Higher Education Policy.
Pressman, J., & Wildavsky, A. (1973). Implementation. Berkeley: University of California Press.
Rhoads, R. A., & Valadez, J. R. (1996). Culture capital and border knowledge: A case study of developmental education. In R. A. Rhoads & J. R. Valadez (Eds.), Democracy, multiculturalism, and the community college (pp. 139-164). New York: Garland.
Richardson, R. C. Jr., Bracco, K. R., Callan, P. M., & Finney, J. E. (1999). Designing state higher education systems for a new century. Phoenix, AZ: The American Council on Education and Oryx Press.
Roueche, J. E., & Roueche, S. D. (1999). High stakes, high performance: Making remedial education work. Washington, DC: Community College Press.
Roueche, J. E., Ely, E. E., & Roueche, S. D. (2001). In pursuit of excellence: The Community College of Denver. Washington, DC: Community College Press.
Salvia, J., & Ysseldyke, J. E. (2004). Assessment in special and inclusive education (9th ed.). Boston: Houghton Mifflin.
Shaw, K. M. (1997). Remedial education as ideological battleground: Emerging remedial education policies in the community college. Education Evaluation and Policy Analysis, 19, 284-296.
Shaw, K. M. (2000). Reframing remediation as a systemic phenomenon: A comparative analysis of remediation in two states. Paper presented to American Educational Research Association, New Orleans, LA.
Shults, C. (2000). Institutional policies and practices in remedial education: A national study of community colleges (Technical Report). Washington, DC: American Association of Community Colleges.
Spann, M. G. Jr. (2000). Remediation: A must for the 21st century learning society (Policy Paper). Denver, CO: Education Commission of the States.
Stake, R. (1995). The art of case study research. Thousand Oaks, CA: Sage.
Swanson, C. B., & Stevenson, D. L. (2002). Standards-based reform in practice: Evidence on state policy and classroom instruction from the NAEP state assessments. Educational Evaluation and Policy Analysis, 24, 1-28.
United Health Foundation (2004). America's health: State health rankings, 2004 edition. Table 21, High School Graduation. Available www.unitedhealthfoundation.org/shr2004/ components/hsgrad.html. Retrieved January 9, 2006.
Weissman, J., Bulakowski, C., & Jumisko, M. K. (1997). Using research to evaluate developmental education programs and policies. In J. Ignash (Ed.), Implementing effective policies for remedial and developmental education (pp. 73-80). San Francisco: Jossey-Bass.
Zeitlin, A. N., & Markus, T C. (1996). Should remediation be mandatory in the community college? Community Review, 14, 27-33.