About This Site

How We Research and Fact Check

Every guide here is built from the same short list of sources in the same order, and every number on the page has to trace back to one of them.

This page is the long version of the accuracy section in our editorial policy. It exists so you can see how a guide here gets made: which sources sit above which, what has to happen before a number is allowed onto a page, and what we do when the honest answer is that nobody knows.

None of it is exotic. Feeding wild birds is not a research frontier: most of what you need to know has already been studied and published by a handful of institutions that have been at it for decades. The work is reading them carefully, resolving where they disagree, and refusing to fill the gaps with confident-sounding padding.

Guides are researched, written and reviewed by the ProBirdFeeder editorial team. As the about page says, we do not invent expert personas or attach stock-photo bylines, which means the only thing you can hold us to is the method.

The source hierarchy#

Every factual claim gets sorted into one of four tiers before it is written. The tier does not just decide where the link goes at the bottom of the page. It decides how strongly the sentence is allowed to be phrased.

Tier 1: the published literature

Peer-reviewed ornithology, plus the Birds of the World accounts that pull it together species by species. This is where measurements, weights, ranges, diet and breeding biology come from on the bird profiles. When a Tier 1 source gives a number, the page gives that number.

Tier 2: the institutions that study feeder birds

This is where most of the site lives, because backyard feeding is exactly what these organizations document. Each is used for what it is genuinely authoritative on, not as a generic citation to sprinkle around:

• Cornell Lab of Ornithology, All About Birds: identification, behavior and general feeding guidance.
• Project FeederWatch, run jointly by the Cornell Lab and Birds Canada: which species show up at feeders, where, when, and on what food. Its counts underpin the bird food chart.
• NestWatch: box dimensions, entrance holes and siting, which is what nest boxes is built on.
• Audubon and Birds Canada: habitat, native plants and regional context.
• USGS National Wildlife Health Center: wildlife disease, including salmonella and house finch eye disease.
• USDA APHIS: avian influenza status and what it means for a backyard feeder.
• US Fish and Wildlife Service: the legal questions, including what the Migratory Bird Treaty Act covers.
• American Bird Conservancy: window collisions and outdoor cats.

Tier 3: state, provincial and extension sources

Many reader questions have no continental answer. Whether you may feed deer, whether bear rules close feeders in your county, whether an ordinance limits what you can hang: that is set locally. Tier 3 is what the regional guides and neighbors and ordinances lean on. A Tier 3 claim is always tied to the place it applies to, and you are told to check your own agency, because we cannot track every jurisdiction in North America and will not pretend to.

Tier 4: practitioner consensus, labeled as such

Some of the most useful advice here has never been the subject of a study, because nobody funds research into how far a tube feeder should sit from a shrub. What exists instead is decades of experience that thousands of people independently report the same way. That is worth something, and it is not evidence. When a page rests on it, the sentence says so, and the glossary marks terms that are practitioner vocabulary rather than scientific ones.

How a guide gets built#

Every page follows the same sequence. It is deliberately boring.

1. Scope the question from what people actually ask. Not a keyword list, but the real form of it: "why did the birds stop coming", "how far from the window", "is this seed moldy".
2. Gather sources, working down the tiers, then look honestly at whether the remaining gaps are consensus or genuine unknowns.
3. Draft against those sources with the numbers in place from the start. Vague drafts hide the spots where nobody knows anything.
4. Resolve conflicts. Higher tier wins. If two sources at the same tier disagree, the page reports the disagreement or gives the range they both fall inside.
5. Check every number against a named source. Each measurement, dilution, dimension and date is read back against the document it came from.
6. Check internal consistency against the rest of the site. Any figure that appears elsewhere is compared with those pages, and a mismatch is resolved before publication.
7. Review against the welfare rule: does any instruction risk harming a bird, and does the page say so where the bird's interests and your convenience pull apart.

Numbers that have to match on every page#

A reference site fails in a specific, embarrassing way: two of its pages answer the same question differently, and you have no way to tell which one to trust. A few figures here turn up in many places, so they are fixed values rather than something each page decides for itself.

• Nectar ratio: 1 part plain white sugar to 4 parts water, by volume. Identical on the nectar recipe, the nectar calculator and every hummingbird page.
• The 5-7-9 squirrel placement rule: at least 5 feet (1.5 m) above the ground, 7 feet (2 m) horizontally from a launch point, 9 feet (2.7 m) below an overhead branch. Same numbers on squirrels and on baffles.
• Window marker spacing: no more than about 2 inches (5 cm) apart, both directions, outside of the glass. Same on window strikes and distance from windows.
• Bleach dilution: 1 part household bleach to 9 parts water, on disinfectants and the cleaning schedule.

How uncertainty gets handled#

The temptation is to answer everything, because a page that says "it depends" feels like a page that wasted your time. We would rather write that the evidence is thin than manufacture a number. Whether feeders create dependency is a fair example: the research that exists is reassuring but limited, and feeder dependency says so instead of ending the argument for you. Same with hot pepper seed, and with the exact date to take a hummingbird feeder down.

• A flat number means a source gives that number and we are confident in it.
• A range means the range is the truth. Feeder height, cleaning interval and seed consumption vary with your yard, and narrowing them to one figure would be a decorative lie.
• "The evidence here is thin" means what it says: somebody has an opinion, possibly a common one, and the support under it is weak.

What this site will not do#

• No product testing claims. There is no lab and no sample cupboard here, and saying otherwise would be a lie. The buying guides teach selection criteria instead, because port geometry and cleanability outlast any particular model.
• No rankings and no "best of" lists, for the same reason.
• No affiliate links, and no recommendation that exists to earn a click. The site is funded by display advertising, as set out on the about page and in the privacy policy.
• No invented statistics: no unnamed studies, no suspiciously precise percentages, no surveys that never happened, no reader numbers.
• No advice that risks harming a bird, even when it works. Greased poles are the standing example: oil on feathers wrecks insulation and waterproofing, so the trick is excluded and the reason given every time it comes up.

How often pages are reviewed#

An update date should mean somebody read the page against current sources, not that a script touched a timestamp. Different pages need that at very different intervals, because a hopper feeder works the way it did last year and an avian influenza situation does not.

Three things pull a review forward: a reader reports an error, one of our standing sources publishes new guidance, or a disease or regulatory situation changes.

AI disclosure#

Drafting and research on this site are assisted by AI tooling. What gets published is then edited, fact checked against the sources above, and structured by the editorial team. We say this in the editorial policy as well, because you deserve to know how anything you read was made.

It changes nothing about the standard. The tiers still apply, every number is verified against a named source that a person opened, consensus is still labeled as consensus, and the model is never the citation. Anything that cannot be traced to a real source does not survive the check, and errors that get through are ours to fix under our name.

How to challenge something here#

Assume we can be wrong, because a site written for a whole continent gets regional detail wrong regularly. If a range looks off for your state, if a bird we describe as absent is on your feeder right now, or if a number here contradicts a source you trust, tell us.

The corrections and updates policy sets out what counts as a correction and what happens after you report one. To send one, use the contact page with the page URL, the exact sentence you think is wrong, what you believe is correct, and a source if you have one. You do not need credentials, and "I live here and that is not what happens" is a perfectly good report.

This site does not give medical, legal or wildlife rehabilitation advice, and the terms explain the limits. If you are looking at a bird in trouble right now, do not wait on us. Call a licensed wildlife rehabilitator.