Anonymity:

Either asks no name or records no name; may help respondent's frankness, but makes follow-up difficult. Also, if your research includes small number of respondents in finite setting, you may not be able to guarantee this because patterns of response, if not specific responses, sometimes reveal respondent's identity.

Confidentiality:

No one attached to specific responses, though the aggregate of names or participating groups may be revealed. To insure as much confidentiality as possible, destroy individual records or return them to respondents as soon as raw data has been aggregated.

Informed consent:

Respondent knows s/he is being studied and has agreed to participate, even if responses can be traced back.

Your most important concern: That you offer respondents what you honestly believe you can deliver from among the above choices.

JWK, 1/96 About the Delphi Method: Variation on questionnaire and survey use. Developed at Rand Corporation in early 1950s. Since being declassified in 1960s, had been used to wide range of purposes involving goal projections, strategy determination, group preference assessment, and other issues where group concensus is desirable. Step 1: Identify population of respondents. Little agreement about how many this should be. General criterion:people who are knowledgeable about issues but not in constant contact with one another. Step 2: Determine willingness of all to participate. There will be several rounds of responses required, and participation of most respondents to each roundis necessary. Step 3: Gather individual input on the issue, normally through some open-ended questions which do not precategorize data. Crucial: Guarantee anonymity of each participant so that opinions attached to one or more influential respondents do not influence others. Step 4: Categorize responses into basic categories which do not repeat or overlap. Opportunity for considerable bias here. Several persons doing categorization of unstructured responses, then comparing notes, helps to reduce. Step 5: Each respondent is given assembled group responses, and asked to react privately and individually to the alternatives offered. Reaction may be a simple statement of agreement or disagreement, ranking on a scale, etc. Step 6: Researcher(s) analyze new input, indicating distribution of responses, and return this analysis to respondents.A statistical summary showing distribution of responses is typical; mean or median, standard deviation are good, relatively simple ways of summarizing statistically. (See True, "Descriptive Statistics,"338-348.) Step 7: Each respondent is asked to examine data and, in effect, reassess his/her position relative to group. One alternative is to show distribution of responses, eliminate the least popular, and ask respondents to answer again, using smaller number of response categories. Step 8: Another analysis step, then another iteration. Process continues until concensus is reached or it becomes evident that no concensus is possible. JWK, 1/93 Nominal Group Technique Related to Delphi Method but achieved in group setting where interaction of individuals is assumed. Process: 1. Pose a question to group. 2. Each individual member writes down his or her response(s). 3. Individuals report one response at a time. 4. Group discusses all responses. 5. Group combines all responses, eliminates marginal ones. 6. Individual members pick up to five responses from array, rank them order of preference, and assign values to each. 7. Group accumulates all assigned values. 8. Group discusses these results. In contrast to the highly "individualized" group process of the Delphi technique, this process relies on the dynamics of interaction; insists that every member's concerns are heard by everyone else at least once; provides each member an equal opportunity to influence the entire group; and induces a sense of responsibility among participants to achieving some kind of group concensus. Simple Sharing and Corporate Listening Model developed by the sisters who engaged in conversation with the Vincentian Fathers from time to time. Basic premise: The uncritical sharing of ideas and assumptions The group identification of similarities and differences Each person in turn shares his or her ideas, responses, or proposals with the group. There is no discussion or debate concerning these ideas, although questions of clarification may be asked. A recorder takes notes about individual contributions on a flipchart so that individual may insure that his or her contribution is represented accurately and fairly. After each individual presentation, the recorder facilitates group discussion to identify shared understandings, assumptions, and values; shared identification of differences; and underlying themes which tie individual presentations together. Again, there is no discussion or debate. The premise of this approach is that the honest sharing of ideas and concerns in an uncritical environment leads to common understandings and a sense of shared direction toward common goals.