Data on the prices of beverages were collected from May 15 to June 10, 2012. Four geographic areas on the island of O‘ahu (Hawai‘i Kai, Manoa, Waimanalo, and Wai‘anae) were selected for assessment. Hawai‘i Kai and Manoa were the higher per capita income areas, and Waimanalo and Wai‘anae were the lower per capita income areas that were selected based on convenience.1
Within each area, at least one supermarket (a store having at least five aisles of household staple foods), convenience store (a 7-eleven or other gas station/mart), and quick serve restaurant (L&L Hawaiian Barbeque, McDonald's, Burger King, or Zippy's) were assessed. In total, 8 supermarkets (2 in Hawai‘i Kai, 2 in Manoa, 1 in Waimanalo, and 3 in Wai‘anae), 5 convenience stores (1 in Hawai‘i Kai, 2 in Manoa, 1 in Waimanalo, and 1 in Wai‘anae), and 10 quick serve restaurants (4 in Hawai‘i Kai, 1 in Manoa, 2 in Waimanalo, and 3 in Wai‘anae) were selected based on availability.
Data collection was conducted throughout the week at various times by a University of Hawai‘i at Manoa researcher using data collection forms. Beverages were first classified as healthy or unhealthy based on the 2010 Dietary Guidelines for Americans relating to nutrient density.14
Healthy beverages included milk (2% white milk, 1% white milk, chocolate skim milk), 100 % orange juice, unsweetened tea, diet soda, and unsweetened coffee. Unhealthy beverages (beverages containing added sugar) included regular soda, fruit drinks, sweetened tea, sports drinks, and flavored water. Ready to drink sweetened coffee and energy drinks were treated as separate categories consistent with the beverage industry classification of these as functional drinks, and water was also considered separately consistent with the beverage industry's classification of bottled water.15
Up to three brands of beverages were chosen in each category; for each beverage category, the three cheapest brands were selected based on the prices of the smallest available size of that particular beverage. However, because multiple brands and sizes of soda and water were available, soda and water brands were selected based on two criteria. First, the selection process was limited only to soda brands that were available in at least three different sizes; then, among these, the three with the lowest cost at the smallest size were included in the assessment. Similarly, the criteria used to select the brands of water were first availability in at least two different sizes, and then lowest cost at the smallest size. Once the beverage brands were selected, the prices and ounces of all available sizes (bottled, canned, and fountain drinks) of these beverage brands were collected. The goal was to ascertain the regular prices of these beverages (not sales prices) by checking the price labels and/or menus.
Mean price per 20 ounces was the primary unit used to compare the data since 20 ounces was the most common beverage size. Since both the price and size of each beverage was recorded, a beverage's price per ounce was easily calculated by dividing its price by its size in ounces; the calculated price per ounce was averaged across all available sizes for a particular beverage. A beverage's price per 20 ounces was then calculated by multiplying 20 by the previously calculated average price per ounce value.
The mean prices per 20 ounces of healthy and unhealthy beverages were compared in general and by per capita income area using analysis of variance (ANOVA), independent samples t-tests, and descriptive statistics. These data analyses were performed using SPSS Statistics version 20, and P ≤.05 was considered statistically significant.