The first half of this paper concentrates on the need for the institution of evaluation methods which stress the systematic use of both quantitative techniques and qualitative, direct observation methods. In this way, a more comprehensive picture of the events under consideration may be presented, and changes in student learning and behavior may be defected when new educational practices are introduced. It is suggested that the essentially quantitative, dominant evaluation methodologies have not revealed significant differences in the results of new education programs because they do not take into account potential influences on the evaluation process. Five problems which commonly arise when it is suggested that sophisticated direct observation techniques be adopted, are identified and discussed: (1) unfamiliarity with methods, (in response to this, various methods are outlined and further references are provided); (2) the methods' validity and reliability; (3) the time factor involved; (4) observer training; and (5) ethicality of the methods. This section is followed by a bibliography, intended as an introduction for those unfamiliar with various direct and participant observation methodologies. The research reports and reviews included span six decades of study of observation and empirically grounded, theoretical evaluation techniques. (AEF)