PUBH 8211 Week 7 Analysis Plan Example

Course	PUBH 8211 Research Methods for Public Health
Week	Week 7
Paper type	Analysis plan brief
Length	About 660 words, 5 pages
Format	APA 7 student paper
School	Walden University
Program	PhD in Public Health
Updated	October 2026

Free sample paper for PUBH 8211 Week 7

Fixed Before the First Survey Is Scored: An Analysis Plan Naming the Model, Covariates, Missing-Data Rules, Coding Approach and Integration Steps for a Survey-Then-Interview Study of Unworn Hearing Aids

Student Name

Walden University

PUBH 8211: Research Methods for Public Health

Instructor Name

Month Day, Year

What this page is doingThe title stresses timing, because an analysis plan written before data are seen protects the study from letting patterns in the data choose the test.

Fixing Decisions Before the Data

An analysis plan written after data arrive invites the analyst to choose whatever test produces an interesting result. This plan is written before the first survey is scored, for the Tall Pines hearing aid study. It fixes the primary model, covariates, missing-data handling, interview coding approach and the steps that will integrate the two strands. Primary and exploratory analyses are labeled in advance.

What this page is doingExplaining the risk of choosing tests after the fact justifies the level of detail that follows.

Quantitative Strand

Primary outcome

Low daily use, defined as a response of "none" or "less than one hour a day" to the daily-use item of the International Outcome Inventory for Hearing Aids at six months (Cox & Alexander, 2002).

Primary model

A logistic regression of low use on prespecified predictors: degree of hearing loss in the better ear, out-of-pocket cost paid for the aids, one-way travel time to the clinic, living alone and the inventory's benefit item. Clinic is entered as a fixed effect, since there are only three. Age and sex are included as covariates. These predictors were chosen because a scoping review of non-use identified perceived value, cost, fit and comfort, and situational factors among the main reasons people give (McCormack & Fortnum, 2013).

Sample size check

With about 250 respondents and about 50 low users, the model can support roughly five predictors under the common guideline of about ten events per predictor. The prespecified list is held to five main predictors plus clinic, age and sex, and the report will note that the events-per-predictor ratio is modest.

Missing data

Complete-case analysis is used for any predictor missing under 5% of values. If more are missing, multiple imputation with 20 imputed datasets will be used, with all model variables and clinic in the imputation model. The proportion missing will be reported for each variable.

Secondary analysis

For participants with data-logged wear time, self-reported and logged daily use will be compared, and the primary model will be rerun with logged use as the outcome. This is labeled secondary because only a subset will have logged data.

Exploratory analyses

Interactions between distance and living alone, and comparisons among the three clinics, are labeled exploratory. They will be reported with that label and will not be used to support main conclusions.

Qualitative Strand

Approach

Constructivist grounded theory, which treats analysis as an interpretive process shaped by researcher and participants together and develops categories through comparison (Charmaz, 2014).

Coding

Initial coding line by line of the first several transcripts, using gerunds to capture actions such as "putting them away for visitors" or "giving up after the battery died." Focused coding then selects the most significant initial codes to organize larger segments. Memo writing after each interview records emerging categories and questions for later interviews.

Theoretical sampling

After about eight interviews, later participants are chosen to develop or test emerging categories, for example inviting a participant who lives near a clinic but stopped anyway.

Software

Transcripts are coded in qualitative software, with a codebook dated at each revision.

Integration

At sampling

Survey associations determine the dimensions along which interview participants are varied.

At interpretation

A joint display will place each survey association in one column and the qualitative categories that explain, extend or contradict it in the next, with a third column stating the integrated inference. Mixed methods researchers describe joint displays as a way to make integration visible rather than leaving results side by side (Fetters et al., 2013).

Discordance

Where interviews contradict a survey association, for example if cost is associated with low use but interviewees rarely mention cost, the discordance will be reported and examined, not resolved by favoring one strand.

What the Plan Rules Out

The plan rules out adding predictors after seeing the data, switching outcomes or reporting only significant associations. Departures go into the deviation log, each with a stated reason.

References

Charmaz, K. (2014). Constructing grounded theory (2nd ed.). Sage.

Cox, R. M., & Alexander, G. C. (2002). The International Outcome Inventory for Hearing Aids (IOI-HA): Psychometric properties of the English version. International Journal of Audiology, 41(1), 30-35. https://doi.org/10.3109/14992020209101309

Fetters, M. D., Curry, L. A., & Creswell, J. W. (2013). Achieving integration in mixed methods designs: Principles and practices. Health Services Research, 48(6 Pt 2), 2134-2156. https://doi.org/10.1111/1475-6773.12117

McCormack, A., & Fortnum, H. (2013). Why do people fitted with hearing aids not wear them? International Journal of Audiology, 52(5), 360-368. https://doi.org/10.3109/14992027.2013.769066

PUBH 8211 Week 7 instructions, in plain terms

Analysis plan assignments ask students to specify, before data collection, how each strand's data will be analyzed and how strands will be integrated. Prompts typically require the outcome definition, statistical model, predictors and covariates, missing data handling, qualitative analysis approach and integration method, with primary and exploratory analyses distinguished. Instructors reward plans that justify predictors from literature, check model capacity against expected events, commit to rules before seeing data and describe integration concretely, for example with a joint display. Plans that leave choices open until data arrive, or that treat qualitative analysis as an afterthought, usually lose credit. Several instructors ask students to register or date the plan with their advisor before any data are collected, so later readers can see it was fixed in advance.

How the PUBH 8211 Week 7 example is put together

The brief opens by explaining why the plan is fixed in advance. The quantitative section defines the outcome, names the model and five predictors with a source, checks events per predictor and sets missing data rules, secondary and exploratory analyses. The qualitative section describes the approach, coding stages, theoretical sampling and software. The integration section covers sampling, a joint display and discordance. A final section lists practices the plan rules out. Secondary and exploratory analyses are labeled in their own subsections so readers can see which results will carry the main conclusions. A short section lists practices the plan forbids.

Where the marks sit in the PUBH 8211 Week 7 rubric

Graders evaluate whether the plan is complete, justified and prespecified. Credit is given for a clear outcome definition, predictors drawn from literature, attention to model capacity and explicit missing data rules. The qualitative approach should match the stated tradition. Integration should be concrete. Labeling exploratory analyses and committing to report deviations show research integrity. Naming the software and version used shows the plan can be reproduced. Plans that address the capacity of the model, given the expected number of events, show statistical awareness that graders value. A defined rule for multiple imputation, set before data arrive, adds transparency. Describing how discordant findings will be handled shows maturity in mixed methods reasoning.

PUBH 8211 Week 7 help from the desk

Students often write analysis plans after collecting data. Write yours first and date it. Define your outcome exactly. Limit predictors to what your expected number of events can support. Decide missing data rules now. Describe your qualitative coding steps. Draw an empty joint display to show how strands will meet. To prespecify the analysis for your own study, share your questions and expected sample. Share the dated plan with your advisor before data collection. Keep a deviation log if anything changes. Build the joint display template now and fill it later. Name the software and version you will use. Write out your coding steps as you would teach them to a new research assistant.

Get a PUBH 8211 Week 7 example written to your instructions

Attach your research questions and the measures each strand collects, along with the Week 7 plan instructions and rubric. The plan names a primary model with reasoned covariates, labels exploratory work, matches a qualitative tradition to the question and sketches integration, all in 24-48h, and a first request carries no fee. Its study exists only as an illustration.

More PUBH 8211 papers and PhD in Public Health samples

PUBH 8211 Week 7 questions, answered

Where can I find a free PUBH 8211 Week 7 sample paper?

The Week 7 analysis plan is shown above, fixing methods for a study of hearing aid non-use before data arrive.

Why write the analysis plan before seeing data?

To prevent patterns in the data from steering the choice of tests and outcomes.

How many predictors can the model support?

About five, given roughly 50 expected low users and a guideline of about ten events per predictor.

What is a joint display?

A table placing quantitative results beside qualitative findings to show how they explain, extend or contradict each other.

What happens when the strands disagree?

The discordance is reported and examined rather than resolved by favoring one strand.