Chapter 6: Research Methods: Surveys and Observations

Chapter 6: Research Methods: Surveys and Observations

Main Takeaway

  • The document outlines key principles for designing effective surveys and conducting observational research, focusing on:

    • Question formats,

    • Common pitfalls such as respondent biases (acquiescence, social desirability, response sets),

    • Challenges in observation (observer bias, observer effects, reactivity),

    • Strategies to maintain construct validity,

    • Ethical guidelines for both methods.

1. Surveys and Polls

1.1. Definitions

  • Survey/Poll:

    • A method of posing questions to people through various means, such as:

    • Telephone,

    • Personal interviews,

    • Written questionnaires,

    • The Internet.

1.2. Question Formats

  • Open-ended question:

    • Allows respondents to answer in any way they wish, encouraging rich detail in responses.

  • Forced-choice question:

    • Requires respondents to select the best option from two or more given choices.

  • Likert scale:

    • A rating scale providing multiple response options, typically anchored by specific terms such as:

    • Strongly agree, agree, neither agree nor disagree, disagree, strongly disagree.

  • Likert-type scale:

    • A rating scale that does not conform strictly to the traditional Likert scale format.

  • Semantic differential format:

    • A response scale where numbers are anchored by contrasting adjectives (e.g., good to bad).

1.3. Problematic Question Types (Threats to Construct Validity)

  • Leading question:

    • Wording that encourages respondents to favor one specific response over others.

  • Double-barreled question:

    • A single question that asks two different things simultaneously, resulting in confusion for the respondents.

  • Negatively worded question:

    • A question containing negative phrasing, making it complicated or confusing for respondents.

1.4. Respondent Shortcuts & Non-Differentiation (Response Sets)

  • Response set (Nondifferentiation):

    • A shortcut used by respondents in lengthy surveys where they do not engage with each specific item.

  • Acquiescence (Yea-saying):

    • The tendency of respondents to answer