Participatory Action Research (PAR): A form of Whyte's (1991) "participatory action research" (PAR) was used to identify communication, learning, teaching, architectural and acoustical issues from the perspectives of the teachers and students. This approach is widely used in the appraisal of qualitative aspects of environments to assist in programming new facilities and post occupancy evaluation of existing buildings. It was hypothesized that the complex interactions among the physical environment of the classrooms, the teaching techniques of the faculty, the behavior of the students, and the quality of the learning that occurs in the rooms can be identified using these methods. Once the broad range of issues were identified, a series of case studies were conducted in actual elementary schools to gather the data for the present project.

Room and Student Selection: Once the major issues were identified through the PAR studies, a series of rooms and students were evaluated for case studies. Initially, the study focused on 3 groups: Pre-kindergarten to grade 1, grades 2-3, and grades 4-6. Further studies will include middle school, high school; and college students as well as adults. Children were initially selected and screened for normal hearing. Follow-up work will address these issues relative to the special populations noted above.

Schools were chosen to gain representative samples of different socioeconomic classes, races, and genders in the students. The classrooms were chosen in several localities including urban, suburban and rural areas. Buildings constructed at different times in different conditions with different acoustical qualities were also chosen.

Acoustical measurements: A series of acoustical measurements were made at multiple locations in each class room from multiple source locations using the impulse response technique. A TEF analyzer, JBL's Smart system and two custom systems were available for this work. Acoustical measures of reverberation time, early reverberation time, early to late energy ratios, loudness (or relative strength), Alcons, speech transmission index, background noise levels and signal-to-noise ratios as well as binaural measures were made to document the acoustical conditions. In addition, the architectural features such as dimensions, volume, seat locations, materials, shape, furnishings, etc., were recorded in detail.

Speech recognition tests: Speech recognition of children in the rooms was assessed by nonsense syllables and monosyllabic word tests. The stimuli were presented in the classroom via a calibrated digital tape recorder through a loudspeaker that closely approximates the directional characteristics of the human voice. The source was placed at the front of the classroom and at other locations identified in the PAR studies where the teacher might frequently stand. Monosyllabic nonsense stimuli were used because: (1) practice/learning effects are minimal; (2) children as young as 5 years of age can perform this task; (3)closed set testing responses can be utilized; and (4) consonant confusion matrices can be produced from the data. The child's task was to circle the correct choice on an answer sheet. On each speech-recognition form, the location/position of the child in the room (distance from the speaker, position in the classroom, etc.) was documented.

Learning qualities tests. A survey instrument and participatory action techniques were used to assess acoustic and non-acoustic factors that might affect speech recognition and learning.