7.6 Digital Media, Message Testing, and Evaluation

Choosing channels and learning before launch

Digital media support reminders, education, peer connection, service navigation, and rapid updates. Channels include websites, email, text messaging (SMS), social platforms, video, webinars, patient portals, apps, podcasts, and online ads. The CHES decision is never whether a channel is modern; it is whether the channel fits the audience, objective, privacy needs, resources, and evaluation plan.

Before launch, confirm the message is:

• Noticeable and understandable to the intended audience.
• Credible, respectful, and tied to a clear action.
• Accessible and safe for the channel being used.

Reach is not access, and trust varies

Reach is not the same as access. A platform may have millions of users while the intended audience does not use it for health information. A patient portal is secure but misses people who never activated an account. SMS works for reminders only if messages are short, private enough for shared phones, and linked to one clear action.

Trust differs by channel and messenger. A local clinic may be trusted in one community while a government account is trusted, or distrusted, in another. Community organizations extend reach and credibility when they already hold relationships with the audience. The exam may test whether you pick a messenger and channel for trust rather than popularity.

Privacy and consent

Digital health messages can expose sensitive information on shared devices or public screens. A reminder reading "Your HIV appointment is tomorrow" can cause real harm if someone else sees it. A privacy-protective version says: "You have an appointment at 2 p.m. tomorrow. Call this number with questions." Follow organizational policy and applicable privacy rules such as the Health Insurance Portability and Accountability Act (HIPAA) when protected health information is involved.

Pretesting methods

Pretesting is a small-scale check before full release. The goal is to learn whether the audience notices the message, understands it, finds it credible, sees the action, and can use the channel.

Include people similar to the intended audience. Staff review checks accuracy but cannot substitute for audience testing: a flyer can be medically correct yet fail because the call to action is hidden, and a video can be appealing yet fail because captions are missing or the first seconds never state the benefit.

Interpreting metrics and evaluating outcomes

Digital metrics need careful reading. Impressions mean the content was displayed; reach estimates how many people saw it; click-through rate shows interaction; completion rate shows whether people finished a video or module; comments may reveal questions or misinformation. None of these alone proves behavior change. They are process measures.

Match the measure to the objective:

• Objective = awareness, then measure recall and recognition.
• Objective = scheduling, then measure completed appointments, not likes.
• Objective = skill use, then measure demonstration or self-efficacy.

Process data still matter, because poor reach can explain weak outcomes. For exam items, do not choose a channel simply because it is digital. Read the audience profile, the objective, and the constraints; the best answer is often a mix, such as text reminders plus community radio plus printed materials plus partner outreach. Strong planning pretests messages, protects privacy, ensures accessibility, and evaluates whether the message actually moved people toward the intended action.

Process vs. outcome evaluation

The exam draws a sharp line between process (formative/implementation) evaluation and outcome/impact evaluation. Process evaluation asks whether the campaign was delivered as planned and reached the audience: impressions, reach, dose, fidelity, and cost. Impact and outcome evaluation ask whether knowledge, attitudes, behaviors, or health status changed. A digital dashboard full of clicks and views answers a process question, not an outcome question. When a stem asks whether a campaign "worked" or "changed behavior," the correct measure is an outcome measure tied to the objective.

A/B testing and digital pretesting in practice

A/B testing sends two message versions to comparable groups and compares which drives more of the target action, isolating the effect of one change such as subject line or call-to-action wording. It is most credible when the two groups are randomized and the metric matches the objective. For new channels, run a small pilot post with feedback before scaling, and watch the comment stream for emerging misinformation, which is itself an evaluation signal that may require a correction message.

Privacy, equity, and the digital divide

Digital strategy must account for the digital divide: differences in device ownership, broadband access, data cost, and digital skills that can leave the highest-need groups least reachable online. A campaign that lives only on a social platform or a portal may widen inequities even as its dashboard looks healthy, because the people who most need the information are not in the data at all. This is why a blended approach, pairing digital with community radio, print, and trusted in-person outreach, is often the strongest exam answer.

Add privacy safeguards (no diagnoses in shared-device messages, HIPAA compliance) and accessibility (captions, alt text, mobile layout), and the digital plan meets the CHES standard: it reaches the intended audience, protects them, and is evaluated against the behavior the program set out to change, not merely the views it collected.