Eligibility at two levels, a recruitment path, and a power calculation scripted around a stepped-wedge rollout make up the HS840 sampling plan for Unit 5. Searches like "hs 840 unit 5 assignment example", "hs840 unit 5 sample" and "hs840 unit 5 example" land here.
What a finished HS840 Unit 5 sampling plan looks like
A code appendix follows six pages of plan. Clinics come first: the twelve in the network, all included, with criteria stated so a replication elsewhere could select comparable sites. Clinicians follow. Inclusion requires at least [four] half-day patient sessions per week and a panel of their own; residents, locum clinicians and anyone with planned leave over [three] months are excluded, each for a stated reason. The recruitment path names who contacts clinicians, when, and what they are told, kept separate from the clinic-level randomization. The power section lists every input with its source: the expected reduction in after-hours time, its standard deviation, the intraclass correlation and the within-clinic correlation over time, all drawn from the network's own audit logs for the prior year. The script's result is reported with a sensitivity table.
How a HS840 Unit 5 example is structured
Population precedes number, because a sample size is meaningless until it is clear whom it counts. Criteria are written to be applied without judgment calls, and each exclusion carries a rationale tied to measurement or ethics; locum clinicians are excluded because their audit logs follow them across systems. Recruitment is kept separate from randomization since clinicians are enrolled after clinics are assigned, a sequence that invites identification bias in cluster trials, and the plan says how that risk is contained. The power analysis follows the Hussey and Hughes framework for stepped-wedge trials, run in R so that every assumption is visible and changeable. Assumptions are sourced to prior-year data rather than borrowed from unrelated studies. A sensitivity table shows power across plausible intraclass correlations, so a reader can see how fragile the conclusion is.
Clinics, all twelve
Every clinic in the network enters, which removes site selection but limits generalization; the plan describes the clinics' size, payer mix and location so readers elsewhere can judge resemblance.
Clinicians by criteria
Regular panel, minimum session load, no planned extended leave; residents and locums excluded with reasons. Each criterion is phrased as a check someone could run against a staffing roster.
Enrollment after assignment
Clinics are randomized before clinicians consent, so the plan freezes the invitation list from the roster on one date, preventing selective enrollment after sites learn their step.
Inputs, each with a source
Effect size, variance, intraclass correlation and cluster autocorrelation, taken from the prior year's audit logs, with the query that produced each value described in a sentence.
The script and its sensitivity
R code implementing the stepped-wedge variance formula, a reported power of [0.85] at the planned schedule, and a table showing how power falls as the intraclass correlation rises.
Where marks go in HS840 Unit 5
A number without its basis is the most reliably penalized feature of sampling plans, and at doctoral level graders expect every input to the power calculation named and sourced. Borrowing an effect size from an unrelated study, or choosing one because it produces a convenient sample, reads as reverse engineering. Clustered designs raise the stakes: ignoring the intraclass correlation overstates power, sometimes dramatically, and graders check for it. Criteria should be operational rather than categorical; 'full-time clinicians' invites argument, while a session count does not. Recruitment deserves its own paragraph, because in cluster trials the order of randomization and consent can bias who enters. Plans that show sensitivity to their assumptions earn credit for honesty about what the calculation depends on and where it would fail.
Get a HS840 Unit 5 example written to your instructions
Power depends on the design, so include the question and design already settled, the Unit 5 prompt, the rubric, and any pilot figures or prior-year data. Eligibility becomes criteria, recruitment stays separate from assignment, and the size is justified by script with every input sourced. First custom samples are free; turnaround is 24-48h.
HS840 Unit 5 questions, answered
Do I have to use R for the power analysis?
No. G*Power, Stata, SAS and online calculators all work for simple designs, but clustered and stepped-wedge designs often need specialized functions or simulation, which scripts handle well. The sample uses R because the code documents every assumption. Whatever software you use, report the inputs and their sources so the result can be reproduced.
Where do effect size assumptions come from?
From the best available evidence about your outcome in a similar population: prior data from the setting, a pilot, or published studies with comparable measures. Many designs also use a minimally important difference, the smallest change worth detecting. The sample draws variance from the network's own logs and sets the target reduction at a size leaders said would justify the tool's cost.
What if the calculation says I need more clinics than exist?
Then say so, and consider changing what can change: a longer study, more steps, a more precise outcome, or a larger detectable effect. If none suffices, the study may be underpowered for its primary outcome and should be framed as a feasibility or estimation study. Reporting that honestly is better than adjusting assumptions until the number fits.