many different levels. One result of these changes is that the average book length has increased two to four times in the past several decades.
In response to the need for quality science textbooks for all students, not just science majors, some authors are returning to descriptive and historical approaches. Generally, books for science literacy courses describe important ideas and discoveries, present a limited number of fundamental concepts, and emphasize the links among different facts and principles. Others (e.g., Trefil and Hazen, 1995) take an interdisciplinary approach, by covering a range of science disciplines in a coherent, connected manner.
Research on the effectiveness of textbooks has focused on two general areas: text structure and layout. The study of text structure has focused on how the reader builds cognitive representations from text. Recent work categorizes the structure of science text as either a proof-first or a principle-first organization (Dee-Lucas and Larkin, 1990). The proof-first organization develops a proof or argument that builds to a conclusion, usually in the form of a fundamental concept, principle, or law. In principle-first organization, a concept or principle is stated explicitly, then the evidence needed to support it is presented. The prevalence of the proof-first structure in contemporary textbooks may be due to the fact that most college science textbooks are written by scientists with little formal training in education. They present science the way it is practiced by experts. However, studies by Dee-Lucas and Larkin (1990) indicate that the principle-first structure is more effective for long-term retention and understanding by novice readers.
Layout and illustrations are important predictors of a text's effectiveness. One of the most effective types of illustration, especially for students with low verbal aptitude, is a simple multicolor line drawing (Dwyer, 1972; Holliday et al., 1977). Although more visually appealing, and more prevalent in the current textbook market, realistic drawings or photographs are less effective at enhancing student learning. The organization of information on a page also affects student learning (Wendt, 1979).
Before selecting a text, it is important to know what books are currently on the market. Colleagues who teach the same or a similar course (in your department or at other institutions) are good sources of ideas and information. Your campus bookstore's textbook manager can provide the name and phone number for textbook sales representatives from many different companies. Science education publications (see Appendix B) carry advertisements from major publishers, and some feature a book review section or annual book buyer's guide. Professional society meetings also provide a chance to talk to publishers and see their new textbooks. Many companies will supply review copies to potential textbook adopters, in return for information about the course in which it might be used.
There are a number of factors to consider when selecting a textbook. To be of greatest value to students, the objectives of a textbook must be consistent with those of the course. Authors often try to meet particular objectives in their books, and these may differ among the choices. Skim the preface to see whether you share the author's approach to the subject.