Observations

Observations

1. Definition

   • Systematic technique for documenting behaviors, events, and settings as they occur in their natural context. It may be structured, using predefined categories, checklists, or unstructured, relying on rich, exploratory field notes. In both cases, the goal is to reduce observer influence and capture behavior as authentically and reliably as possible.

2. Planning

   • Purpose & questions: What behavior or process do you need to understand?
   • Unit of analysis: individuals, groups, events, spaces.
   • Design: Participant vs. non-participant. Structured vs. unstructured. Covert vs. overt. Naturalistic vs. simulated
   • Sampling: Time sampling (fixed intervals). Event sampling (every occurrence of X). Subject sampling (predefined sub-groups)
   • Instruments: Checklist with observable indicators (yes/no, like). Category guide with operational definitions. Field-note sheets (context, verbatim quotes, critical incidents)
   • Observer: Calibration with examples. Manual with operational definitions. Pilot to refine categories
   • Ethics: Informed consent as applicable. Confidentiality. Avoid unnecessary recording

3. Conducting an observation

   • Procedure: Arrive early, choose vantage points, blend into setting. Use short codes for real time notation. Record context, participants, materials
   • Recording: Time sampling: mark presence per interval. Event sampling: count specific events. Thick description: separate objective description from interpretation
   • Quality control: Double observer subsample to estimate inter-rater reliability. Frequent short pauses to tidy notes. Daily checks for legibility
   • Wellbeing & safety: Do not disrupt routines; intervene only for risk per protocol

4. Issues

   • Reactivity: behavior shifts when observed. Mitigate: familiarization periods; consistent observer presence
   • Observer bias: expectations, halo, primacy and recency. Mitigate: clear rubrics, training, blinding where possible
   • Reliability: inter-rater (Cohen’s kappa, ICC)
   • Validity: Content (coverage of construct). Ecological (real-world authenticity)
   • Ethics: permissions, secure storage, minimal identifiers
   • Practical constraints: time, access, sight lines, noise, crowding

5. Data analysis

   • Quantitative: Frequencies, proportions, means/SD. Pre–post comparisons. Inter-rater reliability (kappa/ICC). Visuals: bar charts, run charts, time series
   • Qualitative: Open: axial coding to themes. Matrices (behavior, context). Triangulation with other sources
   • Mixed methods: Integrate counts (how often) with narratives (how/why) Joint displays to align quantitative indicators with themes

6. Uses

   • Education and programs: monitor routines, interactions, safety practices, evaluate interventions and events
   • Research: understand processes in natural settings
   • Quality improvement: feedback for stakeholders, inform changes to space, materials, or procedures