Development and assessment of an alternative to the narrative observation

For seven years, the narrative observation system has been the principal tool of the Behavioral Ecology Research Team in its study of health care at the Texas Institute for Rehabilitation and Research. When the program of research was begun, the narrative observation approach to data collection and analysis (separation of observation from coding and data reduction) was chosen because nothing was known in an explicit and quantitative sense about the behavioral trajectory of spinal cord injured persons. Although costly in terms of necessary equipment and person hours, this approach proved highly successful. The treatment staff has asked to receive the most up-to-date behavioral measures possible for their patients, so that they can use the data to plan patient programs. The purpose of this study was to develop an alternative system for data collection and management that is more efficient than the narrative system, yet which yields the same derived measures of interest. A pencil-and-paper shortform (which combines and eliminates various steps) was devised for observers. Two sets of ten paired observations were completed. The first set paired ten narrative observations with ten shortform observations in an attempt to assess the convergent validity of the two systems. The second set paired shortform observations in an attempt to assess the reliability of the new system. At several levels of detail, the shortform system displayed a great deal of congruence with the narrative system and the shortform also proved to be very reliable. The shortform also afforded considerable savings in person hours and other resources. Methodological implications, recommendations for other applications, and guidelines for further research are discussed.