Survey Research Methods

• Survey research is a form of inquiry that assumes meaningful information is obtainable by asking parties of interest what they know, believe, and how they behave. Balthazar & Vendrely (2022)
• Surveys can be used in descriptive to experimental studies as a way to collect data.
• Completing questionnaires to gather information about participants, such as client satisfaction, is an example of survey research.
• Evaluating and designing the survey involves the same issues, regardless of the type of study.
• Survey research collects information through broad sampling strategies and usually targets self-reported knowledge from a large number of participants.
• There are low levels of control over sample characteristics and administration.
• It is useful when knowledge exists outside of scholarly literature and discourse.
• Census is a comprehensive survey method.
• Marketing surveys and polls are examples of survey research.
• Surveys and questionnaires involve written responses and can be self-administered (often online) or conducted as interviews (face-to-face or telephone).
• Surveys are self-reports, not directly observed, using standardized procedures.
• Surveys gather information on facts, knowledge, behaviours, opinions, and personal characteristics.

Advantages

• Surveys are efficient, allow for large samples, and are valid for self-observations like attitudes.
• Anonymity in surveys encourages honesty.

Disadvantages

• Response rates are often low.
• Self-selection can lead to unrepresentative responders.
• Misinterpretation of questions can occur.
• Respondents may lack motivation for accuracy.
• There can be potential recall/memory bias in self-reports, especially with past or sensitive information.
• Established questionnaires have often been tested to mitigate these issues.

Survey Instrument Choice

• Consider using an existing instrument after reviewing its psychometric properties, applicability, adaptability, and potential impact of adaptations on validity.
• Designing a survey instrument is easy to do, but not easy to do well.
• Wording and ordering of questions affect responses.
• Researchers should provide information on the development, reliability, and validity of surveys.
• Demographic questions determine who is answering the survey which aids interpretation, allows grouping of responses, and generalization.
• The survey also needs content questions which are what you want to know/document.

Steps for Designing a Survey Instrument:

• Draft
• Expert review
• Revise
• Pilot test
• Revise
• Repeat steps as necessary

Open-Format Survey Items:

• Responses are subjective and vary, making them difficult to code and analyze.
• They allow for uninterpretable or incomplete answers.
• Open-Format surveys are effective for interviews with follow-ups, but not as useful for broad surveys.

Closed-Format Survey Items:

• Provide quantitative data, with each question requiring a single response.
• Questions should be exhaustive, including options like "don't know," "not applicable," or "other"
• Questions should be mutually exclusive in most cases.
• Respondents may need instruction to choose the "best" or "most important" answer.
• Some closed-format questions allow selecting all applicable options or a specific number of items.
• The scope of the response may need to be defined, such as identifying a particular period or time-frame.
• Respondents may be asked to calculate an average or estimate typical use.

Scales Used in Surveys:

• Categorical which is Nominal
• Scales with characteristics on a continuum, such as ordinal scales or continuous data (interval or ratio).
• Types of scales include Likert, Likert-like, and visual analogue scales.

Likert Scale

• This scale is often used for attitudes or values.
• The original scale has 5 choices representing levels of agreement.
• An example question with answers: How important are grades to you? Not important - Somewhat Important - Neutral - Important - Very Important
• Respondents without strong feelings can choose a neutral option.

Likert-Like Scales

• Likert-like scales modify the original Likert format and may omit the neutral option to force a response.
• Options include different numbers of choices (e.g., 7 rather than 5) and may use anchors.
• Many treat these as interval which need to judge if reasonable
• Central tendency and variation are measured and stats are used

Semantic Differential or Visual Analog Scale

• It is a simple method to measure the intensity of a one-dimensional characteristic.
• This scale uses adjective pairs representing two ends of a continuum with a line in between (e.g., organized-disorganized).
• Respondents mark on the line, measure in mm
• It's a simple method to measure the intensity of a one-dimensional characteristic.

Q-Sort Technique

• Used for forced rank ordering of alternatives.
• Participants sort cards with items relevant to the research question into levels.
• The distribution of items is preset for each level (e.g., critical, desirable, optional, unimportant).

Administering Surveys Includes:

• Identify a population of interest and use probability sampling, stratified or cluster sampling, or convenience samples.
• Mailing lists and email lists are often used.
• Restrictions to particular geographic areas may be appropriate.
• A cover letter introduces the survey, provides instructions, and addresses ethical issues such as anonymity and implicit consent.
• Sending a follow-up reminder increases response rates, often via email to all recipients, not just those who haven't responded.

Analyzing Survey Responses Involves:

• Collating responses and calculating frequencies (absolute and relative) of the participants who select the response.
• Central tendency and variability, such as mean ratings and standard deviation, need to be calculated.
• Referencing the sample size is very important when analyzing survey results.
• Additional descriptors entail correlations and multiple regressions.
• If group participants based on characteristics (e.g., demographics) can compare responses

Explore survey research methods, a form of inquiry that gathers information by asking individuals about their knowledge, beliefs, and behaviors. Surveys are versatile tools for collecting data in various studies. Understand their application, evaluation, and design for effective data collection.