“…Thus, one of the earliest demands on applied researchers in behavior analysis was for accurate and reliable description and observation of the dependent variable (e.g., Heyns & Lippitt, 1954). The classic literature of observational technology (e.g., Arrington, 1943;Bijou, Peterson, & Ault, 1968;Rosenthal & Rosnow, 1969;Webb, Campbell, Schwartz, & Sechrist, 1966;Weick, 1968) was joined in later years by a new literature, concerned with problems of dependent variable observation such as observer reactivity (e.g., Hanley, 1970;Romanczyk, Kent, Diament, & O'Leary, 1973), observer bias (e.g., McNamara & MacDonough, 1972; O'Leary, Kent, & Kanowitz, 1975), observer drift (e.g., Reid, 1970), observational complexity (e.g., Kazdin, 1977a) and 477 1982, 15) [477][478][479][480][481][482][483][484][485][486][487][488][489][490][491][492] NUMBER 4 (WINTER 1982) various sources of inflation of observer reliability estimates (e.g., Hartmann, 1977). The number of articles in the Journal of Applied Behavior Analysis (JABA) that contain reliability estimates of the dependent variable have steadily increased in recent years (Hayes, Rincover, & Solnick, 1980).…”