NOTE: This publication is out of print and no longer available for purchase.
This publication highlights trends related to the growth of information, increased professional specialization, collisions between education professionals and nonprofessionals, and tensions between educational innovators and certain political and religious groups. It examines issues that relate to information dissemination, and to the different orientations toward knowledge expressed by scientists, educators, and clinicians. A selected ERIC bibliography on educational dissemination is also included.
Trends and Issues in the Dissemination of Child Development and Early Education Knowledge
Table of Contents
- Current Trends in Education
- The Information Age
- Trends in Collision
- The Blame Drain
- Issues in Dissemination
- The Optimum Information Hypothesis
- Optimum Redundancy in Information Dissemination
- Optimum Conceptual Size
- The Vividness Problem
- Orientations to Knowledge
- Two Cultures within the Field
The launch of a new journal to serve the international community of early childhood educator seems an appropriate occasion to discuss trends and issues surrounding the development, dissemination and application of knowledge in the field. Though the ideas presented here are based largely on experience in the United States, it is hoped that they are relevant to the trends and issues facing early childhood specialists in other countries. Observations form my experience working with colleagues outside of the U.S. suggest that if some of these ideas do not apply to them currently, they are very likely to do so in the near future.
The ideas presented in this paper have developed from my 25 years experience as director of the ERIC Clearinghouse on Elementary and Early Childhood Education (ERIC/EECE) at the University of Illinois in Urbana-Champaign. Most readers of this journal are familiar with the ERIC database. Many are frequent contributors to and users of its vast document collection. They may not, however, realize that ERIC/EECE is only one of 16 ERIC Clearinghouses, each of which is responsible for identifying, selecting, abstracting, and indexing the report and journal literature in its own assigned specialty, and many of which overlap the specialty of ERIC/EECE. All 16 clearinghouses add the information they collect and process to the ERIC database. The ERIC system summarizes the report literature in its printed index Resources in Education and makes the full texts available on microfiche in hundreds of libraries and other information centers around the world, and through document delivery services. The ERIC system also lists abstracts of more than a thousand journal articles in each monthly issue of Current Index to Journals in Education, similarly available around the world.
In addition to collecting reports, conference papers and other documents and journal articles related to children from birth to age twelve for parents, teachers, policymakers, and others involved in their education, ERIC/EECE disseminates this information through newsletters, digests, bibliographies, monographs, journal articles and other publications. A large proportion of my own time is allocated to sharing this information with teachers, parents, and many others in the field, through seminars, workshops and formal presentations at conferences and conventions.
With these experiences as background, this paper is presented in two parts. First, some observations are presented on current trends in the environment to which the dissemination of child development and early education is directed. Second, some broader issues related to the dissemination of knowledge are discussed. I will close with a discussion of some implications of the trends and issues.
The Information Age
It is often noted that information and knowledge are expanding exponentially in all fields. So it is with the field of child development and early education. The ERIC database currently includes more than 800,000 summaries of documents and journal articles, of which early childhood education specifically is only a part. Our clearinghouse alone has been responsible for contributing nearly 35,000 abstracts of unpublished articles, reports, papers, and journal articles from 1967 through 1992. Other clearinghouses also contribute thousands of such documents about young children in special populationse.g. bilingual and non-English-speaking children, children with special needs, rural and urban children, and on topical areas, including literacy, reading, mathematics, and so forth. My best guess is that well over 100,000 of the documents in the collection are directly related to our field and our work.
At ERIC/EECE we abstract articles from nearly two dozen monthly and quarterly journals and similar publications. Since 1980, 10 major new journals have been added to our field, and more are expected in the next year or two. In addition, we regularly monitor and peruse more than 200 newsletters, magazines, and other journals from a wide variety of associations, publishers, and organizations whose work is related to early childhood education.
Another strong trend in the information environment is the pace at which specialization increases. New subcategories of specialization in the field emerge virtually daily: computing in the early years, early intervention programs, public policy, sub-disciplines related to language acquisition, multiculturalism, play, play-grounds, art, family issues, parent involvement and education, and so forth. At the current rate of journal development we will soon need a journal about the journals in the field!
Trends in Collision
The trend toward increasing specialization, which produces specialized expertise as the knowledge base expands, seems to collide with strong public and political sentiments against professionals and experts 1 that appear to be widespread in some of our countries. It is manifested in the United States by the frequent inclusion of individuals from the business and corporate world on panels, task forces, commissions, and advisory boards. Frequently they serve as chairpersons, and frequently they are chosen as a way to circumvent the professionals and experts thought to have merely a vested interest in the status quo. This strategy betrays an assumption that the obvious interest that business and corporate leaders have in the outcomes of education can make up for their lack of expertise and experience in the field for which they are supposedly generating better ideas, and that this ignorance puts them at an advantage over professionals and experts.
Among educational innovators there is also some concern about what they perceive as anti-intellectual activities on the part of certain fundamentalist or far-right groups. There is a sense in which the certainty and conviction with which these groups present their views is unsettling to scholars and other intellectuals trained to qualify their claims with statements of probability and caution. In times of rapid change in the midst of complex forces, statements of conviction are often far more appealing than reminders of uncertainties.
Perhaps as a result of some of these trends in collision, many of the school reforms introduced in the U.S. in the last decade seem punitive in tone. Some also appear to be at cross-purposes. Political pressures have led to the encouragement of decentralized decision-making about schooling through such strategies as school-based decision-making, local (especially parental and business) input, and participation in hiring staff and setting curricula. On the other hand, national resources are now allocated to national test development and assessment, and pressure mounts in the U.S. in some quarters for a national curriculum. The collision of these opposing trends can be seen at work in the UK today. In the United States, they are perhaps best exemplified in a recent reform act in the state of Kentucky that adopted a statewide (developmentally appropriate) curriculum in the primary schools, but also mandated site-based decision making!
Michael Katz (1992), discussing such reforms as those implemented in the city schools of Chicago, points out that these reforms
challenge historic processes of professionalization and the ascendance of experts...Although reformers have criticized teachers harshly, they have expected them to transform their practiceby themselves, with guidance from outsiders, or under pressure from lay persons lacking professional knowledge and skill. Reform, in fact, frequently places tremendous burdens on teachers, whose effective workload expands with no compensating increase in authority or pay (p. 58).
But, says M. Katz, while professional expertise seems to be regarded with suspicion among reformers and activists, "they continue to require datanot necessarily to illuminate the issues and problems, but to support their points of view" (p. 58).
A similar collision seems to me to be the frequent allusion especially on the part of politiciansto parents as experts, often phrased as "parents are the best teachers," or "the first teachers," or "parents know their children best." At the same time, enormous resources are allocated at all levels to educate parents about how to prepare their children for school, rather than how to prepare the school to be ready for the children. Opinion makers and politicians seem unable to admit or even to acknowledge the possibility that some parents know their children well, and some do not, or that some parents know some aspects of their children and not others, as is true of their teachers.
The Blame Drain
There is also something of a "blame drain" in the field of public education, at least in the United States. Often parents complain to teachers that they don't want their children tested so frequently; teachers respond that they don't want to do it, but that the school authorities require it; school authorities blame district regulators, who blame the state, who blame the legislators, who say the taxpayers demand it; and, of course, a large proportion of taxpayers are the parents! This cycle of blame seems to me to drain energy and resources away from attending to the complex problems of contemporary education.
As we continue our work as scholars in the field of early childhood education, I ask you to consider six issues related to the dissemination and adoption of the knowledge in our field that we propose to share.
The Optimum Information Hypothesis
West (1981) has suggested that the more information confronting individuals, the more likely they are to be selective in what they attend to, and the more selective they are, the more likely they are to attend to information that is agreeable, compatible, or even identical with what they already hold to be true or right. Information scientists similarly claim that despite innovations in storage and retrieval of information, in the long term the level of accessibility of information remains more or less constant. Thus, the more information available, the less likely we are to attend to what is new or to what departs from what we already know and accept anyway. Although this formulation of the issue refers to the amount of information offered, it also refers to the rate at which information is made available (i.e., the amount of information provided in a given unit of time).
West's (1981) suggestion can be stated as the hypothesis that there is an optimum amount of information below and above which new information in the form of ideas, concepts, knowledge, facts, etc. is not acquired, adopted or used. That is to say, having both too little or too much information available is equally unlikely to advance the acquisition of new ideas. It is likely that individuals vary in what constitutes an optimum quantity or rate of information availability. It would be difficult to ascertain empirically an optimum flow of information. However, this hypothesis may serve to remind us to inhibit our dissemination impulses, and to be wary of simply churning out more pieces of information. At the very least, we might be encouraged to temper our expectations of the impact we can have and to be more modest in our expectations of getting new know-ledge adopted and applied. As we project the development of this new journal, I recommend that contributions that synthesize and integrate disparate sources of information and research findings be given particular preference.
Optimum Redundancy in Information Dissemination
Klapp (1986) points out that the term redundancy, though often used as a criticism of an experience, has positive attributes that are not always appreciated. He suggests that there is both "good" and "bad" redundancy (p. 71). Klapp identifies four functions of good redundancy: continuity, communication, identity, and social resonance. He points out that in the social realm, redundancy provides a kind of reassuring continuity. Klapp's characterization of communication redundancy is quite relevant to this discussion of issues in dissemination of child development knowledge. He points out that
whether by repeating information within a message or series of messages coming from several channels, or cues in the context telling us what we may assume about the source and intention of the message: that they will not be lost or misunderstood, or the memory tape erased (p. 73).
Lieb-Brilhart (1989) suggests that people acquire information best when it is available redundantly through a variety of channels. She points out that when there is knowledge worthy of dissemination "and that we believe to be relevant to educational improvement, planned, coordinated redundancy is key to the implementation" (p. 8). However, there is an optimum or appropriate level of redundancy; too little increases the chances that the message in question will be lost in the larger information environment; too much increases the chance that so much time will be taken to access and absorb the messages that the energy and time available for implementing the actions recommended in the messages may be diffused or lost.
Optimum Conceptual Size
Occasionally, on reviewing reports of research, especially in the developmental and educational psychology literature, I suspect that one attribute that may influence applicability and adoptability of new knowledge is the "size" of the ideas and concepts they address. If, for example, we inspect the tables of contents of such journals as Child Development, Developmental Psychology and the Journal of Educational Psychology, we are very likely to find that the majority of articles deal with relatively small-scale phenomena, or with molecular or even micro-level variables.
On the other hand, many of the concepts and ideas found in the practitioner's literature seem too large in size to make them useful, applicable or adoptable. Ideas like "all children are individuals," "Each child has his or her own learning style," and so forth seem very large. What are the practical implications of such statements? Should the teacher say "good morning" in 20 different ways? Use X number of different story-telling strategies? Of course, all children are individuals! The assertion is so true, that in a sense, it is uninformative: It does not help a practitioner determine to which events, situations, learnings or other phenomena it does or does not apply.
The concepts and ideas in developmental and psychological journals in which we write for each other are relatively small in size. But ideas or concepts of the micro-level size are unlikely to stimulate new practices directly, which is the ultimate goal of most dissemination efforts as represented by the ERIC system and by journals for practitioners. On the other hand, the very large concepts in much practitioner literature seem to serve as doctrine, or ideological reminders, strengthening practitioners' sense of certainty (correctly or incorrectly).
The role of doctrine and ideology in the nature of practice should not be dismissed lightly. It seems reasonable to assume that in any field in which the database is unreliableespecially in terms of its validitya vacuum is created, and this vacuum is filled by ideologies and doctrines (see Katz, 1977). My hypothesis is that the weaker the knowledge base of a profession, discipline, or subject, the more ideology-bound it is. The field of early childhood education is especially susceptible to data weakness for two reasons.
First, the object of inquiry and investigationsthe young childis, by definition, immature. This immaturity means that the organism is unstable. By comparison, mature organisms are relatively stable, if not rigid. But observed changes in young children may signal instrument insufficiency, construct weakness, maturation, growth, or all of the above in unspecifiable proportions.
The second reason for unavoidable data weakness in early childhood education is that the definitive or critical experiments which might settle its central issues empirically cannot be performed. As long as there are any reasons to believe that something is "good" for children, it would be unethical to withhold it from some of them just for the sake of the advancement of science. Similarly, as long as there is any reason to suspect an experience is less than optimal for children, it would be unethical to subject them to it just for the sake of a scientific experiment. Thus the field of early childhood education is marked by the typical problems associated with heavy reliance on clinical cases, clinical data and insights.
Taken together, these constraints on the development of reliable data and reproducible results, plus the pressure to protect the powerless (i.e., children) through commitments of sentiment, set an ideal stage for ideological battles. Thus the specialized knowledge that should constitute the bases for rationalizing professional practice is constantly in dispute, frequently marked by resistance to counter-evidence, and often imbued with ideological and doctrinaire passions.
The Vividness Problem
Another issue in dissemination is what might be called the "after they've seen the movie they'll read the book" phenomenon. It appears that, at least when it comes to entertainment, reading follows exposure to ideas presented through a more vivid medium. Following a film adaptation, there is frequently a rush on books that had been gathering dust on library shelves for years (note the current large displays in book shops of E. M. Forster's Howard's End, and the current high sales of Maya Angelou's work, unprecedented until she appeared at the inauguration of President Clinton). There also exists the process of "novelization," or writing a book after a film or television series has been produced when none existed before! Thus, ideas, images, thoughts, and feelings stirred through the visual media whet appetites for further exposure to the original material.
The vividness problem is not, strictly speaking, a particular problem of the adoptability or applicability of child development knowledge per se. Rather, this issue is related to some of the problems of getting information, ideas, and concepts attended to through one medium in order to get them followed up in another.
If the pattern of reading the book "after they've seen the movie" is generalizable to non-entertainment reading, then we might consider what vivid media are most effective as mechanisms for the dissemination of child development and early childhood education knowledge. It has often been observed that, following formal lectures that include references to articles and books, many members of the audience ask for complete citations and references and show strong interest in obtaining copies of the paper presented, presumably with the intention of following up what has been made vivid through an aural medium. The extent of follow-up is difficult to guess. I know of no empirical studies of this phenomenon. However, if this is a common phenomenon, and the probability of studying research reports increases following a "live" or "lively" presentation, what dissemination practices should we consider?
Inasmuch as we remain largely dependent on the print medium, we might be wise to present new knowledge and insights through a "case method" of reporting. A case would be a vivid description of a standard predicament common to the experience of the practitioners we wish to influence. The new information and its implications for practice can be fleshed out in terms of the case in a way that may help make it vivid and more easily adopted.
The popularity of the expression "an idea whose time has come" is probably well-deserved. Not just in terms of cosmic chronology, but in more specific and concrete terms, child development knowledge is wanted and presumably utilized when a particular predicament occurs when a teacher or parent has reached the end of his or her patience. Making ideas, concepts, and facts available when people are at the beginning of a process appears to be ineffective. At that point, the information is merely taken for granted, or cast aside, until suddenly it is of almost life-or-death importance. Again, this problem suggests that much new information might be better utilized if explicated in the context of standard predicaments, or case study formats, that most teachers confront frequently.
Orientations to Knowledge
Another issue in the adoption of new knowledge in early education concerns the possibility that those who produce it and those expected to adopt and apply it have different ways of apprehending and using such knowledge. If this is so, we will have to take those differences in orientation to knowledge into account.
Potential differences in orientation to knowledge can be examined by adopting the framework presented by Freidson in his study of the sociology of knowledge in the medical profession. Freidson classifies subsets of members of the medical profession in terms of two distinct types of "mentalities," or orientations to knowledge and research.
According to Freidson's analysis, persons within the field of medicine can be classified into at least two groups: those with the scientific mentality, embodied in the professors in the medical faculties, who are usually scientists rather than physicians; the other group has what he calls a clinical mentality, reflected in the perspective of the medical practitioner. These contrasting orientations to knowledge are depicted on five interrelated dimensions, as shown in Table 1 below.
Orientations to Knowledge and Research
|Scientific Orientation||Teacher Educators, etc.||Clinical Orientation|
|1. reflective||?||1. active|
|2. conceptual||?||2. pragmatic|
|3. theoretical||?||3. subjective|
|4. skepticism||?||4. faith|
|5. determinacy||?||5. indeterminacy|
The first dimension, reflective versus active, suggests that scientistsdevelopmental psychologists, or other scholars and researchers related to the fieldare well served by the disposition to be reflective and to consider alternative courses of action, explanation, theories, interpretations, and so forth. On the other hand, practitionersteachers, or child care workers, or parent educatorsneed the disposition to act in situations of high ambiguity, even in situations in which no relevant reliable information is available.
The second dimension, a conceptual versus pragmatic orientation, suggests that psychologists and other scholars in the field seek concepts to explain how something works, whereas the practitioner will settle more quickly for what works, even without explanations. For the latter, the premium is on doing something, on responding to the practicalities of the pressing situation at hand.
The third dimension, theoretical versus subjective orientation, similarly suggests that the scholar is looking for theories that help organize observations and build a system of understandings and concepts that hang together coherently and sensibly. According to Freidson, the practitioner is more disposed to be reassured by direct, first-hand experiences, or even by the second-hand reports of other practitioners' first-hand experiences, than by a theory. Practitioners are likely to accept scientists' data if, and only if these data correspond to their own subjectively derived views (i.e., if they know it from their own experience or at "gut level"); they seem likely to abandon their subjectively acquired views when the scientifically developed knowledge gainsays them.
Perhaps this disposition to rely heavily on subjectively derived views and insights is what is alluded to in statements about the importance of "owning" one's own program, innovation, or idea. The importance of ownership is often emphasized in discussions about the limitations of the so-called "trickle-down" theory of innovation and information dissemination. However, it would be a pity if such ownership were required in engineering or medicine for innovations to be applied!
The fourth dimension, skepticism versus faith, refers to such dispositional or orientational differences as the scientists' tendencies to prize doubt, to be concerned about the reliability and generalizability of the results of experiments, and so forth. For practitioners, however, doubt and skepticism may be dysfunctional. For example, strong conviction about the rightness of a course of action may influence its effectiveness: uncertainty in teaching may result in giving children mixed signals, which may in turn cause the children to resist the adult, which may further undermine the adult's confidence and hence the effectiveness of the action taken.
In a recent study comparing the differential effects of three preschool curriculum models on inner city four-year-olds (Marcon, 1992), the data showed that the differential effects were related to the convictions of the teachers. Marcon (1992) concluded that the children in classrooms where teachers strongly believed in the curriculum model they were implementing "did better on standardized measures of development than children whose teachers were torn between opposing models" (p. 527).
This finding suggests that strong beliefs may be related to the clarity with which adults express their values, expectations and ideas to children, and that such clarity affects children's participation in the activities provided. It may be that teachers and parents should strive for some optimum level of faith in the rightness of their actions so that children accept adult direction mostbut not allof the time. Observations of teachers and parents interacting with young children suggests to me that either excessive uncertainty or faith can cause much interaction to be consumed by power and management issues rather than the more important issues related to children's intellectual and social growth.
The term "faith" as used here refers to the practitioners' need to believe in the appropriateness of the action chosen even when no supportive evidence is available. Indeed, sometimes counter-evidence is dismissed on the grounds that it is based on experimental, unrealistic, contrived, un-lifelike or non-naturalistic treatment conditions. (However, when the results of such contrived experimental treatments support strongly held beliefs, concern about methodology is usually set aside!)
The fifth dimension, determinacy versus indeterminacy, refers to the scientist's search for lawfulness, or discovery of the underlying laws and theories upon which he or she hopes to be able to make good predictions. The practitioner, on the other hand, tends to hold that events in the real world are far too complex to make discovery of the operating laws possible.
If we accept Freidson's formulation of the two mentalities within the same field as generalizable to our own2--and many other fieldsthen its implications are worthy of our consideration as we develop a tradition for the new journal.
The two groups within the field of early childhood educationthe scholars and the practitionerscan be thought of as two subcultures, perhaps, each with its own orientations or habits of mind, and to some extent, each with its own language or technical jargon. Table 1 depicts the two orientations in opposing columns. I have suggested elsewhere (Katz, 1987) that those who train practitioners might occupy the column between the scientist and the practitioner, the psychologist and the teacher. What would the positions of this middle group be on the five dimensions? What positions would be desirable ones?
According to the formulation of the problem presented here, a profession contains within it at least two subcultures: Those characterized by scientific and practitioner/clinical mentalities. It seems reasonable to suggest that those who actually train professionalslike teacher educators, advisors and consultantshave to be bicultural and bilingual, and to feel comfortable, accepted and conversant in both cultures and languages.
Perhaps among readers of this journal are many who belong in more than one of the three columns; perhaps some operate in all three at the same time or are sometimes in one and sometimes in another. Those of us responsible for training and supporting the development of practitioners the logical adopters and appliers of child development know-ledgeare not really welcomed either by the pure scientist, whose findings we wish to transmit, or by the pure practitioner, whose practices we wish to influence. Each of these groups is likely to see us as deficient in its own specializations (Katz, et. al. 1982). Shulman (1986) points out that
to conduct research, scholars must necessarily narrow their scope, focus their view, and formulate a question far less complex than the form in which the world presents itself in practice. This holds for any piece of research; there are no exceptions (p. 6).
Given the deep differences in ways of thinking in the scientific and practitioner sub-cultures, there may be an important role to play for those in advisory and training positions in helping each of the two cultures to understand each other.
It seems that, if we are to succeed in getting child development and new early education knowledge applied, we will have to attend to the rate at which practitioners receive information, or at least the extent to which we must compete with all other sources of information to which they are subjected. We will also have to concern ourselves with how to speak with confidence and clarity about ideas large enough to be useful and small enough to help them make decisions about significant aspects of practice.
As scholars in the field, we are also faced with a dilemma concerning the ethics of speaking with caution and appropriate qualifiers about what we do know and what we do not know. When decision-makers ask us for clear, unequivocal conclusions from research about what courses of action to take, we may have to say simply that whatever decisions are taken, all carry with them their own errors, and that the problem is to select which errors are preferred. The "choice of preferred error" strategy requires moral, philosophical and ultimately political decisions rather than technical or theoretical ones.
The mission of improving practice, by which we hope to improve the environments and experiences offered to all our children is complex, but surely worthy of the hard work ahead.
Dunn, Jr., R. M. (1993). Economists of the Macro Sort Don't Do Much Good. Herald Tribune. Thursday, March 25, 1993. p. 4.
Freidson, E. (1972) The Profession of Medicine. New York: Dodds Mead.
Katz, L. G. (1977). Early Childhood Programs and Ideological Disputes. In L. G. Katz. Talks with Teachers. Washington, D. C.: National Association for the Education of Young Children.
Katz, L. G., Raths, J. D., Irving, J., Mohanty, C., Kurachi, A., & Sani, M. (1982). Reputations of teacher educators among members of their role set. Paper presented at the annual conference of the American Educational Research Association, New York. 1982.
Katz, L. G. (1987). The Nature of Professions: Where is Early Childhood Education? In L. G. Katz. Professionalism, Child Development and Dissemination: Three Papers. Urbana, IL: ERIC Clearinghouse on Elementary and Early Childhood Education.
Katz, M. B. (1992). Chicago School Reform as History. Teachers College Record. 94(1). pp. 56 - 72.
Klapp, O. E. (1986). Overload and Boredom. New York: Greenwood Press.
Lieb-Brilhart, Barbara. (1989). What is the appropriate level of redundancy in federally sponsored education dissemination programs? Paper presented at the American Educational Research Association Annual Meeting, San Francisco, CA, 1989.
Marcon, R. A. Differential effects of three preschool models on inner-city 4-year-olds. Early Childhood Research Quarterly. 7(4). pp. 517-530.
Shulman, L. S. (1986). Those who understand: knowledge growth in teaching. Educational Researcher, 15(2), 4 - 21.
West, C. K. (1981). The social and psychological distortion of information. Chicago, IL: Nelson-Hall.
(1) Doubts about the practical value of expertise are not limited to education. Dunn (1993) makes the case that hose countries with the most winners of Nobel Prizes in economics have the worst economies; those with substantial numbers of active economists have the next worst economies, and "There have been no Nobel prizes for economists from Germany or Japan. Nor have there been prizes for the economists from fast-growing Asian "tiger" such as Taiwan, South Korea and Singapore." (p. 4). He further states that macroeconomics cannot be observed directly and therefore must be understood through models, and models are never defeated by facts, but by competing models. He adds "This sometimes leaves governments with little alternative to basing policies on theories that fly in the face of experience, until better theories are developed. As a last resort, they might want to look around and see what is actually happening. But that will never win you a Nobel prize." (p. 4).
(2) The distinctions between the two 'mentalities' -- scientific and clinical -- within early childhood education might be even sharper than those within the field of medicine in as much as the differences in training, education, socioeconomic and cultural backgrounds between the scholars and practitioners in the field might be even greater in early childhood education than in medicine.
This publication was prepared with funding from the U.S. Department of Education. The opinions expressed in this publication do not necessarily reflect the positions or policies of the U.S. Department of Education.