Search Filter
- Publications
- Abstract of Theses and Dissertations
- Database
- The Values and Validity of Group Informant Survey Technique in Program...
The Values and Validity of Group Informant Survey Technique in Program Planning
Dissertation Abstract:
The study was conducted to determine the validity of the group informant survey (GIS) technique o collecting data in a barrio in comparison with the complete enumeration survey (CES) method.
Group informants consisted of the barrio captain and six barrio councilors of each of seven barrios in Pila, Laguna, while CES respondents included 518 individual household heads. The same structured interview schedule was used in both survey methods. Information on the barrio was gathered on 213 items under seven community variables: transportation, population, water and sanitation, agricultural conditions, economic factors, community improvement, and people’s aspirations for and problems in better living. The t-test, Spearman’s rank-difference coefficient correlation and Pearson’s coefficient of linear correlation were used in data analysis.
Findings showed the GIS informants to be community leaders, older in age, higher in educational attainment, longer residents in the barrio, and more familiar with local conditions than the average resident.
Of the 213 informational items, 175 (82.16%) of the GIS data significantly correlated with the CES data. The validity of the GIS depended on the amount of information collected, that is, the more the data involved, the bigger the estimate bias. The GIS was more valid in gathering popular, familiar or widely-shared information in the barrio and data actually kept in record. However, the GIS tended to be less valid in ranking the most important data on highly personal matters, as on people’s aspirations and need problems. Although barrio size was not correlated with data validity, a bigger bias was observed inlarger communities.
The study suggests that in program planning, the GIS may be used to obtain information with acceptable validity level, but data that are highly personal in nature should better be secure by using the CES or other sampling surveys.