Canonical: https://www.playbookrg.com/brand/book/content-scorecard/ Description: Draft survey questions, comprehension tasks, sample planning and a reporting framework for a defined content evaluation. No universal score thresholds or demonstrated… > Static reading copy of the published page, including authored disclosures. Interactive controls, personalized state, and account data are not reproduced. Use the canonical page for interaction and citations, and the dedicated content feeds for complete resource inventories. Source terms still apply; this format does not extend the Brand CC0 license to other products or third-party material. # Content scorecard Draft survey questions, comprehension tasks, sample planning and a reporting framework for a defined content evaluation. No universal score thresholds or demonstrated outcomes. 9 min read · Free to copy and adapt · CC0 On this page 15 sections Plan a test of whether readers understand a particular resource and can find its intended next action. Use the example surveys below to diagnose the experience, then revise and retest. **This is proposed evaluation guidance, with no Playbook study results.** These adapted questions are not a validated measure of gambling harm, safer behavior or treatment effectiveness. Favorable ratings alone do not establish any of those outcomes. Start with the [evaluation planning worksheet]() and the [printable quick reference](). ## Start with a question Name the content version, intended audience, language, placement and decision the test will inform. For example: “After reading this card, can a reader explain what the quoted probability covers and find the full rules?” Observe that explanation and next action. A rating of how clear the card *feels* answers a different question. Check the facts, links, accessibility and service details before asking participants to use the material. A popular resource with an incorrect probability or a broken support link needs correction. ## Why NPS? A recommendation question can provide optional feedback about a resource. It is not the required primary measure of educational content. Decide whether recommending this particular information makes sense for the audience and context; omit the question if it does not. The example below adapts the familiar recommendation scale to content. It does not establish comparability with an operator’s customer survey, a sportsbook app or a streaming service. This scorecard supplies no industry benchmark or universal publication threshold. ## Methodology and sources Methodology and sources: Component, What this page supports, What it does not establish | Component | What this page supports | What it does not establish | | --- | --- | --- | | Recommendation question | An adapted content question with the NPS scoring convention | A validated predictor of educational or gambling outcomes | | Five rating pairs | Draft prompts for discussing the content experience | Cross-cultural equivalence, reliability or validated pass marks | | Recall and knowledge checks | Content-specific tasks with an answer key and coding plan | A validated literacy scale or evidence that the content caused improvement | | Follow-through question | Participants’ reports of selected actions | Verified behavior, appropriateness of an action or reduced harm | [Bain’s scoring explanation]() supports the NPS calculation below. It does not validate these Playbook questions. [AAPOR’s survey guidance]() supports pretesting understandable questions and keeping methods consistent when measuring change. Pretest each language with intended readers; translated wording alone does not establish comparable scores. ## Plan the sample and procedure Choose the study design before recruitment. A small formative session can reveal a specific misunderstanding without estimating its prevalence. A quantitative comparison needs a sample rationale tied to its main question, intended precision or detectable difference, and analysis. Obtain suitable research expertise for that design. There is no fixed minimum of 25 or 30 responses that makes every study adequate. Record recruitment, inclusion criteria, incentives, field dates, missing answers and follow-up losses. An opt-in sample may differ from the intended audience. More responses alone do not resolve that difference. Agree voluntary participation, withdrawal, data access, retention, contact handling and required reviews before collection. This worksheet does not supply an approved protocol or data policy. Keep support access independent of participation, and avoid intercepting someone who is seeking immediate help. Research involving people at risk or seeking support needs an appropriate specialist protocol beyond this general-content worksheet. ## Pulse survey **Three draft prompts for initial feedback.** Completion time should be checked in a pilot. 1. If this question fits the context: “How likely would you be to recommend this information to someone who wants to understand \[topic\]?” Use 0–10, anchored “Not at all likely” and “Extremely likely.” Permit skipping or “Not applicable.” 2. “How would you rate this information?” Use 1–7, anchored “Forgettable” and “Memorable.” Permit “Not applicable.” 3. “What, if anything, would you change?” Optional open text. Add a separate task if you need to test comprehension or navigation: “In your own words, what does this figure describe?” or “Show where you would look for the rules.” Record assistance and errors. A memorability rating is not a memory test. ## Full survey **Seven draft response items:** the recommendation question, five ratings and optional open feedback. This expands Pulse by adding four rating pairs. Use the same 1–7 scale with each pair’s endpoints clearly displayed: Full survey: Low endpoint, High endpoint, Intended discussion | Low endpoint | High endpoint | Intended discussion | | --- | --- | --- | | Forgettable | Memorable | Perceived memorability | | Boring | Engaging | Interest in the information | | Preachy | Respectful | Tone and treatment of the reader | | Confusing | Clear | Perceived clarity | | Generic | Made for me | Perceived audience fit | Keep “Not applicable” separate from the midpoint. These endpoints may not form equivalent opposites for every reader. Ask during pretesting what they mean, revise when needed, and retain the exact version used in each study. Do not combine the pairs into a validated overall scale without evidence supporting that use. ## Recall survey **Four draft items for a separate follow-up.** Choose a delay that answers the study question. Seven to fourteen days is one possible planning choice, not a validated universal interval. Record actual elapsed time and whether the participant saw the content again. Ask unprompted recall before showing statements or the resource: 1. “What, if anything, do you remember from the information about \[topic\]?” Open text. 2. “True, false or not sure: \[one scoped statement\].” 3. “True, false or not sure: \[a second scoped statement\].” 4. Optional: “Since seeing the information, which, if any, of these have you done?” Offer relevant actions, “None of these,” “Not sure” and “Prefer not to answer.” Make those last three exclusive of action selections. For knowledge checks, create the answer key before collection. Keep the game and rule conditions explicit. For example, a question about independent random spins should say that they are independent; a deposit-limit question must refer to a verified feature in the actual product. Avoid guessing what an operator offers. See the [current game guides]() for scoped facts. Define how recall will be coded, including partial, incorrect and absent recall. If using more than one coder, agree how disagreements will be resolved. Report knowledge answers separately, including “Not sure.” A correct true/false answer can occur by guessing. Self-reported follow-through is not proof of benefit; choosing no action may be appropriate. Do not promise a follow-up response rate. Report how many people were invited, reached and completed each stage, and explain what is known about loss to follow-up. ## Cultural fit testing Start by checking comprehension and relevance with intended readers in each language. The “Generic–Made for me” item is one feedback prompt, not a validated test of cultural fit. If comparing two versions, define the primary outcome and assignment procedure in advance. Random assignment can support a causal comparison when the rest of the design supports it. Record what changed between versions and keep other conditions comparable. Without random assignment, differences in audience, timing or placement may explain an observed difference. Plan sample size, uncertainty, exclusions, stopping rules and any multiple-comparison handling before viewing the results. A higher average, a half-point difference, two weeks of exposure or a fixed impression count does not by itself establish that an adaptation works. “Inconclusive” is a useful result. Consult the [cultural adaptation guide]() for design considerations; validate the actual version with its audience. ## Interpreting results ### NPS For valid 0–10 responses, calculate **the percentage rating 9–10 minus the percentage rating 0–6**. Ratings of 7–8 remain in the denominator. Missing, skipped and “Not applicable” responses are excluded from that denominator and reported separately. The score ranges from −100 to +100; it is not an average rating or a percentage of improved players. [Scoring source](). If nobody gives a valid answer, report “No valid responses,” not zero. Report counts and the score’s uncertainty using a method appropriate to the design. A negative score means more valid answers fell in 0–6 than 9–10; it does not prove active discouragement. A positive score does not establish actual sharing or learning. ### Rating pairs The midpoint of 1–7 is 4. It is a scale position, not a validated pass mark. Show the distribution and valid count for each item; an average can conceal opposing experiences. Inspect open feedback and observed task failures. Avoid declaring the material successful just because every mean exceeds a chosen number. ### Recall and observed tasks Report the answer-key criteria, task successes, errors, assistance and missing data. Differentiate an observed incorrect explanation from an unfavorable opinion. To claim improvement or an effect of the content, the comparison and analysis must support that claim. Use an appropriate outcome study before making claims about safer behavior or reduced harm. ## Running a test 1. Define the primary question, audience, content version and intended decision. 2. Select or adapt questions and tasks; pretest wording, accessibility and delivery. 3. Agree the sample rationale, recruitment, data handling and analysis plan. 4. Collect according to the plan. Document deviations and reasons; do not stop because a favorable score appears. 5. Review errors, distributions, limitations and negative findings alongside the main result. 6. Record the decision, make revisions and define the next check. The questions can be copied into a suitable survey tool. This page does not provide response storage or a deployed study. Setup effort and fieldwork duration depend on the design; no fixed time estimate is promised. ## Reporting format Use the following worksheet. [AAPOR’s disclosure guidance]() is a reference for transparent reporting of methods and limitations. ``` Question and intended decision: Resource version, language and placement: Population, recruitment, eligibility and incentives: Study design and comparison/assignment: Sample rationale and planned stopping rule: Dates, follow-up interval and repeat exposure: Invited / reached / completed at each stage: Exact questions, scales, task script and answer key: Valid and missing counts for each item: Results, distributions and appropriate uncertainty: Task errors, assistance and unfavorable findings: Selection, measurement and follow-up limitations: Deviations from the plan: Decision, responsible owner, revisions and next check: ``` Keep participant identifiers and sensitive responses out of public reports. Only share quotations or data within the agreed consent and disclosure arrangements. A public project issue is not a place to collect identifiable participant records. ## Segment notes Recruit people who can answer the defined question. Do not assume every audience interprets the same scale or uses the same device. Record the relevant language and access needs and check the survey’s reading order, labels, keyboard operation and mobile layout. General-content feedback, staff training evaluation and support-service research may need different protocols. ## Quick-reference decision tree Quick-reference decision tree: Finding, Next decision | Finding | Next decision | | --- | --- | | Incorrect fact, broken essential action or harmful misunderstanding | Correct the issue and review before wider use | | Unclear wording or observed task failure | Revise the relevant content and retest | | Results too uncertain or sample does not answer the question | Report the limitation and decide whether more or different evidence is needed | | Results meet criteria agreed for this specific study | Record the scoped conclusion and next monitoring or evaluation step | No row awards a general safety approval or certification. [Open the one-page reference]() or [find its downloadable formats](). ## Survey templates Copy the Pulse, Full or Recall prompts above into a tool suited to the agreed data plan. Preserve endpoint labels, missing-answer options and the order of unprompted recall before recognition. Test branching and mutually exclusive choices. Run a sample export to confirm what each column and blank value means before collection. This scorecard does not include a verified embeddable survey widget. It does not automatically gather responses, calculate benchmarks or approve publication. ## References - Bain & Company. [Measuring Your Net Promoter Score](). Primary explanation of the scoring convention; not validation of this content questionnaire. - American Association for Public Opinion Research. [Best Practices for Survey Research](). Professional guidance on design, questionnaire development, pretesting and analysis. - American Association for Public Opinion Research. [Disclosure Standards](). Professional guidance on reporting methods and limitations. Sources checked September 6, 2026. The content-specific protocol suggestions on this page are proposed Playbook guidance. No source above establishes Playbook effectiveness or universal NPS, sample-size or rating thresholds.