PRA Technotes (technical papers) are available for download in Adobe Acrobat ".pdf" format.
History of survey research
Conducting surveys seems simple enough: ask people questions, record their answers, and report the results. Even Roman Emperors polled their citizens in ancient times.
Quality issues in survey research
Although the science of survey sampling has reached the point where we can talk about error rates in terms of fractions of a percent, considerably less is known about issues relating to questionnaire design.
The most important element of a telephone survey is the data collection instrument - the interviewer.
The focus group is a common form of social and market research. Groups usually involve around 5–12 participants who are guided through a series of structured questions by a moderator.
Response rates on mail surveys
The mailed questionnaire has an undeserved reputation as a poor method of survey research.
Ordinary conversation is the most common form of information collection. The in-depth interview merely extends and formalizes everyday conversation.
Primer on polls
Public opinion polls are pervasive. For the most part, published reports by pollsters fail to provide sufficient information to allow expert appraisal of the overall validity of the results.
It seems strange that a sample of 1,000 for Canada has about the same theoretical accuracy as a sample of 1,000 for Toronto or Red Deer when there is such a difference in the populations.
Conjoint analysis is a method for consumers to assign value to various features of a product or service. This research technique simulates the consumer choice decision, usually in a written questionnaire.
Economic impact multipliers
Economic impact analysis is an important tool for predicting the effects of a variety of economic decisions on a community, town, city, or country.
Strategic pricing using discrete choice models
Pricing a new product or service, or deciding whether to adjust the price up or down, is one of the most important decisions made in business. Price too low, and you give up revenue. Price too high, and your customers leave.
Comparing two respondent selection methods in telephone surveys
The problem: how do First Birthday versus First Answer respondent selection processes differ?
Follow-up study of home purchase intentions
PRA, on behalf of Canada Mortgage and Housing Corporation (CMHC), conducted an annual survey asking Canadians about their home purchase and home renovation intentions for the coming year. In 2002, PRA re-contacted respondents to the 2000 and 2001 surveys to conduct a follow-up interview. This technote summarizes the results obtained from an analysis of the data gathered.
Reporting interview and focus group findings
Reports on qualitative data, such as interviews and focus groups, often appear imprecise and vague. Terms such as "some" respondents or "a majority" of participants (in a focus group) convey the impression that the writer wishes to obscure the exact count.