Methodology
This page describes how the prices on RhodyShelf are collected, what they do and do not represent, and where they are known to be wrong. It exists so you can decide what these figures are suitable to be used for before you use them.
What RhodyShelf measures
RhodyShelf records the listed price of cannabis products offered for sale by licensed adult-use dispensaries in Rhode Island, together with the pack size, brand, category and whether the item was in stock at the time of observation.
It does not measure what anyone paid. Listed prices are observed; transactions are not. RhodyShelf has no access to quantities sold, so it cannot weight prices by what people actually buy. Any average or middle value taken across these listings is a summary of shelf prices, not a consumer price index, and must not be presented as one.
It also does not measure quality. There is no rating, no review, no panel and no laboratory testing of our own. Where potency is shown it is the figure the dispensary itself published.
Coverage
All 9 of 9 licensed adult-use dispensaries operating in Rhode Island are read. This is a census, not a sample: there is no sampling frame, no selection of shops, and therefore no sampling bias. Most published price indices are samples; this one is not, and that is the strongest thing that can be said for it.
Coverage of shops is complete. Coverage of transactions is not, for the reason given above. A typical day holds roughly 4,400 in-stock listings across those 9 shops; the exact count moves as menus change and is shown on each shop's own page.
Medical-only dispensaries, delivery services and unlicensed sellers are out of scope entirely.
Collection schedule
Menus are read twice a day, at approximately 09:15 and 18:15 America/New_York. Both runs cover every shop.
A listing that has not been seen in the most recent 24 hours is removed from the site rather than shown with an old price. A product that sells out disappears; it does not linger at a stale price.
The site itself trails the underlying record by roughly 30 to 60 minutes after each run, because pages are regenerated on their own schedule rather than the moment new prices land.
How a listing becomes a figure on this site
Every price shown is a price a dispensary published. Nothing is estimated, interpolated or filled in. Where a figure cannot be computed honestly it is left blank instead of guessed — see the next section.
Markdowns are derived by comparing a listing's current price to its own stated original price. The discount fields published alongside them are not used: they carry an error rate of roughly 0.1%, which was enough to once headline a $1 saving as “100% off”. Two prices and a subtraction are more reliable than a field that claims to have already done the subtraction.
Where we decline to compute a figure
The single most useful thing to know about this data is where it stops being trustworthy, so those places are listed rather than papered over.
Price per gram is refused for some rows. The stored weight field carries more than one unit depending on the product, and two of those units render identically. Where the unit cannot be resolved with confidence, no unit price is shown. A blank is correct; a confidently wrong price per gram is worse than none.
Edible dose sizes are not comparable across shops. For some edibles the stored figure is total THC in milligrams; for others it is the net weight of the confection itself. Both print as “mg”. Because the two cannot be told apart reliably, RhodyShelf does not publish price-per-dose comparisons for edibles at all.
Pack sizes are normalised before comparison. Menus write an eighth of an ounce variously as 3.5 g and 3.54 g; a quarter as 7 g and 7.09 g. These are the same retail unit and are grouped together. Sizes that do not fall close to a standard retail unit are excluded from size comparisons rather than rounded into one.
Potency figures are filtered, not corrected. Implausible values are dropped. They are never adjusted toward what they “should” be.
Matching the same product across shops
Dispensaries do not share product identifiers, so the same item at three shops usually arrives as three unrelated records. RhodyShelf matches on brand, normalised product name, pack size and category.
This links roughly 12% of the catalog across shops. Matching on the published identifier alone would link about 4%. Both numbers are low, and the honest reading is that most products cannot be price-compared across shops at all. Where a comparison is not shown, it is because no confident match was found — not because the product is only sold in one place.
Strengths and limitations
Strengths. Every licensed adult-use shop in the state is covered, so there is no sampling bias. Observations are twice daily, so a figure is at most about half a day old. Every price is one a dispensary published rather than one RhodyShelf inferred.
Listed, not transacted. These are shelf prices. They exclude tax, exclude loyalty pricing and bundle offers that do not appear on the menu, and are not weighted by sales volume.
Between-scan drift. A price can change, and a product can sell out, minutes after it is observed. A figure here is a statement about a moment, not about the day.
Dead product links. Roughly 23% of product pages return a “not found” response over time as items churn off menus. This is expected in a catalog with this much turnover, and those pages say so plainly rather than pretending the product still exists.
Menu errors pass through. When a dispensary publishes a wrong price, RhodyShelf shows the wrong price. There is no independent verification step, and no claim is made that a menu is correct — only that this is what it said.
No price history is published. Figures are point-in-time observations. There is no back-revised series. A citation from this site is only meaningful with its observation time attached.
Revisions and corrections
Scheduled. The catalog is re-read twice daily and largely rewritten each time. Prices are point-in-time observations and earlier observations are not restated.
Unscheduled. If a figure is wrong, write to rhodyshelf@gmail.com. Corrections that change a published figure are noted in the table below rather than made silently.
How to cite this data
Cite the specific page you read the figure from, together with the observation time printed on it. A figure without its as-of time is not a claim this site supports — an unqualified “flower costs $X in Rhode Island” is not something RhodyShelf has said.
RhodyShelf, “Flower prices at Reef Wellness,” https://rhodyshelf.com/dispensary/reef-wellness/flower, observed 20 September 2026.
Changes to this methodology
| Date | Change | Why |
|---|---|---|
| 2026-09-20 | First published. | Collection rules existed in code but had never been stated. |
| 2026-09-20 | Edible price-per-dose comparisons withdrawn. | The stored figure means total THC on some rows and net product weight on others, and both print as “mg”. |
Questions
Corrections, questions about these rules, or a licensed dispensary that should be covered: rhodyshelf@gmail.com. More about the project on the about page.