Definition
Survey research aims to collect information about the preferences, thoughts, and behaviours of people. The main methods of data collection are interviews and questionnaires.
Questionnaires are systematic tools that ask a set of questions and collect written responses from research participants.
Questionnaire may utilize questions that are:
Structured and require the participant to select from a set of given responses
Unstructured and allow the participant to write an original response
Questionnaires are commonly administered through:
Self-administered mail surveys: questionnaires are mailed to participants, participants complete the questionnaire and mail back to researchers
Group-administered questionnaires: participants are gathered to a common location to complete and return the questionnaire to researchers
Online/web surveys: questionnaires are completed over the internet using an interactive forum
Structured response formats:
Dichotomous response: provides two responses ie: yes/no
Nominal response: provides more than two unordered responses ie: blue, pink, yellow, green
Ordinal response: provides more than two ordered responses ie: never, rarely, sometimes, often, always
Interval-level response: allows participants to select a response from a scale where there is no value for zero ie: Likert scale (very dissatisfied, dissatisfied, neutral, satisfied, very satisfied)
Continuous response: allows participants to provide a numeric response ie: age- 37
Characteristics of good survey questions:
Uses clear and understandable language
Gives participants all the necessary information to answer the question
Asks specific questions with one focus at a time
Ie: 1. Do you consume alcohol? 2. Do you consume Marijuana?
NOT: Do you consume alcohol and Marijuana?
Uses appropriate amount of detail
Avoids ambiguity
Ie: How many members live in your household?
NOT: How big is your family?
Promotes understanding by avoiding use of negatives and double negatives
Ie: Should the government increase healthcare funding?
NOT: Should the government not increase healthcare funding?
Avoids biased language
Avoids making presumptions
Avoids imaginary questions
Questions may be sequenced:
Chronologically
Based on increasing difficulty
Grouped by topic
Using contingency questions
Ie: if you answered yes proceed to question 2, if you answered no proceed to question 3
Using open ended questions at the end of the survey
Other considerations for questionnaires:
Keep surveys short and concise to limit time needed to complete them
Keep responses confidential
Send participants a copy of the final results
Thank participations who complete your questionnaire
Pretest the questionnaire before use in the field setting
Statistical Analysis:
Quantitative analysis of data may include the use of descriptive or inferential statistics
Qualitative analysis of data may include use of grounded theory, content analysis, or hermeneutic analysis