Investigación SEP

1. What is a Delphi method?, What are its characteristics?

It was created in Santa Monica at 1950´s

• Design of the cuestinoary, to colect information for your  research, the information need to be gived by an expert un that topic.
• This methodology consist in doing surveys to experts on the topic, so the coordinators can process the information and get conclusions
• The feed back its very important.
• This methodolgy is repetitive or iterative.
• It is important to the method to mantain it annonymus.
• The obective of this is to do a consensus of the topic.

Preparation

• It’s fundamental to choose the number for the experts panel and the criteria needed to include or exclude people from the panel.
• Some criteria are:

• Profession and Charge
• Experience
• Link with the topic
• Specific capacitation type.

• More elevated criteria rea:

• Research quantity
• Publications

• The minimum number of experts is seven, and adding more experts diminishes the information error; however, it’s not recommendable to choose more than 30 experts because of increases in costs.
• It’s recommended to begin with a greater number than wished since people may leave during all the sessions; it’s possible to include new experts in other rounds.
• Steps for selecting panel members:

• Identify the most appropriate categories of experts or stakeholder groups for the panel.
• Populate the stakeholder groups with names derived from: previous research participation, publications on topic, professional email lists, professional organization board and advisory panel involvement.
• Contact individuals
• Create sub-lists for each stakeholder groups: Rank experts based upon criteria of representation of professional role/state or territory/specialty practice area/public or private health service.
• Invite experts according to their ranking for each stakeholder group. Stop soliciting experts when group size is reached or total population invited.

Instrument (Questionnaire) Preparation

• Its characteristics depend on the research objective.

• It can be an existing one
• It can be constructed with the participation of the experts when there has not been much research about it.

• “Classic Delphi” starts by choosing the questionnaire in consult with the experts by individual interviews.

• The first round must start with open questions to avoid bias in the study.

• If the questionnaire is selected first, the expert selection must be the second activity.
• The questionnaire must have precise, independent and quantifiable attributes: this determines the type of statistical analysis and feedback.
• There are four kinds of questions:

• Yes-No; Agree or Disagree
• Ordinal Scale Question
• Evaluate in an Interval
• Open Questions (Comments)

• It’s common to use a mixture of questions in a questionnaire.

Consulting way decision

• The communication with the experts may be used with e-mail, fax or conventional mail. The direct deliver may be possible when the panel and their fearless allow it.
• There are online tools that may help develop the Delphi Method:

• SurveyMonkey

Consult

1. The first step is crucial: you need to set the number of rounds to be made. you need to let the experts know how many there will be.
2. Another crucial point is that every expert opinion must be secret.
3. Another important thing to take into consideration is if the members who have not replied to the first round are to be invited to participate again. (you have to give them a chance to reply within two or three reminders)
4. Experts need to understand a minimum of basic statistics.
5. The most typical number of rounds goes from two to three (but there can be more, depending on the project). There should not be many rounds, but also not so little.
6. The second part of the first step is to make contact with the experts. You present the objectives of the Delphi project, and also the conditions of the questionnaire development. You need to also give precision about the topic to evaluate.
7. The second part is when the experts receive the questionnaire solved with statistical methods from the first round. you also give them a new questionnaire if the first was edited.
8. The objective of making more rounds is to decrease dispersion and increase convergence of the opinions and delimit consensus opinion.
9. Feedback may always be qualitative.  

Consensus

Construction of the consensus

Objective: General Agreement

Level of agreement: 75%-85%

(ethical situations 100%)

The most important thing in this method is the final report.

Results report

wide and precisely in the description of the process

• How the tasks are going to be executed
• Present the statistical results in tables and graphics
• Honesty and exposed the difficulties
• Final result of the panel of experts (description of each question)

Application

We saw how the previous information is applied in a real case in Havana Cuba.

...