Questions, answered first

Usually it needs a determination rather than a full review. Most WGU MSN capstones are single-site quality-improvement projects, and the most common honest outcome is a written determination from WGU's IRB office that the project is not human-subjects research. You still file to get that determination; it is not something you declare yourself. Misclassifying in either direction is costly.

The two errors are symmetrical. Call a research design "QI" and you collect data under a claim the office never reviewed, so the branch you walked was never approved and the data can be unusable from the root. Call a QI project "research" and you volunteer into a review it never needed: a heavier folder, a board's calendar, a consent file for a project that only ever meant to fix one unit's handoff. The route and the form for the determination request are WGU's to set; check the current handbook or portal for the exact submission type, and read the intent statement as the office will read it before you pick it.

A determination is the IRB office's written finding that a project is not human-subjects research, so it does not need exempt, expedited or full board review. You get one by filing a short, precise request through WGU's current portal describing the aim, the site, what data will be recorded and why nothing in it identifies a person. The office issues it; you do not declare it.

The request is short but it is not casual. The intent statement has to say, plainly, that the aim is to improve practice at the one site and to report to that site's leadership; the data description has to show that nothing recorded points back at a person, including any key you might be tempted to keep; and the two halves have to agree with each other and with whatever instrument you attach. A clean request is a document many candidates can file themselves, and the free application review will say so when it is true. Keep the written determination; it is the document that lets collection begin.

Four questions decide it: whether the project draws information from or about living people; whether anything you record identifies them; whether the aim is to improve practice at one site or to produce generalizable knowledge; and whether a vulnerable population or more than minimal risk is involved. Topic and method matter far less than those four answers.

Walk them in order. No living people, or people but nothing identifiable recorded and a single-site improvement aim, lands at the determination. People, nothing identifiable, but an aim meant to travel, reads as an exempt category. Identifiable information at minimal risk is expedited territory. Identifiable information plus a vulnerable population or real risk is talk-first territory, where the convened board is possible and a small truthful redesign is worth discussing before anything is drafted. The identifiers question does the most work: a linkage key you keep, wherever stored, is an identifier.

Exempt means the project fits a named low-risk category, confirmed by the reviewer rather than declared by you. Expedited means identifiable information at minimal risk, read closely by a designated reviewer instead of the convened board. Full board means the convened committee meets on it, usually because of a vulnerable population or more than minimal risk. A fourth outcome sits before all three: a determination that the project is not human-subjects research at all.

The word "exempt" misleads people: it does not mean exempt from filing. You file a complete application and the reviewer confirms the category, and what decides it is the file. Expedited is where the folder carries the weight, because identifiability is exactly what the consent file, the data-security plan and the site letter exist to protect. Full board is rare for a WGU capstone and worth designing away from when the question allows it, and worth walking honestly when it does not.

No. The tree is a reading, not a ruling. It asks the four forks a reviewer asks and shows where a project answered honestly is likely to land, including the honest landing where you can file a determination request yourself. The IRB office's written determination or approval is the only ruling, and the board's decision is its own.

The tree is most useful as a mirror: answer it for the project as you actually plan to run it, not as you hope it reads, and notice where the answers disagree with your draft. "Nothing identifiable" on fork 2 while the methods section describes a coded key is the kind of drift the board will circle. When the leaf surprises you, the free application review exists to read the same forks with a second pair of eyes.

No. Collection begins when the approval or the written determination exists, not before, and there is no retroactive approval. Chart pulls to see what is there, pilot runs of a survey and notes on conversations you plan to quote all count as collection.

The rule spans every leaf, including the lightest. On the determination branch the clean sequence is written determination first, data second; the letter is quick precisely because the category is light, so there is nothing to gain by starting early. Data gathered ahead of the letter generally cannot be blessed afterwards, and a project built on it can be unusable however good the work. It is the cheapest mistake on the tree to avoid: it costs a date on a calendar and nothing else.

Plan on it. WGU's current handbook or IRB portal names the human-subjects training it expects and which modules; CITI is the usual vehicle, and the completion record goes into the file. Finish it before filing, because a missing training record is one of the simplest reasons a file stalls.

Training is about the person filing, not the project's risk level, so it applies on the determination and exempt branches as much as on the heavier ones. Save the completion record where you can find it, keep it current for the life of the project, and check WGU's current portal for whether anyone else named on the application, a site collaborator for instance, needs a record of their own.

If the project touches a facility's people, records, rooms or data, plan on a letter on the facility's letterhead signed by someone with authority to grant that access. It is routinely the slowest document in the folder because it moves at the speed of someone else's inbox, so request it first.

The letter has teeth. It names the project, names what you may access (people, charts, rooms, data) and comes from a role that can actually grant it: a director or administrator, not a friendly colleague on your unit. Many facilities route the request through their own machinery first, a nursing research council or an education office, and that internal loop is the long pole nobody budgets for. On our files we draft the letter for the signer's letterhead and carry the request, so the only thing left to move is the signature.

Sometimes. Larger facilities, especially on the identifiable branch, may want their own review or their own research council to read the project before they sign a permission letter, and some have an IRB of their own. Ask the site early, because its loop runs alongside WGU's and is often the slower of the two.

What WGU's IRB office requires is set by WGU's current handbook and portal; what the facility requires is the facility's own. They are two boards, and a file that satisfies one does not automatically satisfy the other. The practical move is to ask the person who will sign the letter a single question at the start: does this facility need to review the project itself before granting access? The answer decides whether there is one branch to walk or two in parallel, and the site letter should acknowledge the site's own loop when it exists.

Often yes, but the relationship changes the file. Your own patients, supervisees or direct reports can feel they cannot decline, so the board reads recruitment wording, voluntariness language and anonymity claims closely. Surveying your own patients can also move the project toward the talk-first branch, so plan that before drafting.

The design choices that help are small and honest: an anonymous instrument; distribution by someone outside the reporting line, a department coordinator rather than the manager; a plain line that declining changes nothing about care, employment or standing; and no reporting cell so small that one person is identifiable. Your own patients are a different matter from your own colleagues, because the care relationship itself is the pressure; that is the case to talk through before a document exists.

Plan on some form of it whenever living people answer questions, even on the light branches. For an anonymous QI survey that is usually an information sheet or a consent statement at the top of the instrument rather than a signed form; for identifiable interviews it is a fuller consent file. Which form WGU's office expects for your category is set by WGU's current handbook or portal.

Whatever its form, the consent material has known load-bearing parts: what the project is and who runs it; that taking part is voluntary and declining changes nothing; what is collected, where it lives, who sees it and when it is destroyed; the honest risks, even when the honest answer is inconvenience and little else; how to withdraw; and a real person to contact. The recruitment email, the flyer and the huddle script are read as part of the same file, and must promise nothing the sheet retracts.

Often, and it is one of the cleanest designs for a single-site improvement project, provided what you record is not identifiable and the facility has granted access in writing. Aggregate counts with no names, no record numbers, no linkage key and no dates tied to a small unit keep the project on the left side of the tree; a key you keep moves it to the right.

Two things decide it. The first is what exists in your dataset after extraction, not what you saw on the screen; a spreadsheet with a medical record number column is identifiable, and so is one with a date of service joined to a unit small enough to name the patient. The second is the facility's permission: chart access is the facility's to grant, so the site letter should name it specifically. Pulling charts "just to look" before the board's letter is collection; wait for the letter.

A return is a set of notes, not a rejection. You answer each note specifically, revise every document the answer touches so the file still tells one story, and resubmit. Most returns trace to a mismatch: recruitment that promises what the consent sheet retracts, an intent statement in research language over a QI design, or a data plan that quietly keeps an identifier.

The reply has a shape reviewers recognise: each note quoted, answered in the board's own order, and followed by exactly where the file changed. Arguing with a note rarely helps; fixing the page does. The trap is the partial revision, where the consent sheet is corrected and the flyer still says the old thing. Ruth Ellery owns this station on our files, and the same desk answers amendments after approval and continuing-review notices until the project closes.

We do not quote durations, because they are the board's and they vary with the branch, the season and the file. What you control is how many times the file goes round: a complete, consistent file that gives the reviewer no reason to pause is the only reliable way to shorten the walk. WGU's current handbook or portal is the place for any stated expectations.

Two practical truths. The determination branch is quick precisely because the category is light, which is one more reason not to dress a QI project as research. And the slowest document in the folder is usually not yours: the site permission letter moves at the speed of someone else's inbox, so it is the limb to start first. We never promise our own turnaround either; what we promise is that the file is whole when it goes in and that every note is answered.

No. We handle the entire IRB process, which means the determination and plan, every document, the submission and every reply until approval, while the research itself stays yours: your project, your data, your findings. Pricing is quoted in writing after the free application review as a fixed fee, in three packages: Determination & Plan, The Complete File, and Full Handover, the premium package that takes the whole branch off your desk.

The free review is the entry, and it sometimes ends with "file it yourself". When it does not, the reading names your branch and the quote follows for the package that fits it, with no hourly billing. The pricing page sets out each package station by station; how it works walks the process; and on every package the board's decision stays the board's own.