The evaluation was conducted at two sites in southern California, with a winter study beginning in January and a summer study beginning in July. The intervention site was the San Diego Zoo, located near downtown San Diego. The comparison site, which received only the evaluation procedures, was the San Diego Zoo's Wild Animal Park, which is located in a relatively rural area approximately 29 miles from the zoo. The sites are operated by the same zoological society and sell similar items in their gift shops.
A nonequivalent control group design was used to evaluate the effects of a multi-component intervention on hat wearing by children and on the purchasing of sunscreen and select hats in gift shops. During the winter study, data were collected for a consecutive 10-week period. The first 4 weeks consisted of evaluation only at both sites (baseline phase), during which time the researchers documented hat and sunscreen use. This was followed by a 6-week period of intervention at the zoo only, with continued evaluation at both sites (intervention phase). The summer study consisted of 4 weeks each of a baseline and intervention phase, for a total of 8 weeks of data collection.
Two primary outcome measures were used: (1) direct, unobtrusive observations of hat use by children who appeared 12 years or younger as they exited the zoo and park sites and (2) sales of sunscreen and targeted hats at the zoo and park gift shops. For the observational measures, trained observers were stationed near the exit of the zoo and park on a portion of baseline and intervention phase days of both the winter and the summer study. Observers attempted to record data on all children appearing to be 12 years or younger as they came through the exit gate, noting the child's estimated age, gender, hat use (none, visor, cap/bonnet, flap hat, 2-in. brim, 3-in. brim, stroller/umbrella cover, or hood/backward cap), and sunburn level. When an infant's gender could not be determined, the "baby" category was used. Observations were conducted from 2:00 to 4:00 p.m. in the winter study and from 3:00 to 5:00 p.m. in the summer study. On approximately 51% and 60%, respectively, of the winter and summer observation days, a second observer also recorded data in order to obtain reliability estimates.
For the sales measure, the numbers of units of the targeted items sold each day of the winter and summer studies were obtained from the merchandising buyer. Data were also obtained from the marketing department on daily attendance rates; the researchers adjusted for this variable in the analyses.