Thesis (MPhil (Information Science))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: National Statistical offices (NSOs) are the sources of wide ranges of socio-economic, demographic and agricultural data and information that are used to monitor and evaluate development programs and formulate policies. The data generated by NSOs is used as basis for making decisions and also used to assess the extent and causes of poverty. Various stakeholders such as researchers, the World Bank, the International Monetary Fund, the UN, and various NGOs prepare and release research materials and annual reports using data and information obtained from NSOs and line ministries. For example, Deneulin and Shahani state that one of the intentions of the annual Human Development Report (HDR) prepared by the UNDP is “to assess the quality of life of a population and be an advocacy tool for its improvement with a political purpose of raising awareness and generating debate on public issues and concerns which would otherwise not be on the political agenda”1. Based on the different approaches to poverty, different sets of data and information are produced and used for poverty measurement. Mostly, poverty is measured using data obtained from nationally representative household surveys which focus on income and expenditure, ownership, access to and use of some basic services. Another approach uses data on mental satisfaction; still others assume poverty to be multi-dimensional and argue that income alone is not enough. They view poverty as deprivation of basic capabilities due to high rates of mortality, illiteracy, malnourishment, unemployment, ill health, lack of education and social exclusion, etc2. The quality of data and information (such as integrity, methodological soundness, accuracy and reliability, serviceability and accessibility) generated by data-producingagencies therefore needs to be preserved and improved in order to obtain meaningful results from the measurement of poverty in any of the approaches and to satisfy the growing data quality demands of stakeholders. Loshin states that “strategic decisions based on untrustworthy information are likely to result in poor decisions”3. This study focuses on the role played by national statistical offices in poverty reduction in general. It examines the various activities, players, interactions, and ICTs used at the various stages of the statistical process in the Ethiopian Central Statistical Agency (CSA) to generate poverty-related data and information and how the quality of this data can be preserved and improved. The purpose of this research is therefore to identify poverty related data quality problems with respect to the IMF’s DQAF and assess where in the statistical process specific types of ICTs can improve data quality. For this reason interpretive case study method with the researcher as participant observer was adopted to study how poverty related data and information is produced. It was found out that some of the data quality problems can be addressed using appropriate ICTs with the availability of reliable power infrastructures. / AFRIKAANSE OPSOMMING: Nasionale Statistiekkantore (NSOs) is die bron van ‘n wye reeks sosio-ekonomiese, demografiese en landboukundige data en inligting wat gebruik word om ontwikkelingsprogramme te monitor en te evalueer. Die data wat deur NSOs geskep word, word aangewend as grondslag vir besluitneming. Die data word ook gebruik om die omvang en oorsake van armoede te bepaal. Verskeie betrokkenes soos navorsers, die Wêreldbank (WB), Internasionale Monetêre Fonds (IMF) en die VN en NSOs skep en versprei verskillende navorsingsuitsette en jaarverslae wat gebruik maak van die data en inligting wat verkry word van NSOs en ministeries. So konstateer Deneulin en Shahani dat een van die doelstellings van die Verslag op Menslike Ontwikkeling (HDR), soos opgestel deur die VNDP, is om “die lewensgehalte van ‘n bevolking te skat en om as werktuig en voorspraak vir die verbetering daarvan op te tree, met die politiese doelwit om bewustheid te verhoog en debatvoering oor openbare sake en kwessies, wat andersins nie op die agenda sou verskyn nie, aan te voor”.4. Na gelang van die verskillende benaderings tot armoede word verskillende stelle data en inligting geproduseer en gebruik vir die meting van armoede. Armoede word gewoonlik gemeet deur data te gebruik wat bekom word van landswye opnames van huishoudings en ingestel is op inkomste en besteding, besitreg, toegang tot en die gebruik van ‘n paar basiese dienste. ‘n Ander benadering gebruik data oor geestelike bevrediging; ander weer aanvaar dat armoede multidimensioneel is en voer aan dat inkomste alleen nie genoeg is nie. Hulle beskou armoede as die ontbering van basiese vermoëns weens ‘n hoë sterftesyfer, ongeletterdheid, ondervoeding, siekte, gebrekkige onderwys, sosiale uitsluiting en dies meer5. Die gehalte van data en inligting (soos integriteit, metodologiese deeglikheid, akkuraatheid en betroubaarheid, bruikbaarheid en toeganklikheid) wat deur agentskappe opgelewer word moet dus bewaar en verbeter word ten einde ‘n beduidende resultaat te bekom uit die meting van armoede deur enige van die benaderings en ook om belanghebbendes se groeiende aandrang op datagehalte te bevredig. Loshin beweer dat “strategiese besluite gebaseer op onbetroubare inligting waarskynlik swak besluitneming tot gevolg sal hê”.6. Hierdie ondersoek konsentreer op die rol wat gespeel word deur nasionale statistiekkantore in die algemene bekamping van armoede. Dit ondersoek die verskillende aktiwiteite, rolspelers, interaksies en ICTs wat op verskeie stadiums van die statistiese proses deur die Etiopiese Sentrale Statistiekagentskap (CSA) gebruik word om data en inligting oor armoede te skep en hoe die betroubaarheid van data behou en verbeter kan word. Die doel van hierdie navorsing is dus om kwaliteitsprobleme wat verband hou met armoededata ten opsigte van die IMF se DQAF te identifiseer en om te besluit waar in die statistiese proses bepaalde soort ICT’s die gehalte van data kan identifiseer. Om hierdie rede is die interpretiewe gevallestudiemetode aanvaar om te bepaal hoe armoede-verwante data en inligting geskep word. Die slotsom was dat sommige van die probleme in datagehalte aangespreek kan word deur die gebruik van gepaste ICT’s met die beskikbaarheid van betroubare mag-infrakstrukture.