Method at a glance
- Modality: semantic self-report (conceptual, not perceptual)
- Measures: How strongly a given property (e.g. an interoceptive sensation like ‘heart racing’, a behavior like ‘clenched fists’, or a situation like ‘insulted’) is associated with a concept (e.g. an emotion category like anger) — revealing the content and structure of a person’s mental representation of that category, not any perceptual ability or bodily state.
- Procedure: Adapted from cognitive-psychology property-verification tasks (Kan et al. 2003; Pecher et al. 2004). On each timed trial a category–property pair is shown (e.g. ‘How much does HOT come to mind when you think about ANGER?’) and rated on a Likert scale (0 = did not come to mind at all → 9 = immediately came to mind). Properties are pre-validated in a pilot so that each is strongly and uniquely associated with one category and belongs to one modality. Trials are randomized, attention-check foils inserted, and reaction time recorded (to exclude over-deliberated trials and covary cognitive speed). Ratings are multilevel — many category–property items nested within person and within item — and analysed with cross-classified mixed models.
- Also called: property verification task, property generation/rating task, emotion property association
Strengths
- Isolates the modality of a representation. By crossing interoceptive vs behavioral vs situational properties against the same emotion categories, it can ask whether a manipulation (e.g. age) moves the interoceptive content of emotion concepts specifically, holding the categories fixed — the comparison that lets MacCormack et al. (2021) claim an interoceptive-specific effect rather than a general one.
- Cheap, scalable, and semantic. Runs online at scale; measures conceptual knowledge directly rather than inferring it from a perceptual task, which suits constructionist questions about the content of emotion categories (Situated conceptualization (concepts as predictions)).
- Timed and instruction-constrained to push participants toward automatic semantic association and away from effortful reasoning — the ratings are meant to reflect what comes to mind, not what one endorses on reflection.
Limitations
- It measures conceptual knowledge, not perception or physiology. A property-association rating says how strongly ‘racing heart’ is linked to ‘anger’ in the mind — not whether the person’s heart races, nor whether they can perceive it. It is [[interoceptive-sensitivity|interoceptive knowledge]], not accuracy or sensibility, and results must not be read as an interoceptive ability. This is the key caveat on MacCormack et al. (2021).
- Emotion foils are unreliable, so no true/false verification. Unlike classic property-verification (GRASS–red is reliably false), emotion category–property pairs have no consensus false foils (ANGER–cold may be valid across people/cultures), so only attention-check foils can be used and the task cannot score accuracy against a ground truth.
- Cannot separate individual experience from cultural stereotype. A rating can reflect either the participant’s own emotional experience or culturally learned beliefs about what emotions ‘should’ feel like — the recall/summary-judgment problem constructionists flag (Robinson & Clore 2002), and the reason MacCormack et al. paired it with an experience-sampling study.
- Item-set dependent and power-asymmetric. Results depend on which properties survive pilot validation, and unequal numbers of items per modality can bias power toward detecting effects in the better-populated modality (interoceptive, in the 2021 study).
Connections
- Key studies: MacCormack et al. (2021)
- Related methods: Emotion-word rating norms (multidimensional semantic norming), Ecological Momentary Assessment (EMA)
A semantic self-report method imported into emotion research from cognitive psychology’s property-verification paradigm (Kan et al. 2003; Pecher et al. 2004), where participants rate how much a property (red) relates to a concept (apple) to probe the structure of their category representations. The wiki meets it through MacCormack et al. (2021), who used it (following Oosterwijk et al. 2012, 2015) to ask how much interoceptive, behavioral, and situational properties come to mind for five negative emotion categories, and whether that changes with age.
What it is for, and what it is not
The task’s value is that it reads the content of an emotion concept directly. If you want to know whether “the body” is part of what someone’s concept of anger contains — as opposed to whether they can feel their body, or what their body actually does — this is the instrument. That makes it a natural fit for constructionist questions, where an emotion is a categorization built partly from stored interoceptive knowledge (Situated conceptualization (concepts as predictions)): the property-association task measures that stored knowledge.
The corresponding limit is the one to keep foregrounded everywhere it appears: a high “heart racing → anger” rating is a fact about a concept, not about a heart or its perception. On the wiki’s Interoceptive taxonomy (Farb et al. Box 1), this is closest to interoceptive knowledge (what one knows about emotion-related bodily sensations) — a construct distinct from the accuracy the Heartbeat detection task measures and the sensibility the Multidimensional Assessment of Interoceptive Awareness (MAIA) measures. Reading a property-association result as evidence about interoceptive ability is the standing error the method invites, and the reason MacCormack et al. ran a second study by a completely different method (the Day Reconstruction Method) to check whether the conceptual pattern tracked lived experience.
Where it sits among the wiki’s methods
It is the conceptual/semantic pole of the emotion-measurement spectrum, adjacent to Emotion-word rating norms (multidimensional semantic norming) (both norm the content of emotion words rather than inducing a state) but trial-based and modality-crossed rather than a static per-word norm. It trades away the ecological validity of Ecological Momentary Assessment (EMA) and the physiological ground-truth of lab induction (Affective picture viewing) in exchange for a clean read on what an emotion category contains and the ability to isolate one modality of that content by comparison — the trade that made the interoceptive-specificity claim in MacCormack et al. (2021) possible.