Lec 56: Questionnaire Design & Scaling Techniques

Value of the Information & Strength of the Argument

The lecture provides a valuable, structured framework for questionnaire design, emphasizing the critical role of item writing in data quality. The argumentation is logically coherent: it starts with the principles for individual items, then moves to response formats, then to multi-item scales and reliability, culminating in a hands-on example. The instructor makes effective use of concrete examples (e.g., a smartwatch trust scale) to illustrate abstract concepts like double-barreled questions and reverse coding. Practical guidance, such as when to use odd vs. even Likert scales and the trade-offs of ranking, is evidence-based and actionable. The explanation of Cronbach’s alpha and its interpretation is clear, and the distinction between reliability and validity is well articulated. The course is designed for a design-research audience, and the content is delivered in an accessible yet rigorous way.

Scientific Rigor, Source Quality, Title Accuracy

The lecture demonstrates strong scientific rigor in its methodological prescriptions, consistent with established psychometric standards. The instructor correctly identifies common item-writing pitfalls and uses appropriate terminology (e.g., acquiescence bias, latent constructs). The use of Cronbach’s alpha as a measure of internal consistency is standard practice, and the guidance on pilot testing aligns with recommended research protocols. However, the lecture does not cite specific external sources or empirical studies; it relies on general methodological knowledge. The title accurately reflects the content, which strictly adheres to questionnaire design and scaling techniques. No public comments were provided, so no analysis of viewer feedback is possible.

The title accurately represents the content: the lecture covers questionnaire design, item construction, response formats, and scaling techniques.

8/10

Clear pedagogical lecture on questionnaire design and scaling, grounded in established psychometric principles (e.g., construct validity, Cronbach's alpha). Instructor is an academic professional at IIT Guwahati. Content is consistent with standard research methods literature, but no direct citations of specific studies are provided beyond general references. Practical examples and piloting illustration strengthen credibility.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 00:30Introduction: lecture objectives and connection to previous lecture on item writing principles.
• 02:00Four principles for every item: one construct per item, neutral wording, clear and specific, answerable and relevant.
• 05:00Example of a double-barreled question and how to split it into separate items.
• 07:00Discussion of Likert scales: symmetric scale with neutral midpoint, odd vs. even scales, and scale length.
• 09:00Semantic differential and visual analog scales: capturing associations and fine-grained continuous judgments.
• 10:00Ranking as a way to reveal priorities and force trade-offs, with caution against decision fatigue.
• 12:00Introduction to multi-item scales and latent constructs; need for several items to triangulate.
• 14:00Reverse-coded items: purpose in controlling acquiescence bias and detecting careless responders.
• 16:00Cronbach's alpha: interpretation, acceptable threshold of 0.7, and distinction from validity.
• 18:00Worked example: building a smartwatch trust scale, piloting with n=30, alpha from .45 to .82.
• 20:00Key takeaways: measurement error is permanent; format follows construct; reliability precedes validity; pilot and repair.

• NPTEL course page: User Research Methods — Course overview and syllabus for the lecture series.
• Playlist of the course lectures — Full set of videos for the User Research Methods course.

• Cronbach's alpha - Wikipedia — Consistent with the use of alpha >0.7 as acceptable internal consistency.
• Likert scale - Wikipedia — Supports the discussion of symmetric scales and neutral midpoint.