Upload
giles
View
47
Download
3
Embed Size (px)
DESCRIPTION
Data sources and collection methods. Ken Mease Cairo, June 2009. What types of Data?. A thorough assessment may well include: Archival and secondary data Survey data Quantitative and qualitative approaches and data It also will likely include de jure and de facto information. - PowerPoint PPT Presentation
Citation preview
Data sources and collection methods
Ken Mease
Cairo, June 2009
What types of Data?
A thorough assessment may well include:• Archival and secondary data• Survey data • Quantitative and qualitative approaches and
data• It also will likely include de jure and de facto
information
Qualitative and Quantitative data
• There are basically two types of data: qualitative and quantitative
• Qualitative data are usually text or words and quantitative data - numbers
• Qualitative approaches, if conducted in a rigorous manner, require more skill than most quantitative approaches
Qualitative Data and Approaches
• Qualitative approaches provide text data, but increasingly audio, video and images
• They are more time consuming to analyze
• Text management software, such as NVIVO, AtlasTI and AnSWR (available free at http://www.cdc.gov)
• Coding is a very subjective process and open to various problems, such as investigator bias or a lack of inter-coder reliability
Quantitative Data and Approaches
• Quantitative approaches provide increased rigor by investigating relationships at known levels of probability
• They are easier to analyze because researchers use standard, replicable techniques
• Common software includes SPSS, Stata and SAS
Sources and Types of Governance Data
• Country reports/desk studies • Cross-country (region)comparative surveys• Expert assessments• Government data• Household surveys• Mass opinion surveys• Key Stakeholder Surveys• Media Indicators• Business surveys
Combining different sources and methods
• Combining methodologies and types of data often provides the most useful results.
• Combining archival information and administrative data with original qualitative and quantitative survey data allows for triangulation
• This approach can increase the level of professionalism, credibility and legitimacy.
Administrative, archival and secondary
These data, both qualitative and quantitative, objective, reported events, perception and proxies come from a variety of sources:
•Narrative reports, administrative data and other information routinely collected by government ministries
•The constitution, laws, policies and legislation
•Statistics and data gathered by NGOs, international organizations and academics.
Data Mapping
• It is the best way to identify existing data - accessibility, quality and gaps
• One tool available to assist with this process is the IMF’s Data Quality Assessment Framework
• A senior academic conducted the data mapping exercise in Zambia and was very valuable
Survey Data
• Qualitative and Quantitative
• Sample sizes can range from 20 to 20,000 or more
• Costs can range accordingly
• There are a range of options for who collects the data – independent surveys firms, academics, NSOs
• Perception and reported events data
Different Types of Interviewing
• Structured interviews use an identical instrument for each respondent,
• Interviewers are trained and have explicit instructions.
• This technique uses primarily structured questions with fixed response sets and very few open-ended questions
Types of Interviewing
• Semi-structured interviews use a written list of questions that need to be covered in a particular order
• The questions are often developed from informal discussions and focus groups.
• They can include open-ended and/or more structured questions
• Ideal when working with elites, managers, bureaucrats and other people who have limited time
Types of surveys and data collection methods
• Face-to-face data collection is likely the best option in most developing countries
• It is also the most expensive and time-consuming
• It requires professional management of trained interviewers, sampling and other aspects of the study
Mail surveys
• Work well only if the postal system is reliable
• The questionnaire must be carefully designed for self-administration, and there should not be too many language issues
• The cost is usually quite reasonable
• It misses the homeless and other vulnerable groups who may not have a valid mailing address
Internet surveys
• Require special planning and design
• Very good for certain populations
• These surveys can suffer from low response rates.
• Developing and implementing internet surveys has become very reasonable in terms of cost
• Very reasonably priced internet services are available - Survey Gizmo at www.surveygizmo.com
Telephone surveys
• May have coverage issues in most countries.
• In many countries, people do not have a phone in their home, and if they do have a phone, it is usually a mobile or cell phone
• Most cell phone providers do not make telephone numbers available for use in random-digit-dial surveys.
Focus groups
• Can generate information about the background conditions surrounding governance issues.
• Focus groups are usually efficient in terms of time and money.
• They are highly participatory and have the potential of generating solutions to the problems identified by the group members
• It is very demanding and requires highly skilled coordinators.
Focus groups
• Make individual ratings insignificant
• Accuracy can suffer, as some individuals may not feel comfortable to speak up in public, while others are hard to keep quiet
• Focus groups yield less systematic results
• Focus groups are best used to identify issues and develop surveys rather than as the only source of data
Who collects the data
• Outside contractors – surveys and desk studies
• Local survey researchers
• Local academics for desk studies
• National Statistical Offices
• Government ministries
• NGOs/CSOs
Data collection and dissemination exercise
• Pick a Governance issue in your country
• Choose the levels and types of the data you will collect – Macro, Micro, De jure and De facto
• Identify the types and sources of data
• Choose the data collection method/s
• Identify who will collect the data
• Select several options for sharing the results with different audiences