DIFFERENCES IN OPINION-SURVEY RESPONSE PATTERNS AS A FUNCTION OF DIFFERENT METHODS OF SURVEY ADMINISTRATION

FOR A POPULATION OF MALE TECHNICAL EMPLOYEES IN A LARGE NATIONAL ORGANIZATION THE METHOD OF RESPONDING TO AN OPINION SURVEY MADE NO DIFFERENCE IN THE LEVEL OF RESPONSES GIVEN. HOWEVER, THE CONDITIONS UNDER WHICH THE SURVEY WAS ADMINISTERED DID HAVE AN EFFECT ON RESPONSE. WHEN EMPLOYEES WERE SURVEYED ON THEIR JOB LOCATIONS UNDER THE SUPERVISION OF A COMPANY REPRESENTATIVE THERE WAS A TENDENCY TO RESPOND MORE FAVORABLY TO A SIGNIFICANT NUMBER OF GENERAL-OPINION QUESTIONS, PARTICULARLY QUESTIONS DEALING WITH THE "COMPANY IN GENERAL," THAN WHEN THEY WERE PERMITTED TO RESPOND IN A NONOFFICE LOCATION. THE UNDERLYING DYNAMICS OF THIS RESULT ARE NOT CLEAR, BUT THE SIGNIFICANT POINT IS THAT THE METHOD OF QUESTIONNAIRE ADMINISTRATION CAN SERVE AS A POTENTIAL SOURCE OF DISTORTION IN EMPLOYEE-OPINION-SURVEY DATA.