The six items and what each one is
Every item is written as a separate indicator with its own meaning. None of them is a proxy for another, and there is no construct that all six are supposed to measure together.
- 1 · Perceived interaction knowledge
- Whether the respondent believes cannabis products may interact with medicines. This is a perception item. It does not test factual knowledge, and a confident wrong answer and a confident right answer are indistinguishable here.
- 2 · Perceived disclosure rationale
- Whether the respondent understands why product, amount, route, frequency, and timing might matter in a clinical conversation. Again a perception, not a comprehension test.
- 3 · Discussion comfort
- Anticipated affect. Comfort reported in a browser is not comfort experienced in an appointment.
- 4 · Anticipated clinician response
- An expectation about someone else's behavior. It measures the respondent's prediction, not what any clinician would actually do.
- 5 · Conversation preparedness
- Self-assessed readiness. The item most exposed to response shift, because learning what a good disclosure contains can lower a person's rating of their own readiness.
- 6 · Discussion intention
- A stated intention. The intention–behavior gap is the standard caution here: an intention recorded on a form is not an action taken in a clinic, and this item is also the most exposed to social-desirability pressure.
Response format, and why nothing is added up
Each item offers four ordered labels — Disagree, Somewhat disagree, Somewhat agree, Agree — plus an explicit “Not sure / prefer not to answer”, which is stored as a missing value rather than a midpoint. There is no neutral middle option, so the opt-out is not doing double duty as “neither”.
The four labels are ordinal: they have an order but no established spacing. Storing them as 0–3 is a convenience for validation, not a measurement claim, and the tool never adds, averages, or differences those integers. Analysing ordinal responses with metric models is a documented source of confident error, including inflated and deflated effects and outright sign reversals in which the analysis reports the opposite of the pattern in the data (Liddell & Kruschke, 2018). The export declares ordinal: true and noCompositeScore: true so that a downstream reader inherits the constraint rather than discovering it.
What pre/post pairing supports
Pre/post linkage permits only descriptive item-label transitions for exact pairs. It does not estimate effectiveness or causation and does not control for exposure, timing, attrition, response shift, testing effects, or social-desirability bias.
The threats are specific and each has a literature:
- Response shift — the respondent's internal standard for the question changes between measurements, so the two answers are not on the same yardstick (Howard & Dailey, 1979; Sprangers & Schwartz, 1999).
- Testing effects — completing the questionnaire is itself an exposure that can change the second response.
- Social-desirability bias — sensitive-topic items pull toward the answer the respondent believes is wanted (Krumpal, 2013).
- Regression to the mean — extreme first responses tend to be followed by less extreme ones for purely statistical reasons (Barnett et al., 2004).
- Attrition — codes with a PRE and no POST are visible in this tool, but their absence is not random and is not correctable here.
A fuller treatment, with a worked example of reading a single pair, is on the design rationale page.
Clinical framing
Cannabis interactions vary by cannabinoid, product, amount, route, frequency, timing, concomitant medicine, and person. Discussion may help a clinician assess interactions, adverse effects, sedation or impairment, and care context; this site gives no clinical advice, and nothing here should delay or replace a conversation with a qualified professional.
Authoritative background
Methodological references
- Howard, G. S., & Dailey, P. R. (1979). Response-shift bias: A source of contamination of self-report measures. Journal of Applied Psychology. doi:10.1037/0021-9010.64.2.144
- Sprangers, M. A. G., & Schwartz, C. E. (1999). Integrating response shift into health-related quality of life research: a theoretical model. Social Science & Medicine. doi:10.1016/S0277-9536(99)00045-3
- Krumpal, I. (2013). Determinants of social desirability bias in sensitive surveys: a literature review. Quality & Quantity. doi:10.1007/s11135-011-9640-9
- Barnett, A. G., van der Pols, J. C., & Dobson, A. J. (2004). Regression to the mean: what it is and how to deal with it. International Journal of Epidemiology. doi:10.1093/ije/dyh299
- Liddell, T. M., & Kruschke, J. K. (2018). Analyzing ordinal data with metric models: What could possibly go wrong? Journal of Experimental Social Psychology. doi:10.1016/j.jesp.2018.08.009
- Regulation (EU) 2016/679 (General Data Protection Regulation), Article 4(5), definition of pseudonymisation. Official Journal text
Sources and wording reviewed 29 August 2026.