Original Data Collection Reliability and Validity in Original
Original Data Collection
Reliability and Validity in Original Data Collection • Reliability: consistency of measurement (Bollen 1989) • Validity: Validity is concerned with the meaningfulness of research components…whether (researchers) are measuring what they intended to measure
Measurement in Original Data Collection
Benstead 2014 • What are the main findings of the paper? • What are the lessons learned about “enumerator dress” for other contexts?
Enumerator Effects • Affect responses to attitudinal, sensitive, complex, and open-ended questions • that interviewers’ own opinions affect survey responses • gender, level of experience, and personality also appear to affect both responses and the willingness of potential respondents to participate (Michelitch and Lupu 2017) What role does positionality play in data collection?
Bystander Effects • Afrobarometer data says 44% of interviews are conducted with someone else present (Michelitch and Lupu 2017) • Simultaneous household interviews (Bleck and Michelitch) • Self Administered Audio-MP 3 Questionnaires (Chauchard 2013)
High Quality Responses • Discomfort, unfamiliarity with data collection method can generate non-responses and satisficing (Krosnick 1991) Kramon and Weghorst 2012
Behavioral Measures • Why do some social scientists prefer behavioral measures to attitudinal responses? • What are some potential trade-offs/problems with behavioral measures?
Habyarimana et al 2007 • Why is there lower level of public provision when there are higher levels of ethnic diversity? • Evaluate mechanisms: preferences, technology, and strategy selection • What are the findings? What is the body of evidence?
Bleck and Michelitch 2015
Open ended messages to President Obama by Category
Concepts in Context “Most importantly, we have heard that there agricultural subsidies for farmers, but we have not received any subsidies. I request that President Obama ask his counterpart in Mali if (village name) is not a part of Mali, because the authorities do nothing for us. (#341)” “We are not in Mali. We don’t eat, we don’t sleep, we don’t have any health or sanitation infrastructure, no schools, no water. …The state has abandoned us. There is nothing for us but poverty and the misery of peasants. “ Interview with Village Council Member, Village 5, July 2012
*Paper Assignment 1* Papers should be 6 -8 pages (1 -1. 25 inch margins, 12 pt font) with page numbers. Chose your preferred citation format, but make sure that you cite relevant sources. The paper should be submitted in pdf format via email to ggoertz@nd. edu by February 24 th at midnight. Please place a hard copy in the green box in front of Jaimie’s office door (2130 C Jenkins Nanovich) by Monday Feb 26 th at 5 pm. Prompt: In this paper, examine a key variable/concept/treatment/indicator/data that might be relevant for your research. This could be an explanatory variable or an outcome variable. Start by describing a concept that is relevant to your research. How did you select this concept? How is it relevant to your research question? Here are some ideas for exploring your concept-variable-treatment-indicator-data. a. Choose a guideline(s) from the February 7 th lecture that is most relevant or interesting and use it to analyze your concept. b. Discuss how you could measure your concept using existing data. What is the assumption behind the data that you plan to use? c. Discuss issues that you would consider as you collect original data. Examples include how and if your respondents’ understanding of your concept affects your research, what you need to consider when trying to measure your concept, and how the research process affects the measurement of your concept. d. Discuss the aggregation procedures of a key concept-indicator data-indicator. e. Evaluate the "indicators" of your concept typically used.
- Slides: 13