Natural Experiments, Impact on Community Activities

For many decades, most North Americans have habitually chosen watching television over other in-home leisure activities. TV viewing also displaces many leisure activities away from home, either in other dwellings or out of doors. A natural experiment in Canada provided the opportunity to compare the effects of television viewing in three towns assigned the pseudonyms of Notel, Unitel, and Multitel. Although their communications profiles were different, the three towns were similar in size, demographic variables such as socioeconomic status (SES), the cultural backgrounds of the residents, and the types of industry in the area. Each town had a population of about 700 but served an area four times as large through its schools and services. The purpose of the experiment was to compare Notel residents' participation in other leisure activities during the year before they first obtained television reception (Phase 1) with their participation 2 years later (Phase 2). The study also compared Notel's data with information from Unitel, with one Canadian public channel, and Multitel, with four (the same Canadian public channel and three American private channels).

This opportunity to conduct research on the effects of television was unusual and important because causal inferences can be made in natural experiments, provided that alternative possible explanations of the results (threats to internal validity) can be ruled out. This encyclopedia entry focuses primarily on the study conducted by Tannis MacBeth Williams and Gordon Handford (1986) with students in grades 7 through 12 in the three communities, with some results for adults included for comparative purposes.

The main goal of the study was to find out the extent to which residents of Notel, Unitel, and Multitel participated in their town's community activities. A method called ecological psychology, or behaviorsettings analysis, was particularly well suited to this goal. In the 1950s, Barker and Wright (1955/1971) did a behavior-settings analysis of a town in the United States that they called Midwest. They theorized that each unit of the environment, or behavior setting, places limits on the range and type of behavior likely to occur there, sometimes for physical reasons but also because of social and other conventions. Their system provides a method for specifying environmental units that can be applied to entire communities such as Notel, Unitel, and Multitel, towns about the same size as Midwest.

In an initial visit to each town during Phase 1 and then again 2 years later, several people in each of the following categories were interviewed: retailers, town clerks and elected officials, officers of community clubs/organizations, recreation-commission personnel, school teachers, newspaper editor, police, clergy, and children. The interview information was used to determine the public behavior settings. Researchers also obtained copies of the previous year's community newspaper to find items referring to community activities, organizations, meetings, special events, and so on.

The list of activities and events unique to each town was organized into the 12 categories Barker and Wright used to describe Midwest. These were sports; open areas, such as parks, playgrounds, and swimming holes; businesses, such as stores and offices; civic activities, such as the post office and town hall; out-of-school educational activities such as music lessons and adult classes; meetings of clubs and other nonsports organizations; medical activities, such as visits to the hospital or doctor's office; community dances, parties, and suppers; special days, such as weddings, funerals, and elections; religious activities; entertainment, such as special movies, parades, and bingo; and other activities, such as fund-raising events, town cleanup campaigns. Information on these 12 categories was obtained using a questionnaire with about 275 items for each town in each phase. Another questionnaire assessed participation in 58 private leisure activities such as bicycling, reading books, and exercising. Whereas separate community activity behavior-setting questionnaires were developed for each phase of the study in each town, the same private leisure activity questionnaire was given in all three towns in both phases.

...