SURVEYS at Vincennes University
Survey Resources
The following links provide additional information regarding survey design.
-
Internet, Phone, Mail, and Mixed-Mode Surveys: The Tailored Design Method (2014), 4th Edition, Don A. Dillman, Jolene D. Smyth, and Leah Melani Christian
-
Survey Research Methods (Applied Social Research Methods), 5th Edition, Floyd J. Fowler
-
Margin of Error Calculator
-
Sample Size Calculators
Survey Terminology
-
Aggregate data – Data combining all information into a single dataset for analysis, e.g. all responses
to a given question in a survey are summarized.
-
Anonymity – Information that is provided by a respondent without means to identify the individual.
Differs from confidentiality in that survey responses cannot be tracked back to a specific individual.
-
Closed-ended question - Respondent chooses from a list of options provided; may give single choice or multiple-choice
options.
-
Completion rate – The number of respondents who successfully completed a given survey.
-
Confidentiality – Protocols established to safeguard and maintain the privacy of any personally-identifiable
information. Differs from anonymity in that survey responses may be kept confidential, e.g. aggregated or limited sharing
publicly, but not anonymous, i.e. researcher has access to information provided by
the respondent.
-
Disaggregate data – Data that separates information into smaller groups for analysis, e.g. gender or
employment status.
-
Demographic questions – Questions designed to provide more information about the respondent, e.g. gender,
age, or race/ethnicity.
-
Double-barreled question – Question that contains two or more thoughts (parts), but requires a single response
thereby creating confusion and increasing the likelihood of skewed analysis.
-
Leading question – A question that guides the respondent to a desired answer, which introduces bias
into the survey results and compromises the validity of the analysis.
-
Likert-type scale – A type of scale that offers multiple levels of agreement or disagreement, satisfaction
or dissatisfaction, etc.; typically for a series of statements about a given topic.
-
Loaded question – Question worded in such a way that the respondent has no option but to agree with
the implied answer sought. This term is similar to a leading question in that bias is introduced into the survey results.
-
Margin of error – Degree of error (uncertainty) in survey results received when using random sampling.
A lower margin of error generally indicates a greater likelihood that results are
representative of the population. See Margin of Error Calculator.
-
Multivariate analysis – A means of analysis by which two or more variables are analyzed together, e.g.
gender and age.
-
Non-Probability sampling - Individuals are selected based on non-random characteristics or from a specific
population.
-
Open-ended question - Respondent uses their own words to answer the question. May be a single word, short
phrase, or longer response.
-
Probability sampling - Sometimes referred to as random sampling whereby a random selection of the population
is selected.
-
Qualitative research – Research that uses non-numerical data, i.e. text responses to analyze respondents’
opinions about a specific topic or subject matter.
-
Quantitative research – Research that uses numerical data to analyze responses in a survey, e.g. counts
and/or percentages of a given response.
-
Random sampling - Sometimes referred to as probability sampling whereby a random selection of the
population is selected.
-
Reliability – How easily a survey can be replicated and still achieve the same results thus increasing
the credibility of the results by means of a consistent process.
-
Response rate – The number of individuals who complete a survey divided by the number of people
invited to take the survey.
-
Sample - Selection of a representative portion of a larger population, which is designed
to determine a fair analysis of the characteristics and/or opinions of the whole population.
-
Sample size – The number of individuals included in a survey or study. It is important to find
the appropriate sample size for a survey in order to obtain the most accurate results.
A Sample Size Calculator is useful for determining the appropriate sample size needed.
-
Statistical significance – The likelihood that responses from two or more variables (or groups) are not due
to chance or coincidence.
-
Univariate analysis - A means of analysis by which only one variable at a time is analyzed.
-
Validity – Assesses how well the questions in a survey are measuring what they intended to
measure, thus providing an accurate representation of an individual’s beliefs, opinions,
etc.
