NR-585 · Week 6 of 8 · The analysis plan and data management

NR-585 Week 6 Analysis Plan and Data: How to Write It

The short answer

A telehealth service wants to know whether a scheduled reminder call reduces no-shows. The data will arrive as a spreadsheet of appointments with a yes or no in one column and a group label in another, and the entire analysis question becomes: what test compares two proportions, and what will the result look like when it is written up. That is the honest scale of a graduate analysis plan. It is not statistical virtuosity. It is choosing tests that match your data, saying what each will tell you, and describing how the data will be cleaned, stored and protected on the way there. NR-585 Week 6 grades the plan, not the arithmetic.

Your section may print this as NR 585 or NR585; it is the same course. Chamberlain publishes no syllabi outside Canvas. The placement here is our teaching judgment from the course's catalog arc; your section's rubric decides what your week actually asks.

NR-585 Week 6 grading scale at Chamberlain, the criterion levels this assessment is scored on, from Chamberlain Tutors
How Chamberlain grades NR-585 Week 6, visualized by Chamberlain Tutors.

What NR-585 Week 6 asks for

An analysis plan answers four questions in order, and writing it in that order prevents most of the trouble. What will you describe? What will you test? What assumptions does each test carry and how would you check them? And what would count as an answer to your question, stated before any data exists?

Descriptive statistics come first because they are how a reader meets your sample. Means and standard deviations for interval and ratio variables that are roughly symmetric, medians and interquartile ranges where the distribution is skewed, counts and percentages for categorical variables. Skewed data is the norm rather than the exception in clinical work, because length of stay, wait times and number of contacts all pile up at one end, and a plan that promises means for every variable has not thought about the shape of what it will collect.

Inferential choices follow directly from the level of measurement you set in the previous stage and from the structure of the comparison. Two independent groups with a continuous outcome point toward a t test; the same comparison measured twice in the same people points toward a paired test. More than two groups points toward analysis of variance. Two categorical variables point toward a chi-square test. A continuous outcome predicted from several variables at once points toward multiple regression. Relationships between two continuous variables point toward correlation, with the reminder that correlation describes association and never establishes cause. Non-parametric alternatives exist for every one of these and are the appropriate choice when assumptions fail, which in nursing samples they often do.

Qualitative proposals have a parallel obligation and it is graded just as hard. Name the analytic approach, describe the coding process in steps, say who codes and how disagreements are resolved, and describe how rigour would be established through auditability, member checking, or whatever your approach uses. Data will be analysed for themes is the qualitative equivalent of naming no test at all.

The written product at this stage is usually an analysis and data management section of roughly 900 to 1,300 words, sometimes with a table linking each research question to its planned test. Expect a rubric row on protection and handling of data as well as on the statistics themselves.

The NR-585 Week 6 method, step by step

Six moves for writing an analysis plan that matches the study you designed.

  1. Tabulate each question against its variables and their levels

    One row per question: the variables involved, the level of measurement of each, the number of groups, and whether measurements are independent or repeated. Nine tenths of test selection falls out of this table without any statistical reasoning at all.

  2. Select the test from the structure, then name its non-parametric alternative

    Choose the parametric test that fits, then name the alternative you would use if assumptions fail. Writing both, with the condition that would switch you between them, demonstrates the reasoning a single test name hides.

  3. State the assumptions and how each would be checked

    Normality, equality of variance, independence of observations, adequate expected cell counts for categorical tests. Say how you would examine each: histograms, a formal test, an inspection of cell counts. This is the paragraph that separates a plan from a list.

  4. Decide what a meaningful result looks like before you have one

    Commit to a significance level and, more importantly, say what size of difference would matter clinically. Four minutes off a wait time may be statistically detectable and clinically irrelevant, and saying so in advance is the mark of someone who understands what tests do.

  5. Write the missing data and cleaning rules in advance

    How records with incomplete fields would be handled, how out-of-range values would be checked, whether anyone would be excluded after enrollment and on what rule. Decisions written before data exists are method. The same decisions made afterwards are choices a reader cannot verify.

  6. Describe storage, access and de-identification concretely

    Where data lives, who can reach it, how identifiers are separated from the analytic file, how long records are kept and what happens at the end. This paragraph carries into the ethics stage and is far easier to write here than to retrofit later.

A layout and word budget for an analysis and data management section

Our frame for an analysis section of roughly 1,100 words. It is our own outline rather than anything the university issues, and your week's rubric outranks it wherever they disagree.

SectionWhat belongs in itWord target
Preparation and cleaningEntry checks, range checks, how errors are detected, and who performs the verification.140 to 180
Descriptive planWhich summary statistic applies to which variable, with a note on distributions you expect to be skewed.170 to 210
Inferential plan by questionEach question paired with a named test, its non-parametric alternative, and the reason the test fits.250 to 300
Assumptions and checksThe assumptions carried by your chosen tests and the specific method used to examine each.180 to 220
Missing data and significanceThe handling rule for incomplete records, the alpha level, and what magnitude of difference would matter clinically.160 to 200
Storage and protectionLocation, access control, separation of identifiers, retention period and destruction plan.170 to 210

Evidence craft for analysis plans

Cite a statistics or methods source when you name a test. Test selection is a methodological claim like any other. One citation when the test first appears converts a preference into a supported decision.

Keep statistical language exact. Significant means a result was unlikely under the null hypothesis, not that it was large or important. A non-significant result is not evidence of no difference, particularly in small samples. In a research methods course this precision is directly graded.

Ask for effect sizes alongside tests. Say which effect size measure you would report for each analysis and why it helps a clinical reader. A plan that reports only p values is describing an output nobody can act on.

Match the analysis to the design you already defended. If your design compares two units at one point in time, a paired test has no place in the plan. Internal contradictions between the design section and the analysis section are among the easiest errors for a grader to spot because both are in the same document.

Write qualitative rigour in operational terms. Name the analytic approach and its source, describe coding as a sequence, say how many coders and how disagreement is resolved, and name the specific strategy that supports trustworthiness rather than listing four criteria without method.

Treat de-identification as a method, not a promise. Say how the link between identifiers and study numbers is held, who holds it, and when it is destroyed. Data will be kept confidential is an intention. A separated key file with named access is a procedure.

Five mistakes that cost points in this week's territory

  • A test that does not fit the data. Proposing a t test for a categorical outcome is the single most common mismatch, and it usually traces back to a variable whose level of measurement was never stated.
  • Assumptions unmentioned. A plan naming three tests without naming a single assumption has skipped the part of the section that demonstrates understanding.
  • Statistical significance treated as importance. Without an effect size and a stated clinical threshold, a plan cannot say what result would matter.
  • Missing data ignored. Every clinical dataset has gaps, and a plan that never mentions them is describing an idealized study rather than the one proposed.
  • Themes will emerge. In a qualitative plan this phrase substitutes for the coding procedure, the coders and the rigour strategy, all of which carry marks.

Before you submit

  • Each research question is paired with a named analysis and a reason it fits
  • Levels of measurement from the previous stage drive every test choice
  • Assumptions are named with the method that would check them
  • An alpha level and a clinically meaningful difference are both stated
  • Missing data handling is written before any data exists
  • Storage, access and de-identification are described as procedures with named responsibility

Writing the analysis plan for NR-585?

Send the rubric out of Canvas with your questions and variable list. A premium original draft comes back in 24 to 48 hours with each question matched to a defensible test, assumptions named and the data management plan written as procedure, and revisions run until the grade lands.

Questions students ask about this stage

I have not taken statistics in years. How much do I actually need to know here?
Enough to match a test to a data structure and to explain the match in two sentences. You are not being asked to run an analysis or to derive anything. Build the question-by-variable table first, because once you can see that you have two independent groups and a continuous outcome, the test almost names itself, and the same is true for every other combination you are likely to propose. Then write what the test would tell you in plain language: whether the average time differs between the two groups by more than would be expected by chance. Graders in this course reward accurate plain language over statistical vocabulary used decoratively, and misused vocabulary is easy to spot from the other side of the page.
Does my plan need a power analysis if I already gave a sample size?
The sample size justification usually lives in the sampling section, and this section's job is to be consistent with it rather than to repeat it. If you justified your number from comparable published studies, say here that the planned analyses assume a sample in that range and note which analysis is most demanding of sample size, because that one sets the real requirement. Regression is the usual culprit: each additional predictor raises the number of cases needed, and a proposal planning four predictors on 60 participants is stretching further than it admits. One honest sentence acknowledging that constraint, and stating what you would drop if recruitment fell short, is worth more than a confident claim that the sample is adequate.
What software should I name?
Name whatever you would realistically use and be consistent about it. Common statistical packages are the expected answer for quantitative work, and spreadsheet software is defensible for straightforward descriptive analysis if you say what it would and would not handle. Qualitative proposals can name a coding package or state that coding would be done manually with an audit trail maintained, which is entirely acceptable at this scale. What matters is not the brand but the sentence after it: what the software would be used for, who would operate it, and how the analytic decisions would be recorded so that another person could follow them. A named package with no described use adds nothing to the row.

Keep going

Online now