Experimentation Intro – Week 11 (Emotion Measurement & Course Logistics)

Course Overview

  • Instructor: Uwe Serdült

    • Affiliation: Digital Governance Systems Lab, Department of Information Science and Engineering

    • Contact: serdult@fc.ritsumei.ac.jp

  • Session: “Introduction to Experimentation – Week 11” (June 18, 2025)

  • Objectives of Week 11

    • Provide context for the remainder of the semester

    • Re-introduce Background and logistics of Exercise 2 (Emotion Measurement)

    • Deliver a concise theoretical primer on measuring emotions that will underpin Exercise 2.

Semester Timeline (Remaining Weeks)

  • Week 11 – Exercise 2: Data Collection (in-class, observation based)

  • Week 12 – Exercise 2: Online, on-demand follow-up

  • Week 13 – Finalize and submit Exercise 2

  • Week 14 – End-of-Term Assessment (culminating evaluation)

    • Implied: Assessment likely covers theory, methodology, and findings from Exercises 1–2.

Background to Exercise 2

  • Theme: Emotion Measurement

  • Emphasis on collective, observational data-gathering at the OIC campus

  • Aligns with course learning goals:

    • Apply experimental and quasi-experimental techniques in a real-world setting

    • Critically evaluate multiple emotion-measurement methods

    • Address construct and content validity concerns when operationalizing complex psychological concepts.

Measuring Emotions

  • Multi-dimensional construct; no single metric suffices.

  • Three broad methodological families introduced:

    1. Self-Reported Measures

    • Strongest validity when tied to recently experienced emotions.

    • Limitations:

      • Differences in awareness, ability, and willingness to articulate internal states.

      • Subject to social desirability and memory biases.

    1. Physiological Measures

    • Autonomic Nervous System (ANS) indicators

      • Focus: skin conductance and cardiovascular activity.

      • Practical issue: Hard to map one physiological change to a single emotional dimension.

    • Startle Response Magnitude

      • Trigger: Sudden, intense stimulus → Neck/back muscle tension or eye-blink reflex.

      • Provides reliable data chiefly for high-arousal, negative stimuli.

      • Cannot differentiate discrete emotions (e.g., anger vs. fear).

    1. Neuro-Cognitive & Behavioral Measures (Mauss & Robinson 2009)

    • Brain States

      • Electroencephalography (EEG) – high temporal, low spatial resolution.

      • Neuroimaging (e.g., fMRI) – high spatial, lower temporal resolution.

    • Behavioral Proxies

      • Vocal characteristics – pitch, loudness, tone.

      • Facial behavior – micro-expressions, action units.

Validity & Correlation Insights (Mauss & Robinson 2009)

  • Cross-method correlations are moderate to low

    • Empirical takeaway: r≈0.2 to 0.4r \approx 0.2 \text{ to } 0.4 is typical.

  • Content validity issue: One indicator captures only a slice of the construct.

  • Construct validity risk: We might inadvertently measure a different phenomenon entirely (e.g., general arousal instead of discrete emotion).

  • Conclusion: Triangulation is essential; employ multiple converging indicators.