Evaluation of a Seven-Week Web-Based Happiness Training to Improve Psychological Well-Being, Reduce Stress, and Enhance Mindfulness and Flourishing: A Randomized Controlled Occupational Health Study

The researchers tested a structured, seven-week web-based "happiness training" program against a waitlist control group (no active intervention during the study period). The training was delivered entirely online and consisted of weekly modules based on positive psychology principles.

**The intervention included:**

Weekly exercises drawn from positive psychology (e.g., gratitude journaling, identifying personal strengths, practicing acts of kindness, savoring positive experiences, mindfulness meditation exercises)

Short video lectures and written materials delivered through a web platform

Participants were asked to complete exercises for approximately 15–30 minutes per day, several times per week

The program was self-paced within each weekly module

**The comparator was:**

A waitlist control group that received no training during the study period but was offered the training after the final follow-up assessment

**Primary outcomes measured:**

Happiness and satisfaction (visual analog scales, VAS, 0–100)

Psychological well-being (WHO-5 Well-Being Index)

Perceived stress (Stress Warning Signals scale)

Mindfulness (Freiburg Mindfulness Inventory)

Flourishing (Flourishing Scale)

Recovery experience (Recovery Experience Questionnaire)

**Secondary/exploratory outcomes:**

Salivary cortisol (stress hormone)

Salivary alpha-amylase (marker of autonomic nervous system activity)

Attention network performance (Attention Network Test, ANT)

**Sample size:** 147 employees total (76 in the intervention group, 71 in the waitlist control group)

**Population:** Employees from a German health insurance company (AOK) and a medical device company (Dräger)

**Age range:** Mean age approximately 41 years (range not explicitly stated, but working-age adults)

**Gender:** Approximately 60% female, 40% male

**Setting:** Workplace-based, with participants completing the training on their own time or during work hours (the study was conducted as an occupational health initiative)

**Inclusion criteria:** Employed adults with access to the internet; no specific mental health exclusions were reported

**Exclusion criteria:** Not explicitly detailed, but participants were presumably healthy enough to be working full-time

**Important note:** This was not a clinical sample. Participants were generally healthy employees, not people diagnosed with depression, anxiety, or other mental health conditions. The study was designed as a prevention/wellness intervention, not a treatment for mental illness.

The study used a combination of self-report questionnaires and objective biological/performance measures:

**Self-report instruments:**

**Visual Analog Scales (VAS) for happiness and satisfaction:** Two single-item scales, each a 100mm line where participants marked their current level (0 = not at all happy/satisfied, 100 = extremely happy/satisfied). Simple but well-validated for capturing subjective states.

**WHO-5 Well-Being Index:** A 5-item questionnaire measuring positive psychological well-being over the past two weeks. Scores range from 0 to 100, with higher scores indicating better well-being. A score below 50 suggests poor well-being.

**Stress Warning Signals Scale:** A 10-item scale measuring physical and psychological stress symptoms (e.g., tension, irritability, sleep problems). Higher scores indicate more stress.

**Freiburg Mindfulness Inventory (FMI):** A 14-item scale measuring mindfulness (attention to the present moment, non-judgmental awareness). Higher scores indicate greater mindfulness.

**Recovery Experience Questionnaire:** A 12-item scale measuring how well participants recover from work stress (psychological detachment, relaxation, mastery experiences, control during off-work time). Higher scores indicate better recovery.

**Flourishing Scale:** An 8-item scale measuring psychological flourishing (sense of purpose, positive relationships, self-acceptance, competence). Higher scores indicate greater flourishing.

**Objective measures:**

**Salivary cortisol:** Collected via saliva samples at specific times (morning and evening) to measure the cortisol awakening response and diurnal cortisol slope. Cortisol is a primary stress hormone; elevated or flattened diurnal patterns indicate chronic stress.

**Salivary alpha-amylase:** An enzyme that increases with sympathetic nervous system activation (the "fight or flight" response). Higher levels indicate acute stress arousal.

**Attention Network Test (ANT):** A computerized task measuring three components of attention: alerting (maintaining readiness), orienting (selecting information), and executive control (resolving conflict). Reaction times and accuracy are recorded.

**Timing of measurements:**

Baseline (before the 7-week training)

Post-intervention (immediately after the 7-week training)

Follow-up (4 weeks after the training ended)

**Study design:** Randomized controlled trial (RCT) with two parallel groups: intervention (immediate happiness training) and waitlist control (no training during the study).

**Randomisation:** Participants were randomly assigned to either the intervention or control group. The paper states "randomized controlled trial" but does not provide details on the randomisation method (e.g., computer-generated random numbers, sealed envelopes). This is a minor weakness — without knowing the method, we cannot fully assess whether allocation was truly unpredictable.

**Blinding:** This was an open-label study. Participants knew whether they were in the training group or the waitlist group. The researchers administering the assessments were likely also aware of group assignments (no mention of blinding of outcome assessors). This is a significant limitation because self-report measures are highly susceptible to expectation effects — people who know they're receiving a "happiness training" may report being happier simply because they expect to be.

**Duration:**

Training period: 7 weeks

Total study duration: approximately 11 weeks (7 weeks training + 4 weeks follow-up)

Assessments at 3 time points: baseline, week 7 (post), week 11 (follow-up)

**Statistical approach:**

Used analysis of covariance (ANCOVA) to compare groups at post-intervention and follow-up, controlling for baseline scores

Reported Cohen's d effect sizes (standardized mean differences)

Used intention-to-treat analysis (participants analyzed in the groups they were assigned to, regardless of how much training they completed)

Reported p-values and effect sizes for all primary outcomes

**What this design can and cannot prove:**

*What it CAN prove:*

That the happiness training caused changes in self-reported outcomes compared to no intervention, assuming randomisation worked properly

That the effects persisted for at least 4 weeks after the training ended

The magnitude of these effects (large for happiness/satisfaction/well-being, moderate for stress/mindfulness/flourishing)

*What it CANNOT prove:*

That the training is better than an active placebo or alternative intervention (e.g., a different type of online training, a book, a support group). Without an active control, we cannot rule out that the benefits came from simply doing *something* structured and expecting improvement.

That the training causes objective physiological changes (the biological measures were null)

Long-term effects beyond 4 weeks of follow-up

That the training works for clinical populations (depression, anxiety disorders)

That the training works in non-German, non-corporate settings

**Major methodological weaknesses:**

1. **No active control group** — This is the most critical weakness. The waitlist design cannot control for placebo effects, attention effects, or demand characteristics. Participants who signed up for "happiness training" and received it likely expected to become happier.

2. **No blinding** — Both participants and researchers knew group assignments.

3. **Self-report only for primary outcomes** — The objective measures (cortisol, alpha-amylase, attention) showed no significant effects, raising questions about whether the self-report improvements reflect real change or just perceived change.

4. **Attrition not fully reported** — The paper mentions 147 participants at baseline but does not clearly state how many completed all assessments. Differential dropout between groups could bias results.

5. **Single geographic/cultural context** — German corporate employees; results may not generalize to other cultures or work environments.

**Primary outcomes (self-report):**

**Happiness (VAS):** Significant improvement in the intervention group compared to control. Effect size d = 0.93 (large). P = 0.000 (reported as p < 0.001). The intervention group's happiness scores increased from approximately 62/100 at baseline to 72/100 post-intervention, while controls remained stable around 60/100.

**Satisfaction (VAS):** Significant improvement. Effect size d = 1.17 (very large). P = 0.000. Satisfaction scores increased from approximately 58/100 to 70/100 in the intervention group.

**Quality of life/well-being (WHO-5):** Significant improvement. Effect size d = 1.06 (large). P = 0.000. WHO-5 scores increased from approximately 58/100 to 68/100 in the intervention group.

**Perceived stress (Stress Warning Signals):** Significant reduction. Effect size d = 0.64 (moderate). P = 0.003. Stress scores decreased from approximately 22/40 to 18/40 in the intervention group.

**Mindfulness (Freiburg Mindfulness Inventory):** Significant increase. Effect size d = 0.62 (moderate). P = 0.006. Mindfulness scores increased from approximately 38/56 to 42/56.

**Flourishing (Flourishing Scale):** Significant increase. Effect size d = 0.63 (moderate). P = 0.002. Flourishing scores increased from approximately 44/56 to 48/56.

**Recovery experience:** Significant increase. Effect size d = 0.42 (small-to-moderate). P = 0.030. Recovery experience scores increased modestly.

**Secondary outcomes (objective measures):**

**Salivary cortisol:** No significant differences between groups in morning cortisol, evening cortisol, or the cortisol awakening response. One isolated cortisol value (evening cortisol at follow-up) showed a significant difference, but this is likely a chance finding given multiple comparisons.

**Salivary alpha-amylase:** No significant differences between groups.

**Attention Network Test:** No significant differences between groups in alerting, orienting, or executive control reaction times or accuracy.

**Follow-up (4 weeks post-training):**

All significant self-report improvements were maintained at the 4-week follow-up, with effect sizes remaining similar or slightly reduced compared to post-intervention.

To translate these numbers into plain English:

**Happiness and satisfaction improved by about 10–12 points on a 100-point scale.** This is roughly equivalent to moving from "moderately happy" to "quite happy" — a noticeable shift in daily experience. The effect size (d = 0.93–1.17) is considered large in psychological research. For context, this is about the same magnitude of improvement seen in some antidepressant trials for depression, but in a non-clinical population.

**Perceived stress dropped by about 4 points on a 40-point scale (10% reduction).** This is a moderate effect (d = 0.64). To put it in perspective: if your stress level was at a 7/10, it might drop to about a 5.5/10. Noticeable, but not transformative.

**Mindfulness increased by about 4 points on a 56-point scale (7% increase).** This is a moderate effect (d = 0.62). Equivalent to going from "sometimes mindful" to "often mindful" on a few more items.

**Flourishing increased by about 4 points on a 56-point scale (7% increase).** Moderate effect (d = 0.63). A meaningful but not dramatic shift in overall psychological thriving.

**The objective measures showed zero meaningful change.** This is important: people *felt* better, but their stress hormones and attention performance did not change. This could mean the self-report improvements were partly placebo, or that the biological measures are less sensitive to these kinds of changes, or that 7 weeks is too short to produce physiological change.

**Acknowledged by authors:**

The study used a waitlist control rather than an active control, limiting the ability to attribute effects specifically to the happiness training content

Self-report measures are subject to social desirability and expectation bias

The sample was predominantly female and from two specific companies, limiting generalizability

The follow-up period was only 4 weeks

**Critical additional limitations:**

**No blinding whatsoever** — This is arguably the most serious limitation. When people know they're receiving a "happiness training," they are highly motivated to report being happier. The large effect sizes could partly reflect this expectation.

**No measurement of adherence** — The paper does not report how many exercises participants actually completed, or whether those who completed more exercises showed larger effects. Without this, we cannot know the minimum effective "dose."

**No active control for time/attention** — The control group received nothing. Any structured activity (even reading a book, taking a walk, or joining a weekly online group) might produce similar benefits.

**Objective measures were null** — The fact that cortisol, alpha-amylase, and attention showed no changes suggests the effects may be limited to subjective experience rather than underlying physiology.

**Multiple comparisons** — With many outcome measures, some significant results would be expected by chance. The authors did not adjust for multiple comparisons.

**Industry context** — The study was conducted in corporate settings where employees might feel pressure to report positive outcomes, especially if their employer was promoting the program.

**No long-term follow-up** — Four weeks is not long enough to know if effects persist or fade.

For someone running their own n=1 experiment:

**What to test:**

A structured 7-week positive psychology program combining: gratitude journaling (3 times/week, listing 3 things you're grateful for), identifying and using your character strengths (1 exercise/week), performing acts of kindness (2–3 times/week), savoring positive experiences (daily 5-minute reflection), and brief mindfulness meditation (10 minutes daily).

The specific exercises used in this study are common in positive psychology; you can find similar protocols in Martin Seligman's work or through free online resources like the Greater Good Science Center.

**Minimum meaningful duration:**

7 weeks appears to be sufficient for self-report changes. However, the null biological results suggest that physiological changes may require longer (12+ weeks) or more intensive practice.

For a self-experiment, commit to at least 8 weeks. Track weekly to see if changes are linear or appear suddenly.

**What to measure (specific metrics):**

**Primary:** A daily 0–100 happiness rating (simple VAS). Also a weekly 0–100 satisfaction rating. These are the most sensitive measures from the study.

**Secondary:** A brief stress scale (e.g., Perceived Stress Scale, 10 items, free online). A mindfulness scale (e.g., Freiburg Mindfulness Inventory short form). A flourishing scale (e.g., Flourishing Scale, 8 items).

**Optional but valuable:** Track something objective if possible — sleep quality (via a wearable), heart rate variability (HRV) via a chest strap, or daily step count. The study found no objective changes, but you might.

**Measure at baseline, weekly during the intervention, and at 4 weeks post-intervention.**

**Key confounds to control for:**

**Expectation effects:** You know you're doing a "happiness experiment," which biases your ratings. To partially control for this, add a "dummy" measure you don't expect to change (e.g., "how creative do you feel today?") to see if you're just rating everything more positively.

**Life events:** Major positive or negative events (new job, relationship change, illness)