Glossary
Special | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | ALL
S
Signature domain (for health facilities)The health facility signature domain from MEASURE Evaluation contains all the information necessary to identify a facility uniquely, and should be explicitly included in all health facility surveys. While each element in the Signature Domain may or may not uniquely identify a facility (and indeed some elements may not exist), the collection of as many elements as possible will more reliably result in correct matching of facilities across surveys. (MEASURE Evaluation 2007) | |
Spatial autocorrelationSpatial autocorrelation refers to the dependence on spatial location of observed values of a phenomenon. If spatial autocorrelation exists, values are considered to be clustered or grouped geographically. If no autocorrelation exists, values are considered to be randomly distributed geographically. For more information, see glossary entry for Moran's I. | |
Spatial rate smoothingRate smoothing is a procedure to address the variance instability related to estimating rates in areas with widely varying populations. Variance instability is particularly pertinent in areas with small population numbers. Raw rates and smoothed rates will differ less as underlying population numbers in areas increase. Smoothing increases the precision of risk estimates. (GeoDa Center) | |
SPOT 5High spatial resolution satellite owned and operated by the French company SPOT Image (www.spot.com). For a technical specifications sheet and a gallery of SPOT imagery, see www.satimagingcorp.com | |
StovepipingTerm referring to the practice of keeping data separate from, and not connected to, other data. Stovepiping of data can lead to isolated, one-time solutions and narrow interpretations of data. | |
Strategic information system (SIS)The hardware, software, facilities, data, and personnel needed to achieve the most cost-effective system for satisfying [an] organizations [information] needs (U.S. GAO 1992) | |
Supervised classificationA procedure for identifying spectrally similar areas on an image by identifying "training" sites of known targets [known via ground truthing conducted during fieldwork] and then extrapolating those spectral signatures to other areas of unknown targets. In contrast, an unsupervised classification is the categorization of digital image data by computer processing based solely on the image statistics without availability of training samples or a-priori knowledge of the area. (CCRS) | |
Surveys and surveillanceBiological and behavioral surveys and surveillance are essential to determine the drivers and the spread of the HIV epidemic in a country. HIV surveys and surveillance may focus on the general population, most-at-risk populations, or both. Protocols and data collection tools for surveys should be based on international standards, such as the Demographic and Health Survey, the AIDS Indicator Survey, and the Multiple Cluster Indicator Survey. (UNAIDS 2008) | |