Skip to main content

Table 8 Updated sample size calculations for the adolescent outcomes using the between-cluster coefficient of variation from the baseline survey (with missing responses multiply imputed)

From: The Well London program - a cluster randomized trial of community engagement for improving health behaviors and mental wellbeing: baseline survey results

Outcome Baseline prevalence or mean (across all intervention and control clusters) Km Minimum detectable effect size Expected effect size in original study design
Adolescents     
Healthy eating 56% eat fruit daily or almost daily 0.003 17% increase in prevalence 30% increase in prevalence
Unhealthy eating scoreb Mean unhealthy eating score = 3.0 0.06 Decrease of 0.26 standard deviations  
Physical activityc Mean PAQ-A score = 2.7 0.04 Increase of 0.25 standard deviations -
Mental health and wellbeing Mean PANAS-positive score = 29.8 (ranges 11 = lowest positive affect, to 55 = highest positive affect) 0.03 Increase of 0.23 standard deviations -
  Mean PANAS-negative score = 18.0 (ranges 11 = lowest negative affect, to 55 = highest negative affect) 0.02 Decrease of 0.21 standard deviations  
  Mean SDQ = 12.4 (range 0–15 =  normal, 16-19 = borderline, 20- 40 abnormal) 0.03 Decrease of 0.22 standard deviations 30% increase achieving key thresholds for mental health
  74% Have normal SDQ scores 0.12 31% increase in prevalence  
  29% have borderline or abnormal SDQ scores 0.04 14% decrease in prevalence  
  1. aEffect sizes for binary outcomes are relative increases in prevalence.
  2. aCalculations are based on multiply imputed datasets; comparison to complete cases showed no substantial differences except in Km for adolescent fruit consumption which decreased from 0.05 to 0.004.
  3. bPossible range 1 to 5; higher score indicates more frequent consumption of unhealthy food items (chips, chocolate or sweets, and sugar sweetened beverages).
  4. cRange 1 to 5; 1, very inactive; 5, very active.