Measure the conversation, not merely whether people finished a game. A defensible evaluation combines pre/post measures of communication satisfaction and competence, a baseline and follow-up of workplace intergenerational climate, coded behavior during play, and interviews or focus groups. Wong et al. describe the logic precisely: “Given that the intervention was explicitly designed to structure and enhance intergenerational dialogue, this study focuses on proximal communication outcomes, specifically communication competence and communication satisfaction, as direct indicators of interaction quality.”
Completion, points, smiles, productivity and retention are useful secondary signals, but none proves that employees from different generations communicate better. A single escape-room score or a 96% completion rate cannot establish that claim. The protocol below shows how an HR or learning team can test a specific, modest proposition: a defined game improves the quality and ease of cross-generational interaction.
What should a workplace gaming evaluation measure first?
Start with proximal outcomes that can plausibly change during or soon after a session: communication satisfaction, communication competence and observable interaction quality. Treat isolation, loneliness, productivity, retention and customer outcomes as distal outcomes requiring longer follow-up and separate designs. Wong et al. define social isolation as “the objective absence or insufficiency of meaningful and sustained communication”; one game session cannot establish that a lasting condition has changed.
Use a written claim such as: “Among employees placed in mixed-generation teams, this cooperative game will increase communication satisfaction and competence from baseline to immediately after play, compared with same-generation teams.” That statement identifies the people, intervention, comparison, outcomes and timing. “Gaming will fix ageism and raise productivity” is too broad to test with one event.
The measurement stack
| Layer | Instrument or method | When to use it | What it can support |
|---|---|---|---|
| Communication satisfaction | Four selected Hecht Interpersonal Communication Satisfaction Inventory items, 6-point scale | Before and immediately after play | Whether conversations felt enjoyable, smooth and easy to enter |
| Communication competence | Six selected Interpersonal Communication Competence Scale (ICCS) items, 4-point scale | Before and immediately after play | Perspective-taking, emotional awareness, topic management, flow and sensitivity to unspoken cues |
| Workplace climate | Workplace Intergenerational Climate Scale (WICS), or the earlier Workplace Intergenerational Atmosphere scale | Baseline and several weeks after the intervention | Perceptions of stereotypes, contact, affect, inclusiveness and retention across age groups |
| In-game behavior | Video or trained-observer event sampling with a defined codebook | During play | Who speaks, who withdraws, turn-taking, biography-related talk, conflict and shared laughter |
| Implementation and engagement | Attendance, missions completed, platform events, short game-experience items and a 0–10 engagement item | During and immediately after play | Whether participants were exposed to and accepted the intervention; not proof of skill change |
| Qualitative experience | 45–60-minute focus groups or interviews | Within about 10 days | Why conversation felt easier or harder and which mechanics helped or hindered |
Which survey instruments should we use?
Keep the core questionnaire short enough to complete twice, but use instruments with a clear construct rather than an improvised “did you like the game?” score.
Communication satisfaction: four Hecht items
Ask respondents to rate each statement from 1 (extremely disagree) to 6 (extremely agree): “I felt I could talk about topic with the other person” (Wong et al. changed “anything” to “topic”); “I am generally satisfied with the conversations”; “I enjoy the conversations”; and “These conversations flow smoothly.” Sum the items for a total score. Report the mean or total at both time points, the paired change and its 95% confidence interval.
Communication competence: six ICCS items
Use a 1 (extremely disagree) to 4 (extremely agree) response format. The selected items cover whether a participant can put themselves in others’ shoes; whether they know what others are feeling (the published item is reverse-direction and must be scored consistently); whether others think they understand them; smooth shifts between topics; negotiating which topics are discussed; and noticing what people say as well as what they do not say. These map to perspective-taking, emotional awareness, conversational flow, topic management and nonverbal sensitivity.
WICS for the organizational scoreboard
The Workplace Intergenerational Climate Scale was designed to measure attitudes and perceptions about workers of different ages in the workplace. Its original form has 20 items across five subscales: lack of generational stereotypes, positive intergenerational affect, intergenerational contact, workplace generational inclusiveness and workplace intergenerational retention. A Spanish higher-education adaptation reduced it to 16 items in a sample of 1,116 university teachers and supported its structure in teaching, research and management contexts.
Administer WICS as a workplace pulse before the program and again weeks later. It is not a 20-minute post-game satisfaction sheet. Validation evidence still needs to be extended across industries, countries and workforce compositions, so report the version, language and sample rather than treating every score as universally comparable.
Optional measures that should stay secondary
If the game teaches health content, Wong’s study also used HLS-EU-Q47 health-literacy items and the Patient Enablement Instrument. Those measures are unnecessary for a communication-only workplace game. Add knowledge, productivity or retention measures only when they are explicitly part of a preregistered, longer-term outcome plan.
How do I run a four-week evaluation?
A four-week schedule gives you a baseline, a controlled exposure, immediate reaction data and enough time to see whether a climate signal persists.
- Week 0 — define the intervention and claim. Name the format (tabletop dialogue game, escape-room simulation or digital learning journey), the communication outcome and the comparison. Record whether the mode is creative, cooperative or competitive; mode changes the behavior you should expect to see.
- Week 0 — assign teams deliberately. Create mixed-generation teams and same-generation comparison teams where operationally possible. Record age bands, role, tenure, location, language and prior relationships. If half of the mixed teams receive a diversity primer, treat that as an experimental factor rather than assuming the primer works.
- Week 0 — collect baseline data. Administer the Hecht items, ICCS items and WICS (or WIA). Add a neutral wording check such as separate ratings for “game,” “friendly competition,” leaderboards, badges, prizes and avatars. Capture prior experience with games and any accessibility needs.
- Week 1 — run the session and capture exposure. Record attendance, team composition, start and finish, missions or cards completed, hints, withdrawals and completion. For in-person play, obtain consent for video or use two trained observers. For a digital journey, export event logs rather than relying on a single final score.
- Immediately after play — repeat the proximal scales. Repeat the same Hecht and ICCS items, then add Wong’s implementation block: fun, encouragement of intergenerational communication, satisfaction with mission-card design, overall satisfaction and a 0–10 engagement item. Keep these implementation items separate from the pre/post outcomes.
- Within 10 days — conduct qualitative sessions. Hold separate or mixed focus groups lasting 45–60 minutes. Ask what made it easier to start a conversation, whose ideas were heard, when age assumptions appeared, and which game mechanic changed behavior. Wong interviewed within 10 days, used two independent coders and discussed the codebook to agreement; that approach limits recall error and makes interpretation auditable.
- Week 4 — repeat the climate pulse. Re-administer WICS or WIA if the claim concerns the workplace rather than one event. A climate shift should be interpreted alongside exposure, staffing changes, other inclusion initiatives and response rate.
What should observers code during the game?
Self-report captures perception, but event sampling shows what actually happened. Define a short interval (for example, every 20 seconds or at each conversational event), identify the speaker and listener, and code the interaction immediately.
- Verbal content: game-related talk, work-related talk, biography or personal-history talk, clarification, explanation and disagreement.
- Interaction pattern: turn-taking balance, direct address across generations, interruptions, topic shifts, invitations to speak and responses to invitations.
- Friction and withdrawal: unresolved conflict, digression, silence, visible withdrawal, side conversations and one person taking over.
- Nonverbal behavior: shared laughter, encouragement, eye contact where observable and signs of confusion or disengagement.
- Power balance: whose suggestions are adopted, who handles the rules or technology and whether older or younger participants are treated as the default expert.
Zahn et al.’s field study illustrates why mode belongs in the analysis. In a Swiss retirement-home trial with 10 older participants (mean age 84.8 years, SD 5.9), 30 twenty-minute sequences produced more verbal communication in creative mode than cooperative mode (P=.04, Cohen f=0.611), more biography-related communication (P=.03, r=0.707), and more shared laughter in competitive mode (overall nonverbal difference P=.02, Cohen f=0.758). Those are setting-specific findings, not workplace effect estimates, but they show that “communication” is not one behavior.
A 2025 Acta Psychologica sequential analysis of intergenerational digital-game-design collaboration similarly found that high-performing groups exchanged ideas more directly and evenly. Lower-performing groups showed digression and conflict-related withdrawal, especially among older adults. Code those process failures instead of hiding them behind an overall enjoyment score.
How should the results be analyzed?
Analyze the pre/post change, not only the post-session average. For each scale, report the baseline mean, post mean, mean difference, 95% confidence interval, response count and effect size. Paired t-tests are suitable for a simple pre/post design; use a mixed model or analysis of covariance when you have comparison teams, repeated sessions or substantial baseline imbalance.
- Report Cohen’s d for change and apply a Hedges correction when a subgroup has fewer than 20 people. Wong used 0.2, 0.5 and 0.8 as small, medium and large reference points.
- Show mixed-generation and same-generation results separately. A favorable overall mean can conceal a null or negative result in one age band.
- Code qualitative interviews thematically, preserving examples of easier conversation, stereotype threat, silence and unexpected expertise. Have two coders examine a subset and reconcile the codebook.
- Model exposure separately from outcome: attendance and completion explain who received the intervention; they do not show that communication improved.
- Pre-register any productivity, retention or customer metric and follow it for a period long enough to plausibly change. Do not attribute a quarterly business result to a one-hour game without a comparison and a mechanism.
Wong’s Hong Kong tabletop study planned a minimum sample of 63 to detect a medium effect of 0.5 at alpha .05 and 80% power. It enrolled 81, with 68 completing the intervention and evaluation (84% response): 23 young adults aged 18–39, 14 middle-aged adults aged 40–64 and 31 older adults aged 65 or above. Nineteen groups contained three to five people with at least one person from each age band, and 52 of 68 completers (76.5%) joined qualitative sessions. These figures are a useful planning reference, not a guarantee that a small corporate pilot will have adequate power.
What do the workplace and community studies actually show?
Wong et al.: structured dialogue can improve proximal measures
In the Hong Kong Healthy Aging in Dialogue Tabletop Game/Card Game evaluation, communication satisfaction improved with Cohen d=0.745 and communication competence with d=0.377. Mean satisfaction moved from 18.57 (SD 3.31) to 21.34 (SD 1.87), a mean difference of −2.76 (95% CI −3.66 to −1.87); the negative sign reflects baseline-minus-post scoring, while the authors describe improvement. Competence moved from 17.62 (SD 2.73) to 18.66 (SD 2.33), mean difference −1.04 (95% CI −1.71 to −0.37). Health literacy also improved (d=0.409).
Effects were not identical by age. Middle-aged participants had a satisfaction effect of d=1.131, while older adults were the only subgroup with a statistically significant competence change (mean difference −1.32). Patient enablement was higher among older adults (mean 6.10, SD 2.18) than young adults (mean 4.22; P=.04). The sample was community-based, Cantonese-speaking and 82.4% female, so it should not be presented as a corporate trial.
Participants explain the mechanism better than a score does. One young participant said: “Sometimes, when talking to middle-aged and older individuals, you may not always know what to say. Not everyone is skilled in discussing various topics with others. On the contrary, this approach raises topics for conversation on which everyone can share their opinions. It’s not too tense, and it makes the conversation with them much easier.” An older participant said: “The gamification provides a welcoming space to share personal matters (anonymously), which helps players to soothe and relax, especially on sensitive topics that might be difficult to bring up with friends, as you may be uncertain if the friend may spread the word unintentionally.”
Richards, Becker and Stollings-Holder: a game can expose friction
The closest workplace experiment was an escape-room simulation with 60 employees—10% of a Southeastern U.S. utility’s workforce. Intergenerational teams were less successful at problem-solving than same-generation teams. A 30-minute generational-diversity primer did not change intergenerational outcomes. The study measured problem-solving success more directly than psychometric communication change, so it should not be converted into a claim that escape rooms damage communication.
Use the finding as a warning: a mixed-age game may surface unequal participation, conflict or withdrawal before it improves anything. Measure those behaviors, and consider longer diversity training, reverse mentoring and repeated practice rather than treating a primer as a complete intervention.
Sanofi’s Generations Journey: completion is exposure, not communication gain
Sanofi’s “Jornada de Gerações,” co-developed with Fábrica de Cursos on the playNwhere learning platform, put four-person mixed-age teams through three browser-based games about age prejudice, the benefits of collaboration and initiatives for employees over 50. Browser access was intended to reduce barriers for older employees, including factory workers. Platform analytics plus participant reflections produced a 96% completion rate.
The authors, Pistono and Baptista, conclude that gamified interventions such as the “Jornada de Gerações” can serve as powerful tools to address workplace ageism and promote inclusion, while also calling for “mixed-method evaluations.” Treat 96% as evidence that participants completed the journey, not as a communication effect size.
Acceptance differs by generation and by mechanic
A 2023 study of 367 employees in a company of more than 17,000 found statistically significant generational differences in gamification acceptance. Generation Z (ages 16–24) represented 46 respondents, Generation Y (25–40) 180 and Generation X (over 40) 141; Generation Z reported the greatest enjoyment. Separate 2023 survey data found that 42% of Boomers, 33% of Gen Xers, 45% of Millennials and 57% of Gen Z respondents rated “gamification” as an effective motivator. Yet leaderboards were rated effective by 57% of Boomers and 64% of Gen Z, while physical prizes were rated effective by 63% of Boomers and 64% of Gen Z.
That gap is a wording problem. Maxwell Foxman observes: “Many older generations associate the word ‘game’ as being frivolous and not being associated with ‘real work.’” He suggests reframing light competition as competition rather than relying on the word game. Ask about each mechanic instead of forcing respondents to accept or reject the label “gamification.”
Which KPIs belong on an HR dashboard?
| KPI | Interpretation | Do not claim |
|---|---|---|
| Hecht satisfaction change | Conversations felt easier, smoother or more enjoyable | Long-term trust or reduced isolation |
| ICCS competence change | Self-reported perspective-taking, flow and sensitivity improved | Objective communication skill without behavioral corroboration |
| WICS/WIA change | Workplace climate perceptions shifted over weeks | A causal game effect without a comparison and other-context controls |
| Cross-generation speaking balance | Talk was distributed more evenly during play | Equal airtime always means better communication |
| Biography-related talk and shared laughter | The mode elicited particular social behaviors | These behaviors alone prove inclusion |
| Attendance, completion and mission progress | Participants received the intended exposure | Communication improvement |
| Mechanic-level acceptance | Which elements—leaderboards, badges, prizes, avatars or streaks—are acceptable to each generation | That one mechanic works for every age group |
Operational analytics can also reveal sustainability. A 5 June 2026 snapshot of roughly 13,000 active boards from about 10,000 creators found 73% were a single ranked list; in a September 2025 cohort, 17% were still receiving score updates at week 12. Those vendor figures describe platform activity, not independent evidence of cross-generational communication.
What mistakes invalidate the conclusion?
- Calling engagement communication. A fun score, leaderboard click or 96% completion rate measures exposure or acceptance.
- Using only a post-event pulse. Without a baseline, you cannot estimate change.
- Pooling all generations. Report each age band and team composition; older adults may improve competence while another group does not.
- Ignoring game mode. Creative, cooperative and competitive designs elicit different talk and nonverbal behavior.
- Using “gamification” as the only attitude item. Older workers may reject the label but welcome a leaderboard, prize or friendly competition.
- Generalizing across settings. Community participants in Hong Kong, utility employees in the Southeastern United States, retirement-home residents in Switzerland and Sanofi employees are not interchangeable populations.
- Overclaiming from small studies. Zahn’s behavioral sample was 10 older residents; Richards used 60 employees in one utility; Sanofi reported completion rather than a communication effect.
- Assuming a primer is sufficient. The 30-minute diversity primer in the escape-room study did not change mixed-team outcomes.
No large, multi-firm randomized workplace trial was identified that combines WICS with Hecht and ICCS before and after a corporate gaming program and follows communication for months. Report that evidence boundary plainly.
Do escape rooms improve intergenerational teamwork and communication?
They can create a rich observation window, but the available workplace evidence does not show a guaranteed improvement. In the Southeastern U.S. utility study, mixed-generation escape-room teams were less successful at problem-solving than same-generation teams, and the short diversity primer did not reverse that result. An escape room is therefore best used as a structured stress test: pair its outcome with coded turn-taking, invitations, conflict, withdrawal and post-session communication scales.
If the goal is dialogue rather than pressure, a tabletop conversation game with prompts may be a better first exposure. The format should follow the claim: use creative prompts when you need biography and verbal exchange, cooperative tasks when joint planning is central, and competition only when shared laughter or energetic engagement is an explicit, safe outcome.
How should a final report state the verdict?
Separate four conclusions: whether people participated, whether they enjoyed the format, whether proximal communication changed, and whether workplace climate changed later. A sound report might say that satisfaction rose immediately with a specified effect size, that observed speaking balance improved or did not improve, and that WICS was unchanged at four weeks. It should identify the setting, sample, response rate, team assignment, game mode, language, accessibility measures and missing data.
Use the wording “consistent with improved communication satisfaction in this sample” unless the design supports a stronger causal statement. Keep productivity, retention and isolation in a future-outcomes section unless they were measured with an appropriate comparison and follow-up. As Foxman puts it, workplace games can build “trust, social cohesion, and camaraderie,” but those social outcomes still need a design that measures them rather than assuming them.
Frequently Asked Questions
How large should a workplace gaming pilot be?
Base the sample on the smallest effect and subgroup you need to detect, not on the number of people who can fit in one room. Wong planned for at least 63 participants for a medium effect of 0.5, while Richards’ 60-person single-company study was useful but limited in generalizability. If age-band comparisons matter, power each band or label those analyses exploratory.
Should employees be told the study is about generations?
Tell participants the purpose, consent requirements and data handling, but avoid wording that primes stereotypes during the game. Ask neutral baseline items about conversation and specific mechanics, then explain the generational comparison in the report. Protect video and demographic data through restricted access and a deletion schedule.
Can a digital learning journey measure communication without video?
It can measure exposure, pacing, choices, completion and written reflections, but platform logs rarely reveal who listened, interrupted or withdrew. Add chat or discussion coding, short interviews, and pre/post Hecht and ICCS items; use WICS later if the claim concerns workplace climate.
What if the post-game survey response rate is low?
Compare responders and nonresponders on age band, team, attendance and completion, then report the missingness rather than treating completers as the whole workforce. Use an immediate, mobile-friendly survey and schedule the follow-up before the session ends. A high completion rate for the game does not repair a biased evaluation sample.