National Coral Reef Monitoring Program: Socioeconomic surveys of human use, knowledge, attitudes, and perceptions in Tutuila, American Samoa from 2014-01-28 to 2014-02-22 (NCEI Accession 0167228)

Data were collected via telephone using random digit dialing and via face to face interview, and the total sample size for this survey was 448. See file "NCRMP-Socio-AmericanSamoa-2014_Questionnaire" for full questionnaire. Surveys were conducted on Tutuila Island during February 2014. To obtain a stratified sample of respondents, villages in American Samoa were categorized by degree of urbanization (urban, semi-urban, and rural) and location on Tutuila Island (east, west, northeast, or northwest). Villages were randomly selected from each category, and the percent population of the sample villages in each category approximated the percent of the total island population in that category. Residents were sampled proportionate to the total population in each village. Sampled villages included: Amanave, Amouli, Aoloau, Faga'itua, Fagasa, Failolo, Faleniu, Ili'ili, Lauli'i, Leone, Malaeimi, Nu'uuli, Onenoa, Pago Pago, and Vaitogi. In total, 448 residents were surveyed, yielding a confidence level of 95% and a confidence interval of 4.6% for the island-wide results. Prior to surveying, information regarding the monitoring effort was presented to the Office of Samoan Affairs as well as to mayors of the villages included in the target sample. A local survey team was trained in survey administration and protocol. Survey questions were carefully translated to capture the meaning of each question in the English version. A Samoan word can have several meanings, so multiple Samoan-speaking project team members reviewed and revised the translation to ensure consistency with the English version of the survey. This team conducted surveys in-person in each of the selected villages on weekday afternoons and early evenings, as well as on Saturdays. Residents were sampled opportunistically due to the difficulty of pre-selecting random households (e.g., no street addresses for homes, large numbers of unoccupied structures), and the challenges of using a random sampling approach in the field (e.g., vicious dogs). The survey team made an effort to survey different regions of each village and, to the extent possible, to target respondents from different demographic categories. Respondents were offered the choice to conduct the survey in either English or Samoan, based on their own language preference. Eighty-one percent of surveys were conducted in Samoan and 19% conducted in English. Data analysis included descriptive analyses (e.g., measures of central tendency, examination of distribution), as well as examinations of relationships between variables (e.g., cross tabulations, correlation, regression analyses). For a complete description of the data collection and analysis methods, see Levine et al. (2016).