What Does Integration imply in Choosing a Unit

What Does Integration imply in Choosing a Unit of Enumeration: Enterprise, Holding or Individual? Does it Matter? Perspectives from Africa Willis Oluoch-Kosura, CMAAE / U. o. N

Outline • • • Introduction Importance of statistical integration Drawbacks of integration Limitations to statistical integration Opportunities

Introduction Statistics applications development & scientific sectors: Describe reality of people’s lives; § Evidence to develop & monitor development policies; § Management & delivery of services; § Tracking progress & assessing policy impacts; § Enhance resource management

• Statistical application faces challenges; • Solution: Statistical integration üframework of applications, techniques and technologies ücombining disparate data to provide the user with a unified view of the data success based on rapid adoption of databases with capacity to share or merge existing data repositories for bridging the information gap

Importance of Statistical Integration • Richer databases for detailed & extensive analyses; • Reduced cost of planning & data collection; • Lower burden on respondents; • Promote definitions, standards & classifications – data accuracy & coherence • promote data scrutiny and quality

• Harmonize methods for comparative analysis between units or time scales; • Promote use of common tools & processes in statistical analysis; • Promote methods of data storage (databases); • Economical use of human capacity to prepare, analyze & interpret data - reduces cost of specialized training; • Detailed view of the situational analysis than usingle data source;

Drawbacks of Integration • Lower flexibility of data collection, storage & management: ü Time Adjustment for collection – seasons annually; data than ü Sampling unit size adjustment – zones than holdings, hectares than acres ü difficult to integrate household & agricultural data if sample frames don’t correspond

Limitations to Statistical Integration (A) Related to data collection process • Varying data needs and sources - Different problems, needs, variations in questions, questionnaires, depth of coverage & data collected • Use of varying local units – agric household than holding itself, varying/no regard to sampling unit size, Varying units for estimating production

• Varying production methods farming techniques, spacing & management; • Surveys based on recalling power annual than seasonal, hence, subjectivity; • No standard sampling unit ü Household (single, nuclear, extended) ü Holding (hectare, acre, etc) • Diseases – wipes out sampling units

• Varying sampling methods – cater for problem, data needs & available resources (random, stratified, multi-stage sampling etc) • Migration & cultural barriers – inaccessibility to sampling units or main respondents (men) • Insecurity & conflicts – deaths, migration, hostile environment for data collection

(B) Institutional & technical constraints • Policy & legal constraints – copyrights, data protection Act 1998, security reasons, commercial value, varying national data policies • Weak national statistical systems – not well developed, low skilled manpower, loose attachment with data producers • Poor IT infrastructure - inefficient IT connectivity - hard to adopt new technologies

• Varying technologies & requirements – varying user interfaces, development environment, metadata repository to document, security & management frameworks • Data inconsistency and poor quality – varying methods of collection, poor storage formats & inadequate presentations (e. g. no metadata), lack classifications

• Weak coordination & collaboration among data producers üno reference to NS offices, üno co-ordination between them.

Opportunities • Growing IT technologies – (internet, CD-ROMS, databases and GIS) ü allow new ways of production by easing calculations and comparisons ü reduce costs dissemination of production & ü establishment of data dissemination centres ü wider & faster data dissemination than through printed publications

• New financing opportunities to support NSS üWorld Bank's Trust Fund for Statistical Capacity Building üSTATCAP üAfrican Development Bank & designing & updating of NSDSs DFID - üAFRISTAT for enhancing consistency & comparability of statistical data üPARIS 21 to implement NSDSs for better national statistics by 2010

• Increased data demand by research agencies and nations • Increasing cooperation between nations or agencies üfor resources mobilization üstandards & procedures for integrated systems üconflict resolution collection to enhance data

• Existing data from national international research bodies &

Thank you