MPH-505 · Week 8 of 8 · Evaluation and the finished plan

MPH-505 Week 8 Evaluation and the Finished Plan: How to Write It

The short answer

The closing stage asks for the evaluation and, usually, the complete plan assembled into one document. Evaluation here has three layers that must not be confused: process, which asks whether the campaign happened as designed; outcome at the construct level, which asks whether beliefs, confidence or norms moved; and behavioral or health outcome, which asks whether the action changed. A logic model ties inputs and activities to those outputs and outcomes, and its value is that it exposes any promise your plan cannot keep. Your section may print this as MPH 505 or MPH505; 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.

MPH-505 Week 8 grading scale at Chamberlain, the criterion levels this assessment is scored on, from Chamberlain Tutors
How Chamberlain grades MPH-505 Week 8, visualized by Chamberlain Tutors.

What MPH-505 Week 8 asks for

A campaign to raise on-time completion of the two-year well-child visit ran for five months across a family practice network, and the report submitted at the end said the campaign was successful because two thousand text reminders were delivered and the practice's social posts reached a large audience. Nothing in that sentence is about children attending appointments. The evaluation that would have answered the question was available all along in the scheduling system: the share of children turning two in the campaign period who completed the visit within a defined window, compared with the same share in the equivalent period a year earlier, with attention to whether anything else changed in between. Building that comparison before launch, rather than reconstructing it afterwards, is what this stage teaches.

The three layers each need their own indicators. Process indicators cover delivery and fidelity: how many messages went out, whether they went out on schedule, whether partners carried what they agreed to carry, and how many people were exposed. Construct-level outcomes cover the things your theory said had to move, measured with items you can name, ideally from a validated instrument. Behavioral outcomes cover the action itself, and the strongest source is almost always a record system that already exists rather than a survey you would have to run.

Design matters as much as indicators. A single measurement after the campaign tells you almost nothing. A before-and-after comparison in one group is better and remains vulnerable to everything else that happened during the period, including seasonality and any concurrent programme. A comparison group, a stepped rollout across sites, or a series of measurements before and after that lets you see the underlying trend all strengthen the inference substantially, and each is often feasible in real settings. Say which you chose and be explicit about what your design cannot rule out.

Where the final deliverable assembles the whole plan, integration is graded. That means the objective in section one matches the indicator in the evaluation section, the constructs named in the theory section are the ones the messages carry and the survey items measure, and the segments are consistent from the formative work through to the message table. Most lost marks in final submissions come from internal contradictions rather than from any single weak section, so read the assembled document once from front to back purely for consistency.

The MPH-505 Week 8 method, step by step

Six moves for an evaluation section that could actually be run.

  1. Build the logic model backward from the outcome

    Start at the behavior change, work back through construct change, exposure, activities and inputs. Built forward, logic models drift into lists of everything the programme will do.

  2. Name the data source before the indicator

    Scheduling systems, immunization registries, pharmacy records, attendance logs, existing surveys. An indicator with no source is a measurement you will not take.

  3. Choose a comparison you can defend

    A prior period, a similar site, a stepped rollout, or several measurements before and after. Then write one sentence naming what your comparison still cannot rule out.

  4. Measure construct change with named items

    Two or three items per construct, drawn from a published scale where one exists, asked at defined points. This is what lets you explain why the campaign worked or did not.

  5. Set a threshold and a decision it triggers

    What level of change counts as success, what level triggers redesign, and what would stop the programme. A number attached to a decision is what turns evaluation into management.

  6. Read the assembled plan for contradictions

    Objective against indicator, constructs against messages against items, segments against channels. Fixing internal inconsistency is the cheapest set of marks available in the final week.

Assemble the final plan and budget the words

Our frame for a closing evaluation and integrated plan, sized for roughly 1,800 to 2,200 words. It is our own outline rather than anything the university issues, and your week's scoring guide outranks it wherever the two disagree. Scale proportionally if your assigned length differs.

SectionWhat belongs in itWord target
Objective restatedThe behavioral objective in the same words used in the opening, with its baseline and target.120 to 160
Logic modelInputs, activities, outputs, short-term and behavioral outcomes, built backward and shown as a figure or table.250 to 300
Process evaluationDelivery, fidelity and reach indicators, each with its source and who records it.250 to 300
Construct-level outcomesThe items measuring each construct, the instrument they come from, and the measurement points.280 to 340
Behavioral outcome and designThe indicator, the record system, the comparison chosen, and what it cannot rule out.300 to 360
Thresholds and decisionsSuccess, redesign and stop levels, each tied to a review point and a named decision maker.200 to 250
Limitations and sustainabilityWhat the evaluation will not establish, and what happens to the work when the initial funding or attention ends.200 to 250

Evidence craft for evaluation writing

Keep the verbs matched to your design. A before-and-after comparison in one group supports increased alongside and was followed by. Only a design with a comparison group or a strong quasi-experimental structure supports produced or caused. This one discipline protects more evidence marks in this course than any other habit.

Cite the instruments you borrow. Validated scales for efficacy, intention, norms and readiness exist for many behaviors, and using one with attribution is both stronger and faster than writing your own items. Say what the scale measures and in what population it was validated.

Report expected effects with their base. If comparable campaigns moved a behavior by a few percentage points, say so, with the study and the population. A plan promising a large change when the literature reports small ones has set a threshold it will fail.

Say what the evaluation costs. Staff time to pull records, incentives for a follow-up survey, analysis time. Evaluation sections that assume measurement is free are the ones that get cut first in practice, and naming the cost is a mark of professional writing.

Five mistakes that cost points in this week's territory

  • Process measures presented as results. Messages sent and people reached describe activity. They do not tell you whether anyone did anything differently.
  • An evaluation with no comparison. A single post-campaign measurement cannot distinguish your effect from the trend that was already running.
  • Constructs never measured. If the theory was the argument for the design, not measuring the constructs means you can never explain why the result came out as it did.
  • Thresholds with no decision attached. A target that triggers nothing is a hope. Say what happens at each level.
  • Internal contradictions across the assembled plan. An objective about one behavior and an evaluation measuring another is the single most common defect in final submissions.

Before you submit

  • The final objective matches the opening objective word for word
  • The logic model runs backward from the behavioral outcome and holds together
  • Every indicator names the record system or instrument that produces it
  • Construct-level items are attributed to a published scale where one exists
  • The design is named along with what it cannot rule out
  • Each threshold is tied to a review point and a decision maker

Finishing MPH-505 this week?

Send the final scoring guide and whatever you have assembled so far. A premium original paper comes back in 24 to 48 hours with a logic model built backward, indicators tied to real data sources and the whole plan checked for internal consistency, and revisions run until the grade lands.

Questions students ask about this stage

How do I evaluate a campaign that is entirely hypothetical?
Write it as a plan you intend to execute, with the same specificity you would need if someone approved it tomorrow. Name the indicator, the system that already holds the data, the person whose job would include pulling it, the comparison period, the measurement points and the analysis you would run. Hypothetical describes the status of the programme, not the precision of the writing. The version that loses marks promises to assess effectiveness through surveys and outcome data, which is the same sentence in every mid-band paper. The version that scores names an existing record system, which incidentally demonstrates that you understand what measurement actually costs and who would have to do it.
What if I cannot get a comparison group?
Say so and strengthen the design in the ways still available to you. Several measurements before the campaign begins let you see the existing trend, which is the single most valuable addition when no comparison group exists, because a rising line before launch explains a rising line afterwards. A stepped rollout across sites or neighbourhoods gives you a comparison without denying anyone the programme, since later groups serve as controls for earlier ones. Seasonal matching to the same months in a prior year helps for behaviors with strong seasonality. Whatever you choose, name the specific rival explanations that remain, such as a concurrent programme, a policy change or a national news event, and say how you would detect them.
How much of the earlier weeks should the final document repeat?
Enough for the document to stand alone, rewritten rather than pasted. Check your section's policy on reusing your own submitted work, since some programmes require self-citation, and then treat the earlier material as evidence rather than as prose. Pull the objective, the segments, the constructs and the messages into the new structure and write the connective tissue fresh, because pasted sections show up as changes in register and carry framing that belonged to a different question. A useful discipline is to draft the final document from an outline built out of your own earlier conclusions, without the old files open, and then check back against them for accuracy and citations once the draft exists.

Keep going

Online now