Chapter 7 - Research Methods

Purpose of surveys

Gather many types of information.

Survey samples

May be randomized, assigned, stratified, open, well-defined, ill-defined, or convenience samples.

Survey challenges

Respondent reliability, interpretation, bias, nonresponse, and extrapolation can affect findings.

Survey methods

Mail, telephone, internet, and interviews.

Survey planning

Define specific goals, objectives, or aims before developing research questions and hypotheses.

Improving response rates

Use a well-designed, short instrument; consider an inducement and follow-up.

Email survey design

Begin with interesting items, limit length, and send notices or reminders.

Survey introduction

Make the introduction clear, informative, appropriately worded, and carefully proofread.

Survey follow-up

Reminders may increase response rates; follow-up must protect confidentiality.

Web survey advantages

Can reach large, dispersed samples, collect substantial data, and provide downloadable results.

Web survey limitations

Low response rates, item misinterpretation, and lack of internet access can be challenges.

Phone surveys

Offer timely personal contact while allowing anonymity, but require phone access.

Personal interview surveys

Allow deeper exploration and qualitative or quantitative data, but interviewers may influence respondents.

Survey item construction

Each answerable item should help address the research question or hypothesis.

Demographic data

Information that describes the sample.

Demographic item design

Use specific, relevant, understandable questions with nonoverlapping ranges.

Survey instrument quality

Items should be clear and answerable; the instrument should be reliable and valid.

Developing survey items

Write initial ideas, then edit, define, delete, combine, and categorize them.

Structured items

Allow the investigator to control responses.

Unstructured items

Can be difficult to quantify.

Ranking items

Specify ranks, provide clear instructions, and use a consistent format.

Likert scales

Gather information about opinions.

Verbal frequency scales

Response choices should capture needed responses and balance positive and negative options.

Survey assembly

Place the introduction first; organize the survey clearly and proofread it.

Pilot testing

Tests whether the survey is understandable; a large group is unnecessary.

Pilot-testing review

Check length, clarity, directions, needed data, bias, appearance, and suggested improvements.

Pilot-testing administration

Obtain warranted IRB approval and keep good records.

Small populations

Consider surveying the entire population.

Convenience samples

May not represent the population; a large sample does not guarantee accuracy.

Confidence level and interval

Common confidence levels are 95% or 99%; the interval expresses a margin of error.

Statistical analysis planning

Choose analyses during planning, consult a statistician, and understand the tests used.